Vol. 10 N. 3 SEPTEMBER 2022 ISSN: 2346-3775 Published by: In association with: International Center for The Tropical Crops Genetic Resources Tropical Agriculture Institute of The Chinese Academy of (CIAT), Cali, Colombia Tropical Agricultural Sciences (TCGRI-CATAS), Haikou, Hainan, P.R. China www.tropicalgrasslands.info International Center for Tropical Agriculture (CIAT) retains copyright of articles with the work simultaneously licensed under the Creative Commons Attribution 4.0 International License (to view a copy of this license, visit creativecommons.org/licenses/by/4.0/). Accordingly, users/readers are free to share (to copy, distribute and transmit) and to remix (to adapt) the work under the condition of giving the proper attribution. i CIAT received International Science and Technology Cooperation Award for scientific cooperation with CATAS Ambassador Li Junhua presented certificate to Juan Lucas Restrepo, Director General of the Alliance Bioversity-CIAT. Photo: CATAS On May 31st, 2022, the ceremony for the International Science and Technology Cooperation Award of the P.R. China was held at the Chinese Embassy in Italy. The Chinese Academy of Tropical Agricultural Sciences (CATAS), as a cooperation unit, attended the award ceremony online. On behalf of the Chinese government, Ambassador Li Junhua presented the medal and certificate to the International Center for Tropical Agriculture (CIAT). Li Junhua congratulated CIAT (now the Alliance Bioversity-CIAT) on winning the China International Science and Technology Cooperation Award, and fully affirmed the productive cooperation and fruitful achievements that CIAT, as an international organization, has carried out with CATAS for a long time. He looked forward to more cooperation between the Alliance and Chinese partners such as CATAS. He stressed that China will deepen international scientific and technological cooperation with the Alliance, share scientific public goods, strengthen global scientific research cooperation and jointly meet major global challenges. Juan Lucas Restrepo, Director General of the Alliance of Bioversity International and CIAT thanked the Chinese government for this award, reviewed the fruitful cooperation between CIAT and CATAS, and said that the Alliance would continue to vigorously deepen cooperation with China and jointly contribute to the sustainable development of global agriculture. On behalf of the Chinese cooperation unit, Liu Guodao introduced the cooperation achievements made by CATAS and CIAT in the cultivation of new varieties, technology promotion, platform co-construction and personnel training around the two major crops, tropical forages and cassava since 1982, and the co-publication of the Online Journal Tropical Forages-Forrajes Tropicales since 2012. He said that CATAS will further deepen cooperation with the Alliance, expand cooperation fields, carry out jointly research on agricultural biodiversity and sustainable development in tropical areas, and make new contributions to promote the higher level development of global tropical agriculture. Dr. Rainer Schultze Kraft, CIAT Emeritus Scientist and winner of the 2016 Friendship Award of the Chinese government, appreciated the cooperation experience, achievements and the profound friendship with CATAS staff over the past 40 years. He looked forward to the further cooperation between the two sides in the future in research on tropical forages and the continued support of the Tropical Grasslands-Forrajes Tropicales online journal. *Edited from catas.cn/EN/contents/1304/216020.html ii Michael Peters, Leader of CIAT Tropical Forages Program; Rainer Schultze-Kraft, then Editor of the TGFT Journal and the late Professor Bai Changjun, Head of the CATAS Tropical Pasture Research Center during the inception ceremony of the Journal. Chinese stylo pioneers: Liu Guodao, Chaozu He and the late Bai Changjun with Prof. Rainer Schultze-Kraft in Stylosanthes guianensis plots, Danzhou, Hainan, December 2015. While visiting CIAT, CATAS researcher Huan Hengfu receiving a box with several hundred of forage germplasm samples provided by the CIAT genebank. Palmira, August 2019. iii Editors Jean Hanson, Danilo Pezo, International Livestock Research Institute (ILRI), Tropical Agriculture Research and Higher Education Ethiopia Center (CATIE), Costa Rica Management Committee Robert J. Clements, Danilo Pezo, Agricultural Consultant, Australia Tropical Agriculture Research and Higher Education Center (CATIE), Costa Rica Liu Guodao, Chinese Academy of Tropical Agricultural Sciences Rainer Schultze-Kraft, (CATAS), P.R. China The Alliance of Bioversity International and CIAT, Colombia Jean Hanson, International Livestock Research Institute (ILRI), Cacilda B. do Valle, Ethiopia Empresa Brasileira de Pesquisa Agropecuária (Embrapa), Brazil Asamoah Larbi, Agricultural Consultant, Ghana Lyle Winks, Former editor of “Tropical Grasslands”, Australia Michael Peters, The Alliance of Bioversity International and CIAT, Kenya Editorial Board Caterina Batello, Huan Hengfu, Food and Agriculture Organization of the United Nations Chinese Academy of Tropical Agricultural Sciences (FAO), Italy (CATAS), P.R. China Robert J. Clements, Mario Herrero, Agricultural Consultant, Australia Commonwealth Scientific and Industrial Research Organisation (CSIRO), Australia Albrecht Glatzle, Iniciativa para la Investigación y Transferencia de Masahiko Hirata, Tecnología Agraria Sostenible (INTTAS), Paraguay University of Miyazaki, Japan Orlando Guenni, Peter Horne, Universidad Central de Venezuela (UCV), Venezuela Australian Centre for International Agricultural Research (ACIAR), Australia Jean Hanson, International Livestock Research Institute (ILRI), Johann Huguenin, Ethiopia Centre de Coopération Internationale en Recherche Agronomique pour le Développement (CIRAD), France Michael David Hare, Ubon Ratchathani University, Thailand iv Muhammad Ibrahim, T. Reginald Preston, Tropical Agriculture Research and Higher Education University of Tropical Agriculture Foundation (UTA), Center (CATIE), Costa Rica Colombia Asamoah Larbi, Kenneth Quesenberry, Agricultural Consultant, Ghana University of Florida, USA Carlos E. Lascano, H. Max Shelton, Universidad Nacional de Colombia - Sede Bogotá, The University of Queensland, Australia Colombia Werner Stür, Robert Paterson, Australian Centre for International Agricultural Research Agricultural Consultant, Spain (ACIAR), Australia Bruce Pengelly, Cacilda B. do Valle, Agricultural Consultant, Australia Empresa Brasileira de Pesquisa Agropecuária (Embrapa), Brazil Principal Contacts Jean Hanson International Livestock Research Institute (ILRI), Ethiopia Email: jeanhanson2010@gmail.com Danilo Pezo Tropical Agriculture Research and Higher Education Center (CATIE), Costa Rica Email: danilo.pezo@catie.ac.cr Tiziana Pagnani Statistical analysis The Alliance of Bioversity International and CIAT Colombia Email: t.pagnani@cgiar.org Technical Support José Luis Urrea-Benítez and Anny Isabella Yedra The Alliance of Bioversity International and CIAT Colombia Email: CIAT-TGFT-Journal@cgiar.org v Table of Contents Research Papers The influence of seed structures on dormancy in seeds of Urochloa hybrid cultivar ‘Mulato Ⅱ’ 156 Jinhai Liu, Hehua Wang, Fucheng Luo, Yan Wang, Cui Xu, Jinjuan Jiang Forage characterization of Carajás grass (Cenchrus purpureus × C. americanus) fertilized with a range of doses of protected urea under irrigation during the growing season 164 Francisco Gleyson da Silveira Alves, Maria Socorro de Souza Carneiro, Marcos Jacome Araújo, Rafael Felippe Ratke, Barbara Silveira Leandro de Lima, Nayrlon de Sampaio Gomes, Rafael Rodrigues da Silva, Ricardo Loiola Edvan Physiological, morphological, and biochemical characterization of Cratylia argentea (Desv.) Kuntze seeds 172 Eduardo Pacca Luna Mattar, Daniel Teixeira Pinheiro, Wander Douglas Pereira, Bruno Portela Brasileiro, Walter José Rodrigues Matrangolo, Paulo César Hilst, Paola Andrea Hormaza Martínez, Denise Cunha Fernandes dos Santos Dias Seasonal nutritive value and in vitro fermentation kinetics of foliage of some multipurpose shrub species in northeastern Mexico 184 Miguel Chávez Espinoza, Hugo Bernal Barragán, Maribel Guerrero Cervantes, Israel Cantú Silva, Mauricio Cotera Correa, Humberto González Rodríguez, Andrés Eduardo Estrada Castillón Benefit of feeding Urochloa hybrid cultivar ‘Cobra’ on milk production in Tanzania 195 Solomon Mwendia, An Notenbaert, Beatus Nzogela, Angello Mwilawa Effects of adding agro-industrial by-products and bacterial inoculant at ensiling on nutritional quality and bacterial colonization of Tifton 85 [Cynodon dactylon (L.) Pers.] silages 204 André Sanches de Avila, Maximiliane Alavarse Zambom, Andressa Faccenda, Marcela Abbado Neres, Luana Muxfeldt, Cibele Regina Schneider, Marcelo Martini Stum, Ricardo Dri, Pâmela Rosana Schneider Influence of plant population density of Chamaecrista rotundifolia on its value for hay making in the Eastern Amazon, Brazil 214 Angélica Lucelia da Silva Nascimento, Natan Lima Abreu, Raimundo Vagner de Lima Pantoja, Ingrid Stefanie Queiroz de Oliveira, Josilene do Nascimento Gomes, René Jean Marie Poccard Chapuis, Letícia de Abreu Faria The role of leucaena in cattle fattening and breeding production systems in Eastern Indonesia 222 Fahrul Irawan, Dahlanuddin, Michael J. Halliday, Roger S. Hegarty, Frances C. Cowley The impact of COVID-19 on the sustainable intensification of forage-based beef and dairy value chains in Colombia: a blessing and a curse 237 Stefan Burkart, Manuel Díaz, Karen Enciso, Andrés Charry, Natalia Triana, Martín Mena, José Luis Urrea-Benítez, Irieleth Gallo Caro, Rein Van Der Hoek Predicción del valor nutricional de sorgo para forraje mediante espectroscopia de reflectancia en el infrarrojo cercano (NIRS) y ecuaciones empíricas 249 Sonia Pereira-Crespo, Adrián Botana, Marcos Veiga, César Resch, Laura González, Roberto Lorenzana, Valentín García- Souto, María del Pilar Martínez-Diz, Gonzalo Flores-Calvete Ingestive behavior and dry matter intake of dairy cattle grazing Kikuyu grass (Cenchrus clandestinus) pastures 261 Yesid Avellaneda-Avellaneda, Edgar Mancipe-Muñoz, Juan Vargas-Martínez Digital imaging outperforms traditional scoring methods for spittlebug tolerance in Urochloa humidicola hybrids 271 Luis Hernández, Paula Espitia, Juan Andrés Cardoso vi Regional Communications Nitrogen and phosphorus fertilizer application to Elephant grass (Cenchrus purpureus syn. Pennisetum purpureum) cultivar ‘Cameroon’ in an arenosol in Rio Grande do Norte, Brazil 280 Luiz E.C. Oliveira, Fábio H.T. Oliveira, Gualter G.C. Silva, Marcio G. Silva Bezerra, Éric G. Morais, Gabriel F.R. Bezerra, Giovana S. Danino, Ermelinda M.M. Oliveira, Francisco V.S. Sá Forage production and quality of Urochloa decumbens cultivar ‘Basilisk’ in Okinawa, Japan 288 Takashi Hanagasaki Short Communications Effects of feeding dried olive (Olea europaea) leaves with wheat straw-concentrate rations on feed conversion efficiency in Awassi rams 297 Mazen Alomar, M. Rateb Al-Masri, Moutaz Zarkawi vii Tropical Grasslands-Forrajes Tropicales (2022) Vol. 10(3):156–163 156 doi: 10.17138/TGFT(10)156-163 Research Paper The influence of seed structures on dormancy in seeds of Urochloa hybrid cultivar ‘Mulato Ⅱ’ La influencia de las estructuras de semillas sobre la dormancia en semillas de Urochloa híbrido cultivar 'Mulato Ⅱ' JINHAI LIU1,2, HEHUA WANG1,2, FUCHENG LUO1, YAN WANG1, CUI XU1 AND JINJUAN JIANG1 1College of Animal Science and Technology, Yunnan Agricultural University, Kunming, China. ynau.edu.cn 2College of Animal Science, Xichang University, Xichang, China. xcc.edu.cn Abstract This study determined the effects of seed structures on seed dormancy and tested methods to break dormancy in seeds of Urochloa hybrid cultivar ‘Mulato II’. Seeds stored for 10 months in indoor ambient conditions were studied to determine effects of seed structures on seed germination and their water permeability. Results showed that seed structures presented a barrier to water permeability. Removal of lemmas, puncturing the seed coat, seed structure removal and sulfuric acid immersion all reduced seed dormancy. Water and alcohol extracts from different parts of seeds inhibited seed germination of Brassica pekinensis seeds. There were 3 mechanisms responsible for seed dormancy; first, the mechanical barrier of seed structures, which excluded water and reduced gas exchange as well as restricting growth of the embryo; second, an endogenous germination inhibitor mainly found in lemmas; and third, water permeability of the seed coat (including pericarp and testa). The mechanical removal of lemmas and immersion in concentrated sulfuric acid reduced seed dormancy, although mechanical removal of the lemma alone was effective, convenient and safer. Keywords: Seed extracts, seed germination, vitality index, water permeability. Resumen Este estudio determinó los efectos de las estructuras de semillas en su dormancia y probó métodos para romper la dormancia del cultivar híbrido de Urochloa 'Mulato II'. Se estudiaron semillas almacenadas durante 10 meses en condiciones de ambiente interior para determinar los efectos de las estructuras de semillas sobre su germinación y permeabilidad al agua. Los resultados mostraron que las estructuras de las semillas presentan una barrera a la permeabilidad al agua. La eliminación de los lemas, la perforación de la cubierta, la eliminación de la estructura de la semilla y la inmersión de ácido sulfúrico redujeron la dormancia de la semilla. Extractos de agua y alcohol de diferentes partes de semillas inhibieron la germinación de semillas de Brassica pekinensis. Hubo 3 mecanismos responsables por la dormancia de las semillas: En primer lugar, la barrera mecánica de las estructuras de las semillas, que excluía el agua y reducía el intercambio de gases, además de restringir el crecimiento del embrión; En segundo lugar, un inhibidor endógeno de la germinación que se encuentra principalmente en los lemas; Y tercero, la permeabilidad al agua de la cubierta de la semilla (incluyendo pericarpio y testa). La eliminación mecánica de los lemas y la inmersión en ácido sulfúrico redujo la dormancia de las semillas, aunque la sola eliminación mecánica del lema fue efectiva, conveniente y segura. Palabras clave: Extractos de semillas, germinación de semillas, índice de vitalidad, permeabilidad al agua. Correspondence: Fucheng Luo, College of Animal Science and Technology, Yunnan Agricultural University, Kunming, 65220061, China. Email: lfc-999@126.com Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) Reasons for seed dormancy of Urochloa hybrid cultivar‘Mulato Ⅱ’ 157 Introduction reported that U. brizantha cultivar ‘MG-5’ seeds aged at 41 °C for 96 h with 0.2 % KNO3 or 0.2 % calcium nitrate Grasses of the genus Urochloa include annual or as primers exhibited a seed germination rate of 94 %. perennial forages for cut and carry or grazing systems, Usberti (2007) reported that dormancy was reduced as well as for water and soil conservation, in tropical in seeds of U. brizantha by soaking in concentrated and subtropical regions of the world (Thomas et al. 1987; sulfuric acid for 15 min and storing at 40 °C while 15 Li and Guo 1990). The cultivated species of Urochloa min of concentrated sulfuric acid treatment also reduced originated from African savannas and are now widely the dormancy of U. decumbens seeds (Duan et al. 2015). grown in tropical and subtropical areas in Africa, the Martins and da Silva (2001) reported that treatment Americas, Oceania, and south-east Asia (Kobayashi and of Marandu seeds with concentrated sulfuric acid Kato-Noguchi 2015; Simeão et al. 2016; Lozano et al. for 15 min reduced seed dormancy while Garcia and 2017; Castañeda-Pimienta et al. 2017; Moreira et al. 2018). Cícero (1992) showed that the best method to break the The primary cultivated species include U. ruziziensis, dormancy of Marandu seeds was to soak in sulfuric acid U. brizantha, U. decumbens, U. humidicola, and for 15 min followed by soaking in 0.2 % KNO3 solution. U. dictyoneura. The CIAT-bred Urochloa hybrid cultivar Batista et al. (2016a) showed that after 5 min of treatment ‘Mulato Ⅱ’ (CIAT36087, U. ruziziensis × U. decumbens with concentrated sulfuric acid and 3 h of hydration with × U. brizantha) became widely used after promotion 0.5 mg gibberellic acid/L, seeds of U. brizantha showed by Semillas Papalotla S.A. seed company (Argel et al. higher physiological potential and seedling emergence 2005; Argel et al. 2007; Phaikaew et al. 2008). Urochloa rate. Hare et al. (2014) reported that seeds of Mulato II seed production is mostly in Thailand. Mulato Ⅱ was treated with concentrated sulfuric acid for 10 min had a introduced to provinces in China (Hainan, Fujian and germination of 80 % and Pereira et al. (2017) reported Yunnan) in 2005 (Li et al. 2009; Zeng et al. 2009; Liu et that when scarified seeds of Mulato II were aged at al. 2013; Deng et al. 2013). 42℃ and 98 % humidity for 48 hours, their germination Mulato Ⅱ shows good spittle bug resistance with ability was enhanced. high yields of good quality forage, strong tillering The seed structure of Mulato II consists of a lemma ability, a well-developed root system and creeping and and a seed coat (including pericarp and testa) (Figure erect growth characteristics. Despite excellent forage 1). There are few studies focussing on the relationship attributes, adoption is hindered by seed dormancy. between seed structures and seed dormancy. In this Better understanding of dormancy mechanisms and study, we sought to measure the effects of seed treatments development of dormancy-breaking technologies should on the dormancy of Mulato II seeds after storage for 10 promote adoption of Mulato Ⅱ. months. We aimed to study inhibitory effects of ‘Mulato Baskin and Baskin (2004) proposed a seed dormancy Ⅱ’ seed structure extracts on Brassica pekinensis seeds, classification, which divided seed dormancy into five reveal the influence of the seed structures on seed types: physiological, morphological, morphological germination and highlight the most appropriate method physiological, physical and compound dormancy. Studies to break seed dormancy, using treatments including have been reported on chemical treatment, osmotic mechanical scarification and concentrated sulfuric acid. regulation and aging treatment for breaking dormancy in Urochloa seeds (Whiteman and Mendra 1982; Materials and Methods Câmara et al. 2002; Bonome et al. 2006; Batista et al. 2016a). Whiteman and Mendra (1982) reported that the Seed characteristics germination rate of U. decumbens was 72 % after 20 min of treatment with concentrated sulfuric acid (H2SO4) Urochloa hybrid cultivar ‘Mulato II’ seeds were harvested compared to 40 % of intact stored seed. Costa et al. (2011) in SiMaoGang, Simao, Pu’er, Yunnan, China (22°30' reported that the germination rate of shelled U. brizantha N, 100°35' E; 1,200–1,250 masl). The seeds were dried, cultivar ‘Marandu’ seeds was less than 20 %, while the cleaned and stored indoors under ambient conditions for germination rate of denuded seeds reached 60 %. Bonome 10 months (from November 2018 to August 2019). et al. (2006) showed that the germination uniformity of Random seed samples were prepared for the study Marandu seeds immersed in potassium nitrate (KNO3) using a riffle seed divider. Shape and colour of seeds were solution for 12 h was better than for untreated seeds or observed using a magnifying glass. Three replicates of seeds immersed in other solutions. Batista et al. (2016b) 100 randomly selected seeds were measured for length, Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) 158 J. Liu, H. Wang, F. Luo, Y. Wang, C. Xu and J. Jiang width and thickness using a Vernier calliper. Thousand- The top of paper germination method was used with 3 seed weight was determined by weighing 10 replicates replicates of 100 seeds in 90 mm petri dishes (Yan 2017). of 100 randomly selected seeds. The moisture content Seeds were germinated at alternating temperatures of of seeds was determined using the International Seed 30/20 °C (16/8 h) with 12/12 h light in an incubator for Testing Association (ISTA) constant-temperature oven 14 d with regular watering. Dishes were observed daily drying method at 130 ⁰C (Fan et al. 2016; Yan 2017). to assess germination, which was defined as when the radicle protruded to 1 mm. Germination was counted Seed treatments/scarification after 7 and 14 days. After germination, the seedlings were placed in an oven for 15 min at 105 °C and dried Seeds were mechanically scarified as: at 65 °C for 24 h until a constant weight was reached. Punctured seeds with lemmas (A): the seed was The dry weight was noted and the germination percent, held with forceps and the lemmas and seed coat were germination potential, germination index, and vitality punctured with a dissecting needle. index of the seeds were calculated according to the Seeds without lemmas (B): the seed was held with following formulas (Hu et al. 1992): forceps and the seed coat and lemma were separated with Germination potential (GP) = total number germinated a dissecting needle and the lemma removed with forceps. within 7 days/number of seeds tested × 100 % Punctured seeds without lemmas (C): after removing Germination precent (GR) = total number germinated the lemma, the seed coat was punctured with a dissecting within 14 days/number of seeds tested × 100 % needle. Germination index (GI) = ∑Gt / Dt Naked seeds (D): the seed structures were removed Vitality index (VI) = GI × S by peeling the seed structures (lemma and seed coat) where: using tweezers. Gt is the number germinated in T days Control: intact seeds with full seed structures. Dt is the corresponding number of germination days For acid scarification, 3 replicates of 100 intact seeds S is the dry weight of a single seedling. were scarified with 98 % sulfuric acid for 5, 10 or 15 min, or left untreated as control. After repeated rinsing Effect of extracts from different parts of seeds on with distilled water, the germination test was conducted. germination of Brassica pekinensis Determination of seed permeability Effect of extracts from different parts of the seeds as germination inhibitors were assessed uing the methods of The water absorption rate of 100 seeds each of the Liu (2015) and Yan (2017). The seed parts were separated control, treatments A, B, C and D was assessed by into lemmas, seed coat (including pericarp and testa) and placing them in different beakers, adding 3 ml of caryopsis. Each sample of 1 g was crushed in a mortar and distilled water to each and soaking at room temperature 10 ml of distilled water was added to each in beakers which (15 °C - 25 °C) for 0, 2, 4, 6, 8, 12, 16, 20, 24, 36, and 48 were sealed with plastic wrap at 4 °C. After leaching for h. After soaking, surface water was removed by blotting 24 h, the residue was filtered and extracted twice more with filter paper and seeds were immediately weighed. by rinsing with distilled water. The 3 leachates for each The water absorption rate of the seeds was calculated seed part were merged and made up to a volume of 50 ml according to the method of Luo et al. (2014). with distilled water. The concentration of the leachates Water absorption rate of seeds (%) = (W2-W1)/W1×100 was 20 mg/ml. A 15 ml sample was diluted with distilled where: water to a concentration of 10 mg leachate/ml for later W1 is the weight of the seeds before water absorption, use. Another set of 1 g samples of separated lemmas, seed and W2 is the weight after water absorption. coat and caryopses were crushed, extracted with 20 ml of methanol and leached at 4 °C for 24 h as above. The residue Germination was filtered, extracted twice more with methanol and the 3 leachates for each seed part were merged. The methanol Seeds from each of the control and treatments A, B, C was vaporized in a furnace in a fume hood to obtain solid and D were soaked and disinfected with a 2 % copper crystals. The crystals were re-dissolved in 50 ml of distilled sulphate solution for 10 min, washed repeatedly with water and 15 ml of the solution was taken and diluted with distilled water and air-dried before germination. distilled water to 10 mg leachate/ml for later use. Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) Reasons for seed dormancy of Urochloa hybrid cultivar‘Mulato Ⅱ’ 159 Three replicates of 100 Brassica pekinensis seeds 1). The average seed length, width, height, thousand seed per treatment were placed on filter paper moistened weight and water content are shown in Table 1. Seed with 3 ml of the different leachates and a control using vitality was 84.38 and initial germination rate was 21 %. 3 ml of distilled water in 9 cm petri dishes. The dishes were placed in an incubator at alternating temperatures Seed permeability of 30/20 °C (16/8 h) with light cycles of 12/12 h for 4 days. Observations were made every day until the end of Water absorption rate increased rapidly for 2 h followed the fourth day. If distilled water was added to the petri by a slow rising trend up to 16 h and reached saturation dish to prevent drying out, the number of germinated after 24 h (Figure 2) and was affected by treatment. B. pekinensis seeds was counted after 48 h. The radicle length of B. pekinensis was measured using Vernier Effects of removing seed structures on seed germination callipers after 72 h. The inhibition rate was calculated according to the method of Wang (2017): The 4 mechanical seed treatments all significantly Inhibition rate (%) = (germination of control group – improved seed germination compared to the control germination of treatment group)/germination of control (P<0.05) (Table 2). Removing the seed structures produced group × 100 %. the highest germination rate (82.67 %) and seed vitality (84.38 %). The germination potential, germination rate, Data analysis germination index and vitality index of treatments were significantly different from the control (P<0.05). Excel 2010 and SPSS 21.0 were used for data collation and statistical analysis. SPSS 21.0 was used for one-way Effects of concentrated sulfuric acid treatment on seed variance (ANOVA) analysis and Duncan's test was used germination to compare the differences among treatments. Seeds soaked in sulfuric acid for 5, 10 and 15 min showed Results significantly increased seed germination potential, germination rate, germination index and vitality index Seed characteristics compared to the control (Table 3). Soaking in concentrated sulfuric acid for up to 15 min eroded the lemma while The seeds of Mulato Ⅱ were flat, ovoid and light yellow, the seed coat was still well-preserved. Five minutes of showing a ventral uplift and abaxial flat surface (Figure soaking was sufficient to increase seed germination rate, a = Back of spikelet; b = Ventral spikelet; c = Caryopsis appendage; d = Caryopsis; e = Naked seed (without shells); f = Pattern of spikelet slitting along dorsal ventral axis. 1 = Lemma; 2 = Glume; 3 = Rachilla; 4 = Palea; 5 = Seed coat (including pericarp and testa); 6 = Endosperm; 7 = Embryo. Figure 1. The morphology and appendages of Urochloa hybrid cultivar ‘Mulato Ⅱ’ seeds. Table 1. Seed characteristics of Urochloa hybrid cultivar ‘Mulato Ⅱ’ Length (mm) Breadth (mm) Height (mm) 1000 seed weight (g) Water content (%) Vitality (%) Germination percent (%) 5.29 ± 0.21 1.96 ± 0.14 1.21 ± 0.12 6.35 ± 0.02 8.96 ± 0.06 84.38 ± 1.93 21.00 ± 2.67 Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) 160 J. Liu, H. Wang, F. Luo, Y. Wang, C. Xu and J. Jiang germination rate and vitality index. The increases in seed all had inhibitory effects on the seed germination germination rate, germination potential rate, germination of B. pekinensis (Table 4). The degree of inhibition index and vitality index of seeds decreased slightly, but decreased with decreasing concentration. The 20 not significantly, with the extension of the soaking time mg leachate/ml water extract from the lemmas had with concentrated sulfuric acid. Significant differences the strongest inhibitory effect on germination of were only seen with soaking for 15 min (P<0.05). B. pekinensis and was significantly different from the control (P<0.05). The caryopsis alcohol extract (10 Inhibitory effect of extracts from different parts of mg/ml) showed the weakest inhibitory effect and was seeds on germination of Brassica pekinensis seeds not significantly different compared with the control, although the root length was significantly different Water and alcohol extracts from different parts of the (P<0.05). 50 a a a a a a a a Intact seeds with 45 a lemmas 40 b b b b b b Punctured seeds 35 a b b c c c with lemmas 30 b c c b c c c d d d Seeds without 25 b b c c c c d d d lemmas 20 c c d d Punctured seeds 15 c d without lemmas 10 d d Naked seeds 5 d without shell 0 d 0 2 4 6 8 10 16 20 24 36 48 Soaking time (hours) Note: Different letters in same soaking time are significantly different (P<0.05). Figure 2. The change in water absorption by seeds of Urochloa hybrid cultivar ‘Mulato Ⅱ’. Table 2. Effect of mechanical treatment on germination of Urochloa hybrid cultivar ‘Mulato Ⅱ’ seeds. Treatment Germination potential (%) Germination percent (%) Germination index Vitality index Control 22.67±2.40 e 26.67±1.76 e 6.20 ±1.20 e 0.044 ±0.004 e A 32.67±2.40 d 32.67±2.40 d 12.44 ±0.97 d 0.060 ±0.003 d B 64.00±1.15 b 64.00±1.15 b 30.97±1.23 b 0.176 ±0.007 a C 50.00±2.31 c 50.00±2.31 c 24.03±1.03 c 0.130 ±0.005 b D 82.67±2.40 a 82.67±2.40 a 40.28± 1.53 a 0.084 ±0.003 c Note: Values are given as mean ±SE. Values with different letters in the same column are significantly different (P<0.05). Table 3. Effect of chemical treatment on germination of Urochloa hybrid cultivar ‘Mulato Ⅱ’ seeds. Treatment Germination potential (%) Germination percent (%) Germination index Vitality index Untreated control 22.67 ± 2.40 c 26.67 ± 1.76 c 6.20 ± 1.20 c 0.044 ± 0.004 c Soaking in concentrated 78.00 ± 3.16 a 78.00 ± 3.16 a 37.06 ± 0.83 a 0.110 ± 0.002 a sulfuric acid for 5 min Soaking in concentrated 74.00 ± 2.46 a 74.00 ± 2.46 a 36.16 ± 0.89 a 0.104 ± 0.004 a sulfuric acid for 10 min Soaking in concentrated 71.33 ± 1.67 b 71.33 ± 1.67 b 34.12 ± 0.62 b 0.092 ± 0.003 b sulfuric acid for 15 min Note: Values are given as mean ±SE. Values with different letters in the same column are significantly different (P<0.05). Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) Warter absorption (%) Reasons for seed dormancy of Urochloa hybrid cultivar‘Mulato Ⅱ’ 161 Table 4. Effect of extracts from different parts of Brachiaria hybrid cultivar ‘Mulato Ⅱ’ on germination of Brassica pekinensis seeds. Extracts from different parts Concentration (mg/ml) Germination inhibition (%) Root length inhibition (%) Distilled water (control) 0 c 0 f Water extract of lemmas 20 19.33 ± 6.36 a 68.98 ± 1.73 a 10 2.00 ± 0 bc 57.13 ± 3.57 ab Alcohol extract of lemmas 20 7.33 ± 2.40 b 49.18 ± 3.04 bc 10 2.67 ± 1.33 bc 42.98 ± 0.92 c Water extract of seed coat 20 2.33 ± 1.33 bc 48.76 ± 4.20 bc 10 2.00 ± 1.15 bc 45.58 ± 3.51 bc Alcohol extract of seed coat 20 1.33 ± 1.33 bc 9.77 ± 1.18 e 10 0.67 ± 0.67 bc 8.69 ± 2.23 e Water extract of naked seeds 20 2.00 ± 0 bc 45.04 ± 4.61 bc 10 0.67 ± 0.67 bc 28.59 ± 2.95 d Alcohol extract of naked seeds 20 1.33 ± 0.67 bc 26.09 ± 0.51 d 10 0.67 ± 0.0.67 bc 8.75 ± 1.77 e Note: Values are given as mean ±SE. Values with different letters in the same column are significantly different (P<0.05). Discussion Lemmas, seed coat and caryopses all contain substances that inhibit germination in higher concentrations. Water This study showed that seed structures caused water extract from lemmas was most effective with extract absorption barriers. While the seed coat did not hinder from caryopses the least effective, indicating that the water absorption, the presence of lemmas promoted endogenous inhibitor was mainly in the lemmas. This is water absorption to the seed overall and helped the seed consistent with results of Duan et al. (2015). Ajala-Luccas bind more water. The finding that seed structures were et al. (2018) showed that dormancy of U. humidicola seed involved in seed dormancy is consistent with research is due to synergistic effects between age, GA/ABA balance from Whiteman and Mendra (1982), who suggested and residual structure of the panicoid spikelet covering that the mechanical barrier of the U. decumbens seed caryopses. Mechanical removal of lemmas effectively tegument, which restricts access of oxygen and water, is reduced the dormancy of Urochloa hybrid cultivar a cause of dormancy. Câmara et al. (2002) also believed ‘Mulato II’ seeds and resulted in a higher seed vitality that the seed shell of U. brizantha cutlivar ‘Marandu’ index than use of concentrated sulfuric acid for 5 min. It restricted gas exchange and inhibited seed germination, is more convenient, simpler and safer to remove lemmas resulting in seed dormancy. Duan et al. (2015) showed mechanically than to immerse the seeds in sulfuric acid. that impermeability of the seed coat was the main reason Lemmas can be removed by rubbing seeds on rough for seed dormancy in U. decumbens. Therefore, the surfaces, with coarse sand or beating. permeability of seed structures may be one of the causes of dormancy, but not the main cause. Conclusions The results showed that mechanical treatment promoted seed germination, with removal of lemmas This study showed that removal of lemmas and soaking giving the highest germination. While puncturing seeds with concentrated sulfuric acid for 5 min was more the seed coat and leaving the lemmas intact promoted effective at breaking seed dormancy than soaking for 10 seed germination, it was not as effective as treatments or 15 minutes. After 5 min, the germination percent, involving removing lemmas, indicating that lemmas germination index, and vitality index of Mulato II seeds were the main cause of dormancy in this species. decreased with the increase in soaking time. Soaking seeds in concentrated sulfuric acid promoted There were 3 main reasons for the dormancy of seed germination because acid eroded the lemmas and Urochloa hybrid cultivar ‘Mulato Ⅱ’ seed: mechanical increased the permeability of the seed coat. Similar barrier of the seed coat and lemmas, endogenous inhibitor results have been found in other studies (Whiteman and (mainly found in lemmas) and water permeability of the Mendra, 1982; Câmara et al. 2002; Duan et al. 2015). seed coat (including pericarp and testa). Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) 162 J. Liu, H. Wang, F. Luo, Y. Wang, C. Xu and J. Jiang Acknowledgments Costa CJ; Araújo RB de; Bôas HDCV. 2011. Treatments for seed dormancy release in Brachiaria humidicola (Rendle) This work was supported by the Modern Agricultural Schweick. Pesquisa Agropecuaria Tropical 41(4):519–524. Grass Industry Technology System Construction Project (In Portuguese) doi: 10.5216/pat.v41i4.15100 in Yunnan Province (2017KJTX0018). Deng NJ; Dong ZS; Li DR; Pu GB; Liu JP; Liu JX; Li M; Guan HW; Zhang JH. 2013. A comparative experiment of 4 References introduced species of Brachiaria genus in Xishuangbanna. Pratacultural Science 30(5):821–825. (In Chinese). bit.ly/ 3QbPqmJ (Note of the editors: All hyperlinks were verified 5 September 2022). Duan XH; Luo FC; Zhong S; Li QX; Xue SM; Yang GR; Huang Ajala-Luccas D; Ribeiro-Oliveira JP; Duarte Silveira LE; BZ. 2015. A preliminary study on causes of seed dormancy Da Silva EAA. 2018. An integrative insight on dormancy of Brachiaria decumbens. Grassland and Turf 35:49–52. alleviation in diaspores of Urochloa humidicola (Rendle) (In Chinese). doi: 10.13817/j.cnki.cyycp.2015.01.009 Morrone & Zuloaga, a tropical grass with great economic Fan SG; Wang YR; Baskin CC; Baskin JM; Hu XW. 2016. A and ecological impact. Plant Biology 20(2):252–262. doi: rapid method to determine germinability and viability of 10.1111/plb.12655 Agriophyllum squarrosum (Amaranthaceae) seeds. Seed Argel PJ; Miles JW; Guiot JD; Lascano CE. 2005. Cultivar Science and Technology 44(2):410–415. doi: 10.15258/ mulato (Brachiaria híbrido CIAT 36061). Gramínea sst.2016.44.2.01 de alta producción y calidad forrajera para los trópicos. Garcia J; Cícero SM. 1992. Overcoming seed dormancy in Centro Internacional de Agricultura Tropical (CIAT), Brachiaria brizantha cv. Marandu. Scientia Agricola Cali, Colombia. hdl.handle.net/10568/72142 49(1):9–13. (In Portuguese). doi: 10.1590/S0103- Argel PJ; Miles JW; Guiot JD; Lascano CE. 2007. Cultivar 90161992000400003 mulato II (Brachiaria híbrido CIAT 36087): Gramínea de Hare MD; Phengphet S; Songsiri T; Sutin N; Stern E. 2014. alta calidad y producción forrajera, resistente a salivazo y Germination of tropical forage seeds stored in ambient adaptada a suelos tropicales ácidos bien drenados. Centro and controlled temperature and humidity conditions. Internacional de Agricultura Tropical (CIAT), Cali, Tropical Grasslands-Forrajes Tropicales 2(1):74–75. doi: Colombia. hdl.handle.net/10568/69705 10.17138/TGFT(2)74-75 Baskin JM; Baskin CC. 2004. A classification system for Hu J; Shen JP; Ruan SL. 1992. Application of material seed dormancy. Seed Science Research 14(1):1–16. doi: efficiency in the determination of rice seed vigor. Seed 10.1079/SSR2003150 11:58–59. doi: 10.16590/j.cnki.1001-4705.1992.04.054 Batista TB; Binotti FFS; Cardoso ED; Costa E; Nascimento Kobayashi A; Kato-Noguchi H. 2015. The seasonal variations DM do. 2016a. Appropriate hydration period and chemical of allelopathic activity and allelopathic substances in agent improve priming in Brachiaria seeds. Pesquisa Brachiaria brizantha. Botanical Studies 56:25. doi: Agropecuaria Tropical 46(3):350–356. doi: 10.1590/1983- 10.1186/s40529-015-0105-7 40632016v4638422 Li BH; Guo ZY. 1990. The growth of Brachiaria decumbens in Batista TB; Cardoso ED; Binotti FFS; Costa E; Sá ME de. Simao. Yunnan Journal of Animal Science and Veterinary 2016b. Priming and stress under high humidity and Medicine 3:36–37. bit.ly/3dogO3A temperature on the physiological quality of Brachiaria Li ZD; Gao GX; Bai CJ. 2009. Comparison trial of 6 cultivars brizantha cv. MG-5 seeds. Acta Scientiarum Agronomy of Brachiaria spp. Chinese Journal of Tropical Crops 38(1):123–127. doi: 10.4025/actasciagron.v38i1.26021 30(8):1186–1189. bit.ly/3RrGKcO Bonome LTS; Guimarães RM; Oliveira JA; Andrade Liu JH; Wang HH; Zou YM; Huang BZ; Chang N; Liu GD; Zhou VC; Cabral PS. 2006. Effect of osmoconditioning in C. 2013. Evaluating the adaptation of 14 Brachiaria varieties seeds of Brachiaria brizantha cv. Marandu. Ciência in southern subtropical areas of Yunnan. Acta Prataculturae e Agrotecnologia 30(3):422–428. (In Portuguese). Sinica 22(3):60–69. (In Chinese). bit.ly/3cKzJ8n doi:10.1590/S1413-70542006000300006 Liu WY. 2015. The study on seed dormancy and breaking Câmara HHLL; Stacciarini-Seraphin E. 2002. Germination mechanisms of Medicago Truncatula. PhD Thesis, Gansu of seeds of Brachiaria brizantha cv. Marandu under Agricultural University, Lanzhou, Gansu Province, different storage periods and hormonal treatment. Pesquisa P.R.China. bit.ly/3xwqIqO Agropecuária Tropical 32(1):21–28. (In Portuguese). bit.ly/ Lozano MC; Martinez NM; Diaz GJ. 2017. Content of the 3RdB4DG Saponin protodioscin in Brachiaria spp. from the Eastern Castañeda-Pimienta L; Olivera-Castro Y; Wencomo- Plains of Colombia. Toxins 9(7):220. doi: 10.3390/ Cardenas HB. 2017. Agronomic evaluation and selection toxins9070220 of Brachiaria spp. accessions on moderate fertility soils. Luo FC; Beng J; Duan XH; Xu WH; He C; Zhang YD; Fan SR; Pastos y Forrajes 40(4):272–276. bit.ly/3Rr8YUY Guo FG. 2014. Study the seed permeability and embryo Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) Reasons for seed dormancy of Urochloa hybrid cultivar‘Mulato Ⅱ’ 163 dormancy of Setaria sphacelata cv. Narok produced in Simeão R; Silva A; Valle C; Resende MD; Medeiros SP. Yunnan Province. Grassland and Turf 34(4):19–24. (In 2016. Genetic evaluation and selection index in tetraploid Chinese) doi: 10.13817/j.cnki.cyycp.2014.04.004 Brachiaria ruziziensis. Plant Breeding 135(2):246–253. Martins L; da Silva WR. 2001. Dormancy performance of doi: 10.1111/pbr.12353 Brachiaria brizantha seeds submitted to thermal and Thomas D; Grof B; Li GD; Chen LW. 1987. Grass of the chemical treatments. Pesquisa Agropecuária Brasileira savanna regions of South America: Antropogon gayanus, 36(7):997–1003. (In Portuguese) doi: 10.1590/S0100- Brachiaria decumbens, and Panicum maximum. Animal 204X2001000700010 Science Abroad 6:11–19. (In Chinese) doi: 10.13817/j.cnki. Moreira EA; Souza SM de; Ferreira AL; Tomich TR; cyycp.1987.06.005 Azevêdo JAG; Sobrinho FS; Benites FRG; Machado Usberti R. 2007. Performance of tropical forage grass FS; Campos MM; Pereira LGR. 2018. Nutritional (Brachiaria brizantha) dormant seeds under controlled diversity of Brachiaria ruziziensis clones. Ciência Rural storage. Seed Science and Technology 35(2):402–413. doi: 48(2):e20160855. doi: 10.1590/0103-8478cr20160855 10.15258/sst.2007.35.2.15 Pereira DDS; Carvalho MLM de; Guimarães RM; Oliveira Whiteman P; Mendra K. 1982. Effects of storage and seed AS; Reis GA. 2017. Physiological changes in hybrid treatments on germination of Brachiaria decumbens. Brachiaria cv. Mulato II after accelerated aging to Seed Science and Technology 10(2):233–242. overcome dormancy. Journal of Seed Science 39(3):244– Yan F. 2017. Studies on seed dormancy mechanism of 253. doi: 10.1590/2317-1545v39n3171313 Daphne giraldii Nitsche. PhD Thesis. Gansu Agricultural Phaikaew C; Pholsen P; Nakamanee G; Poonpipat W. 2008. University, Lanzhou, Gansu Province, P.R. China. bit.ly/ Seed production and seed quality of Brachiaria hybrids 3DwHyte (Mulato and Mulato Ⅱ) at different sites of Thailand. Zeng RQ; Lin YS; Hong JJ; Huang YB. 2009. Study on the growth Proceedings of the XXI International Grassland Congress characteristic and relative feeding value of four varieties of and the VIII International Rangeland Congress, Hohhot, Brachiaria decumbens in Southern Fujian. Pratacultural P.R. China, June 29-July 5, 2008. bit.ly/3KJoTvV Science 26(8):107–111. (In Chinese). bit.ly/3cJoGMK (Received for publication 03 January 2021; accepted 30 August 2022; published 30 September 2022) © 2022 Tropical Grasslands-Forrajes Tropicales is an open-access journal published by International Center for Tropical Agriculture (CIAT), in association with The Tropical Crops Genetic Resources Institute of The Chinese Academy of Tropical Agricultural Sciences (TCGRI-CATAS). This work is licensed under the Creative Commons Attribution 4.0 International (CC BY 4.0) license. Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) Tropical Grasslands-Forrajes Tropicales (2022) Vol. 10(3):164–171 164 doi: 10.17138/TGFT(10)164-171 Research Paper Forage characterization of Carajás grass (Cenchrus purpureus × C. americanus) fertilized with a range of doses of protected urea under irrigation during the growing season Caracterización de forrajes de híbridos de Cenchrus purpureus fertilizados con dosis de urea protegida en diferentes estaciones FRANCISCO GLEYSON DA SILVEIRA ALVES1, MARIA SOCORRO DE SOUZA CARNEIRO1, MARCOS JACOME ARAÚJO2, RAFAEL FELIPPE RATKE3, BARBARA SILVEIRA LEANDRO DE LIMA1, NAYRLON DE SAMPAIO GOMES2, RAFAEL RODRIGUES DA SILVA1 AND RICARDO LOIOLA EDVAN2 1Programa de Doutorado Integrado em Zootecnia, Universidade Federal do Ceará, Fortaleza, CE, Brazil. pdiz.ufc.br 2Programa de Pos-Graduacao em Zootecnia, Universidade Federal Do Piauí, Bom Jesus, PI, Brazil. posgraduacao.ufpi.br//ppgz 3Departamento de Agronomia, Universidade Federal de Mato Grosso do Sul, Chapadão do Sul, MS, Brazil. cpcs.ufms.br Abstract The objective of this study was to assess the agronomic and nutritional responses of Carajás grass (Cenchrus purpureus × C. americanus, syn. Pennisetum purpureum × P. glaucum, cultivar ‘Carajás’) fertilized with protected urea. The experimental design was completely randomized blocks in split-plot arrangement over time. The treatments consisted of 5 levels of nitrogen (0, 100, 200, 400 and 800 kg N/ha/year) and measurements were made over 2 seasons (spring 2015 and summer 2016), with 8 replicates. Leaf and stem elongation and senescence rate of tillers increased as N dosage increased, while tiller density, leaf:stem ratio, live:dead material ratio and phyllochron declined. Forage biomass increased with N dosage reaching 47 t DM green forage/ha at 800 kg N/ha but DM production per unit of N applied declined dramatically as level of N applied increased. There was no effect of season. For crude protein (CP) and fiber concentrations, a positive effect was observed with increasing N application, with maximum CP% of 172 g/kg with 800 kg N/ha in spring. Further studies are warranted to determine if economics indicate that the higher fertilizer levels are justified and even protected urea should still be applied on a number of occasions, but still less often than conventional urea, rather than as a single dose at the beginning of spring. Keywords: Biomass, chemical composition, elephant grass, nitrogen fertilizer, tropical pasture. Resumen El objetivo de este estudio fue evaluar las respuestas agronómicas y nutricionales del pasto Carajás (Cenchrus purpureus × C. americanus) fertilizado con urea protegida. El diseño experimental fue de bloques completos al azar en un arreglo de parcelas divididas en el tiempo. Los tratamentos consistieron en cinco dosis de nitrógeno (0, 100, 200, 400 e 800 kg/ha/año) y las mediciones se realizaron durante 2 temporadas (primavera de 2015 y verano de 2016), con 8 repeticiones. Las tasas de senescencia de macollas y de elongación de hojas y tallos aumentaron junto con las dosis de nitrógeno. La densidad de macollos, la relación hoja/tallo, la relación material vivo/muerto y el filocrón se redujo junto con las dosis de nitrógeno. La biomasa del forraje se incrementó con las dosis de N, obteniendo 47 t/ha de biomasa total del forraje verde y no mostró ningún efecto debido a las estaciones para la dosis de 800 kg/ha/año, pero la producción de materia seca por unidad de N aplicada disminuyó drásticamente a medida que aumentó el nivel de N aplicado. No se detecto efecto de la temporada en que se efectuó la cosecha. Para el contenido de proteína cruda y fibra, se observó un efecto positivo al aumentar la dosis de nitrógeno, con Correspondence: Francisco G. da S. Alves, Federal University of Ceará, Department of Animal Science, Av. Mister Hull, 2977, CEP 60440-554, Fortaleza, CE, Brazil. Email: gleysonufc54@gmail.com Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) Forage characterization fertilized with protected urea 165 un PC% máximo de 172 g/kg a una dosis de 800 kg N/ha/año en la primavera. Se justifican más estudios para determinar si la economía indica que los niveles más altos de fertilizantes están justificados e incluso si la urea protegida debe aplicarse en varias ocasiones, pero con menos frecuencia que la urea convencional, en lugar de una dosis única al comienzo de la primavera. Palabras clave: Biomasa, composición química, nitrógeno, pasto elefante, pastura tropical. Introduction Materials and Methods Carajás grass is a hybrid generated by crossbreeding Location and experimental design of Cenchrus purpureus (Schumach.) Morrone (syn. Pennisetum purpureum Schumach.) and Cenchrus The study was carried out at the Experimental Farm americanus (L.) Morrone (syn. Pennisetum glaucum) and of the Federal University of Piauí, Cinobelina Elvas is cultivated through seeds, a distinct advantage when Campus, located in Alvorada do Gurgueia, Piauí, Brazil compared with vegetative propagation of other varieties (8°22'30" S, 43°50'48" W; 239 masl). The region has a of C. purpureus. The cultivar was recently released by climate classified as Tropical Dry (Nunes 2011). The the company MATSUDA® and there is a need to study climatic data observed during the experimental period its productive potential. were recorded at a meteorological station located in that Most grasses respond significantly to application of region (Table 1). nitrogen (N) fertilizer. Urea is one of the main sources of N used in forage production due to its lower cost per unit Table 1. Mean values of climatic conditions throughout the of N and high concentration of N compared with other study. N sources (Mota et al. 2015). However, N losses via Climatic 2015 2016 volatilization in the form of ammonia can be a problem, variable Oct Nov Dec Jan Feb Mar as they decrease the efficiency of use by crops (Tasca et Maximum 38.4 37.0 38.0 31.1 34.6 34.8 al. 2011), especially if urea is supplied in large quantities temperature to the soil at any given time. (°C) In order to increase efficiency of use of N from urea, Mean 31.5 30.3 30.7 26.3 27.7 28.2 especially at high dosages, urease and nitrification temperature inhibitors and polymer coating are being used (Okumura (°C) and Mariano 2012). Release of N from such protected urea Minimum 24.1 23.5 22.9 22.5 21.2 22.5 in the soil for uptake by plants is slow, so this fertilizer temperature can be applied in a single dose during the grass growth (°C) cycle, reducing application costs and allowing the use Precipitation 31.1 144.8 18.1 348.4 38.7 108.4 of high application rates. This contrasts with application (mm) of conventional urea, which normally occurs in smaller Relative 38.0 48.2 43.9 78.8 62.7 66.4 doses at intervals during the grass growth cycle to avoid/ humidity (%) reduce N losses (Silva et al. 2011), resulting in higher labor usage. The experimental design was completely randomized We hypothesized that applying protected urea would blocks with split-plots in time. Treatments consisted of 5 result in efficient usage of N with high quality of forage nitrogen doses (0, 100, 200, 400 and 800 kg N/ha/year) from Carajás grass for animal feeding and would allow high and subplots were 2 evaluation seasons (spring 2015 application rates at any one time with minimal seasonal and summer 2016), with 8 replications (4 blocks and 2 differences in growth, production and chemical composition replications in time, 2 harvests, for each season). Spring of the grass. Therefore, the objective was to evaluate the (characterized by irregular rains and high temperatures) morphogenic and structural characteristics, production and extended from the beginning to the middle of the rainy chemical composition throughout the growing season of season of the region (October‒December 2015), while Carajás grass fertilized with varying doses of protected urea summer (characterized by much heavier rainfall and in a dry tropical environment. The study is a complement to lower temperatures) represented the second half of an experiment in which fertilization of Carajás grass with the rainy season (January‒March 2016). The source protected urea was compared with that of non-protected of nitrogen used was protected urea (FH Nitro Mais®, urea (Alves et al. 2018). developed by the company Fertilizantes Heringer®), Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) 166 F.G.S. Alves, M.S.S. Carneiro, M.J. Araújo, R.F. Ratke, B.S.L, de Lima, N.S. Gomes, R.R. da Silva and R.L Edvan. applied in a single dose in October 2015. This fertilizer difference between the total length of the leaf blade and consists of urea granules coated with boron and copper the remaining green portion. Length of stem was the containing inhibitors. distance from its ligule to the base of the tiller. From Before starting the experiment, a compound soil data obtained the following indexes were calculated: sample was collected from the 0‒20 cm horizon, leaf elongation rate (LER); stem elongation rate (SER), representative of the area, for soil chemical analysis referring to the mean daily elongation on half of the and characterization. Results obtained in the analysis sheaths and true stem of the tiller; total leaf senescence were: pH (H2O) - 5.40; P and K - 9.6 and 21.2 mg/dm3, rate (TSR); and phyllochron. respectively; Ca, Mg, Al and H + Al - 2.4, 0.6, 0.0 and Grass structure was recorded before each harvest. The 3.5 cmol/dm3, respectively; sum of bases - 3.1 cmol/dm3; following evaluations were carried out at each harvest: effective cation exchange capacity (CEC) - 3.1 cmol/ population density of tillers per clump (PDT), obtained dm3; CEC at pH 7.0 - 6.5 cmol/dm3; and base saturation by counting the number of live tillers in a clump from - 46.8 %. Based on these results, soil correction and the central row of each plot; leaf:stem ratio (L:S); and fertilizer application were performed according to Vilela live material:dead material ratio (L:D). Weights were et al. (2002), applying dolomitic limestone (PRNT 80 %) obtained using a digital electronic scale, with capacity to elevate soil saturation levels to 60 % and applying 40 of 1 g to 5 kg, model Sf-400 UNICASA®. kg P/ha (as single superphosphate, 18 % P2O5) and 60 kg At each harvest forage from a linear meter of the K/ha (as potassium chloride, 48 % K2O). central row was collected to determine: total forage The experimental area of Carajás grass was planted biomass (TGB); green leaf biomass (GLB); and green in 2013. Seed was sown in rows at 0.64 g/m, with inter- stem biomass (GSB) by separating harvested material row spacing of 80 cm, as recommended by the company into live and dead material, before live material was MATSUDA®. In October 2015 a uniformity cut was separated into leaf blade and stem. The L:S ratio was performed at 20 cm above ground level. The grass was determined by dividing GLB by GSB, while L:D was irrigated every 3 days with a 10 mm water blade through determined by dividing green biomass by dead forage a sprinkler system, following the recommendations of biomass. Following each harvest, remaining forage on the company which supplied the hybrid. Manual weeding all plots was cut at 20 cm above ground and removed. was conducted to remove undesirable plants. The experimental area was divided into 15 plots Chemical composition of 4 m2 (4 × 1 m), with 5 rows of grass per plot. Plots were separated by uncultivated spaces of 1 m with To determine the chemical composition of Carajás grass, 2 m between blocks. Forage was harvested 4 times forage samples were dried in paper bags in a forced-air during the evaluation period, twice during each season circulation oven at 60 ± 5 °C for 72 hours and then milled (spring and summer) every 40 days at a height of 20 cm with a Willey Mill using a 1 mm sieve. above ground, according to the recommendation of the Dry matter (DM), organic matter (OM), ether extract company MATSUDA® for Carajás grass. (EE), crude protein (CP) and mineral matter (MM) were determined according to the methods of AOAC (2005). Measurement of morphogenesis, structural variables Neutral detergent fiber (NDF) and acid detergent fiber and biomass production (ADF) were determined according to Van Soest et al. (1991). For measuring morphogenesis 2 grass clumps were identified at random in each plot and colored rings were Statistical analysis applied to 3 tillers/clump. Evaluations commenced 3 days after the uniformity cut and were repeated every 3 Results were submitted to analysis of variance, mean days throughout. Appearance, elongation and senescence comparison test and regression analysis. Interactions of leaf blades and elongation of stems were recorded. were determined when significant (P<0.05) by the F test. Length of the expanded leaf blade was measured from The means were compared by Tukey’s test (P<0.05). In its ligule to its apex using a rule. regression analyses, the choice of models was based on Length of the emergent leaf blade was measured the significance of the linear coefficient (P<0.05). As a from its apex to the ligule of the last exposed leaf tool, SISVAR® version 5.0 software was used according blade, while length of the senescent portion was the to Ferreira (2011). Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) Forage characterization fertilized with protected urea 167 Results Biomass production Morphogenic and structural characteristics Significant interaction between N dosage and season for GLB (P<0.001) was observed. Nitrogen dosage had a For morphogenic characteristics, there was no significant positive effect (P<0.001) on TGB and for GSB significant interaction (P>0.05) between fertilizer level there were significant effects of N dosage (P<0.001) and and season and no effect of season for any variable season (P = 0.014) (Table 3). evaluated, whereas for nitrogen doses there were For GLB, there was a positive linear effect in both significant responses (P<0.001) for all characteristics seasons with increasing yield as N dosage increased, (Table 2). reaching 26,000 kg DM/ha in spring and 25,800 kg DM/ For LER, an increasing linear behavior was observed, ha in summer at 800 kg N/ha/year (Table 3). Seasonal with an increase of 0.0042 cm/day for each additional differences occurred for only Control and 200 kg N/ha/year kg N applied (Table 2). By contrast, phyllochron treatments with higher yields in summer. TGB increased declined with increasing doses of N, where for each linearly as N application rate increased, with an increase of additional kg N applied there was a reduction of 0.0024 2.29 kg DM/ha for each additional kg N applied (Table 3), days. For SER, a positive linear response occurred with the highest yield of 47,900 kg DM/ha for the highest with increasing nitrogen doses. The same behavior was application rate of 800 kg N/ha/year. Although TGB observed for TSR, which showed an increase of 0.0007 showed this positive response to N fertilizer application, cm/day/kg N applied (Table 2). most of the response occurred with the first 400 kg N/ha, For structural characteristics there was no producing 83.2 % of the total response. significant interaction between fertilizer level and For GSB, there was an increasing linear response season and no significant effect of season was observed with increasing N dose rate, with an increase of 1.21 kg (P>0.05). Fertilizer level had an effect (P<0.05) for all DM/ha for each additional kg N applied. characteristics evaluated (Table 2). For PDT, a decreasing linear response was observed Chemical composition with increasing level of N (Table 2), resulting in a decrease of 18 tillers per clump for an increase from There was significant interaction between N dosage and 0 to 800 kg N/ ha/year. L:S ratio decreased linearly season (P<0.001) for concentrations of CP, EE, NDF, as N level increased, where for each additional kg N ADF and MM. However, DM concentration was affected applied there was a reduction of 0.0015 in L:S ratio. For (P = 0.003) by only N dose rate (Table 4). L:D ratio, there was a decreasing linear response with Dry matter concentration in forage increased by increase in N dosage, with a reduction of 0.0023 in L:D 0.002 g/kg for each unit of additional N applied (Table ratio for each additional kg N applied. 4). Crude protein concentration increased progressively Table 2. Morphogenic values and structural characteristics of Carajás grass (Cenchrus purpureus × C. americanus) following different rates of nitrogen fertilizer application Variable N rate (kg/ha/year) s.e. P-value Equation 0 100 200 400 800 LER1 4.6 6.0 6.8 7.4 8.5 0.20 <0.001 Ŷ=5.4373+0.0042x; r2=86.6 Phy 6.7 6.7 6.1 5.5 4.9 0.15 <0.001 Ŷ=6.7506–0.0024x; r2=95.3 SER 0.2 0.2 0.3 0.4 0.5 0.03 <0.001 Ŷ=0.2415+0.0004x; r2=97.3 TSR 0.7 1.0 1.1 1.2 1.3 0.03 <0.001 Ŷ=0.9025+0.0007x; r2=75.7 PDT 94.8 87.8 86.9 84.4 76.8 4.53 0.002 Ŷ=92.1375–0.0191x; r2=58.7 L:S 2.9 3.0 2.3 2.4 1.7 0.08 <0.001 Ŷ=2.9483–0.0015x; r2=83.3 L:D 9.3 9.5 9.1 8.4 7.7 0.27 0.001 Ŷ=9.5419–0.0023x; r2=94.8 LER = leaf elongation rate (cm/tiller/day); Phy = phyllochron (day); SER = stem elongation rate (cm/tiller/day); TSR = total senescence rate (cm/tiller/day); PDT = population density of tillers per clump; L:S = leaf:stem ratio; L:D = live material:dead material ratio. Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) 168 F.G.S. Alves, M.S.S. Carneiro, M.J. Araújo, R.F. Ratke, B.S.L, de Lima, N.S. Gomes, R.R. da Silva and R.L Edvan. Table 3. Growth rates (kg DM/ha) of Carajás grass (Cenchrus purpureus × C. americanus) in spring and summer following different application rates of nitrogen. Variable Season Fertilizer (kg N/ha) s.e. P-value Equation 0 100 200 400 800 GLB Spring 17,700b 19,600a 21,400b 23,900a 26,000a 1.90 <0.001 Ŷ=19.213+9.2124x; r2=94.9 Summer 18,800a 19,500a 22,600a 23,900a 25,800a <0.001 Ŷ=19.065+9.6358x; r2=84.9 TGB 26,500 36,400 39,300 44,300 47,900 1.19 <0.001 Ŷ=32.028+2.29x; r2=78.5 GSB1 9,500 12,700 15,600 17,300 20,000 0.102 <0.001 Ŷ=11.444+1.21x; r2=86.4 GSB2 Spring 14,800B 0.070 0.014 Summer 15,100A GLB = green leaf biomass (kg DM/ha); TGB = total green forage biomass (kg DM/ha); 1GSB = green stem biomass (kg DM/ha); 2GSB = green stem biomass (t DM/ha). Means for GLB with the same lower-case letter within columns and for GSB with the same upper-case letter within rows did not differ according to Tukey's test (P>0.05). Table 4. Chemical composition of Carajás grass (Cenchrus purpureus × C. americanus) in spring and summer following application of different doses of nitrogen. Season Rate of N (kg/ha) s.e. P-value 0 100 200 400 800 DM1 190 190 191 195 205 2.9 0.003 CP Spring 116bC 122aC 139aC 151aB 172aA 1.5 <0.001 Summer 126aBC 112bC 138aB 145aAB 159bA <0.001 EE Spring 23.7bB 32.1aA 33.2aA 26.2bAB 34.6aA 0.6 <0.001 Summer 34.7aA 27.7bAB 24.3bB 30.3aA 33.2aA 0.003 NDF Spring 604bB 625aB 676aA 672aA 665aAB 2.9 <0.001 Summer 618aAB 626aA 665aA 595bB 583bB <0.001 ADF Spring 364aAB 365aAB 344bB 372aA 388aA 1.8 <0.001 Summer 336bB 340bB 367aA 347bB 354bB <0.001 MM Spring 95.1aB 90.7bC 100.5aA 104.0aA 97.0aAB 2.2 <0.001 Summer 95.2aB 107.5aA 84.9bC 76.2bC 87.6bB 0.123 1Y = 188.3+0.002x; R2=97.3. DM = dry matter (g/kg); MM = mineral matter (g/kg DM); EE = ether extract (g/kg DM); CP = crude protein (g/kg DM); NDF = neutral detergent fiber (g/kg DM); ADF = acid detergent fiber (g/kg DM). Means within columns and parameters with the same lower-case letter and within rows with the same upper-case letter did not differ according to Tukey's test (P>0.05). with increasing N application rate in both seasons. Morphogenic and structural characteristics While significant differences in MM, EE, NDF and ADF concentrations as a result of N application were recorded, During both spring and summer all treatments received there were no consistent patterns in the responses. irrigation at the same level to minimize moisture stress as a variable, which may explain the absence of seasonal Discussion differences in morphogenic characteristics independent of N doses. Despite high temperatures and low relative This study has shown that high doses of protected urea humidity in spring (Table 1), pasture growth in spring can be applied in a single dose at the beginning of the equaled that in summer. The absence of differences in growing season and still produce significant responses structural characteristics of Carajás grass between the in DM production during the summer. Whether or not evaluated seasons (Table 2) was possibly also due to the these levels of N fertilizer are warranted will depend on influence of regular irrigation applied, which eliminated the value of increases in forage production per unit of N effects resulting from moisture deficits commonly applied as the amount of N applied increases. experienced. Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) Forage characterization fertilized with protected urea 169 The increase in LER as N fertilizer application like temperature, soil moisture and light intensity being increased is a reflection of the positive response in cell adequate for growth. According to Valente et al. (2011) division and expansion to additional N supply (Skinner the optimum temperature amplitude which causes most and Nelson 1995). enzymatic carboxylation in C4 grasses is 30‒35 °C. The positive responses in morphogenic characteristics While mean temperature in spring fell within this range, obtained even at high N dosages (400 and 800 kg/ha/ mean temperature in summer fell below the range but year) would be due to the slow release of N provided minimum temperatures in both seasons were similar. In by protected urea, which was also reported by Alves et addition, relative humidity in summer was much higher al. (2018). Lima et al. (2016) found positive effects from than in spring. Application of irrigation prevented the applying N to Urochloa ruziziensis as protected urea up expression of any differences between seasons in terms to the maximum (300 kg N/ha/year) dose evaluated. of available soil moisture levels experienced in the area The reduction in phyllochron as N fertilizer level under rainfed conditions. increased can be explained by the fact that N stimulates The aim of applying protected urea is to reduce the plant growth, conferring greater regrowth capacity rate of N release from the fertilizer so response to added and faster recovery of the photosynthetic apparatus N is extended for a longer period than expected with (Martuscello et al. 2006) following harvesting. Increases conventional urea and response per additional 100 kg N is in SER would be a response to greater shading caused more uniform. When increasing levels of N fertilizer are by increased growth with increasing N doses, promoting applied to crops or pastures there is normally a reduced elongation of stems to access light. Total leaf senescence response to each additional kg N applied, a phenomenon rate also increased with increasing N fertilizer levels known as the Law of Diminishing Returns (Guimarães in line with the positive effect of N on LER, where et al 2011). Despite the urea in this study being protected increased shading of the older leaves may result in more the same principle applied, with responses in green leaf rapid senescence of leaves. production in spring to the initial 200 kg N being 1,850 kg Shading caused by higher growth of plants at DM/100 kg N compared with 425 kg DM/100 kg N when increasing N application rates promoted the reduction the last 400 kg N was applied. In summer the comparable in population density of tillers, because low luminosity responses were 1,900 kg DM/100 kg N and 475 kg DM/100 interferes negatively with tillering potential, since kg N applied. When total green forage production is decreasing light at the base of plants can also suppress considered for the complete study, response to the first 200 the generation of tillers (Martuscello et al. 2015). Despite kg N was 4,675 kg DM/100 kg N, while response to the final lower tiller numbers at higher N application rates, total 400 kg N was 1,175 kg DM/100 kg N applied. We cannot green biomass yield increased and L:S ratio declined conclude from this study how well the urea was protected indicating that individual tillers were larger at high N from rapid release of N but the fact that plants were still rates with greater stem development, which was reflected responding to urea application 6 months after it was applied in the higher NDF concentration as N level increased. and DM responses in summer per kg N applied were very Reduction in live:dead ratio at higher N application similar to those obtained in spring indicates that significant rates would be a function of higher senescence rate protection was certainly provided and N was being released observed with the higher doses of nitrogen fertilizer. throughout the total growth cycle. This effect was reported previously by Alves et al. (2018) for Carajás grass under different N dosages. Chemical composition Biomass production The increase in DM concentration with increasing N application rate is a function of high growth provided Like other C4 grasses Carajás grass showed good by increased N availability. According to Mendonça and responses in forage production following application Rocha (1985), N accelerates metabolism and promotes of N fertilizer (Lopes et al. 2020; Oliveira et al. 2020; maturity in the plant, providing a greater accumulation Domingues et al. 2021). These significant responses of photoassimilates and transformation of these into plant to N fertilizer application in DM production in organs, contributing to an increase in DM concentration both spring and summer and the overall absence of in forage. differences between seasons indicate that the primary In both seasons N fertilizer application produced limiting factor was available N supply, with factors marked increases in CP% in forage from 11.6 % in Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) 170 F.G.S. Alves, M.S.S. Carneiro, M.J. Araújo, R.F. Ratke, B.S.L, de Lima, N.S. Gomes, R.R. da Silva and R.L Edvan. Controls to 17.2 % at 800 kg N/ha in spring and from author’s scholarship and the Foundation for Research 12.6 % to 15.9 % in summer. This was despite large Support of Piauí (FAPEPI) for support in carrying out increases in DM production in treatments receiving N. the experiment. These values indicate that considerable quantities of N were still available in soil at all fertilizer levels despite References the long time elapsed following application of fertilizer. This forage would provide excellent fodder for animals, (Note of the editors: All hyperlinks were verified 18 July 2022). especially given their ability to select a higher quality diet than is on offer. Alves FGS; Carneiro MSS; Edvan RL; Cândido MJD; Furtado RN; Pereira ES; Morais Neto LB; Mota RTM; Nascimento However, application of N increased NDF KS. 2018. Agronomic and nutritional responses of Carajas concentration in spring, probably as a result of increased Elephant grass fertilized with protected and non-protected stem proportions in the forage produced. While this urea. Semina: Ciencias Agrárias 39(5):2181–2194. doi: would lower digestibility of the forage, highest NDF 10.5433/1679-0359.2018v39n5p2181 concentrations recorded were 67 %. In contrast NDF AOAC (Association of Official Analytical Chemists). 2005. concentrations in forage in summer were lower at the high Official Methods of Analysis. 18th Edn. Association of N fertilizer levels than in Controls. The application of N Official Analytical Chemists, Urbana, IL, USA. in high doses along with favorable climatic conditions Domingues AA; Santos AJM; Backes C; Rodrigues LM; can favor the growth and production of tissues that have Teodoro AG; Bessa SV; Ribon AA; Giongo PR; Godoy LJG de; Resende CCF de. 2021. Nitrogen fertilization lower levels of structural carbohydrates (Vitor et al. of Paiaguás grass: Production and nutrition. Revista de 2009), which occurred in summer. Agricultura Neotropical 8(2):e5918. doi: 10.32404/rean. As might be expected, application of N at high doses v8i2.5918 had no significant impact on mineral concentration in Ferreira DF. 2011. Sisvar: A computer statistical analysis forage in spring, while mineral concentration in forage in system. Ciência e Agrotecnologia 35(6):1039–1042. doi: summer was significantly reduced as N level increased. 10.1590/S1413-70542011000600001 This is scarcely surprising given the high yields of forage Guimarães G; Lana RP; Souza MRM; Mendonça BPC; produced at high N application rates, which would have Alves EEM. 2011. Recommendation of fertilizers based resulted in removal of large quantities of minerals from on the law of diminishing return. Revista Brasileira de Agropecuária Sustentável 1(1):52–58. (In Portuguese) bit. the soil. ly/3uW1i4z This study has demonstrated that Carajás grass Lima JES; Nascente AS; Leandro WM; Silveira PM da. responds well to application of protected urea. However, 2016. Urochloa ruziziensis responses to sources and despite the urea being protected, the Law of Diminishing doses of urea. Revista Brasileira de Engenharia Agrícola Returns still operated, with responses per unit of N e Ambiental 20(6):401–407. doi: 10.1590/1807-1929/ applied declining dramatically as amount of fertilizer agriambi.v20n5p401-407 applied increased. Economic assessments are needed Lopes MN; Cândido MJD; Pompeu RCFF; Silva RG da; to determine if the high application rates are justified. Lacerda CF de; Bezerra MA; Morais Neto LB de; Carneiro Similarly, studies seem warranted to assess the overall MSS. 2020. Gas exchange in Massai grass fertilized with nitrogen and grazed by sheep. Bioscience Journal responses when multiple applications of lower doses are 36(1):152–160. doi: 10.14393/BJ-v36n1a2020-42411 applied as opposed to single heavy doses at planting. Martuscello JA; Fonseca DM da; Nascimento Jr D; Santos The reduced responses in DM production to additional PM; Cunha DNFV da; Moreira LM. 2006. Morphogenetic N applied at high doses might mean that overall financial and structural characteristics responses of Massai grass returns could be better with lower application rates and under nitrogen fertilization and defoliation. Revista multiple applications of smaller amounts of fertilizer, Brasileira de Zootecnia 35(3):665–671. (In Portuguese) and even with increased labor costs involved, could be doi: 10.1590/S1516-35982006000300006 more efficient. Economic assessments would provide Martuscello JA; Silva LP da; Cunha DNFV da; Batista ACS; data to prove or disprove these hypotheses. Braz TGS; Ferreira PS. 2015. Nitrogen fertilization in Massai grass: Production and morphogenesis. Ciência Animal Brasileira 16(1):1–13. (In Portuguese) doi: Acknowledgments 10.1590/1089-68916i118730 Mendonça JFB; Rocha GP. 1985. Yield of Elephant grass We thank the Coordination for the Improvement of (Pennisetum purpureum cv. Cameroon) at different cutting Higher Education Personnel (CAPES) for the first ages. Ciência Prática 9:23–29. Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) Forage characterization fertilized with protected urea 171 Mota MR; Sangoi L; Schenatto DE; Giordani W; Boniatti Tasca FA; Ernani PR; Rogeri DA; Gatiboni LC; Cassol PC. CM; Dall'Igna L. 2015. Stabilized nitrogen sources 2011. Ammonia volatilization following soil application as an alternative for increasing grain yield and of conventional urea or urea with urease inhibitor. nitrogen use efficiency by maize. Revista Brasileira de Revista Brasileira de Ciência do Solo 35(2):493–502. (In Ciência do Solo 39(2):512–522. (In Portuguese) doi: Portuguese). doi: 10.1590/S0100-06832011000200018 10.1590/01000683rbcs20140308 Valente TNP; Lima ES; Henriques LT; Machado Neto OR; Nunes CS. 2011. Uses and applications palm forage as a major Gomes DI; Sampaio CB. 2011. Anatomy of forage plants source of economy for the semiarid. Revista Verde de and the availability of nutrients for animal ruminants. Agroecologia e Desenvolvimento Sustentável 6(1):58–66. Veterinária e Zootecnia 18(3):347–358. (In Portuguese) (In Portuguese). doi: 10.18378/rvads.v6i1.551 bit.ly/3bAHRaY Okumura RS; Mariano DC. 2012. Agronomic aspects of Van Soest PJ; Robertson JB; Lewis BA. 1991. Methods urease inhibitor-treated urea. Ambiência 8(2):403–414. for dietary fiber, neutral detergent fiber, and nonstarch (In Portuguese). doi: 10.5777/ambiencia.2012.02.02rb polysaccharides in relation to animal nutrition. Journal of Oliveira JKS de; Correa DCC; Cunha AMQ; Rego AC do; Dairy Science 74(10):3583–3597. doi: 10.3168/jds.S0022- Faturi C; Silva WL da; Domingues FN. 2020. Effect of 0302(91)78551-2 nitrogen fertilization on production, chemical composition Vilela L; Soares WV; Sousa DMG; Macedo MCM. 2004. and morphogenesis of Guinea grass in the Humid tropics. Calagem e adubação para pastagens. In: Sousa DMG; Agronomy 10(11):1840. doi: 10.3390/agronomy10111840 Lobato E, eds. Cerrado: Correção do solo e adubação. Silva DRG; Pereira AF; Dourado RL; Silva FP da; Ávila FW; Embrapa Cerrados, Planaltina. p. 367–382. bit.ly/ Faquin V. 2011. Productivity and efficiency of nitrogen 3SylcNg fertilization in maize under different levels of urea and Vitor CMT; Fonseca DM; Cóser AC; Martins CE; Nascimento NBPT–Treated urea. Ciência e Agrotecnologia 35(3):516– Jr D; Ribeiro Jr JI. 2009. Dry matter production and 523. doi: 10.1590/S1413-70542011000300012 nutritional value of elephant grass pasture under irrigation Skinner RH; Nelson CJ. 1995. Elongation of the grass leaf and and nitrogen fertilization. Revista Brasileira de Zootecnia its relationship to the phyllochron. Crop Science 35(1):4– 38(3):435–442. (In Portuguese) doi: 10.1590/S1516- 10. doi: 10.2135/cropsci1995.0011183X003500010002x 35982009000300006 (Received for publication 10 June 2020; accepted 20 June 2022; published 30 September 2022) © 2022 Tropical Grasslands-Forrajes Tropicales is an open-access journal published by International Center for Tropical Agriculture (CIAT), in association with The Tropical Crops Genetic Resources Institute of The Chinese Academy of Tropical Agricultural Sciences (TCGRI-CATAS). This work is licensed under the Creative Commons Attribution 4.0 International (CC BY 4.0) license. Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) Tropical Grasslands-Forrajes Tropicales (2022) Vol. 10(3):172–183 172 doi: 10.17138/TGFT(10)172-183 Research Paper Physiological, morphological, and biochemical characterization of Cratylia argentea (Desv.) Kuntze seeds Caracterización fisiológica, morfológica y bioquímica de semillas de Cratylia argentea (Desv.) Kuntze EDUARDO PACCA LUNA MATTAR1, DANIEL TEIXEIRA PINHEIRO2, WANDER DOUGLAS PEREIRA2, BRUNO PORTELA BRASILEIRO3, WALTER JOSÉ RODRIGUES MATRANGOLO4, PAULO CÉSAR HILST2, PAOLA ANDREA HORMAZA MARTÍNEZ5 AND DENISE CUNHA FERNANDES DOS SANTOS DIAS2 1Universidade Federal do Acre, Rio Branco, AC, Brazil. ufac.br 2Universidade Federal do Viçosa, Viçosa, MG, Brazil. ufv.br 3Universidade Federal do Paraná, Curitiba, PR, Brazil. ufpr.br 4Embrapa Milho e Sorgo, Sete Lagoas, MG, Brazil. embrapa.br/milho-e-sorgo 5Universidad de Pamplona, Pamplona, Norte de Santander, Colombia. unipamplona.edu.co Abstract Cratylia argentea is a shrub legume native to tropical regions of South America where it is used for animal feed and green manure. In the absence of germination guidelines, the key aim of this study was to define the most suitable temperature for conducting germination and accelerated aging tests. The biochemical attributes of seeds were also assessed. Seeds with 10 % moisture from 4 different seed lots were germinated using the between paper method in a germinator at temperatures of 20, 25, 30 and 35 °C and alternating temperatures of 20/30 °C (16/8 h), with daily counting until germination was stable (seven days without germination). For the accelerated aging test, two temperatures (41 and 45 °C) and six aging periods (0, 24, 48, 72, 96 and 120 h) for seeds with between 10–40 % moisture content were used. Carbohydrates (%), ethereal extract (%), crude protein (%) and macro and micronutrient contents of the seeds were measured. Results showed that C. argentea seeds consist predominantly of starch (22.67 %) and protein (26.45 %) reserves with a low percentage of lipids. For the germination test, the temperature of 30 °C is recommended, allowing greater percentage and speed of germination, with seedling evaluation at 10 and 20 days. For the accelerated aging test, aging for 48 h at 41 °C is recommended to discriminate C. argentea seed lots in terms of quality. Keywords: Accelerated aging, dry environment plant, germination test, seed analysis, tropical shrub legume, unconventional forage crop. Resumen Cratylia argentea es una leguminosa arbustiva originaria de las regiones tropicales de América del Sur, donde se utiliza como alimento para animales y abono verde. En ausencia de pautas de germinación, el objetivo de este estudio fue definir la temperatura más adecuada para realizar pruebas de germinación y envejecimiento acelerado. También se evaluaron los atributos bioquímicos de las semillas. Semillas con 10 % de humedad de 4 lotes de semillas diferentes fueron colocadas entre hojas de papel toalla, en un germinador a temperaturas de 20, 25, 30 y 35 °C y alternando temperaturas de 20/30 °C (16/8 h), con conteo diario hasta que la germinación se estabilizó (siete días sin germinar). Para la prueba de envejecimiento acelerado se utilizaron dos temperaturas (41 y 45 °C) y seis períodos de envejecimiento (0, 24, 48, 72, 96 y 120 h) para semillas con un contenido de humedad entre 10–40 %. Se midieron carbohidratos (%), extracto etéreo (%), proteína cruda (%) y contenido de macro y micronutrientes de las semillas. Los resultados mostraron que Correspondence: Eduardo Mattar, Center of Biological and Natural Sciences, Federal University of Acre, BR - 364 Road, Rio Branco, Acre, Brazil. Email: eduardo.mattar@ufac.br Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) Information about Cratylia argentea seeds 173 las semillas de C. argentea contienen predominantemente reservas de almidón (22,67 %) y proteína (26,45 %) con un bajo porcentaje de lípidos. Para la prueba de germinación se recomienda la temperatura de 30 °C, que permite mayor porcentaje y rapidez de germinación, con evaluación de plántulas a los 10 y 20 días. Para la prueba de envejecimiento acelerado, se recomienda un envejecimiento de 48 h a 41 °C para discriminar los lotes de semillas de C. argentea en términos de calidad. Palabras clave: Análisis de semillas, ensayo de germinación, envejecimiento acelerado, forrajes no convencionales, leguminosas tropicales, planta de ambiente seco. Introduction a defined period. It is an effective test for ranking seed lots for storage capacity and field emergence. Cratylia argentea (Desv.) Kuntze is a shrub legume Therefore, developing the correct methodology for (Fabaceae) native to tropical regions of South America. this test in C. argentea is important for estimating the It has multiple uses, especially as animal feed and as potential of seed lots, as has been observed in different a green manure crop. It has a symbiotic relationship species such as Jatropha curcas (Oliveira et al. 2014), with nitrogen-fixing bacteria (Mattar et al. 2018) and Leucaena leucocephala (Araújo et al. 2017) and is recommended for silvopasture systems (Valles-de Urochloa brizantha (Oliveira et al. 2020). la Mora et al. 2014). The use of C. argentea together Knowledge regarding the morphological aspects with Urochloa brizantha has shown promise for feeding of seeds and seedlings in initial stages of development calves (F1 Holstein × Zebu) (Valles-de la Mora et al. is relevant in development of standard methods and 2017). The species has potential for control of parasitic interpretation of results for the physiological analysis diseases in ruminants (Silva et al. 2017), as well as of seeds. Seed morphology is also useful to identify having a low tannin concentration (Pereira et al. 2018). species under natural conditions (Abud et al. 2010). It shows potential for use as feed in swine production Knowledge of the chemical composition of seeds can (Sarria and Martens, 2013). assist in understanding the biochemical properties of the Germination is the first step in the crop production species and cellular processes involved in seed quality. cycle. Diverse intrinsic and extrinsic factors contribute Vaz Patto et al. (2015) report that the analysis of the to the success of seed germination, including initial mineral composition is also crucial when considering physiological quality, hormonal balance, temperature the nutritional quality of the seeds. Currently, this has and lighting (Lone et al. 2016; Ebone et al. 2019; Wu been evaluated in different species, such as quinoa et al. 2020). Temperature is a key factor for germination (Reguera et al. 2018), beans (Los et al. 2018) and because it determines the ecological limitations for watermelon (Lawal, 2011). These studies bring a new geographic distribution and establishment of the contribution to understanding seed germination and species (Daibes and Cardoso 2018). Therefore, the vigor in these species. Fang and Wang (2007) evaluated definition of optimal temperature(s) for germination changes in biochemical composition and enzyme can be useful for expression of the maximum potential activity during dormancy release in Cyclocarya paliurus of a species, as has been observed in Fabaceae species seeds. Information on lipid content is important because such as Diptychandra aurantiaca (Oliveira et al. oilseeds generally have shorter longevity after harvest 2013), Robinia pseudoacacia (Giuliani et al. 2015) and (Wiebach et al. 2019) related to lipid peroxidation during Calobota sericea (Müller et al. 2019). the deterioration process, being one of the main causes Seed vigor also effects seed quality and crop of loss of vigor in orthodox seeds (Ebone et al. 2019). establishment in the field, governing the capacity of Information regarding seed quality, such as seeds to grow under different environmental conditions germination, vigor, morphology and chemical (Marcos-Filho 2015). Seed vigor can be defined as the composition of C. argentea seeds is scarce. A better sum of the properties of the seed that allows acceptable understanding of germination, storage, control germination and the development of uniform seedlings of seed quality, species ecology, cultivation and under adverse conditions (Finch Savage and Bassel commercialization can help producers develop better 2016). Tests can be used to estimate seed lot vigor, strategies for storage and planting as an alternative forage. including the accelerated aging test, which is used This study aimed to define suitable temperature(s) for to simulate the deterioration process by subjecting germination of C. argentea seeds and to characterize seed seeds to high temperatures and relative humidity over and seedling development, generating technical visual Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) 174 E.P.L. Mattar, D.T. Pinheiro, W.D. Pereira, B.P. Brasileiro, W.J.R. Matrangolo, P.C. Hilst, P.A. Hormaza and D.C.F.S. Días support material. Other objectives were to chemically Seed biochemical analyses characterize seeds and define the best methodology for carrying out accelerated aging tests to discriminate seed A preliminary germination test using the methodologies lots with different physiological quality. described above conducted 90 days after harvest, showed that seed lot 3 had a germination of 89 % Materials and Methods and emergence in sand of 91 % and considered to be of high physiological quality and suitable for Seed sources for the study the analysis. Carbohydrates were quantified by the methodology of Thompson (1990), ethereal extract Four different lots of manually harvested C. argentea and protein by the methodology of Brasil (2009) and seeds were used for the experiments. For the biochemical fatty acids by the methodology of the Associations analyses, biometric characterization of seeds and of Official Analytical Chemists (AOAC 2005). Seed classification of seedlings, a seed lot (approximately samples were ground in a Wiley mill, passed through 0.5 kg of seeds) stored in a refrigerator at 10 ⁰C for 4 a 0.5 mm sieve, and subjected to nitric-perchloric acid months post-harvest was obtained from the seed bank of digestion (Alvarez et al. 2001). Phosphorus (P) was Empresa Brasileira de Pesquisa Agropecuária (Embrapa determined by colorimetry, potassium (K) by flame Milho e Sorgo), Sete Lagoas, Minas Gerais, Brazil. For emission photometry, calcium (Ca) and magnesium the emergence, germination, and accelerated aging tests, (Mg) by atomic absorption spectrometry, nitrogen (N) 4 different seed lots (approximately 1.8 kg of seeds each) by titration after digestion by the Kjeldahl method, and were used (Table 1). sulfur (S) by turbidimetry (Alvarez et al. 2001). Characterization of seeds and development and classifi- Physiological characterization of the seed lots cation of seedlings A seedling emergence test in sand was conducted in a The seeds were imbibed for 24 h on paper towel moistened greenhouse. Four replications of 50 seeds were washed with 2.5 times the weight of the dry paper with water and sown in sterilized sand in polyethylene boxes. The and covered with a sheet of paper towel and rolled. After sand was moistened to 60 % of water retaining capacity. imbibition, seeds were cut longitudinally to use for the Normal seedlings were counted daily until stabilization photographic record. The dimensions of 50 seeds were of emergence, defined as 7 d without emergence, for measured with a caliper and width (mm), length (mm) calculation of percentage of seedling emergence. After and thickness (mm) were recorded. An image was taken stabilization of emergence, the following variables were to illustrate seed structure. obtained: A sample of 45 seeds was sown in sand in a polyethylene Emergence speed index (ESI) according to the tray and maintained for 15 d at 25 °C. Three seeds were equation proposed by Maguire (1962): removed daily for observation. Some seeds were kept intact E1 E2 En for external observation while others were cut longitudinally ESI = + +…+ for internal observation of seedling development. A N1 N2 Nn succession of images was taken daily of the individual where: seeds to track seedling emergence and development over a E1, E2, En = number of emerged seedlings at the first, 15 d period. The seedlings were grouped into “normal” and second and last count; “abnormal” and photographed independently as a reference N1, N2, Nn = number of days of sowing to the first, set (Brasil 2009). second and last count. Table 1. C. argentea seeds used for the analyses. Lot Lot origin Other information 1 Santa Rita Farm of the Empresa de Pesquisa Agropecuária de Seeds stored for 4 months after harvest in a refrigerator Minas Gerais (EPAMIG), Prudente de Morais (MG), Brazil. at 10 °C 2 From the private property of Caetanópolis (MG), Brazil. Ibid 3 Embrapa Milho e Sorgo in Sete Lagoas (MG), Brazil Ibid 4 Embrapa Milho e Sorgo in Sete Lagoas (MG), Brazil Seeds stored for 16 months after harvest in paper bags at room temperature (mean of 22 °C day and 17 °C night). Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) Information about Cratylia argentea seeds 175 Emergence speed (ES) according to the equation seedlings obtained daily (as described above for the proposed by Edmond and Drapala (1958): emergence test), according to the equations of Maguire [(N1E1) + (N2E2) + … + (NnEn)] (1962) and Edmond and Drapala (1958), respectively. ES = E1 + E2 + … + En Accelerated aging test where: E1, E2, En = number of emerged seedlings at the first, Five replications of 40 seeds of each lot were placed on second and last count; suspended metal screens inside plastic boxes (gerboxes) N1, N2, Nn = number of days of sowing to the first, containing 40 mL of distilled water at the bottom. Lids second and last count. were placed on the boxes and they were kept in BOD Shoot dry matter (SDM) and root dry matter (RDM) of seedlings, length of the main root (RL) and shoot length (Biochemical Oxygen Demand) incubators regulated at (SL) were measured. Shoot height and main root length two temperatures (41 and 45 °C) for six periods of aging were measured for each seedling obtained in the emergence (0, 24, 48, 72, 96 and 120 h) (Araújo et al. 2021). After test using a caliper and the results were expressed in cm/ each period, a germination test was conducted on the seedling. After measurement, seedlings obtained from each seeds as described above, using a constant temperature replication of the last count of the emergence test were of 25 °C with evaluations at 10 days after sowing. The placed in paper bags and dried in a forced air circulation number of normal seedlings, abnormal seedlings and oven at a temperature of 65 ºC for 72 h. After this period, dead seedlings were counted and percentages calculated. the material was weighed on a precision balance, with the After seed aging, moisture content was determined results expressed in mg/seedling. according to the method described in Brasil (2009). Germination at different temperatures Experimental design and statistical analysis Five replications of 40 seeds of each lot were treated with The experiments were conducted in a completely fungicide Protreat® at a concentration of 200 mL/100 randomized design. The replicates were randomly kg of seeds and each replicate placed on two sheets of distributed inside the germinators and BODS, and all paper towel moistened with 2.5 times the weight of the the temperatures were conducted simultaneously. The dry paper with water and covered with a sheet of paper germination and accelerated aging tests were conducted towel and rolled. The rolls of paper were kept in seed in a double factorial function. In the germination test, germinators at constant temperatures of 20, 25, 30 and one factor corresponded to the seed lots and another 35 °C (with a photoperiod of 8 h) and at alternating to the temperatures. In the accelerated aging test, one temperatures of 20/30 °C (8h at 30 °C light/16h at 20 °C factor corresponded to the seed lots and another to the dark) until stabilization of germination of all treatments aging periods. Analysis of variance was conducted on (around 30 days). For the constant temperatures the data. The means were compared by the Tukey test (20, 25, 30 and 35 °C), the rolls were maintained in at 5 % probability with use of the R software (R Core a germination chamber (Mangelsdorf) and for the Team 2020). alternating temperatures (20/30 °C), the seeds were placed in plastic bags and kept in a BOD (Biochemical Results Oxygen Demand) incubating chamber without supply of oxygen. The BOD was used for alternating temperatures Biochemical analyses of the seeds because it changes temperatures and light regimes automatically. The use of plastic bags prevents the loss C. argentea seeds consist of predominantly starch and of water and makes it possible to compare the results proteins as reserves and have a low percentage of lipids with constant temperatures conducted in a germination (Table 2). chamber. Evaluations of the number of normal seedlings The nutritional composition was nitrogen (5.59 dag/ were performed daily up to stabilization of the number kg), phosphorus (0.38 dag/kg), potassium (0.56 dag/kg), of normal seedlings germinating. calcium (0.13 dag/kg), magnesium (0.18 dag/kg), sulfur For all treatments, on the date of final count, the final (0.21 dag/kg), copper (9.90 mg/kg), iron (58.80 mg/kg), percentage of normal seedlings was calculated. The ESI zinc (39.95 mg/kg), manganese (25.30 mg/kg) and boron and ES were calculated based on the number of normal (9.07 mg/kg). Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) 176 E.P.L. Mattar, D.T. Pinheiro, W.D. Pereira, B.P. Brasileiro, W.J.R. Matrangolo, P.C. Hilst, P.A. Hormaza and D.C.F.S. Días Table 2. Biochemical composition of C. argentea seeds. and b). The longitudinal section in the central part of the Analyses Unit Results seed revealed the hypocotyl-radicle axis located in the Crude protein % 26.45 region near the hilum and the cotyledons (Figure 1A-c). Ethereal extract % 0.74 Starch % 22.67 Palmitic acid (C16:0) % 0.07 Stearic acid (C18:0) % 0.02 Elaidic acid (C18:1n9c) % 0.11 Linoleic acid (C18:2n6c) % 0.21 Linolenic acid (C18:3n3) % 0.12 Polyunsaturated fats % 0.33 Trans fats % 0.00 Monounsaturated fats % 0.11 Saturated fats % 0.12 Unsaturated fats % 0.44 Fructose mg/kg <50.001 Raffinose mg/kg <50.00 Maltose % <50.00 Free glucose % 0.06 Figure 1. A. Seed in lateral (a), frontal (b), and internal (c) Lactose mg/kg <50.00 perspectives. B. Development of the C. argentea seedling: Root Total carbohydrates % 41.46 emergence at one to two days after sowing (a); development of the primary root from three to five days after sowing (b-c); Fucose mg/kg <50.00 normal seedling with formation of secondary roots, opening Arabinose mg/kg <50.00 of the cotyledons, epicotyl growth and leaf primordia at Galactose % 2.72 approximately six days after sowing (d); from seven to eight Glucose % 31.17 days after sowing (e): from 10 to 11 days after sowing (f); and from 12 to 13 days after sowing (g). Source: own elaboration. Xylose % 1.71 Rhamnose % 5.86 Germination began by day 2 after sowing at room Mannose mg/kg <50.00 temperature (Figure 1B-a). From 3 to 5 days (Figure 1B-b Sucrose % 1.00 and c), the primary root elongated and from day 6 on, Free fructose % 0.01 secondary root formation began. Growth of the epicotyl to Butyric acid (C4:0) % 0.02 around 2 cm length was also seen at this stage (Figure 1B- Carbohydrate results expressed as <50.00 are below the limit d). Plumules or leaf primordia appeared at the tip of the of quantification. epicotyl and a normal seedling can already be observed (Figure 1B-d). Normal seedlings obtained on days 7 and Characterization of seeds, seedling development and 8 are also shown (Figure 1B-e) with a more developed classification epicotyl (around 3 cm). At 10 and 11 days (Figure 1B- f) and at 12 and 13 days (Figure 1B-g), the plumules The C. argentea seeds sampled had a mean width of or already developed primary leaves can be seen. 9.5 mm, mean length of 11.35 mm, and mean thickness Germination is hypogeal, with the cotyledons remaining of 3.16 mm. Based on the observations, the seeds were under the soil surface. No dormancy was identified in the considered exalbuminous, where there is no endosperm seeds. The normal seedlings (Figure 2) consist of a primary and the cotyledons comprise the key storage organ. root, along with the emergence of secondary roots, and The cotyledons are greenish-white (Figure 1A) and upright or slightly curved hypocotyl, with two primary externally, the seeds have a smooth brown seed coat and leaves at the upper tip. A well-developed primary root was a visible hilum (Figure 1A-b). There is also a caruncle, a also observed, an important factor for development of a whitened tissue in the region of the hilum (Figure 1A- a vigorous plant in the field. Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) Information about Cratylia argentea seeds 177 (around 3 days) than the others, confirming the faster germination of this lot (Table 3). Higher values for RDM were observed for lots 2 and 3, which were superior to lot 4, but did not differ significantly from lot 1. For RL, there was no significant difference among the lots. SDM and SL showed a higher value for lot 3, which was significantly superior to lot 4 (Table 3). Germination at different temperatures Germination percentage (%GERM), germination speed index (GSI), and germination speed (GS) at the different temperatures were similar for lots 1, 2 and 3 (Table 4). For these seed lots, there was no difference in %GERM at the temperatures of 20, 25, 30 and 20/30 °C, but significant reduction at 35 °C. Lot 4 was less vigorous and germination was superior at the temperatures of 30 and 20–30 °C, followed by the temperatures of 20 and 25 °C, with the lowest value at the temperature of 35 °C (Table 4 and Figure 3). Figure 2. Abnormal C. argentea seedlings: poorly developed The GSI and the GS under the different temperature root system (a, c, e); damaged root system (b, f); malformation conditions generally showed a similar pattern in the four of the root and shoot system (d). Normal C. argentea seedlings: all essential structures present and good root-shoot ratio (a, b, lots analyzed. The GSI, which is based on the mean c); curvature in the hypocotyl without affecting formation of number of seeds germinated per day, was higher at 30 a normal seedling (d); slightly less developed root system, but °C, followed by 25 and 20/30 °C, which were higher without affecting formation of a normal seedling (e, f). than 20 and 35 °C (Table 4). For GS, the lowest value found was at 30 °C, confirming faster germination at that Physiological characterization of the seed lots temperature. The 4 seed lots exhibited significant differences in initial Accelerated aging physiological quality. Lots 1 and 3 exhibited 89 and 94 % of emergence, respectively; being higher than lots The accelerated aging test showed that 41 °C allowed 2 (75 %) and 4 (62 %) (Table 3). In general, seedling a less drastic reduction in germination (Table 5). At 45 emergence began on day 9 after sowing and stabilized °C there was significantly reduced germination, with on day 30. Lots 1 and 3 exhibited superior values of ESI, germination reduced to 0 % after 48 h in lots 1 and 4. indicating a mean number of approximately 4 emerged This was observed as lower percentages of normal seedlings per day. For lots 2 and 4, these values were seedlings (NS1 and NS) and higher percentages of dead approximately 3 seedlings. For ES, higher values are seeds (DS). Analyzing the results of the combinations found for lot 4 (16 days), i.e. lower physiological quality of temperatures and periods of seed exposure to aging, compared to the other lots. Lot 1 exhibited a lower value at the temperature of 41 °C at the times of 48 h, 72 h Table 3. Characterization of physiological potential of seeds from four lots of C. argentea using the emergence test in sand. Lot E (%) ESI ES (days) RDM (mg/seedling) SDM (mg/seedling) RL (cm/seedling) SL (cm/seedling) 1 89a 3.78a 12.02c 2.93ab 9.56ab 11.61a 6.52ab 2 75b 2.73b 14.60b 3.34a 8.07bc 11.33a 6.75ab 3 94a 3.51a 13.80b 3.48a 10.24a 11.25a 7.29a 4 62b 2.08c 16.02a 2.27b 11.25a 11.09a 6.23b E = total emergence (%); ESI = emergence speed index; ES = emergence speed; RDM = root dry matter; SDM = shoot dry matter; RL = root length; SL = shoot length (cm/seedling). Mean values followed by the same lowercase letters in the columns do not differ from each other according to the Tukey test at 5% probability. Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) 178 E.P.L. Mattar, D.T. Pinheiro, W.D. Pereira, B.P. Brasileiro, W.J.R. Matrangolo, P.C. Hilst, P.A. Hormaza and D.C.F.S. Días Table 4. Germination percentage (%GERM), germination speed index (GSI) and germination speed (GS) of four seed lots of C. argentea incubated at different temperatures (°C). Variable (°C) Lot 1 Lot 2 Lot 3 Lot 4 %GERM 20 90abA 81bA 93aA 37cB 25 89aA 83aA 93aA 46bB 30 89aA 87aA 92aA 58bA 35 29aB 29aB 27aB 4bC 20/30 93aA 84aA 88aA 58bA GSI 20 3.16aC 2.45bC 2.91abC 0.92cC 25 4.74aB 3.60bB 4.27abB 1.83cB 30 7.04aA 5.73bA 6.42aA 3.77cA 35 1.83aD 1.67aD 1.59aD 0.15bD 20/30 4.37aB 3.39bB 3.95abB 2.09cB GS 20 14.67cA 17.96bA 16.24bcA 23.25aA 25 9.60bBC 12.61abBC 11.15abB 13.14aB 30 6.51aC 8.55aD 7.45aC 8.20aC 35 8.07bBC 9.49bCD 8.79bBC 14.62aB 20/30 10.84bB 13.30abB 11.57bB 14.76aB Mean values followed by the same uppercase letters in the columns (among temperatures) and the same lowercase letters in the rows (among lots) do not differ from each other according to the Tukey test at 5% probability. Figure 3. Germination curves of four lots of C. argentea seeds at different temperatures. The curves represent 5 replicates of 40 seeds of each of 4 seed lots. and 96 h, seed lots 2 and 3 maintained their germination in seed moisture content with the aging times at both rates at 0 h (control) and 24 h. However, for lots 1 and temperatures (41 and 45 °C). However, the variation 4, there was a significant reduction in germination of the in seed moisture content was generally low (around 3 seeds that had been placed under accelerated aging under percentage points) among different lots at the same these conditions. There were increases of up to 30 % temperature. Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) 179 E.P.L. Mattar, D.T. Pinheiro, W.D. Pereira, B.P. Brasileiro, W.J.R. Matrangolo, P.C. Hilst, P.A. Hormaza and D.C.F.S. Días Table 5. Normal seedlings at first count (NS1), normal seedlings (NS), dead seeds (DS) and moisture content (MC) in 4 seed lots of C. argentea not under the accelerated aging test (0 h; control) and under the accelerated aging test at two temperatures (41 and 45 °C) and five aging times (24, 48, 72, 96 and 120 h). Conventional test Accelerated aging tests Variable Lot 0 h (control) 41 °C 45 °C 24 h 48 h 72 h 96 h 120 h 24 h 48 h 72 h 96 h 120 h NS1(%) 1 27.0a* 35.5abA 26.5aAB 34.0aA 20.0cB 17.0aB 29.0abA 17.5abB 0.0cC 0.0C 0.0aC 2 18.5a 25.5bBC 29.5aB 44.5aA 47.5aA 14.0abC 21.5bA 20.5aA 9.0bB 0.0C 0.0aC 3 16.5ab 37.5aA 29.5aAB 35.5aAB 33.0bAB 24.0aB 32.5aA 19.5aB 21.5aB 3.5C 0.5aC 4 6.5b 25.0bA 18.0aAB 17.5bABC 8.0dBC 5.0bC 0.0cB 10.5bA 4.5bcAB 0.5B 0.0aB CV(%) 30.01 24.07 43.39 NS(%) 1 70.5a 72.5aA 48.5bB 52.5bB 26.5bC 19.0bcC 61.0bA 25.5bB 0.5cC 0.0aC 0.0aC 2 72.0a 72.5aA 65.5aA 71.0aA 75.5aA 24.0bB 61.0bA 51.0aA 36.0aB 0.0aC 0.0aC 3 72.0a 71.0abA 63.0aAB 68.0aA 66.5AB 53.5aB 71.5aA 60.5aB 44.0aC 5.0aD 1.0aD 4 39.0b 60.0bA 24.0cB 28.0cB 10.0cC 8.0cC 0.5cB 16.0bA 13.5bA 0.5aB 1.0aB CV(%) 10.53 13.63 23.06 DS(%) 1 7.5b 11.0abC 17.5bC 16.0bC 32.0bB 53.0aA 13.0bD 44.0aC 75.5aB 95.5aA 96.5aA 2 9.0b 8.0abB 8.0bcB 12.0bcB 8.5cB 34.0bA 7.5bC 21.0bB 27.0cB 96.0aA 98.5aA 3 4.5b 3.5bA 3.5cA 5.0cA 8.0cA 9.5cA 6.0bC 8.0cC 24.0cB 59.0bA 72.0bA 4 23.5a 16.0aC 30.0aB 28.5aB 48.5aA 41.0bA 54.0aB 45.0aB 55.5bB 85.0aA 83.0bA CV(%) 21.55 27.25 12.29 MC (%) 1 9.9 12.9 38.9 35.8 28.6 40.9 14.4 37.1 26.8 22.8 42.6 2 10.4 13.3 33.1 26.9 28.0 41.5 14.3 30.0 32.5 30.1 41.9 3 10.1 9.4 31.5 28.3 24.4 39.7 13.9 34.1 29.7 26.6 43.8 4 10.3 10.3 22.7 31.7 27.2 42.5 17.0 38.6 31.5 30.9 41.4 *Mean values followed by the same lowercase letters in the columns (among lots) and the same uppercase letters in the rows (among aging times) do not differ from each other according to the Tukey test at 5% probability. Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) Information about Cratylia argentea seeds 180 Discussion large seeds accumulated more N and K in shoots and roots. Specific work on the relationship of seed size and C. argentea seeds were found to have protein-carbohydrate nutrients should be conducted with seeds of C. argentea reserves, following the general pattern of Fabaceae to clarify this. (Marcos-Filho 2016). The low concentration of lipids The germination test is fundamental for determining observed in C. argentea seeds is favorable for longevity the value of seeds for sowing and vigour of different because lipids have lower chemical stability compared seed lots (Marcos-Filho 2016). Although the seed lots to other reserve components due to the lipid peroxidation exhibited high germination percentages, the constant process (Parkhey et al. 2012). The high concentration of temperatures of 20 and 25 °C and alternating temperatures proteins is an important factor for development of the of 20/30 °C led to lower GSI compared to the constant embryonic axis and formation of the seedling (Verma temperature of 30 °C. Therefore, the constant temperature et al. 2015; Finch-Savage and Bassel 2016). However, of 30 °C can be considered better for germination of C. proteins have high affinity with water, which contributes argentea seeds. These results indicate C. argentea is to reduction in storage potential (Marcos-Filho 2016). a truly tropical species and best suited to sowing in the For that reason, storage is recommended under controlled summer months if grown in the subtropics. Based on temperature and relative humidity conditions (Ramos the germination curve at this temperature, day 10 can be et al. 2003). The composition of C. argentea seeds is defined as ideal for the first germination count (attaining similar to the composition of pea (Pisum sativum) seeds 50 % of normal seedlings) and day 20 for final evaluation with a similar quantity of proteins and a low percentage of germination (total stabilization of germination) for the of lipids (Marcos-Filho 2016). 4 seed lots analyzed. The temperature of 35 °C led to low Seeds of C. argentea did not show dormancy as germination percentages in all the seed lots analyzed. previously observed by Montoya et al. (2009), who Temperatures above the ideal during germination can evaluated the effects of cryopreservation in C. argentea cause thermoinhibition, with related oxidative stress and seeds. These authors observed a high initial physiological protein degradation (Liu et al. 2015; Mittler 2017). These quality of seeds and lack of physical or physiological apparent deleterious effects were more pronounced in lot 4, dormancy which drastically reduced germination during which was related to the initial lower germination capacity ambient storage conditions. Although the tegument of low vigor seeds. C. argentea is present in the Amazon, hardness may be responsible for dormancy and contribute Cerrado, and Caatinga biomes (Queiroz and Coradin to longevity during seed storage, from a practical point 1996) and these results were similar to other tree species of view these results confirm that planting can be carried collected from the same region in Brasil (Brancalion et al. out directly after harvest. In legume species (such as 2010). This suggests that germination tests for Brazilian C. argentea) the permeability of the seed coat can be related tree species from the Cerrado and Atlantic Forest should to some pigments, such as proanthocyanidins (Smýkal be conducted at 25 °C and from the Amazon at 30 °C. et al. 2014). According to these authors, the seed coat Accelerated aging at 41 °C for 48 h, 72 h and 96 h provides not just structural and protective functions,but effectively discriminated between seed lots of different has a decisive role in the timing of seed germination by vigor status with lots 2 and 4 showing lower vigor. At regulating water uptake. Another aspect to be considered high temperatures, respiratory rates are increased, causing is the seed size. C. argentea seeds are generally larger higher reactive oxygen species (ROS) production (such than the seeds of other tropical legumes of economic as hydrogen peroxide and superoxide anion), which are importance, such as Leucaena leucocephala, Crotalaria free radicals produced during normal metabolism and are species, Flemingia macrophylla, Tephrosia vogelii, involved in enzymatic reactions, mitochondrial electron Stylosanthes species, Calopogonium mucunoides and transport and signal transduction (Mittler 2017). When tropical kudzu (Pueraria phaseoloides). C. argentea above the basal levels, they cause damage, such as protein seeds have considerable nutrient reserves (as observed for denaturation and lipid peroxidation (Mittler 2017; Ebone starch and proteins), which are important characteristics et al. 2019). It is important to highlight the difference for ensuring greater survival in the field. Seed size can in seed moisture content in the accelerated aging test, be related to a higher amount of nutrients, as observed in especially when comparing 24 and 120 hours of exposure. bean seeds (Perin et al. 2002). These authors stated that Along with high temperatures, high humidities accelerate large seeds increased the plant height, the leaf area index seed respiration and contribute to cellular changes relating and the shoot and root biomass. Plants originating from to seed deterioration (Shu et al. 2020; Pinheiro et al. 2021). Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) Information about Cratylia argentea seeds 181 This is evident when observing the significant reduction plantas. Editora UFV, Viçosa, MG, ISBN: 85-7269-114-6 of normal seedlings and significant increase of dead seeds AOAC International. 2005. Official methods of analysis. 18th mainly in 96 and 120 h aging times. Edn. AOAC Inc., Gaithersburg, MD, USA. The observed results were similar to results with Araújo FS; Félix FC; Ferrari CS; Bruno RLA; Pacheco MV. seeds of L. leucocephala, where the accelerated 2017. Adequacy of the accelerated aging test to evaluate aging test conducted at 41 °C for 96 h was efficient in the vigor in leucena seeds. Revista Brasileira de Ciências Agrárias 12(1):92–97. (In Portuguese). doi: 10.5039/ differentiating seed lots (Araújo et al. 2017). Araújo agraria.v12i1a5422 et al. (2021) concluded the accelerated aging test with Araújo, JDO; Dias DCFDS; Nascimento WM; Martins AO; saturated NaCl solution (76 % RH) conducted at 41 °C Limão MAR. 2021. Accelerated aging test and antioxidant for 48 h is effective for evaluation of chickpea seed vigor. enzyme activity to assess chickpea seed vigor. Journal These test conditions can be used to identify differences of Seed Science 43:e202143038. doi: 10.1590/2317- in storage potential and seedling emergence in the field 1545v43253934 between seed lots (Marcos-Filho 2016). Although 48, 72 Brancalion PHS; Novembre ADLC; Rodrigues RR. 2010. and 96 h at 41 °C were all efficient for the accelerated Optimal temperature for seed germination of brazilian tree aging test, considering the practical implications, the species. Revista Brasileira de Sementes 32(4):15–21. (In time of 48 h is recommended for C. argentea seeds Portuguese). doi: 10.1590/S0101-31222010000400002 Brasil. 2009. Regras para análise de sementes. Ministério because it provides faster results. da Agricultura, Pecuária e Abastecimento (Mapa)/ACS, Brasília, DF. bit.ly/3w8bCa7 Conclusions Daibes LF; Cardoso VJM. 2018. Seed germination of a South American forest tree described by linear thermal time Temperature of 30 °C with the first germination count models. Journal of Thermal Biology 76:156–164. doi: at 10 days and final germination count at 20 days is 10.1016/j.jtherbio.2018.07.019 recommended for conducting germination tests of Ebone LA; Caverzan A; Chavarria G. 2019. Physiologic C. argentea. The combination of the temperature of 41 alterations in orthodox seeds due to deterioration °C with the period of exposure of 48 h is recommended processes. Plant Physiology and Biochemistry 145:34–42. for conducting the accelerated aging test for this doi: 10.1016/j.plaphy.2019.10.028 Edmond JB; Drapala WJ. 1958. The effects of temperature, species. C. argentea seeds are exalbuminous, have sand and soil, and acetone on germination of okra seed. hypogeal germination and do not exhibit seed dormancy. Proceedings of the American Society of Horticultural C. argentea seeds predominantly accumulate starch Science 71:428-434. and protein as reserves and contain a low percentage of Fang SZ; Wang JY. 2007. Changes in the biochemical lipids. All this information is important to optimize the composition and enzyme activity during dormancy release use of the species as a forage. of Cyclocarya paliurus seeds. Forestry Studies in China 9(1):7–13. doi: 10.1007/s11632-007-0002-6 Acknowledgments Finch-Savage WE; Bassel GW. 2016. Seed vigour and crop establishment: extending performance beyond adaptation. We thank the Research Support Foundation of the State Journal of Experimental Botany 67(3):567–591. doi: 10.1093/jxb/erv490 of Acre (FAPAC), Federal Institute of Acre (IFAC), Giuliani C; Lazzaro L; Lippi MM; Calamassi R; Foggi B. 2015. Coordination for the Improvement of Higher Education Temperature-related effects on the germination capacity of Personnel (CAPES) and National Development Council black locust (Robinia pseudoacacia L., Fabaceae) seeds. Scientific and Technological (CNPq) for the funding. Folia Geobotanica 50(3):275–282. doi: 10.1007/s12224- 015-9224-x References Lawal OU. 2011. Effect of storage on the nutrient composition and the mycobiota of sundried water melon seeds (Citrullus (Note of the editors: All hyperlinks were verified 18 August 2022). lanatus). Journal of Microbiology, Biotechnology and Food Sciences 1(3):267–276. bit.ly/3QNfXr9 Abud HF; Gonçalves NR; Reis RGE; Gallão MI; Innecco Liu SJ; Xu HH; Wang WQ; Li N; Wang WP; Møller IM; Song R. 2010. Morphology of seed and seedling of safflower. SQ. 2015. A proteomic analysis of rice seed germination as Revista Ciência Agronômica 41(2):259–265. (In affected by high temperature and ABA treatment. Physiologia Portuguese). doi: 10.1590/S1806-66902010000200013 Plantarum 154(1):142–161. doi: 10.1111/ppl.12292 Alvarez VVH; Dias LE; Ribeiro FS; Souza RB; Fonseca CA, Lone AB; Colombo RC; Andrade BLG; Takahashi LSA; eds. 2001. Métodos de análises de enxofre em solos e Faria RT. 2016. Physical characterization of Rhipsalis Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) 182 E.P.L. Mattar, D.T. Pinheiro, W.D. Pereira, B.P. Brasileiro, W.J.R. Matrangolo, P.C. Hilst, P.A. Hormaza and D.C.F.S. Días (Cactaceae) fruits and seeds germination in different condensed tannins from tropical forage legumes. Pesquisa temperatures and light regimes. Brazilian Journal of Agropecuária Brasileira 53(9):1070–1077. doi: 10.1590/ Biology 76(2):367–373. doi: 10.1590/1519-6984.15914 s0100-204x2018000900011 Los FGB; Zielinski AAF; Wojeicchowski JP; Nogueira A; Perin A; Araújo AP; Teixeira MG. 2002. Effect of seed size Demiate IM. 2018. Beans (Phaseolus vulgaris L.): whole on biomass and nutrient accumulation and on grain seeds with complex chemical composition. Current Opinion yield of common bean. Pesquisa Agropecuária Brasileira in Food Science 19:63–71. doi: 10.1016/j.cofs.2018.01.010 37(12):1711–1718. (In Portuguese). doi: 10.1590/S0100- Maguire JD. 1962. Speed of germination-aid selection 204X2002001200006 and evaluation for seedling emergence and vigor. Crop Pinheiro DT; Dias DCFS; Medeiros AD de; Ribeiro JPO; Science 2(2):176–177. doi: 10.2135/cropsci1962. Silva FL da; Silva LJ da. 2021. Weathering deterioration in 0011183X000200020033x pre-harvest of soybean seeds: physiological, physical, and Marcos-Filho J. 2015. Seed vigor testing: an overview of the morpho-anatomical changes. Scientia Agricola 78(suppl past, present and future perspective. Scientia Agricola 1):e20200166. doi: 10.1590/1678-992X-2020-0166 72(4):363–374. doi: 10.1590/0103-9016-2015-0007 Queiroz LP de; Coradin L. 1996. Biogeografia de Cratylia e Marcos-Filho J. 2016. Seed physiology of cultivated plant. areas prioritárias para coleta . In: Pizarro E; Coradin L. ABRATES, Londrina, PR, Brazil. eds. Taller de Trabajo sobre Cratylia. Potencial del género Mattar EPL; Barros TTV; Brasileiro BP; Mattielo, EM; Cratylia como leguminosa forrajera: Memorias. Centro Coelho MRR; Gama GFV; Dias DCFS. 2018. Response of Internacional de Agricultura Tropical (CIAT); Empresa Cratylia argentea (Desvaux) O. Kuntze to inoculation with Brasileira de Pesquisa Agropecuária (EMBRAPA); Centro Rhizobium sp. and Bradyrhizobium sp. strains. Australian de Pesquisa Agropecuária do Cerrado (CPAC), Cali, Journal of Crop Science 12(06):849–854. doi: 10.21475/ Colombia. p. 1–28. hdl.handle.net/10568/82017 ajcs.18.12.06.PNE524 R Core Team. 2020. R Development Core Team. R: A Language Mittler R. 2017. ROS are good. Trends in Plant Science and Environment for Statistical Computing. r-project.org/ 22(1):11-19. doi: 10.1016/j.tplants.2016.08.002 Ramos AKB; Souza MA de; Pizarro EA. 2003. Algumas Montoya DM; Bonilla CR; Sánchez MS; Cardozo CI; informações sobre a produção e o armazenamento de sementes Escobar R. 2009. Physiological response of Cratylia de Cratylia argentea. Circular Técnica 25. EMBRAPA argentea (Desvaux) O. Kuntze seeds to storage and cryo- Cerrados, Planaltina, DF, Brazil. bit.ly/3CoVeGk preservation conditions. Acta Agronómica 58(3):167–173. Reguera M; Conesa CM; Gil-Gómez A; Haros CM; Pérez- bit.ly/3A0cgHY Casas MA; Briones-Labarca V; Bolaños L; Bonilla I; Müller FL; Raitt LM; Cyster LF; Cupido CF; Samuels MI; Álvarez R; Pinto K; Mujica A; Bascuñán-Godoy L. 2018. Chimphango SBM; Boatwright JS. 2019. The effects of The impact of different agroecological conditions on the temperature, water availability and seed burial depth on nutritional composition of quinoa seeds. PeerJ 6:e4442. seed germination and seedling establishment of Calobota doi: 10.7717/peerj.4442 sericea (Fabaceae). South African Journal of Botany Sarria PI; Martens SD. 2013. The voluntary intake in growing 121:224–229. doi: 10.1016/j.sajb.2018.11.012 pigs of four ensiled forage species. Agricultural and Food Oliveira AKM; Ribeiro JWF; Pereira KCL; Silva CAA. 2013. Science 22(1):201–206. doi: 10.23986/afsci.6957 Effects of temperature on the germination of Diptychandra Shu Y; Zhou Y; Mu K; Hu H; Chen M; He Q; Huang S; Ma aurantiaca (Fabaceae) seeds. Acta Scientiarum: Agronomy H; Yu X. 2020. A transcriptomic analysis reveals soybean 35(2):203–208. doi: 10.4025/actasciagron.v35i2.15977 seed pre-harvest deterioration resistance pathways under Oliveira GL; Dias LAS; Dias DCFS; Soares MM; Silva LJ high temperature and humidity stress. Genome 63(2):115– da. 2014. Accelerated ageing test to evaluate vigour in 124. doi: 10.1139/gen-2019-0094 Jatropha curcas L. seeds. Revista Ciência Agronômica Silva ME da; Araújo JV de; Silva JA da; Carvalho LM. 45(1):120–127. 10.1590/S1806-66902014000100015 De; Chagas E das; Ribeiro RR. 2017. Anthelmintic Oliveira MAS; Nery MC; Ribeiro KG; Rocha AS; Cunha efficacy of Cratylia argentea (Desv.) Kuntze against the PT. 2020. Accelerated aging for evaluation of vigor gastrointestinal nematodes of sheep. Ciências Agrárias in Brachiaria brizantha ‘Xaraés’ seeds. Journal of 38(5):3105–3112. doi: 10.5433/1679-0359.2017v38n Seed Science 42:e202042006. doi: 10.1590/2317- 5p3105 1545v42216691 Smýkal P; Vernoud V; Blair MW; Soukup A; Thompson RD. Parkhey S; Naithani SC; Keshavkant S. 2012. ROS production 2014. The role of the testa during development and in and lipid catabolism in desiccating Shorea robusta seeds establishment of dormancy of the legume seed. Frontiers during aging. Plant Physiology and Biochemistry 57:261– in Plant Science 5:351. doi: 10.3389/fpls.2014.00351 267. doi: 10.1016/j.plaphy.2012.06.008 Thompson JC. 1990. Methods for the determination of Pereira TP; Modesto EC; Nepomuceno DD; Oliveira OF carbohydrates by ion chromatography. Proceedings of de; Freitas RSX de; Muir, JP; Dubeux Jr. JCB; Almeida the Sugar Processing Research Conference (USA), Sugar JCC. 2018. Characterization and biological activity of Processing Research Inc., San Francisco, CA, USA, p. Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) Information about Cratylia argentea seeds 183 381–396. ISSN: 0730-6490 legumes. Critical Reviews in Plant Science 34(1–3):105– Valles-de la Mora B; Castillo-Gallegos E; Alonso-Díaz 143. doi: 10.1080/07352689.2014.897907 MA; Ocaña-Zavaleta E; Jarillo-Rodríguez J. 2017. Live- Verma G; Mishra S; Sangwan N; Sharma S. 2015. Reactive weight gains of Holstein × Zebu heifers grazing a Cratylia oxygen species mediate axis-cotyledon signaling to induce argentea/Toledo-grass (Brachiaria brizantha) association reserve mobilization during germination and seedling in the Mexican humid tropics. Agroforestry Systems establishment in Vigna radiata. Journal of Plant Physiology 91(6):1057–1068. doi: 10.1007/s10457-016-9980-5 184:79–88. doi: 10.1016/j.jplph.2015.07.001 Valles-de la Mora B; Castillo-Gallegos E; Ocaña-Zavaleta Wiebach J; Nagel M; Börner A; Altmann T; Riewe D. 2020. E; Jarillo-Rodríguez J. 2014. Cratylia argentea: A Age‐dependent loss of seed viability is associated with potential fodder shrub in silvopastoral systems. Yield increased lipid oxidation and hydrolysis. Plant, cell & and quality of accessions according to regrowth ages environment, 43(2):303-314. doi: 10.1111/pce.13651 and climatic seasons. Revista Chapingo, Serie Ciencias Wu Q; Bai X; Wu X; Xiang D; Wan Y; Luo Y; Shi X; Li Q; Forestales y del Ambiente. 20(2):277–293. doi: 10.5154/r. Zhao J; Qin P; Yang X; Zhao G. 2020. Transcriptome rchscfa.2013.11.040 profiling identifies transcription factors and key homologs Vaz Patto MC; Amarowicz R; Aryee ANA; Boye JI; Chung HJ; involved in seed dormancy and germination regulation of Martín-Cabrejas MA; Domoney C. 2015. Achievements Chenopodium quinoa. Plant Physiology and Biochemistry and challenges in improving the nutritional quality of food 151:443–456. doi: 10.1016/j.plaphy.2020.03.050 (Received for publication 20 May 2021; accepted 05 August 2022; published 30 September 2022) © 2022 Tropical Grasslands-Forrajes Tropicales is an open-access journal published by International Center for Tropical Agriculture (CIAT), in association with The Tropical Crops Genetic Resources Institute of The Chinese Academy of Tropical Agricultural Sciences (TCGRI-CATAS). This work is licensed under the Creative Commons Attribution 4.0 International (CC BY 4.0) license. Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) Tropical Grasslands-Forrajes Tropicales (2022) Vol. 10(3):184–194 184 doi: 10.17138/TGFT(10)184-194 Research Paper Seasonal nutritive value and in vitro fermentation kinetics of foliage of some multipurpose shrub species in northeastern Mexico Valor nutritivo estacional y cinética de fermentación in vitro del follaje de algunas especies de arbustos multipropósito en el noreste de México MIGUEL CHÁVEZ ESPINOZA1, HUGO BERNAL BARRAGÁN2, MARIBEL GUERRERO CERVANTES3, ISRAEL CANTÚ SILVA4, MAURICIO COTERA CORREA4, HUMBERTO GONZÁLEZ RODRÍGUEZ4 AND ANDRÉS EDUARDO ESTRADA CASTILLÓN4 1Centro Universitario del Sur, Universidad de Guadalajara, Ciudad Guzmán, JAL, México. cusur.udg.mx 2Facultad de Agronomía, Monterrey, Universidad Autónoma de Nuevo León, NLE, México. agronomia.uanl.mx 3Facultad de Medicina Veterinaria y Zootecnia, Universidad Juárez del Estado de Durango, Durango, DUR, México. fmvz.ujed.mx 4Facultad de Ciencias Forestales, Universidad Autónoma de Nuevo León, Linares, NLE, México. fcf.uanl.mx Abstract This study aimed to determine the seasonal variation in chemical composition, metabolizable energy (ME) concentration, in vitro gas production patterns, in vitro organic matter digestibility (IVOMD) and in vitro true organic matter digestibility (IVTOMD) of foliage from 5 native shrub species (Celtis pallida, Croton suaveolens, Forestiera angustifolia, Guaiacum angustifolium and Parkinsonia aculeata) growing in semi-arid areas of northeastern Mexico between July 2018 and June 2019 at 2 research sites. Crude protein (CP) concentrations (>13.2 % DM) found in leaf material should meet or exceed the requirements for maintenance of small ruminants; C. pallida provided the highest CP% (20.8‒29.3 %). While CP% varied with season and site, species had a greater effect than either of those factors overall. ME concentrations ranged between 1.2 and 2.0 Mcal/kg DM. Neutral detergent fiber and acid detergent fiber concentrations ranged from 29.8 to 51.7 % DM and 9.8 to 33.0 % DM, respectively. Data obtained for IVOMD (34.5‒58.8 %) and IVTOMD (64.1‒88.7 %) demonstrate the high nutritive potential of leaf of browse species under study, especially C. pallida, as useful feed supplements for small ruminants in the semi-arid region of northeastern Mexico. Further studies could examine DM yields of browse from the various species, acceptance by small ruminants and their sustainability under regular defoliation by grazing animals. Keywords: Browse, chemical composition, digestibility, gas production, ruminants, season. Resumen Este estudio tuvo como objetivo determinar la variación estacional en la composición química, concentración de energía metabolizable (EM), parámetros de producción de gas in vitro, digestibilidad in vitro de la materia orgánica (DMO) y la digestibilidad verdadera in vitro de la materia orgánica (DVIVMO) del follaje de 5 especies arbustivas nativas (Celtis pallida, Croton suaveolens, Forestiera angustifolia, Guaiacum angustifolium y Parkinsonia aculeata) que crecen en las regiones semiáridas del noreste de México, en dos sitios, entre julio del 2018 y julio del 2019. La concentración de proteína cruda (PC) (>13.2 % de MS) que se encontraron en el material foliar cumple o supera los requisitos para el mantenimiento de los pequeños rumiantes. C. pallida presento el contenido de PC % más alto (20.8‒29.3 %). Si bien el % de PC varió con la estación y el sitio, las especies tuvieron un efecto mayor que cualquiera Correspondence: Humberto González Rodríguez, Universidad Autónoma de Nuevo León, Facultad de Ciencias Forestales, Carr. Nac. No. 85, km 145, Linares, Nuevo León, 67700, Mexico. Email: humberto.gonzalezrd@uanl.edu.mx Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) Nutritive value and leaf digestibility of browse species 185 de de los otros dos factores. Las concentraciones de EM oscilaron entre 1.2 y 2.0 Mcal/kg MS. Las concentraciones de fibra detergente neutra (FDN) y fibra detergente ácida (FDA) oscilaron entre 29.8 y 51.7 % de MS y entre 9.8 y 33.0 % de MS, respectivamente. Los datos obtenidos para DMOIV (34.5‒58.8 %) y DVMOIV (64.1‒88.7 %) demostraron el alto potencial nutritivo de las hojas de las especies arbustivas en este estudio, especialmente C. pallida, como un suplemento alimenticio útil para pequeños rumiantes en las regiones semiáridas del noreste de México. Futuros estudios podrían examinar los rendimientos de MS bajo condiciones de ramoneo de diversas especies, así como su aceptación por parte de los pequeños rumiantes y su sostenibilidad bajo defoliación frecuente por animales en pastoreo. Palabras clave: Composición química, digestibilidad, estación; producción de gas, ramoneo, rumiantes. Introduction Native shrub species grown in the Tamaulipan thornscrub plant ecosystem, in northeastern Mexico, Shrub species have an important role as feed resources have been studied to determine their potential as a for small ruminants in arid and semi-arid agro-pastoral source of minerals, protein and digestible dry matter areas and rangelands (Cerrillo et al. 2006; Habib et al. for small ruminants (Domínguez-Gómez et al. 2011; 2016; Oliveira et al. 2018). The leaves of browse shrubs Guerrero-Cervantes et al. 2012b; Chávez-Espinoza et can be used as protein and fiber supplements by small al. 2020). However, to date there is limited information ruminants grazing according to the season and fodder about the kinetics of ruminal fermentation, as well availability in the rangeland, especially during dry as the metabolizable energy of browse species seasons in extensive areas, such as semi-arid regions of throughout the year. Thus, the objectives of this study northeastern México (Foroughbakhch et al. 2013). Pal were to determine seasonal chemical composition, et al. (2015) and Piñeiro-Vázquez et al. (2017) reported metabolizable energy (ME), in vitro fermentation that leaves of some shrubs and fodder trees may also parameters, in vitro organic matter digestibility (OMD) lower methane production and contribute to methane and in vitro true organic matter digestibility (IVTOMD) mitigation in small-ruminant production systems in plus nutritional potential for small ruminants, of foliar developing countries. However, leaves of leguminous tissue of common native shrub species in northeastern trees and shrubs contain a wide range of plant secondary México, including Celtis pallida Torr. (Cannabaceae), compounds (PSCs), which can represent a challenge Croton suaveolens Torr. (Euphorbiaceae), Forestiera to their value as feed resources (Ramı́rez et al. 2000) angustifolia Torr. (Oleaceae), Guaiacum angustifolium by limiting digestibility of metabolizable energy and Engelm. (Zygophyllaceae) and Parkinsonia aculeata L. protein (Camacho et al. 2010). On the other hand Soliva (Leguminosae). et al. (2008) argued that these components did not affect protein degradation of tropical foliage from shrubs and Materials and Methods trees to an excessive extent. Since chemical analysis does not reflect the effects of Study research sites these PSCs, it is important to choose additional methods of analysis to assess the nutritional quality of native The research was carried out at 2 sites in the counties of shrubs in rangelands. In vitro gas production, measured Linares and Los Ramones, state of Nuevo León, Mexico, using pressure transducers and gas syringes (Getachew briefly described as follows: et al. 2005), can be used to quantify the nutritional Site 1 - Linares: Located in the Experimental quality of foliage of shrubs and fodder trees (Larbi et al. Campus of the School of Forest Sciences, Universidad 1998). It is sensitive to the presence of PSCs in foliage Autónoma de Nuevo León, in Linares County (24°47'45" (Sebata et al. 2011), since those compounds, such as N, 99°32'31" W; 350 masl). The climate is semi-arid condensed tannins (CT), in browse species could limit and subtropical with a warm summer (González- in vitro fermentation (Guerrero et al. 2012a). The DaisyII Rodríguez et al. 2004). Average air temperature during incubator is adaptable and efficient in determining the experimental period varied from 14.1 °C in January forage digestibility (Holden 1999) and is reliable when to 30.4 °C in August and 554 mm rainfall was received. compared with in sacco tests, but with a minimal number Site 2 - Los Ramones: Located in "El Abuelo Ranch" of ruminants needed as rumen liquor donors (Buthelezi in Los Ramones County (25°39'46" N, 99°27'51" W; 200 et al. 2019). masl). The climate is semi-arid with a warm summer Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) 186 M. Chávez-Espinoza et al. (González-Rodríguez et al. 2004). Average monthly air and commercial concentrate (75:25, respectively). To temperature during the study ranged from 14.3 °C in prepare the inoculum, ruminal liquid was mixed with January to 31.5 °C in August and rainfall was 667 mm. a sodium buffer solution and ammonium bicarbonate (35 g NaHCO3 and 4 g NH4HCO3 per liter) in a ratio of Foliage sampling and preparation 1:2 (v/v). Each syringe was inoculated with 40 mL of this buffer solution and placed in an upright position in At each site, 3 representative experimental plots (50 × 50 a water bath at 39 °C. Three syringes containing only m) were randomly demarcated without disturbance, and both 40 mL of inoculum served as Controls. Gas production young and mature leaves of Celtis pallida, Croton suaveolens, was registered at 0, 3, 6, 9, 12, 24, 48, 72 and 96 h after Forestiera angustifolia, Guaiacum angustifolium and inoculation. Data were adjusted using the non-linear Parkinsonia aculeata were randomly and manually equation proposed by Orskov and McDonald (1979), collected, at a browsing height of 1.5 m from 5 representative p = a + b (1-ect), and effective degradability of DM plants (repetitions) of each species in each plot. Leaves (EDDM) was calculated using the equation proposed were sampled monthly in summer (June and August 2018 by McDonald (1981), EDDM = (a+b) c/(c+k) (e-(ct)LT), and July 2019), autumn (October‒November 2018), winter where: (December 2018‒February 2019) and spring (March‒May p is the gas produced at time ‘t’; 2019). The collected leaf tissue was transferred to the a is gas production from the immediately soluble Chemistry Laboratory of the School of Forest Sciences, fraction (mL); where samples were partially dried at 55 °C for 24 h using b is gas production from the insoluble fraction (mL); a forced-air oven (Felisa®, Model FE-292AD, Mexico). c is gas production rate constant for the insoluble fraction (b); Thereafter, leaf tissue was ground in a Thomas Wiley mill t is incubation time (h); (Thomas Scientific Apparatus, Model 3383, USA) using a k is the rumen outflow rate assumed to be 0.05/h, i.e. No. 60 mesh (1 × 1 mm) and stored in zip-lock plastic bags. 5 %/h; and LT is the lag time (h). Chemical analyses Since in vitro production of gas is proportional to degraded DM (Menke et al. 1979), net gas yield Ground leaf tissue was dried at 100 °C for 24 h in a at 24-h incubation of the substrate (mL/200 mg DM) forced-air oven. Organic matter (OM%) was determined was used to calculate metabolizable energy (ME) and by incinerating a dried sample at 550 °C for 3.5 h in in vitro organic matter digestibility (OMD%), using a furnace (Thermo Scientific, Model F48010, USA), equations proposed by Menke et al. (1979) and Menke (Method no. 942.05; AOAC 2011); crude protein (CP%) and Steingass (1988), respectively: from N concentration, obtained by a CHNS/O analyzer ME (Mcal/kg DM) = [2.20 + 0.136(GP24h) + 0.057(CP) (2400 series II, Perkin Elmer) (Method no. 990.03; AOAC + 0.0029(EE2)]/4.184 and OMD (%) = 14.88 + 0.889 × 2011); neutral detergent fiber (NDF%), acid detergent GP24h + 0.45 × CP + 0.0651 × ash, fiber (ADF%) and lignin (%) according to procedures where: described by Van Soest et al. (1991); and ether extract GP24h = gas production after 24 h of incubation (mL (EE%) by extracting lipids with petroleum ether, using gas/200 mg DM); an AnkomXT15 extractor (AOCS AM 5-04). All analyses CP = crude protein (% DM); were performed in triplicate. EE = ether extract (% DM); and ash = ash concentration (%). In vitro gas production determinations In vitro true organic matter digestibility In vitro gas production of collected leaf tissue was determined at the School of Veterinary Medicine, In vitro true organic matter digestibility (IVTOMD) Universidad Juárez del Estado de Durango, Durango, was determined using a DaisyII incubator (ANKOM Mexico, using the technique proposed by Menke Technology, Macedon, NY, USA) in the Laboratory and Steingass (1988). Briefly, 500 mg DM samples in of Animal Nutrition and Feed Quality at the School of triplicate were placed in 100 mL calibrated glass syringes. Agronomy, Universidad Autónoma de Nuevo León, Ruminal fluid was collected from 3 rumen-fistulated Mexico. Foliar samples (approximately 250 mg DM), in Dorper wethers (45 kg body weight), fed with alfalfa hay triplicate, were placed in multilayer polyester filter bags Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) Nutritive value and leaf digestibility of browse species 187 (F57; 5.0 × 5.5 cm, ANKOM Technology Corp., Macedon, (Statistical Package for the Social Sciences) software NY, USA) previously washed with acetone and dried in a package (Version 22.0 for Windows, SPSS Inc., Chicago, forced-air oven at 60 °C for 2 h. Inoculum was prepared by IL, USA) was used for all statistical analyses. diluting ruminal fluid from 2 Saint Croix wethers, provided with a ruminal cannula and fed a concentrate:forage ration Results (80:20) (not containing foliar tissue from the studied browse species), with a buffer solution in a ratio of 1:4. The effects of season, site and species and their respective Inoculum was added to jars containing the filter bags, double and triple interactions on OM, NDF, ADF, lignin, which were purged with CO2 and placed in an incubator for CP and EE concentrations are shown in Table 1. OM, 48 h at 39 °C. Jars were then removed from the incubation NDF, ADF, lignin, CP and EE concentrations were chamber and the bags were washed with distilled water. significantly affected (P<0.05) by both season and species Thereafter, the bags were placed in the Ankom200 fiber (P<0.001). The interaction season × site was significant analyzer (ANKOM Technology Corp., Macedon, NY, (P<0.05) for only lignin and CP concentrations, USA) and treated with neutral detergent solution for 75 while the interaction season × species was significant min, according to the manufacturer's specifications and (P<0.001) for all chemical components. Interaction site guidelines. The bags were rinsed with hot water and × species was significant (P<0.001) for OM, NDF, ADF acetone, before being dried at 55 °C for 24 h. IVTOMD and EE concentrations. For triple interaction, season × was calculated as the difference between OM incubated site × species, significant differences were found for OM and residue after neutral detergent treatment. (P<0.001) and NDF (P<0.05) concentrations. While there was marked variation between species Environmental variables at all sites in most seasons, overall C. pallida showed consistently lowest OM concentration, followed by To measure temperature and precipitation at each site, G. angustifolium, with the remaining species higher. automated HOBO sensors (HOBO Pro Temp/RH Series, On the other hand, C. pallida showed the highest CP Forestry Suppliers Inc., Jackson, MS, USA) were used to concentration of all species in summer and autumn. record environmental variables such as relative humidity Overall CP% was highest in autumn (seasonal mean (%) and air temperature (°C) every hour. Daily rainfall 26.0 %) and lowest in summer (19.7 %). Overall OMD (mm) was measured using a Davis brand automated rain was highest in spring (49.3 %) and lowest in summer gauge, connected to a HOBO Event Onset recorder. (41.4 %), while IVTOMD did not vary markedly between seasons. (75.3‒76.7 %) (Table 2). C. pallida was Statistical analyses consistently highest for both parameters. The 96-h cumulative in vitro gas production (GP96) The effects of site (2), season (4) and species as well as patterns for the 5 shrub species during summer, their double (site × season, site × species and season × autumn, winter and spring at the 2 research sites are species) and triple (site × season × species) interactions shown in Figure 1. Overall, GP96 values in spring on chemical analyses, in vitro gas production parameters samples were higher than in other seasons, while and digestibility data were determined via analysis of C. pallida consistently registered the highest, and variance using a completely randomized design with G. angustifolium the lowest gas production values. a factorial arrangement (Montgomery 2004). Pearson P. aculeata consistently had highest gas production correlation analyses were performed between chemical during the first 12 hours of incubation, while C pallida, composition, digestibility and environmental variables C. suaveolens and F. angustifolia produced much more recorded during the experimental period. The SPSS gas during the following 36 hours. Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) 188 M. Chávez-Espinoza et al. Table 1. Chemical composition (% DM; n = 9) of leaves from 5 shrub species collected at 2 sites in each season during 2018‒2019. Season Species OM1 NDF ADF Lignin CP4 EE S12 S2 S1 S2 S1 S2 S1 S2 S1 S2 S1 S2 Summer C. pallida 80.2 79.4 46.5 51.3 11.0 12.2 2.3 2.7 20.9 24.5 2.7 3.0 C. suaveolens 93.0 91.8 41.5 38.9 23.5 22.3 5.8 5.5 16.3 17.8 3.1 2.3 F. angustifolia 92.2 93.3 29.8 31.3 15.8 16.1 7.6 8.6 13.4 16.2 2.2 2.1 G. angustifolium 83.0 86.3 39.6 39.3 26.0 25.4 14.3 15.2 20.9 21.3 3.6 5.6 P. aculeata 90.4 92.4 45.3 48.9 27.4 30.1 9.6 9.5 18.2 18.8 3.0 2.7 Site mean 87.8 88.6 40.5 41.9 20.8 21.2 7.9 8.3 18.0 19.7 2.9 3.1 s.e.m. 0.3 0.3 0.8 0.8 0.4 0.4 0.3 0.3 0.6 0.6 0.1 0.1 Season mean 88.2 41.2 21.0 8.1 18.8 3.0 s.e.m. 0.2 0.6 0.3 0.2 0.2 0.1 Autumn C. pallida 79.1 79.0 44.4 50.8 11.7 10.7 2.4 2.2 29.3 29.2 2.3 2.8 C. suaveolens 91.7 91.7 48.0 40.9 24.6 26.3 7.3 9.2 25.7 25.9 1.4 1.4 F. angustifolia 93.3 93.2 37.5 38.9 17.7 18.3 8.6 10.1 22.9 23.5 2.5 2.0 G. angustifolium 81.0 85.2 36.6 41.1 24.7 24.6 14.7 14.1 23.7 26.7 4.8 6.0 P. aculeata 90.4 92.8 44.3 49.6 25.5 29.0 9.1 10.7 24.9 27.6 2.4 1.9 Site mean 87.1 88.4 42.1 44.3 20.9 21.8 8.4 9.2 25.3 26.6 2.7 2.8 s.e.m. 0.3 0.3 0.8 0.8 0.4 0.4 0.3 0.3 0.6 0.6 0.1 0.1 Season mean 87.7 43.2 21.3 8.8 26.0 2.8 s.e.m. 0.2 0.6 0.3 0.2 0.2 0.1 Winter C. pallida 76.7 78.7 32.0 47.0 9.9 10.2 2.2 1.9 24.5 20.8 1.8 1.9 C. suaveolens 91.8 92.3 40.4 39.6 22.8 22.1 8.7 7.4 23.4 21.0 1.6 1.7 F. angustifolia 92.2 91.5 34.7 32.1 17.9 16.3 9.2 7.9 20.6 18.1 1.8 1.9 G. angustifolium 82.5 82.4 37.9 34.6 24.1 21.6 13.8 12.7 23.1 22.6 5.0 5.1 P. aculeata 91.3 93.6 47.5 51.5 28.3 33.0 10.2 11.7 21.7 23.8 2.3 2.4 Site mean 86.9 87.7 38.5 40.9 20.6 20.6 8.8 8.3 22.7 21.3 2.5 2.6 s.e.m. 0.3 0.3 0.8 0.8 0.4 0.4 0.3 0.3 0.6 0.6 0.1 0.1 Season mean 87.3 39.7 20.6 8.6 22.0 2.5 s.e.m. 0.2 0.6 0.3 0.2 0.2 0.1 Spring C. pallida 78.2 81.1 38.5 46.8 9.8 10.4 2.0 1.6 23.4 22.1 2.2 2.0 C. suaveolens 91.9 92.0 41.6 43.9 21.9 18.2 5.5 4.5 24.5 23.6 2.7 2.5 F. angustifolia 93.9 93.7 36.2 34.6 20.3 19.3 12.0 11.3 12.8 13.6 2.0 2.1 G. angustifolium 89.1 89.9 40.8 35.3 26.7 23.0 14.1 12.5 21.9 21.6 2.7 3.1 P. aculeata 94.0 94.4 46.2 51.7 24.3 29.1 7.4 8.5 26.3 19.5 2.8 2.3 Site mean 89.4 90.2 40.7 42.5 20.6 20.0 8.2 7.7 21.8 20.0 2.5 2.4 s.e.m. 0.3 0.3 0.8 0.8 0.4 0.4 0.3 0.3 0.6 0.6 0.1 0.1 Season mean 89.8 41.6 20.3 7.9 20.9 2.4 s.e.m. 0.2 0.6 0.3 0.2 0.2 0.1 Factors Season <.001 <.001 0.047 0.002 <.001 <.001 P-value Site <.001 0.001 0.453 0.763 0.955 0.384 Species <.001 <.001 <.001 <.001 <.001 <.001 Season × Site 0.754 0.922 0.238 0.020 0.006 0.657 Season × Species <.001 <.001 <.001 <.001 <.001 <.001 Site × Species <.001 <.001 <.001 0.057 0.891 <.001 Season × Site × Species <.001 0.009 0.176 0.143 0.120 0.069 1OM = organic matter; NDF = neutral detergent fiber; ADF = acid detergent fiber; CP = crude protein; EE = ether extract. 2S1 = Site 1 (Linares); S2 = Site 2 (Los Ramones). Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) Nutritive value and leaf digestibility of browse species 189 Table 2. In vitro gas production at 24 h of incubation (mL/200 mg DM), metabolizable energy (Mcal/kg DM), gas production parameters (a, b, c, a+b and EDDM/500 mg DM), in vitro organic matter digestibility (g/100 g) and in vitro true organic matter digestibility (g/100 g) of leaves of 5 shrub species (n = 9) collected at 2 sites for each season of the year during 2018‒2019. Season / Species GP 1 24 ME (a) (b) (a+b) (c) EDDM IVOMD IVTOMD S12 S2 S1 S2 S1 S2 S1 S2 S1 S2 S1 S2 S1 S2 S1 S2 S1 S2 Summer C. pallida 26.8 24.5 1.7 1.7 -20.5 -19.2 137.3 122.5 116.8 103.3 0.05 0.04 41.4 35.1 49.4 49.1 85.2 84.5 C. suaveolens 23.3 18.5 1.5 1.4 -11.8 -9.0 123.3 107.3 111.6 98.3 0.04 0.03 38.7 30.6 43.4 39.9 71.1 73.4 F. angustifolia 14.6 13.4 1.2 1.2 1.5 1.9 125.8 137.6 128.7 139.5 0.02 0.01 30.1 26.9 34.4 34.5 77.9 78.3 G. angustifolium 14.3 12.0 1.3 1.2 -6.9 -6.4 58.7 55.2 51.8 48.8 0.07 0.06 24.7 20.3 38.1 36.0 76.1 75.1 P. aculeata 23.8 22.8 1.6 1.5 -3.9 -0.7 82.5 76.4 78.7 75.7 0.07 0.07 42.6 42.2 44.8 44.0 72.3 69.9 Season mean 19.4 1.42 -7.5 102.7 95.3 0.05 33.3 41.4 76.4 s.e.m. 0.6 0.0 0.5 1.4 1.4 0.00 1.0 0.6 0.4 Autumn C. pallida 14.2 20.3 1.4 1.6 -12.9 -15.9 124.1 110.5 111.2 94.5 0.02 0.04 24.8 29.6 42.1 47.5 83.4 84.2 C. suaveolens 18.7 16.7 1.5 1.4 -6.3 -4.1 100.3 90.0 94.0 85.9 0.03 0.03 31.5 29.3 43.6 41.9 68.9 66.2 F. angustifolia 14.2 16.6 1.3 1.4 3.2 8.4 141.1 130.9 144.9 138.5 0.01 0.01 27.8 32.1 38.2 40.7 75.2 77.4 G. angustifolium 12.6 13.4 1.3 1.4 -3.0 -1.8 51.0 50.9 48.0 49.2 0.05 0.05 22.1 23.1 38.0 39.8 79.0 78.3 P. aculeata 20.5 16.9 1.5 1.5 -3.4 2.6 83.1 78.8 79.7 82.2 0.05 0.04 36.6 32.6 45.0 42.8 72.7 68.3 Season mean 16.4 1.42 -3.3 96.1 92.8 0.03 28.9 42.0 75.3 s.e.m. 0.6 0.0 0.5 1.4 1.4 0.00 1.0 0.6 0.4 Winter C. pallida 29.4 36.8 1.8 2.0 -14.1 -18.9 118.1 132.3 104.0 113.3 0.05 0.07 46.9 55.8 53.6 58.3 88.2 88.7 C. suaveolens 26.5 31.4 1.7 1.8 -8.7 -9.1 112.8 116.4 104.1 107.3 0.05 0.05 46.3 50.6 49.5 52.7 71.1 70.2 F. angustifolia 19.6 22.0 1.4 1.5 3.0 5.9 120.5 114.2 123.5 120.1 0.03 0.03 43.5 43.3 42.1 43.1 75.4 76.3 G. angustifolium 15.1 16.7 1.3 1.4 -5.7 -5.2 55.5 53.7 49.8 48.5 0.07 0.09 26.0 29.3 39.8 41.0 78.4 79.6 P. aculeata 24.5 26.2 1.6 1.7 0.1 6.1 78.5 71.4 78.6 77.5 0.07 0.09 45.6 51.2 47.0 49.3 70.1 64.1 Season mean 24.8 1.6 -4.7 97.3 92.7 0.06 43.9 47.7 76.2 s.e.m. 0.6 0.0 0.5 1.4 1.3 0.00 1.0 0.6 0.4 Spring C. pallida 31.5 36.8 1.9 2.0 -18.7 -21.2 123.1 135.5 104.4 114.2 0.06 0.07 45.7 55.7 54.8 58.8 85.2 84.5 C. suaveolens 32.7 33.1 1.9 1.9 -10.0 -10.8 115.5 118.0 105.6 107.2 0.06 0.06 54.8 54.6 55.5 55.4 73.3 76.2 F. angustifolia 18.9 19.4 1.3 1.3 13.2 9.6 135.5 136.3 148.7 146.0 0.01 0.02 41.2 40.6 37.8 38.7 74.6 74.4 G. angustifolium 20.4 20.2 1.5 1.5 -1.7 -7.3 60.5 63.6 58.8 56.3 0.10 0.11 38.7 36.4 43.5 43.2 74.1 77.9 P. aculeata 31.4 29.9 1.9 1.8 1.3 2.0 88.4 82.6 89.7 84.6 0.10 0.10 59.5 57.4 55.1 50.6 77.3 69.2 Season mean 27.4 1.7 -4.4 105.9 101.5 0.07 48.5 49.3 76.7 s.e.m. 0.6 0.0 0.5 1.4 1.3 0.00 1.0 0.6 0.4 Factors P-value Season <.001 <.001 <.001 <.001 <.001 <.001 <.001 <.001 0.070 Site 0.233 0.241 0.215 0.073 0.241 0.321 0.666 0.321 0.076 Species <.001 <.001 <.001 <.001 <.001 <.001 <.001 <.001 <.001 Season × Site 0.009 0.081 0.003 0.032 0.081 0.009 0.013 0.111 0.903 Season × Species <.001 <.001 <.001 <.001 <.001 <.001 <.001 <.001 <.001 Site × Species 0.059 0.079 0.001 0.603 0.079 0.470 0.340 0.093 <.001 Season × Site × 0.848 0.882 0.373 0.029 0.882 0.776 0.727 0.904 0.049 Species 1GP24 = In vitro gas production at 24 h; ME = metabolizable energy; (a) = immediately fermented fraction; (b) = slowly fermentable fraction; (c) = constant rate of gas production fraction; (a+b) = potential gas production fraction; EDDM = effective degradability of DM; IVOMD = in vitro organic matter digestibility; IVTOMD = in vitro true organic matter digestibility. 2S1 = Site 1 (Linares); S2 = Site 2 (Los Ramones). Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) 190 M. Chávez-Espinoza et al. 120 Summer 100 Autumn 100 80 80 60 60 40 40 20 20 0 0 0 3 6 9 12 24 48 72 96 0 3 6 9 12 24 48 72 96 Incubation time (h) Incubation time (h) Winter Spring 120 120 100 100 80 80 60 60 40 40 20 20 0 0 0 3 6 9 12 24 48 72 96 0 3 6 9 12 24 48 72 96 Incubation time (h) Incubation time (h) Linares C. pallida C. suaveolens F. angustifolia G. angustifolium P. aculeata Los Ramones C. pallida C. suaveolens F. angustifolia G. angustifolium P. aculeata Figure 1. In vitro gas production at 96 h (n = 9) of leaves of 5 native shrub species collected at 2 sites for each season of the year during 2018‒2019. Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) Gas production (mL gas / 500 mg DM) Gas production (mL gas / 500 mg DM) Gas production (mL gas / 500 mg DM) Gas production (mL gas / 500 mg DM) Nutritive value and leaf digestibility of browse species 191 Discussion should examine biomass production of available forage to assess the potential of these feed sources for maintenance This study has provided valuable information on nutritive of small ruminants in rangelands of northeastern Mexico. value of some common browse species during different In vitro gas production following incubation of leaves seasons in semiarid northeastern Mexico. Results suggest in rumen fluid showed C. pallida produced highest that these species could be a useful source of forage for amounts of gas, while G. angustifolium showed lowest browsing small ruminants in all seasons of the year. gas production, in agreement with reports (Salem et al. Although differences were seen in nutrient 2014) of low total phenol and saponin concentrations concentrations in leaves of C. pallida, C. suaveolens, in C. pallida. Higher levels of tannins and polyphenols F. angustifolia, G. angustifolium and P. aculeata, seasonal may be responsible for lower gas production and in vitro mean CP concentrations in leaf samples ranged from 18.8 digestibility in G. angustifolium. to 26.0 %, which far exceeds the requirements of 7‒10 % In vitro gas production of shrub species at 24 h (GP24) (NRC 2007) for maintenance and growth of white- was positively influenced (r = 0.97; P<0.01) by ME tailed deer, adult range goats and sheep, with C. pallida concentration. GP24 from leaf was inversely proportional providing the highest CP concentration. Ramı́rez et al. to lignin and EE concentrations. Interestingly GP24 values (2000) and Domínguez-Gómez et al. (2011) reported that in the present study (12.0‒36.8 mL/200 mg DM) were this species contains relatively low levels (<7.5 % DM) of lower than those reported by Kara (2019) for fibrous condensed tannins (CT), indicating that CT concentrations feedstuffs (36.7‒75.8 mL/200 mg DM) determined using in these plants should not affect CP digestibility by these rumen fluid of Damascus goats. Negative values for gas animal species. Observed metabolizable energy (ME) production from the immediately fermented fraction (a) concentrations (1.5‒2.0 Mcal/kg DM) in C. pallida were (as low as -21.5 mL/200 mg DM), recorded in this study, somewhat inferior to ME in alfalfa hays (2.1 Mcal/kg DM) are in accordance with those reported (from -3.0 to -1.7 and maize silage (2.2 Mcal/kg DM) as an energy source mL/200 mg DM) in shrubs by Selmi et al. (2010) in north according to Kara (2019). However, it would provide Tunisia. In contrast, F. angustifolia produced positive adequate energy to support small ruminants. values (1.5‒13.2 mL/200 mg DM). Gas production from Mean NDF concentrations in the various seasons the slowly fermentable fraction (b) was similar to or lower ranging from 40.9 to 43.2 % DM and ADF concentrations than values reported by Garcia-Montes de Oca et al. of 20.3‒21.3 % are well below levels at which fiber becomes (2011) (89.5–187.3 mL/500 mg DM) from legume browse a limiting factor for intake. Similar findings have been species in subtropical areas of Mexico, and higher than documented by Hassen et al. (2007) in Indigofera shrub those (31‒179 mL/500 mg) found by Larbi et al. (1998) species and by Domínguez-Gómez et al. (2013) in native in a study of fodder trees and shrubs from West Africa. shrubs of northeastern Mexico. All evaluated species in Fraction (b) was negatively affected by concentrations of this study showed NDF concentration lower than 60 %, ADF (r = -0.62; P<0.01), lignin (r = -0.53; P<0.01) and EE levels which can be adequately managed by ruminants (r = -0.49; P<0.01) and positively by ME concentration without affecting DM intake. (r = 0.23; P<0.01) as reported by Domínguez-Gómez et Low lignin concentrations in C. pallida (range al. (2011). Results for fraction (c) (constant rate of gas 1.6‒2.7 %) support its value as a useful forage source, production) in the present study varied from 0.01 to while high lignin concentrations (12.5‒15.2 %) in 0.11 %/h, and are similar to the averages (0.03‒0.11 %) G. angustifolium suggest that it would be of lower feed reported by El Hassan et al. (2000) for foliage of some value. Bouazza et al. (2012) reported that high lignin African multipurpose trees, but lower than those reported concentrations (>12 % DM basis) were associated with by Cerrillo et al. (2006) (6.1‒6.4 %) for diets consumed by low digestibility in foliage of fodder trees and shrubs from grazing goats in semi-arid regions of north Mexico, where Algerian arid and semi-arid rangelands. While lignin plant species with high CP% and high DM digestibility concentration tended to vary with seasons with lower were present. Potential gas production, fraction (a+b), values in spring, differences between seasons were not varied considerably (48.0‒148.7 mL/500 mg DM) and great and were unlikely to affect intake and nutritive value was higher than reported by Selmi et al. (2010) (26.1‒66.6 of most species in all seasons. Lignin concentrations in mL/500 mg DM) for shrubs in north Tunisia, but similar the current study (range 2.2 % for C. pallida to 13.9 % to values reported by Domínguez-Gómez et al. (2011) for G. angustifolium) range from low to high according (51‒126 mL/500 mg DM), even though polyethylene to Anele et al. (2009) and Kara (2019). Further studies glycol (PEG) was fed with forage in both of those studies. Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) 192 M. Chávez-Espinoza et al. Jančík et al. (2011) suggest that in vitro gas production all seasons in Tamaulipan thornscrub in northeastern can be useful in ranking forages on the basis of nutritional Mexico. Leaf material appears to have efficient rumen value. fermentation thereby providing adequate metabolizable IVOMD in this study was positively associated with energy and high CP for grazing stock. While C. pallida ME concentration (r = 0.98; P<0.01) but negatively affected appears to provide the best quality browse in terms of by lignin concentration (r = -0.50; P<0.01). Results CP, all species evaluated provided good quality forage. reported herein were higher than those observed for leaf Further studies could be devoted to determining biomass of Detarium microcarpum (32.1 %) and Afzelia africana production of the browse species under study and their (49.1 %), browse plants from Nigeria (Okunade et al. sustainability under regular defoliation throughout the 2014), as well as those reported for Quercus rugosa leaves year. Feeding studies would indicate the acceptability to in western Mexico (Carrillo-Muro et al. 2018) and non- small ruminants of the various species. leguminous forage trees in tropical regions of southern Mexico (Rojas-Hernández et al. 2015). High values Acknowledgments recorded for C. pallida, C. suaveolens and P. aculeata may be associated with their NDF concentrations. Future The authors gratefully acknowledge assistance provided studies should consider analyzing NDF disappearance by the late Dr Roque Gonzalo Ramírez Lozano, in rate, since Kara (2019) showed positive correlation planning the research in the initial stages; Facultad between NDF disappearance rate and DM digestibility, de Medicina Veterinaria de la Universidad Juárez del when it is determined with this method. IVTOMD values Estado de Durango for technical assistance; and Consejo were negatively affected by ADF and lignin concentrations Nacional de Ciencia y Tecnología (CONACYT) for (r = -0.82; P<0.01; r = -0.50; P<0.01, respectively), in financial support by Project Grants (A1-S-44878 and agreement with previous studies on litter fall reported 250732) and for providing a doctoral scholarship to the by Rodríguez-Santillán et al. (2015). Trujillo et al. (2010) first author. demonstrated that the DaisyII incubator underestimated NDF disappearance values, compared with in situ References methods. The absence of seasonal variations or spatial differences in IVTOMD indicates that differences among (Note of the editors: All hyperlinks were verified 6 July 2022). species are associated with their individual chemical Anele UY; Arigbede OM; Südekum KH; Oni AO; Jolaosho compositions. Differences between the estimated IVOMD AO; Olanite JA; Adeosun AI; Dele PA; Ike KA; Akinola and IVTOMD values for individual assessments (range OB. 2009. Seasonal chemical composition, in vitro from 14.8 to 44.5 %) could be attributed to the longer fermentation and in sacco dry matter degradation of four incubation period of 48 h with the DaisyII incubator, as indigenous multipurpose tree species in Nigeria. Animal IVOMD is calculated from GP24 data, with only 24 h of Feed Science and Technology 154(1–2):47–57. doi: 10.1016/j. incubation time. According to Norman et al. (2010) leaves anifeedsci.2009.07.007 of some shrubs need to be incubated for more than 72 AOAC International. 2011. Official methods of analysis. 18th h in order to achieve an accurate estimation of in sacco Edn. AOAC International, Gaithersburg, MD, USA. digestibility. According to chemical composition and Bouazza L; Bodas R; Boufennara S; Bousseboua H; López S. 2012. Nutritive evaluation of foliage from fodder trees and nutrient concentration data, IVOMD values appeared to shrubs characteristic of Algerian arid and semi-arid areas. be more realistic, as data for IVTOMD appear to possibly Journal of Animal and Feed Sciences 21(3):521–536. doi: overestimate the feed potential of leaf of browse species 10.22358/jafs/66126/2012 under study. The moderate to high values in both OMD Buthelezi LS; Mupangwa JF; Muchenje V; Nherera-Chokuda and IVTOMD demonstrate the high nutritive potential FV. 2019. Influence of drying technique on chemical of leaf of browse species as a source of forage for small composition and ruminal degradability of subtropical ruminants on smallholder farms in the semi-arid region of Cajanus cajan L. Animal Nutrition 5(1):95–100. doi: northeastern Mexico. 10.1016/j.aninu.2018.03.001 Camacho LM; Rojo R; Salem AZM; Mendoza GD; López D; Tinoco JL; Albarrán B; Montañez-Valdez OD. 2010. In vitro Conclusions ruminal fermentation kinetics and energy utilization of three Mexican tree fodder species during the rainy and dry Browse plants studied appear to have potential for period. Animal Feed Science and Technology 160(3–4):110– contributing significantly to diets of small ruminants in 120. doi: 10.1016/j.anifeedsci.2010.07.008 Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) Nutritive value and leaf digestibility of browse species 193 Carrillo-Muro O; Ramírez-Lozano RG; Castro-Rosales thornscrub shrub species, north-eastern Mexico. Journal AS; Hernández-Briano P; López-Carlos MA; González- of Arid Environments 58(4):483–503. doi: 10.1016/j. Ronquillo M; Rivera-Villegas A; Méndez-Llorente F; jaridenv.2003.12.001 Aguilera-Soto JI. 2018. Influence of landscape and collection Guerrero-Cervantes M; Cerrillo-Soto MA; Ramírez RG; Salem period on yield, nutritive value and in vitro digestibility of AZM; González H; Juárez-Reyes AS. 2012a. Influence of Quercus rugosa leaf litter and its potential use as sheep polyethylene glycol on in vitro gas production profiles and feed. Small Ruminant Research 166:73–77. doi: 10.1016/j. microbial protein synthesis of some shrub species. Animal smallrumres.2018.07.012 Feed Science and Technology 176 (1–4):32–39. doi: 10.1016/j. Cerrillo MA; López OO; Nevárez CG; Ramírez RG; Juárez RAS. anifeedsci.2012.07.005 2006. Nutrient content, intake and in vitro gas production of Guerrero-Cervantes M; Ramírez RG; González-Rodríguez H; diets by Spanish goats browsing a thorn shrubland in North Cerrillo-Soto A; Juárez-Réyes AS. 2012b. Mineral content in Mexico. Small Ruminant Research 66(1–3):76–84. doi: range forages from north Mexico. Journal of Applied Animal 10.1016/j.smallrumres.2005.07.025 Research 40(2):102–107. doi: 10.1080/09712119.2011.607907 Chávez-Espinoza M; González-Rodríguez H; Cantú-Silva I; Habib G; Khan NA; Sultan A; Ali M. 2016. Nutritive value of Cotera-Correa M; Estrada-Castillón AE; Bernal-Barragán common tree leaves for livestock in the semi-arid and arid H; Gómez-Meza MV. 2020. Foliar mineral content of five rangelands of Northern Pakistan. Livestock Science 184:64– shrub species with nutritional potential for small ruminants 70. doi: 10.1016/j.livsci.2015.12.009 in semiarid regions in northeastern Mexico. Ciência Rural Hassen A; Rethman NFG; Van Niekerk WA; Tjelele TJ. 50(10):1–11. doi: 10.1590/0103-8478cr20200202 2007. Influence of season/year and species on chemical Domínguez-Gómez TG; González-Rodríguez H; Guerrero- composition and in vitro digestibility of five Indigofera Cervantes M; Cerrillo-Soto MA; Juárez-Reyes AS; accessions. Animal Feed Science and Technology 136(3– Alvarado MDS; Ramírez-Lozano RG. 2011. Polyethylene 4):312–322. doi: 10.1016/j.anifeedsci.2006.09.010 glycol influence on in vitro gas production parameters in Holden LA. 1999. Comparison of methods of in vitro dry four native forages consumed by white-tailed deer. Revista matter digestibility for ten feeds. Journal of Dairy Science Chapingo Serie Ciencias Forestales y del Ambiente 17:21– 82(8):1791–1794. doi: 10.3168/jds.S0022-0302(99)75409-3 32. doi: 10.5154/r.rchscfa.2010.09.073 Jančík F; Rinne M; Homolka P; Čermák B; Huhtanen P. Domínguez-Gómez TG; González-Rodríguez H; Ramírez- 2011. Comparison of methods for forage digestibility Lozano R.G; Cantu-Silva I; Gómez-Meza, MV; Cantu-Ayala determination. Animal Feed Science and Technology 169(1– CM; Alvarado MDS. 2013. Nutritional profile of four shrub 2):11–23. doi: 10.1016/j.anifeedsci.2011.05.003 species, Northeastern Mexico. International Journal of Bio- Kara K. 2019. The in vitro digestion of neutral detergent fibre resource and Stress Management 4(1):1–8. bit.ly/3JGetN5 and other ruminal fermentation parameters of some fibrous El Hassan SM; Lahlou Kassi A; Newbold CJ; Wallace RJ. 2000. feedstuffs in Damascus goat (Capra aegagrus hircus). Chemical composition and degradation characteristics of Journal of Animal and Feed Sciences 28(2):159–168. doi: foliage of some African multipurpose trees. Animal Feed 10.22358/jafs/108990/2019 Science and Technology 86(1–2):27–37. doi: 10.1016/S0377- Larbi A; Smith JW; Kurdi IO; Adekunle IO; Raji AM; Ladipo 8401(00)00158-9 DO. 1998. Chemical composition, rumen degradation, and Foroughbakhch R; Hernández-Piñero JL; Carrillo-Parra A; gas production characteristics of some multipurpose fodder Rocha-Estrada A. 2013. Composition and animal preference trees and shrubs during wet and dry seasons in the humid for plants used for goat feeding in semiarid Northeastern tropics. Animal Feed Science and Technology 72(1–2):81– Mexico. The Journal of Animal & Plant Sciences 23(4): 96. doi: 10.1016/S0377-8401(97)00170-3 1034–1040) bit.ly/3Q8i8Ws McDonald I. 1981. A revised model for the estimation of protein Garcia-Montes de Oca CA; González-Ronquillo M; Salem degradability in the rumen. The Journal of Agricultural AZM; Romero-Bernal J; Pedraza JF; Estrada JG. 2011. Science 96(1):251-252. doi: 10.1017/S0021859600032081 Chemical composition and in vitro gas production of some Menke KH; Raab L; Salewski A; Steingass H; Fritz D; legume browse species in subtropical areas of Mexico. Schneider W. 1979. The estimation of the digestibility and Tropical and Subtropical Agroecosystems 14(2):589–595. metabolizable energy content of ruminant feedingstuffs bit.ly/3P3LBjT from the gas production when they are incubated with Getachew G; DePeters EJ; Robinson PH; Fadel JG. 2005. rumen liquor in vitro. The Journal of Agricultural Science Use of an in vitro rumen gas production technique to 93(1):217–222. doi: 10.1017/S0021859600086305 evaluate microbial fermentation of ruminant feeds and its Menke KH; Steingass H. 1988. Estimation of the energetic impact on fermentation products. Animal Feed Science feed value obtained from chemical analysis and in vitro and Technology 123–124(Part 1):547–559. doi: 10.1016/j. gas production using rumen fluid. Animal Research and anifeedsci.2005.04.034 Development 28:7–55. González-Rodríguez H; Cantú-Silva I; Gómez-Meza MV; Montgomery DC. 2004. Diseño y Análisis de Experimentos. Ramírez-Lozano RG. 2004. Plant water relations of 2nd Edn. Limusa-Wiley, México DF. p. 79-81. Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) 194 M. Chávez-Espinoza et al. Norman HC; Revell DK; Mayberry DE; Rintoul AJ; Wilmot litterfall as feed resource for grazing animals in semiarid MG; Masters DG. 2010. Comparison of in vivo organic regions. Life Science Journal 12(10):54–61. doi: 10.7537/ matter digestion of native Australian shrubs by sheep to in marslsj121015.06 vitro and in sacco predictions. Small Ruminant Research Rojas-Hernández S; Olivares-Pérez J; Elghandour MMY; 91(1):69–80. doi: 10.1016/j.smallrumres.2009.11.019 Cipriano-Salazar M; Ávila-Morales B; Camacho-Díaz NRC (National Research Council). 2007. Nutrient requirements LM; Salem AZM; Cerrillo-Soto MA. 2015. Effect of of small ruminants: sheep, goats, cervids, and new world polyethylene glycol on in vitro gas production of some non- camelids. National Research Council of the National leguminous forage trees in tropical region of the south of Academies, Washington, DC. Mexico. Agroforestry Systems 89:735–742. doi: 10.1007/ Okunade SA; Isah OA; Aderinboye RY; Olafadehan OA. s10457-015-9796-8 2014. Assessment of chemical composition and in vitro Salem AZM; Kholif AE; Elghandour MMY; Hernandez degradation profile of some Guinea Savannah browse SR; Domínguez-Vara IA; Mellado M. 2014. Effect of plants of Nigeria. Tropical and Subtropical Agroecosystems increasing levels of seven tree species extracts added to a 17(3):529–538. bit.ly/3PaskwL high concentrate diet on in vitro rumen gas output. Animal Oliveira OF de; Lima AF de; Santos MVF dos; Guim A; Cunha Science Journal 85(9):853–860. doi: 10.1111/asj.12218 MV da; Lira MDA. 2018. Chemical composition of hays of Sebata A; Ndlovu LR; Dube JS. 2011. Chemical composition, the Caatinga shrub legumes mororó and sabiá from different in vitro dry matter digestibility and in vitro gas production parts of the plant. Tropical Grasslands-Forrajes Tropicales of five woody species browsed by Matebele goats (Capra 6(2):111–116. doi: 10.17138/TGFT(6)111-116 hircus L.) in a semi-arid savanna, Zimbabwe. Animal Feed Orskov ER; McDonald I. 1979. The estimation of protein Science and Technology 170(1–2):122–125. doi: 10.1016/j. degradability in the rumen from incubation measurements anifeedsci.2011.07.013 weighted according to rate of passage. The Journal of Selmi H; Gasmi-Boubaker A; Mosquera-Losada R; Rekik B; Agricultural Science 92(2):499–503. doi: 10.1017/S002185 Ben-Gara A; Ben-Mahmoud A; Rigueiro-Rodríguez A; 9600063048 Rouissi H. 2010. In vitro gas production by fodder shrubs Pal K; Patra AK; Sahoo A; Kumawat PK. 2015. Evaluation of from the north of Tunisia. Livestock Research for Rural several tropical tree leaves for methane production potential, Development 22(3):1–5. (In French) lrrd.org/lrrd22/3/ degradability and rumen fermentation in vitro. Livestock selm22062.htm Science 180:98–105. doi: 10.1016/j.livsci.2015.07.011 Soliva CR; Zeleke AB; Clément C; Hess HD; Fievez V; Kreuzer Piñeiro-Vázquez AT; Jiménez-Ferrer GO; Chay-Canul AJ; M. 2008. In vitro screening of various tropical foliages, Casanova-Lugo F; Díaz-Echeverría VF; Ayala-Burgos AJ; seeds, fruits and medicinal plants for low methane and Solorio-Sánchez FJ; Aguilar-Pérez CF; Ku-Vera JC. 2017. high ammonia generating potentials in the rumen. Animal Intake, digestibility, nitrogen balance and energy utilization Feed Science and Technology 147(1–3):53–71. doi: 10.1016/j. in heifers fed low-quality forage and Leucaena leucocephala. anifeedsci.2007.09.009 Animal Feed Science and Technology 228:194–201. doi: Trujillo AI; Marichal MDJ; Carriquiry M. 2010. Comparison of 10.1016/j.anifeedsci.2017.04.009 dry matter and neutral detergent fibre degradation of fibrous Ramı́rez RG; Neira-Morales RR; Ledezma-Torres RA; feedstuffs as determined with in situ and in vitro gravimetric Garibaldi-González CA. 2000. Ruminal digestion procedures. Animal Feed Science and Technology 161(1– characteristics and effective degradability of cell wall of 2):49–57. doi: 10.1016/j.anifeedsci.2010.08.001 browse species from northeastern Mexico. Small Ruminant Van Soest PJ; Robertson JB; Lewis BA. 1991. Methods Research 36(1):49–55. doi: 10.1016/S0921-4488(99)00113-3 for dietary fiber, neutral detergent fiber, and nonstarch Rodríguez-Santillán P; Guerrero-Cervantes M; Ramírez- polysaccharides in relation to animal nutrition. Journal of Lozano RG; Bernal-Barragán H; González-Rodríguez Dairy Science 74(10):3583–3597. doi: 10.3168/jds.S0022- H; Juárez-Reyes AS. 2015. Nutritional profile of leaf 0302(91)78551-2 (Received for publication 09 November 2020; accepted 10 June 2022; published 30 September 2022) © 2022 Tropical Grasslands-Forrajes Tropicales is an open-access journal published by International Center for Tropical Agriculture (CIAT), in association with The Tropical Crops Genetic Resources Institute of The Chinese Academy of Tropical Agricultural Sciences (TCGRI-CATAS). This work is licensed under the Creative Commons Attribution 4.0 International (CC BY 4.0) license. Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) Tropical Grasslands-Forrajes Tropicales (2022) Vol. 10(3):195–203 195 doi: 10.17138/TGFT(10)195-203 Research Paper Benefit of feeding Urochloa hybrid cultivar ‘Cobra’ on milk production in Tanzania Beneficio de alimentar con el cultivar híbrido de Urochloa 'Cobra' en la producción de leche en Tanzania SOLOMON MWENDIA1, AN NOTENBAERT1, BEATUS NZOGELA2 AND ANGELLO MWILAWA3 1Tropical Forages Program, Alliance of Bioversity and CIAT, Nairobi, Kenya. alliancebioversityciat.org 2Tropical Forages Program, Alliance of Bioversity and CIAT, Arusha, Tanzania. alliancebioversityciat.org 3Ministry of Livestock and Fisheries, Dodoma, Tanzania. mifugouvuvi.go.tz Abstract Animal genetics, management, diseases, feeds and environment affect milk production in cattle. Feed is the most important and when addressed, cattle show immediate responses. In sub-Saharan Africa, livestock productivity is low largely due to use of low-quality crop residues and natural pastures, often poor in key nutrients for animal performance. In an 8-week on-farm feeding trial with farmers’ participation, milk production under farmers’ practice (FP) was compared with the use of improved Urochloa hybrid cultivar ‘Cobra’ hay (Cobra hay) as an intervention (IN). A crossover design with each cow undergoing FP and IN phases was used. For the initial 2 weeks, the experiment followed FP before shifting to 50-50 FP/IN in week 3 and 100 % IN in week 4 and 5, followed by 50-50 FP/IN in week six and 100 % FP in week 7 and 8. Milk production increased by 15 % under IN and was associated with better feed utilization efficiency of 2 kg DM Cobra hay/L of milk. The use of Cobra hay has potential to increase dairy productivity in Tanzania and other similar tropical ecologies and contexts in sub-Saharan Africa. Keywords: Dairy, feeding trial, feed utilization efficiency. Resumen La genética animal, el manejo, las enfermedades, la alimentación y el medio ambiente afectan la producción de leche en el ganado. La alimentación es lo más importante y cuando se aborda, el ganado muestra respuestas inmediatas. En el África subsahariana, la productividad ganadera es baja en gran parte debido al uso de residuos de cultivos de baja calidad y pasturas naturalizadas, a menudo pobres en nutrientes clave para el rendimiento animal. En una prueba de alimentación de 8 semanas en finca con la participación de los productores, la producción de leche bajo la práctica tradicional (FP) se comparó con el uso de heno del cultivar mejorado Urochloa híbrido ‘Cobra’ como una intervención (IN). Se utilizó un diseño cruzado (crossover) con cada vaca pasando por las fases FP e IN. Durante las 2 semanas iniciales, el experimento utilizó FP antes de cambiar a 50-50 FP/IN en la semana 3 y 100 % IN en las semanas 4 y 5, seguido de 50-50 FP/IN en la semana seis y 100 % FP en la semana 7 y 8. La producción de leche aumentó un 15 % bajo IN y se asoció con una mejor eficiencia de utilización de 2 kg de MS de heno de pasto Cobra/L de leche. El uso de heno de pasto ‘Cobra’ tiene potencial para aumentar la productividad lechera en Tanzania y en otras ecologías y contextos tropicales similares en el África subsahariana. Palabras clave: Eficiencia de utilización de forrajes, lechería, prueba de alimentación. Correspondence: Solomon Mwendia, Alliance of Bioversity International and CIAT, Africa Hub, Off Kasarani Road, P.O. Box 823-00621, Nairobi, Kenya. Email: s.mwendia@cgiar.org Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) 196 S. Mwendia, A. Notenbaert, B. Nzogela and A. Mwilawa Introduction recently received attention for its potential to increase livestock productivity (Mutimura et al. 2016; Cheruiyot Sub-Saharan African livestock productivity is low (Nin et et al. 2018; Mwendia et al. 2021b). Urochloa breeding and al. 2007). The increasing demand for animal source foods cultivar selection has produced hybrids with increased driven by human population growth and dietary change dry matter production and quality, e.g. Urochloa hybrid exerts pressure on livestock feed supply. For example, per cultivar ‘Cayman’ and Urochloa cultivar ‘Basilisk’ capita milk consumption in Tanzania is increasing and (Mwendia et al. 2021a). In Latin America, Urochloa projected to reach 55–100 L/person/yr by 2022 (IFAD contributes significantly to beef production under 2016). Given that arable land is scarce (Jayne et al. 2014), extensive systems (Jank et al. 2014). The Consultative forage production competes with food crops and there Group for International Agricultural Research (CGIAR) is limited or no land for agricultural expansion, more estimates that grasses cover 12 million ha in Latin efficient use of available land becomes key in responding America (Fuglie et. 2021). In eastern Africa, Urochloa to the increasing demand for livestock products. is also gaining attention by livestock producers and Low availability of quality livestock feeds has livestock scientists due to its potential in smallholder resulted in use of crop residues and limited locally agricultural systems (Schiek et al. 2018). produced and poorly formulated concentrate feeds for Urochloa hybrids have been estimated to have the livestock production. Crop residues provide a limited potential to increase milk production by 15 ̶ 40 % in supply of required nutrients for animal maintenance, eastern Africa (Schiek et al. 2018). CIAT (2003) reported growth, reproduction and production (FAO 2018). This that cows fed on Urochloa hybrid ‘Mulato II’ produced leads to low milk and meat production, contributing 11 % more milk during the dry season and 23 % more to high emissions of methane gas per unit of product, during the rainy season compared with those fed on associated with global warming (Makkar 2016). Long Urochloa cultivars ‘Basilisk’ and ‘Xaraes’. Muinga calving intervals and low lactation yields characterize et al. (2016) found that cows fed on Urochloa cultivars livestock on most smallholder mixed farms in Tanzania produced 15 ̶ 40 % more milk compared to those fed and East African countries (Kanuya et al. 2000). with normal farmers’ practice in Kenya. These findings Although animal productivity can be limited by genetics support the need to improve use of Urochloa species and management, feeding is the single most important for improving livestock productivity. This on-farm component that accounts for 70 % of the costs in dairy participatory trial was designed to investigate the effects production (Makkar 2016; Odero-Waitituh 2017) and, if of Urochloa hybrid cultivar ‘Cobra’ hay (Cobra hay) on well addressed, results in positive impact more quickly. milk production in Njombe district, located in Njombe In a recent assessment of livestock feeds in Tanzania, region in the Southern highlands of Tanzania. We chose the estimated requirement for livestock feed is >172 million to demonstrate improved feeding and its impact on t/yr, of which 70 % are roughage based (FAO and IGAD productivity on-farm with farmers’ participation with the 2019). Roughages include Rhodes grass (Chloris gayana), aim to encourage use of this improved cultivated forage. buffel grass (Cenchrus ciliaris), Napier grass (Cenchrus We postulated that the use of improved Cobra hay would purpureus), Guatemala grass (Tripsacum andersonii) increase milk production and forage use efficiency. and natural pastures. There is also localized evidence of livestock roughage shortages in Tanzania, including in high Materials and Methods potential areas. Mwendia et al. (2019) reported farmers’ experience in the southern highlands, where forage Selection of farmers, trial cows and feed availability drops by about 50 % for more than 6 months in a year, with implications of underfeeding livestock. Low Six farmers each with 1 crossbred cow in early lactation forage cultivation in Tanzania (Maleko et al. 2018a; 2018b) (2–3 months after parturition) and in between 2nd to 4th further constrains roughage availability. Use of appropriate calving were selected in Njombe district in February 2019. forages is paramount for improved livestock productivity. Njombe receives bimodal rainfall, mid-February to end Use of forage-based roughages is preferred to grain-based of May in the long rainfall season and mid- November to diets (Scaglia et al. 2014), which not only compete with mid-January in the short rainfall season (NASA, 2021). human food but are more expensive (Makkar 2016). The crossbred cows were from either Friesian, Ayrshire While there are many forages that could fit in dairy or Jersey and were fully stall-fed to allow measurement systems in Tanzania (Cook et al. 2020), Urochloa has of dry matter intake. As a preliminary requirement, the Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) Improving milk production in cattle with Brachiaria 197 Internal Review Board (IRB) of Alliance of Bioversity The feeding trial was done as a crossover design and CIAT scrutinized the process and consent was given with a lactating cow as the experimental unit and each to do the work. cow acting as its control to mitigate against the genetic Cobra hay was selected as a representative of an variation in the selected cows. In practice each cow improved forage grass hybrid with better forage quality underwent feeding phases under FP as well as under traits when harvested at the right growth stage. It is leafy, IN. The selected cows were monitored for 2 weeks with high crude protein and good digestibility (Mwendia under FP before transitioning to a week of 50 % FP and et al. 2021b). Cobra is a hybrid of Urochloa ruziziensis x 50 % intervention feeding (IN), followed by 2 weeks U. decumbens x U. brizantha and is a released forage on IN at 100 % followed by a return to the 50 % each in Latin America and Kenya and is in the process of FP and IN feeding for 1 week, before finishing off with registration in other East African countries. another 2 weeks of feeding under FP. This resulted in Considering a daily dry matter (DM) intake of 15 each of the cows receiving a phase of IN preceded by kg/cow/day, we estimated that 420 kg DM would be FP and followed by FP. The crossover design made it enough to feed a mature cow of about 400 kg live weight possible to randomize for any of the 3 cows having for 4 weeks. Using Cobra mean dry matter yields [3 t FP→IN and the other 3 IN→FP arrangement and avoid DM/ha/cut (Cook et al. 2020)], we estimated that 0.14 bias (Mills et al 2009). ha land was required to produce 420 kg DM of Cobra The cows were offered the estimated feed weight hay. To have sufficient feed if the season has unfavorable (as-fed) to meet the dry matter requirements on a daily weather conditions, the land size was increased to 0.2 basis (Table 1) under 100 % IN feeding. Cobra hay was ha. Selected farmers were assisted to establish 0.2 ha fed to each cow 5-6 times per day until the cow was of Cobra ready for feeding by June 2019. The annual observed not to take any more and the total weight fed mean temperature is 18 °C while rainfall ranges from for the day was recorded. Throughout the 100 % IN, 14 2,700– 4,000 mm. The grass was established in March ̶ 25 kg DM/d was made available (Table 1). During the 2019 when the long rainfall season had set in to provide 50:50 % transition phases, the farmers gave half of what adequate soil moisture. Recommended agronomic they were giving under FP and following that, the other practices were observed in establishing the grass half of the daily feed was given as Cobra hay, about 7 including land preparation (fine soil tilth), seed-rate (6 kg. This phase was to allow the cows to acclimatize to kg/ha), soil fertilization (26 kg P/ ha) and maintaining Cobra hay before being fed 100 % Cobra hay in Week 4 the fields weed-free (Cook et al. 2020). and 5. Throughout the experiment, the cows had access to clean drinking water ad libitum. Farmers’ practice and intervention feeding in trial design Table 1. Experimental cows’ attributes used to estimate daily dry matter requirements. Farmers fed a mix of roughages that were available in their Farm CHGL1 ELW2 EDDMR3 IFFWA4 locality. The roughages included fresh feed harvested at 1 172 420 12.6 22.9 different growth stages, as well as crop residues obtained after crop harvest. All roughage offered was measured 2 153 285 8.6 14.2 with a spring balance (KERN CH 50K50 with 10 g 3 173 427 12.8 25.3 precision) and the weights of offered roughages and their 4 169 399 12.0 24.5 refusals recorded on a daily basis. The initial plan to weigh 5 168 392 11.8 25.1 the different proportions offered under farmers’ practice 6 176 451 13.5 22.0 (FP) was logistically too difficult due to farmer time 1CHGL= Cow heart girth length (cm) constraints. The difference between offered roughages and 2ELW=Estimated live weight (kg) refusals constituted the daily feed intake. Under 100 % 3EDDMR= Estimated daily dry matter requirement (kg) intervention feeding (IN) in weeks 4 and 5, lactating 4IFFWA= Intervention feeding forage weight (kg) as-fed cows were offered Cobra hay daily (Table 1). To gauge the selected cow daily dry matter requirements, heart girth Data collection was measured, converted to live weight, and subsequently the daily dry matter requirement was estimated at 3 % of To address the intensive data collection, farmers who the body live weight (Lukuyu et al. 2012). offered their cows for trials and frontline livestock Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) 198 S. Mwendia, A. Notenbaert, B. Nzogela and A. Mwilawa extension officers were first sensitized on the breadth residues from maize stovers (Zea mays), beans haulms and expectations of the study including the care required (Phaseololus vulgaris) and banana stems (Musa during data collection. Specifically, we developed and spp). Common forage crops included buffel grass printed weekly data sheets. Each sheet had cells to (Cenchrus ciliaris), Napier grass (Cenchrus purpureus), enter daily data on the type and weight of forage offered Rhodes grass (Chloris gayana), Guatemala grass (kg), the forage refusals (kg) and morning and evening (Tripsacum andersonii), star grass (Cynodon dactylon) milk yields (L). Milk was measured with a graduated and to a limited extent leucaena (Leucaena leucocephala) cylinder. Farmers and extension officers were trained and Desmodium (Desmodium intortum) forage legumes in use of the data sheet until they understood it. Each and sorghum (Sorghum bicolor). All the 6 farms fed farmer took records of feeds and milk production on the Rhodes grass and natural pastures, 4 fed star grass, data sheets and extension officers assisted during the Guatemala grass, maize thinnings/stovers and bean start of the experiment and moved around confirming haulms. Four farms offered Napier grass while 2 no farmer had difficulties. farms gave banana leaves. Buffel grass, pearl millet (Pennisetum glaucum), leucaena leaves and Setaria Forage nutritional analysis (Setaria sphacelata) were observed in one farm each. In addition to feed weights and milk yields, a sample Table 2. Forages offered to cows under FP during feeding of approximately 400g of each forage type was taken trial in Njombe District, Tanzania each week for laboratory quality analysis and processed Farm Forages offered under FP at Mpwapwa Tanzania Livestock Research Institute (TALIRI) in Dodoma. Samples of forages collected Rhodes grass, maize thinnings, Guatemala grass, 1 Desmodium, Napier grass, star grass, banana leaves were oven dried to constant weight at 65°C for 48 hand and mixed natural pastures. ground to pass 1 mm sieve and packed in ziplock bags Rhodes grass, Guatemala grass, Napier grass, star labeled accordingly by farm and forage species. The 2 grass, bean haulms, pearl millet leaves and mixed samples were then sent to the International Livestock natural pastures. Research Institute (ILRI) laboratory in Addis Ababa, 3 Rhodes grass, maize stover, Setaria, bean haulms Ethiopia, for analysis. Analysis was done using Near and mixed natural pastures. Infrared Spectroscopy (NIRS) calibrated for tropical Rhodes grass, maize stover, star grass, Desmodium forages. Nutritional parameters included dry matter, 4 intortum, Napier grass, bean haulms, leucaena, ash, crude protein, neutral detergent fiber, acid detergent Guatemala grass, sorghum, buffel grass and mixed fiber, acid detergent lignin, metabolizable energy and in natural pastures. vitro organic matter digestibility. 5 Rhodes grass, maize stover and mixed natural pastures. Data analyses Rhodes grass, bean haulms, star grass, banana 6 leaves, Napier grass, Guatemala grass and mixed Data analyses were performed using GenStat statistical natural pastures. software version 18 by pooling data for feed intake, feed refusal, morning, and evening milk production from the Following forage nutritional analysis, differences six cows. Analysis of Variance following an unbalanced in feed quality across farms were observed in ash, design was applied because the FP phase was 4 weeks crude protein (CP), neutral detergent fiber (NDF) acid while IN feeding and the 50:50 phase took 2 weeks each. detergent fiber (ADF), metabolizable energy (ME) and Least significant difference (lsd) separated the means in vitro organic matter digestibility (IVOMD) (Table 3). and were significantly different at P<0.05. Milk yields Results The six cows significantly increased their milk yields Feeds and feed quality under IN feeding compared to FP except in Farm 5 (Figure 1). The increase ranged from 1.2–21 %. The Under FP, an array of forages was offered to the milk yields among cows were in the order Farm 1>Farm lactating cows (Table 2). This included a mix of crop 3≈Farm 4>Farm 6>Farm 2>Farm 5 (Figure 1). Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) Improving milk production in cattle with Brachiaria 199 Table 3. Nutritional value of feeds offered under FP and IN during 8-week experimental period in Njombe District, Tanzania Farm Feed type DM (%) Ash (%) CP (%) NDF (%) ADF (%) ADL (%) ME (MJ/kg) IVOMD (%) 1 Cobra hay 93.9 15.6 9.8 64.9 36.5 8.0 7.0 50.7 Rhodes grass 93.9 10.8 8.2 73.7 46.7 8.6 6.7 46.9 Star grass 93.7 7.7 8.2 75.6 49.1 10.7 6.3 43.5 Desmodium 92.9 8.0 10.8 56.9 50.9 16.2 6.8 47.9 Banana Leaves 94.1 16.3 13.8 49.3 38.6 10.9 6.7 49.8 Mixed natural pastures 93.7 9.6 7.4 73.2 47.2 9.4 6.6 45.4 Buffel grass 93.9 12.3 8.6 71.6 44.4 8.1 6.6 47.4 Guatemala grass 94.0 11.7 11.9 67.2 43.6 8.5 7.0 50.1 Maize thinnings 93.6 13.6 12.2 64.7 37.8 8.9 7.0 50.0 2 Cobra hay 92.5 8.3 10.4 41.3 20.8 4.9 8.5 58.2 Rhodes grass 93.7 8.2 6.0 76.8 49.8 9.2 6.5 44.4 Star grass 93.4 8.0 8.1 67.6 41.5 9.0 6.9 47.5 Napier grass 93.9 11.0 7.8 72.1 46.3 8.6 6.6 46.4 Bean haulms 93.1 7.1 5.9 73.1 53.6 12 6.8 46.1 Pearl millet 93.6 9.6 8.0 69.8 44.5 8.0 7.0 48.8 Guatemala grass 94.1 12.1 12.2 67.4 43.1 7.8 7.2 51.2 Mixed natural pastures 93.6 9.2 6.3 74.2 44.9 7.8 6.8 47.0 3 Cobra hay 93.5 12.4 7.8 63.2 35.8 8.1 7.0 49.3 Setaria 93.8 13.5 11.7 63.8 38.6 6.1 7.4 53.4 Rhodes grass 94.0 9.9 6.6 75.7 47.6 8.4 6.5 45.1 Bean haulms 92.5 7.3 11.1 51.5 38.1 8.8 8.1 56.4 Mixed natural pastures 93.8 9.5 8.4 72.2 46.7 10.5 6.5 45.0 Maize stover 93.8 13.7 15.5 56.7 34.5 8.0 7.6 55.8 4 Cobra hay 93.8 13.9 7.3 68.7 40.3 8.8 6.6 47.1 Buffel grass 93.9 11.0 6.8 75.3 46.1 7.8 6.7 47.1 Rhodes grass 93.5 8.5 10.0 59.4 35.7 7.5 7.6 52.7 Star grass 93.2 8.0 9.8 62.6 41.9 9.9 7.1 49.1 Desmodium 93.1 11.2 13.0 51.5 51.1 16.6 7.1 52.8 Leucaena 92.4 13.2 19.1 29.9 41.7 20.2 8.1 59.7 Napier grass 93.5 9.3 7.5 65.3 40.6 7.8 7.1 49.0 Bean haulms 92.3 8.2 11.6 29.8 16.7 5.5 8.6 59.2 Sorghum 93.6 8.1 7.1 74.9 46.1 8.1 6.6 45.0 Guatemala grass 94.0 9.6 6.7 76.7 50.6 8.6 6.4 44.7 Mixed natural pastures 93.8 9.5 8.8 72.9 45.7 8.5 6.8 47.6 Maize thinnings 93.1 9.2 8.9 59 33.9 6.9 7.5 51.9 5 Cobra hay 94.1 12.4 10.5 70.2 41.5 7.1 6.9 48.9 Rhodes grass 93.5 9.7 11.6 58.6 35.5 7.0 7.3 51.2 Maize stover 93.2 6.2 3.5 81.0 43.1 6.6 6.9 45.8 Mixed natural pastures 94.0 10.8 9.0 75.4 46 8.8 6.5 45.5 6 Cobra hay 93.9 12.4 6.4 70.1 42.8 8.0 6.8 47.6 Rhodes grass 93.9 8.2 6.8 77.9 51.1 9.8 6.2 42.8 Star grass 93.8 8.3 6.6 76.5 50.3 10.5 6.3 43.4 Banana leaves 93.5 12.9 9.0 59.4 42.6 10.4 7.0 49.5 Napier grass 93.9 7.2 5.3 82.1 51.6 10.1 6.1 41.3 Bean haulms 92.7 6.3 5.2 68.6 54.3 10.2 7.8 52.5 Guatemala grass 93.9 9.3 7.3 74.4 49.4 8.6 6.5 45.5 Mixed natural pastures 93.6 8.8 8.2 71.9 47.8 9.6 6.7 46.4 Overall IN 93.6 12.5 8.7 63.0 36.2 7.5 7.1 50.3 Means Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) 200 S. Mwendia, A. Notenbaert, B. Nzogela and A. Mwilawa Pooling data across the farms under FP, daily feed 2 kg DM contributed to a L of milk unlike 4.5 kg DM/L intake per cow averaged 38.2 kg as fed (Table 4). When under FP (Table 3). The quality of FP feeds compared to IN the cows crossed to 50:50 (FP: IN) intake dropped by 26 % (Table 3) was lower in terms of ash, neutral detergent fiber, followed by a further drop of 24 % when the IN was fed, acid detergent fiber, acid detergent lignin, metabolizable representing a total drop of 50 % comparing FP and IN. energy and in vitro organic matter digestibility. The feed intake values returned significant differences (P< Milk yields over the 8-week experimental period 0.001) among them with FP registering the highest value. (Figure 2) started at a weekly average of 8.8 L milk/d Subsequently, morning milk production was of the order (morning + evening) during the initial 2 weeks under FP < FP: IN < IN with values of 5 < 5.4 < 5.7 L /cow/d FP and increased by 6.8 % in week 3 with 50:50 feeding respectively. The milk yield under IN was significantly before peaking at 10.1 L milk/d (15 % increase) when the higher (P = 0.009) than the yields under FP. Evening milk cows were fed IN. On reverting to 50:50 feeding at week yields followed the same order as that of the morning (Table 6 the production remained at 10.1 L milk/d but when 3) except that the values were lower. Feed use efficiency was completely reverted to FP in week 7 and 8, production greater when Cobra hay was used compared to FP where dropped by 2.6 % and 15.2 % respectively. Figure 1. Mean (± se) of milk production (liters) for six lactating cows under either farmers’ practice (FP), 50:50 of FP and intervention (FP:IN) and intervention alone (IN) in Njombe district, Tanzania. . Table 4. Effects of FP and IN on feed intake and milk yields from experimental cows in Njombe district, Tanzania. Parameter Farmer's practice (FP) Intervention alone (IN) FP: IN P value Lsd Feed intake (kg) 38.2 19.0 28.3 <0.001 4.099` Morning milk production (L) 5.0 5.7 5.4 0.009 0.525 Evening milk productio (L) 4.1 4.5 4.4 0.038 0.372 Feed utilization efficiency (kg DM/L) 4.5 2.0 3.2 <0.001 0.481 Figure 2. Average morning and evening milk production from six experimental cows over 8-week trial period in Njombe distrct, Tanzania. The performance was measured over 3 phases of farmers’ practice (FP), 50:50 of FP and intervention (FP:IN) and intervention alone (IN). Each data point on the lines is an average of 7-daily measurements per cow and for the six cows. Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) Improving milk production in cattle with Brachiaria 201 Discussion increase productivity even with the current crossbred cows, whose productivity potential is underutilized due to poor In livestock production, feed intake is key for feeding (Swai et al. 2014; Maleko et al. 2018a). Tanzania’s maintenance, growth and reproduction. FP comprised low national average lactation of 2,000 L/cow/yr (CSIRO a mix of crop residues, natural pastures and cultivated 2021) could be increased to 2,300 L/cow/yr with Cobra forages across the 6 farms. Crop residues are hay intervention. Assuming half of the estimated 239,237 characterized by low digestibility, energy and crude improved dairy animals in Tanzania are in lactation in a protein content (Methu et al. 2001) as observed by use year, this would translate to about 35.8 million L milk/y of maize stovers in Farm 5 (Table 3) thus limiting cow (CSIRO 2021) and contribute to the projected growth of performance. Although quality of grasses depends on per capita milk consumption (IFAD 2016). stage of growth, young grasses being more nutritious, Use of Rhodes grass and Napier grass, both of which natural pastures are of relatively lower quality than are forage grasses under FP in this study, indicate cultivated forages (Gietema 2005). In the 6 farms, farmers’ awareness about growing forages on farm. In natural grasses had lower digestibility, energy and crude addition to Napier grass’s relatively low crude protein protein compared to the grown fodder grasses (Table 3). content 5.3 ̶7.8 % (Table 3), its cultivation is currently However, there was variation of nutritional parameters negatively affected by smut and stunt diseases (Mwendia in feed samples of the same species from different farms, et al. 2007; Obura et al. 2009). Rhodes grass is also low most likely influenced by farm management e.g. soil in crude protein. Gietema (2005) reports figures as low fertility and possible harvest at different stages of growth as 2.3 %, although the figures of 6.0 ̶ 11.6% we obtained among the farms. Cows in the study ingested more feed in this study are much higher, implying farmers are using under FP than IN. The lower Cobra hay intake compared the grass at much younger growth stages. This implies to FP (Table 4) during IN phase did not depress milk farmers are knowledgeable about forage cultivation and production and meant no abrupt changes in feed type. could benefit from use of better forage crops. Livestock Often abrupt changes in feed types, like the use of extension assistance, access and affordability of seeds different roughage sources as observed under FP, lower of better forages are paramount for adoption to increase the microbial activity in the rumen and consistency in and realize the benefits of improved forages. feeding is of utmost importance in rumen adaptation (Humer et al. 2018). Conclusions The increase in milk yields from feeding Cobra hay demonstrates its potential in supporting milk The study provided emperical evidence on the potential production, despite the genetic differences that may exist of Cobra hay to increase milk production in Tanzania, in the cows involved (Figure 1). The benefit most likely owing to its desirable attributes, including nutrient can be attributed to better quality of Cobra hay than content, high digestibility and high feed utilization feeds offered under FP (Table 3). Milk yield increased efficiency. Cows required more than double the amount with feeding Cobra hay despite reduced feed-intake of feed under FP compared to use of Cobra hay. Involving compared to FP most likely benefiting from the greater farmers in forage feeding trials may contribute to nutritive profile compared to fodder types under FP. To changing their perception about the need for improved produce milk under FP required 4.5 kg feed/L and 2.0 forages for increased productivity. However, for farmers kg Cobra hay/L under IN (Table 4). Where lower yield to adopt improved forage technologies, access to of quality forage is realized per unit of land, compared affordable seeds/planting materials is key. to unimproved forages, it does not therefore necessarily translate into an increased land requirement for more Acknowledgments forage production. This is a key benefit of using improved forages for improving livestock productivity. Farmers The authors appreciate the farmers who willingly take a lot of time cutting and carrying natural pastures allowed their cows to be used in the study and assisted (Paul et al. 2017) and growing improved forages with in data collection and the Tanzania District Councils lower DM requirements would save on time for other authority for their support and interest. Special thanks important activities. for the financial support from the International Fund The increase in milk production (Figure 1, 2 and Table for Agricultural Development (IFAD), through the 4) by changing to Cobra hay confirms the potential to Climate Smart Dairy Project grant number D180 in Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) 202 S. Mwendia, A. Notenbaert, B. Nzogela and A. Mwilawa Tanzania. Equally, the Collaborative Reasearch Program evolution of farming systems, and development strategies (CRP)-Livestock and Environment Flagship under in Africa: A Synthesis. Food Policy 48:1–17. doi: 10.1016/j. the Consultative Group of International Agricultural foodpol.2014.05.014 Research (CGIAR) for driving livestock improvement Kanuya NL; Kessy BM; Bittegeko SBP; Mdoe NSY; Aboud and supporting the research. AAO. 2000. Suboptimal reproductive performance of dairy cattle kept in smallholder herds in rural highland area of References northern Tanzania. Preventive Veterinary Medicine 45(3– 4):183–192. 10.1016/S0167-5877(00)00134-3 Lukuyu B; Gachuiri CK; Lukuyu MN; Lusweti C; Mwendia (Note of the editors: All hyperlinks were verified 24 August 2022). S. 2012. Feeding dairy cattle in East Africa. East Africa Cheruiyot D; Midega CAO; Ueckermann EA; Van den Berg Dairy Development Project, Nairobi, Kenya. hdl.handle. J; Pickett JA; Khan ZR. 2018. Genotypic response of net/10568/16873 brachiaria (Urochloa spp.) to spider mite (Oligonychus Makkar HPS. 2016. Animal nutrition in a 360-degree view and trichardti) (Acari: Tetranychidae) and adaptability to a framework for future R&D work: Towards sustainable different environments. Field Crops Research 225:163–169. livestock production. Animal Production Science doi: 10.1016/j.fcr.2018.06.011 56(10):1561–1568. doi: 10.1071/AN15265 CIAT (Centro Internacional de Agricultura Tropical). 2004. Maleko D; Wai-Tim N; Msalya G; Mwilawa A; Pasape L; Annual Report 2003. Project IP-5. Tropical grasses and Mtei K. 2018a. Seasonal variations of fodder resources and Legumes: Optimizing genetic diversity of multipurpose utilization practices among smallholder dairy farmers in use. CIAT, Cali, Colombia. hdl.handle.net/10568/69061 Western Usambara Highlands, Tanzania. Tropical Animal Cook BG; Pengelly BC; Schultze-Kraft R; Taylor M; Burkart Health and Production 50(7):1653–1664. doi: 10.1007/ S; Cardoso Arango JA; González Guzmán JJ; Cox K; s11250-018-1609-4 Jones C; Peters M. 2020. Tropical forages: An interactive Maleko D; Msalya G; Mwilawa A; Pasape L; Mtei K. 2018b. selection tool; 2nd and Revised Edn. International Center Smallholder dairy cattle feeding technologies and for Tropical Agriculture (CIAT), Cali, Colombia and practices in Tanzania: Failures, successes, challenges International Livestock Research Institute (ILRI), Nairobi, and prospects for sustainability. International Journal Kenya. www.tropicalforages.info of Agricultural Sustainability 16(2):201–213. doi: CSIRO. 2021. Dairy Production in Tanzania. CSIRO, Canberra, 10.1080/14735903.2018.1440474 ACT, Australia. bit.ly/3QU896R Methu JN; Owen E; Abate AL; Tanner JC. 2001. Botanical and FAO. 2018. Ethiopia: Report on Feed Inventory and Feed nutritional composition of maize stover, intakes and feed Balance 2018. FAO, Rome, Italy. bit.ly/3pPgmhv selection by dairy cattle. Livestock Production Science FAO and IGAD. 2019. East Africa Animal Feed Action Plan: 71(2–3):87–96. doi: 10.1016/S0301-6226(01)00212-3 Sustainably developing livestock-dependent livelihoods in Mills EJ; Chan A-W; Wu P; Vail A; Guyatt GH; Altman DG. East Africa. FAO and IGAD, Rome, Italy. bit.ly/3AIAPuF 2009. Design, analysis, and presentation of crossover trials. Fuglie K; Peters M; Burkart S. 2021. The extent and economic Trials 10:27. doi: 10.1186/1745-6215-10-27 significance of cultivated forage crops in developing Muinga RW; Njunie MN; Gatheru M; Njarui DMG. 2016. countries. Frontiers in Sustainable Food Systems 5:712136. The effects of Brachiaria grass cultivars on lactation doi: 10.3389/fsufs.2021.712136 performance of dairy cattle in Kenya. In: Njarui DMG; Gietema B. 2005. Modern Dairy Farming in Warm Climate Gichangi EM; Ghimire SR; Muinga RW, eds. 2016. Zones vol 1. Agromisa Foundation, Wageningen Climate Smart Brachiaria grasses for improving livestock Netherlands. https://bit.ly/3AzWWTZ production in East Africa – Kenya experience. Proceedings Humer E; Petri RM; Aschenbach JR; Bradford BJ; Penner of the workshop held in Naivasha, Kenya, 14-15 September GB; Tafaj M; Südekum KH; Zebeli Q. 2018. Invited 2016. p. 229–237. bit.ly/3K6FHfT Review: Practical feeding management recommendations Mutimura M; Ebong C; Rao IM; Nsahlai IV. 2016. Change to mitigate the risk of subacute ruminal acidosis in dairy in growth performance of crossbred (Ankole × Jersey) cattle. Journal of Dairy Science 101(2):872–888. doi: dairy heifers fed on forage grass diets supplemented with 10.3168/jds.2017-13191 commercial concentrates. Tropical Animal Health and IFAD. 2016. United Republic of Tanzania. Southern Highlands Production 48(4):741–746. doi: 10.1007/s11250-016-1019-4 milkshed development project: Detailed design report. Mwendia SW; Wanyoike M; Wahome RG; Mwangi DM. Report No: 4240-TZ. IFAD, Rome, Italy. bit.ly/3AyJtM5 2007. Effect of Napier head smut disease on Napier yields Jank L; Barrios SC; Valle CB do; Simeão RM; Alves GF. and the disease coping strategies in farming systems in 2014. The value of improved pastures to Brazilian beef central Kenya. Livestock Research for Rural Development production. Crop and Pasture Science 65(11):1132–1137. 19(8):109. lrrd.org/lrrd19/8/mwen19109.htm doi: 10.1071/CP13319 Mwendia S; Mwilawa A; Nzogela B; Kizima J; Mangesho Jayne TS; Chamberlin J; Headey DD. 2014. Land pressures, the W; Loina R; Bwire J; Notenbaert A. 2019. Livestock Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) Improving milk production in cattle with Brachiaria 203 feeds assessment in Southern Highlands in Tanzania. Kenya; a review. Livestock Research for Rural Development Proceedings of TropenTag 2019, Kasel, Germany, 18–20 29(7):139. lrrd.org/lrrd29/7/atiw29139.html September 2019. hdl.handle.net/10568/107044 Paul BK; Birnholz C; Nzogela B; Notenbaert A; Herrero Mwendia SW; Odhiambo R; Juma A; Mwangi D; Notenbaert M; Bwire J; Groot JGJ; Tittonell PA. 2017. Livestock A. 2021a. Performance of Urochloa and Megathyrsus feeding systems and feed gaps across three agro-ecologies forage grasses in smallholder farms in Western Kenya. in Tanzania. Proceedings of the TropAg Conference, Frontiers in Sustainable Food Systems 5:719655. doi: Brisbane, Australia, 20-22 November 2017. hdl.handle. 10.3389/fsufs.2021.719655 net/10568/89476 Mwendia SW; Ohmstedt U; Nyakundi F; Notenbaert A; Peters Scaglia G; Rodriguez J; Gillespie J; Bhandari B; Wang JJ; M. 2021b. Does harvesting Urochloa and Megathyrsus McMillin KW. 2014. Performance and economic analyses forages at short intervals confer an advantage on cumulative of year-round forage systems for forage-fed beef production dry matter yield and quality? Journal of the Science of Food in the Gulf Coast. Journal of Animal Science 92(12):5704– and Agriculture 102(2):750–756. doi: 10.1002/jsfa.11407 5715. doi: 10.2527/jas.2014-7838 Nin A; Ehui S; Benin S. 2007. Livestock productivity in Schiek B; González C; Mwendia S; Prager SD. 2018. Got developing countries: An assessment. In: Evenson R; forages? Understanding potential returns on investment Pingali P, eds. Handbook of Agricultural Economics in Brachiaria spp. for dairy producers in Eastern Africa. 3:2461–2532. doi: 10.1016/S1574-0072(06)03047-7 Tropical Grasslands-Forrajes Tropicales 6(3):117–133. doi: Obura E; Midega CAO; Masiga D; Pickett JA; Hassan M; Koji 10.17138/tgft(6)117-133 S; Khan ZR. 2009. Recilia banda Kramer (Hemiptera: Swai ES; Mollel P; Malima A. 2014. Some factors associated Cicallidae), a vector of Napier stunt phytoplasma in Kenya. with poor reproductive performance in smallholder dairy Naturwissenschaften 96(10):1169–1176. doi: 10.1007/ cows: the case of Hai and Meru districts, northern Tanzania. s00114-009-0578-x Livestock Research for Rural Development 26(6):105. lrrd. Odero-Waitituh JA. 2017. Smallholder dairy production in org/lrrd26/6/swai26105.htm (Received for publication 19 May 2021; accepted 17 August 2022, published 30 September 2022) © 2022 Tropical Grasslands-Forrajes Tropicales is an open-access journal published by International Center for Tropical Agriculture (CIAT), in association with The Tropical Crops Genetic Resources Institute of The Chinese Academy of Tropical Agricultural Sciences (TCGRI-CATAS). This work is licensed under the Creative Commons Attribution 4.0 International (CC BY 4.0) license. Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) Tropical Grasslands-Forrajes Tropicales (2022) Vol. 10(3):204–213 204 doi: 10.17138/TGFT(10)204-213 Research Paper Effects of adding agro-industrial by-products and bacterial inoculant at ensiling on nutritional quality and bacterial colonization of Tifton 85 [Cynodon dactylon (L.) Pers.] silages Efectos de agregar subproductos agroindustriales e inoculante bacterial al momento de ensilar en la calidad nutritiva y la colonización bacterial de ensilajes de Tifton 85 [Cynodon dactylon (L.) Pers.] ANDRÉ SANCHES DE AVILA1, MAXIMILIANE ALAVARSE ZAMBOM2, ANDRESSA FACCENDA1, MARCELA ABBADO NERES2, LUANA MUXFELDT2, CIBELE REGINA SCHNEIDER3, MARCELO MARTINI STUM2, RICARDO DRI2 AND PÂMELA ROSANA SCHNEIDER2 1Instituto de Saúde e Produção Animal, Universidade Federal Rural da Amazônia, Belém, PA, Brazil. novo.ufra.edu.br 2Universidade Estadual do Oeste do Paraná, Marechal Cândido Rondon, PR, Brazil. unioeste.br 3Universidade Estadual de Maringá, Maringá, PR, Brazil. uem.br Abstract The objective of this study was to evaluate effects of adding agro-industrial by-products (soybean hulls and corn- processing residue) and bacterial inoculant to Tifton 85 forage at ensiling on nutritional quality and bacterial colonization of resulting silages. The design was completely randomized in a 3 × 2 factorial scheme, with 6 treatments and 4 replicates. Treatments were: Tifton 85 forage; Tifton 85 + soybean hulls; Tifton 85 + corn-processing residue; Tifton 85 + bacterial inoculant; Tifton 85 + soybean hulls + inoculant; and Tifton 85 + corn-processing residue + inoculant. Inclusion of by-products increased dry matter and organic matter percentages of silages, while addition of soybean hulls improved crude protein concentration in silage. Total digestible nutrients in silages containing by-products were higher than in straight Tifton 85 silage. Addition of by-products increased in vitro dry matter and organic matter digestibilities of resulting silages. Most treatments showed aerobic stability up to 144 hours after exposure to air, except for Tifton 85 + corn-processing residue without inoculant, which became unstable by 120 hours of exposure. Addition of by-products at ensiling of Tifton 85 forage appears beneficial but there seems little benefit in adding bacterial inoculant. More studies on a larger scale are needed to confirm these preliminary results, while feeding studies would determine any improvement in animal performance when fed silage containing by-products. Keywords: Digestibility, fodder conservation, gas production, Lactobacillus, pH, tropical grass. Resumen El objetivo de este estudio fue evaluar los efectos de la adición de subproductos agroindustriales (cáscara de soja y residuos del procesamiento de maíz) e inoculante bacteriano al forraje Tifton 85 al momento de ensilar, sobre la calidad nutricional y la colonización bacteriana de los ensilajes resultantes. El diseño fue completamente al azar en un esquema factorial 3 × 2, con 6 tratamientos y 4 repeticiones. Los tratamientos fueron: forraje Tifton 85; Tifton 85 + cáscaras de soja; Tifton 85 + residuo de procesamiento de maíz; Tifton 85 + inoculante bacteriano; Tifton 85 + cascarilla de soja + inoculante; y Tifton 85 + residuo de procesamiento de maíz + inoculante. La inclusión de subproductos aumentó los porcentajes de materia seca y materia orgánica de los ensilajes, mientras que la adición de cáscaras de soja mejoró la concentración de proteína cruda en el ensilaje. Los nutrientes digestibles totales en los ensilajes que contenían subproductos fueron más altos que Correspondence: André Sanches de Avila, Universidade Federal Rural da Amazônia, Avenida Presidente Tancredo Neves, 2501, CEP: 66.077-830, Belém, PA, Brazil. Email: andre.avila@ufra.edu.br Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) Use of agro-industrial by-products and inoculant in Tifton 85 silages 205 en el ensilaje de solo Tifton 85. Además, la adición de subproductos aumentó la digestibilidad in vitro de la materia seca y de la materia orgánica de los ensilajes resultantes. La mayoría de los tratamientos mostraron estabilidad aeróbica hasta 144 horas después de la exposición al aire, excepto Tifton 85 + residuo de procesamiento de maíz sin inoculante, que se volvió inestable a las 120 horas de exposición. La adición de subproductos en el ensilaje del forraje Tifton 85 parece beneficiosa, pero parece haber poco beneficio en la adición de inoculante bacteriano. Se necesitan más estudios a mayor escala para confirmar estos resultados preliminares, mientras que los estudios de alimentación determinarían cualquier mejora en el rendimiento de los animales cuando se alimentan con ensilaje que contiene subproductos. Palabras clave: Conservación de forrajes, digestibilidad, Lactobacillus, pasto tropical, pH, producción de gas. Introduction objective of this study was to evaluate effects of adding agro-industrial by-products and bacterial inoculant to Grasses of the Cynodon genus are widely used to produce Tifton 85 forage at ensiling on nutritional quality and conserved forages, due to rapid growth rate, high dry bacterial colonization of silage produced after 60 days matter production and elevated nutritional value (Souza of storage. et al. 2006). Among these grasses Tifton 85 Bermuda grass [Cynodon dactylon (L.) Pers.] is of interest, as it Materials and Methods has a higher neutral detergent fiber digestibility and lower lignin concentration than other cultivars of the The experiment was carried out in Marechal Cândido genus (Mandebvu et al. 1999). Rondon (24°31'51" S, 54°01'02" W; 392 masl), in the state Forage from Cynodon species is often ensiled, as an of Paraná, Brazil. The animal experimentation protocol alternative to hay making because of reduced drying time was approved by the local Ethics Committee on Animal (Arriola et al. 2015). However, some characteristics of these Use (case no. 06411). species, such as low levels of soluble carbohydrates and high humidity at the ideal vegetative stage for harvesting, Experimental design and treatments may limit the ability to produce quality silages (Evangelista et al. 2000). In this context, addition of absorbent material, The design was completely randomized with 6 treatments such as ground cereals and agro-industrial by-products, distributed in a 2 × 3 factorial arrangement. Treatments can reduce moisture content, thus avoiding losses due evaluated were silages made from: Tifton 85 forage to undesirable fermentation, effluent production and (TS); Tifton 85 forage + soybean hulls (TSSH); Tifton deterioration (Paziani et al. 2006), as well as improving 85 forage + corn-processing residue (TSCR); Tifton 85 chemical and nutritional composition of silage. forage + inoculant (TSI); Tifton 85 forage + soybean Additives with high pectin levels, such as soybean hulls + inoculant (TSSHI); and Tifton 85 forage + corn- hulls, have high water-absorption capacity, and even processing residue + inoculant (TSCRI). Each treatment in forage with high moisture content, pectin can make had 4 replicates, totaling 24 experimental silos. water unavailable, thus hindering development of Forage was harvested from 1.6 ha of Tifton 85 grown undesirable microorganisms (Rodrigues et al. 2005). on a red clayey eutrophic Latosol with a clay texture By-products obtained during maize cleaning and (Santos et al 2018) and chemical characteristics as milling are also high-quality ingredients (Strazzi shown in Table 1. At 43 days of regrowth forage was 2015), which can be added to forage at ensiling to help harvested at 5 cm from ground level with a shredder reduce moisture concentration and increase nutrient coupled to a tractor and ensiled according to the various concentration of resulting silage. Bacterial inoculants treatments. Soybean hulls and corn-processing residue are also used to improve fermentative characteristics of (Table 2) were added to their respective treatments at silage by increasing prevalence of lactic acid bacteria in 100 g/kg fresh forage, aiming to elevate dry matter the epiphytic population, increasing production of lactic (DM) concentration of ensiled material to 300 g DM/ acid and promoting a drop in pH (Adesogan et al. 2004; kg. Inoculant used consisted of Lactobacillus plantarum Bernardes and Chizzoti 2012). with manufacturer-guaranteed levels of 4.0 × 1010 colony Our hypothesis is that simultaneous addition of agro- forming units (CFU) per gram, Pediococcus acidilatici industrial by-products and bacterial inoculant at ensiling (1.0 × 1010 CFU/g), cellulase and a carrier. It was of forage will improve nutritional characteristics and applied to chopped forage using a pressure sprayer at a bacterial colonization of resultant silage. Thus, the concentration of 2 g inoculant/t fresh forage. Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) 206 A.S. de Avila, M.A. Zambom, A. Faccenda, M.A. Neres, L. Muxfeldt, C.R. Schneider, M.M. Stum, R. Dri and P.R. Schneider Table 1. Chemical analysis of soil in the area used to produce Tifton 85 forage. (H2O) P OM K Ca2+ Mg2+ Al3+ (H+Al) SB CEC BS Clay PH (mg/dm3) (cmolc/dm3) (%) 5.30 36.1 23.2 0.71 5.19 1.89 0.00 4.96 7.79 12.7 61.1 58.1 OM = organic matter; H+Al = potential acidity; SB = sum of bases; CEC = cation exchange capacity; BS = base saturation. Table 2. Chemical composition (g/kg dry matter) of Tifton 85 hydraulic press for juice extraction. Extracted liquid was forage, soybean hulls and corn-processing residue. used to determine NH3-N by the potassium hydroxide Variable Fresh Soybean Corn- distillation method, proposed by Fenner (1965) and Tifton 85 hulls processing adapted by Vieira (1980). forage residue For chemical analysis, samples were dried in a Dry matter (g/kg 247 882 890 forced-air oven (55 °C for 72 h) and ground through a 1 as ensiled) mm sieve screen in a Wiley mill (Star FT-80/2, Fortinox, Organic matter 910 952 895 Piracicaba, SP, Brazil). Samples were analyzed according Crude protein 132 112 109 to AOAC (1990) methodology for DM (method 934.01), Neutral detergent 748 648 491 ash (method 938.08) and crude protein (CP; method fiber 981.10), while neutral detergent fiber (NDF) and acid Acid detergent 382 464 213 detergent fiber (ADF) were determined according to fiber Van Soest et al. (1991) as adapted to the Ankom220 Fiber Analyzer (Ankom Technology, Macedon, NY, USA). Subsequently, the forage was homogenized and Amounts of organic matter (OM) were calculated as the stored in experimental PVC silos (10 cm in diameter and difference between ash and total DM. Acid detergent 50 cm long), equipped with a Bunsen-type valve. A 5 cm insoluble protein (ADIP) was obtained from protein layer of autoclaved, dry sand was placed in the bottom analysis of ADF residue. Total digestible nutrient of the silo covered by a layer of cotton cloth to drain concentration (TDN) was estimated according to the liquids. Compaction was performed with a wooden stick equation described by Kunkle and Bates (1998): to a compaction density of approximately 1.58 kg fresh TDN = OM × (26.8 + 0.595 × IVOMD) forage per silo (0.003925 m3) resulting in a calculated where: specific mass of approximately 402 kg fresh forage/m3. TDN= total digestible nutrients; Caps were applied to the experimental silos and sealed OM = organic matter; and with adhesive tape before silos were stored at room IVOMD = in vitro organic matter digestibility. temperature for 60 days. Dry matter recovery (DMR) was estimated according to the method of weighing silos and dry matter before Data collection and analytical procedures and after ensiling, described by Jobim et al. (2007): (FMop × DMop Silos were opened after 60 days of storage and a 5 cm DMR= ×100 layer of silage from both ends of the silo was discarded FMcl × DMcl ) and remaining material was homogenized and sampled where: for analysis. For evaluation of aerobic stability, 300 g FMop = forage mass at opening; samples of silage from each silo were selected and packed DMop = dry matter % at opening; in plastic flasks at room temperature. Temperatures FMcl = forage mass at closing; and of silage in plastic flasks plus room temperature were DMcl = dry matter % at closing. measured with a digital probe thermometer daily at For in vitro procedures, ruminal fluid was collected from 14:00 h for 6 days (144 hours) after opening. An increase 3 Jersey steers fitted with ruminal cannulae. Animals were of 2 °C above room temperature was considered as loss grazing Tifton grass pasture and received a concentrate of aerobic stability (O’Kiely et al. 2001). mix of ground corn, soybean meal and minerals. Evaluation of hydrogen potential (pH) was performed In vitro gas production was measured using the immediately after silo opening, as described by Cherney technique described by Theodorou et al. (1994) and and Cherney (2003). For analysis of ammoniacal adapted by Mauricio et al. (1999) by means of a wireless nitrogen (NH3-N), 200 g samples were pressed in a computerized system Ankom RF-Gas production Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) Use of agro-industrial by-products and inoculant in Tifton 85 silages 207 system (Ankom Technology, Macedon, NY, USA). where: Pressure of gases produced in the bottles was measured Yijk = temperature value at a given aerobic exposure period; every 10 min for 48 h and then converted to volume. μ = overall mean; Volume of accumulated gas was corrected for fermented Ij = fixed effect of inoculant; DM. Kinetic parameters of ruminal fermentation Bk= fixed effect of by-product; were estimated following the bicompartmental model IBjk = effect of interaction between inoculant and by- proposed by Schofield et al. (1994). product; and In vitro dry matter digestibility (IVDMD) was εijk = random error. estimated using the technique of Tilley and Terry (1963) Data were submitted to analysis of variance and and adapted by Holden (1999) using a DaisyII Incubator when significant, effects of agro-industrial by-product, (Ankom Technology, Macedon, NY, USA). In vitro inoculant and their interaction were compared using the organic matter digestibility (IVOMD) was determined Tukey test. Significance was reported at P<0.05. by burning the remaining residue after incubation. To Silage temperature during aerobic exposure was determine in vitro cell wall digestibility (IVCWD), we analyzed as repeated measurements over time. The used the technique of Goering and Van Soest (1970). mathematical model adopted was: Bacterial population was determined using culture Yijkl=μ+Ii+Bj+IBij+δijk+Tl+BTjl+IBTijl+εijkl; techniques according to Silva et al. (1997). Briefly, 225 where: mL of sterile distilled water was added to 25 g of silage. Yijkl = temperature value at a given aerobic exposure The solution was stirred and considered as dilution period; 10- 1. From this solution, we pipetted 1 mL in successive μ = overall mean; dilutions of 10-2 to 10-8, using test tubes containing 9 Ii = fixed effect of inoculant; mL distilled water. Subsequently, diluted extracts were Bj = fixed effect of by-product; placed in Petri dishes, using 0.1 mL of inoculum per IBij = effect of interaction between inoculant and by- plate seeded on the surface and 1 mL for plates seeded product; at depth. δijk = random effect of silo; For determination of enterobacteria numbers, Tl = fixed effect of aerobic exposure period; samples were seeded at depth on plates with Violet Red BTjl = interaction effect between by-product and Bile Agar and incubated at 35 °C for 24 h. For analysis aerobic exposure period; of Clostridium spp., samples were seeded on the surface IBTijl = interaction effect between inoculant, by- in plates with Reinforced Clostridial Agar and incubated product and aerobic exposure period; and under anaerobic conditions at 35 °C for 24 h, using εijkl = random error. an incubator with a CO2 gas system (TE 399 Tecnal; Covariance structure was chosen by considering Tecnal Laboratory Equipment, Piracicaba, SP, Brazil). the lowest Akaike Information Criterion. Structures of For lactic acid bacteria, samples were seeded on Man, covariance tested included variance compounds (VC), Rogosa and Sharpe Agar and incubated for 48 h at 37 °C. compound symmetry (CS), first-order autoregressive After incubation times were reached, plates containing (AR (1)) and unstructured (UN). 25‒250 colonies were selected for counting. Colony counting was performed using a Quebec Counter (CP Results 608, Phoenix Luferco, Araraquara, SP, Brazil) and values were transformed into log base 10. Chemical composition and in vitro nutritional evaluation Statistical analyses There were significant interactions between by-product and inoculant for concentrations of organic matter All statistical analyses were performed using the (OM) and acid detergent insoluble protein (ADIP) only. MIXED procedure of SAS (Statistical Analysis System, Addition of by-products increased DM concentration version 9.2; SAS Institute Inc., Cary, NC, USA). The in silages (P<0.05), with highest values occurring in experimental design was completely randomized in a TSCR followed by TSSH and then TS (P<0.05; Table 3). 2 × 3 factorial scheme (inoculant × by-products). The Organic matter concentration was greater for TSCR and mathematical model adopted was: TSSH than for TS (P<0.05) without inoculant but in the Yijk=μ+Ij+Bk+IBjk+εijkl, presence of inoculant OM for TSSH was greater than Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) 208 A.S. de Avila, M.A. Zambom, A. Faccenda, M.A. Neres, L. Muxfeldt, C.R. Schneider, M.M. Stum, R. Dri and P.R. Schneider for both TSCR and TS. Crude protein concentration for addition slowed rate of degradation for TS (P<0.05). In TSSH was greater than for TS independent of inoculant absence of inoculant, E for TS exceeded those for TSSH (P<0.05). Concentration of ADIP in the absence of and TSCR, while there was no effect of by-product when inoculant was greater for TS and TSSH than for TSCR, inoculant was added. For TS adding inoculant slowed rate while in the presence of inoculant ADIP was similar for of degradation (P<0.05). In absence of inoculant, lag time all silages. However, adding inoculant reduced ADIP for bacterial colonization (C) for TS was greater than for for TS and TSSH but raised ADIP for TSCR (P<0.05). TSSH and TSCR (P<0.05), while in presence of inoculant, Neutral detergent fiber and ADF were influenced by by-product had no effect (P>0.05). Interestingly adding type of by-product used, with lowest values for TSCR inoculant shortened time for colonization for TS but regardless of inoculant (P<0.05). Total digestible lengthened it for TSCR (P<0.05). nutrient values for TSSH and TSCR were higher than In vitro dry matter digestibility and IVOMD were for TS (P<0.05) with no effect of inoculant. Dry matter influenced by use of by-products (P<0.05), being higher for recovery was not influenced by addition of inoculant or TSSH and TSCR than for TS, with no effect of inoculant by-product (P>0.05). (P>0.05) (Table 4). In vitro cell wall degradability Regarding in vitro gas production (Table 4), production (IVCWD) followed the order TSSH>TS>TSCR (P<0.05), of gas from the rapidly degradable (Fraction A) and with no effect of inoculant. slowly degradable (D) fractions was higher in TSSH than in TS (P<0.05) with no effect of inoculant (P>0.05). Bacterial colonization and aerobic stability However, total gas production (A + D) followed the order TSSH>TSCR>TS (P<0.05). There were interactions There were significant by-product × inoculant interactions between by-product and inoculant for degradation rates for hydrogen potential (pH) and concentration of NH3-N, of rapidly (B) and slowly degradable (E) fractions and plus numbers of lactic acid bacteria, enterobacteria and time for bacterial colonization (C) (P<0.05). For B, by- Clostridium spp. (P<0.05; Table 5). While pH in TSCR product had no significant effect (P>0.05), while inoculant was lower than in other silages regardless of the use or Table 3. Chemical composition and dry matter recovery (g/kg DM) of Tifton 85 silages following addition of agro-industrial by- products and bacterial inoculant at ensiling. Variable Inoculant By-product s.e.m. Inoculant. By-product Inoc. × TS TSSH TSCR By-pr. DM (g/kg Without 237c1 294b 310a 3.44 0.20 <0.01 0.79 fresh silage) With 240c 303b 314a OM Without 898bA 914aA 910aA2 1.08 0.10 <0.01 <0.01 With 901bA 914aA 904bB CP (g/kg Without 110b 119a 118ab 1.74 0.12 0.01 0.27 CP) With 109b 118a 110ab ADIP Without 66.1aA 64.5aA 44.6bB 1.91 0.13 <0.01 <0.01 With 56.4aB 58.2aB 53.2aA NDF Without 675a 689a 601b 5.96 0.18 <0.01 0.36 With 693a 684a 618b ADF Without 397b 422a 346c 3.43 0.89 <0.01 0.48 With 402b 415a 345c TDN Without 577b 646a 649a 8.73 0.76 <0.01 0.45 With 586b 662a 634a DMR Without 911 956 988 18.6 0.13 0.24 0.06 With 979 986 962 1Means within rows followed by different lower-case letters are different (P<0.05). 2Means within columns and variables followed by different upper-case letters are different (P<0.05). TS = Tifton 85 silage; TSSH = TS + soybean hulls silage; TSCR = TS + corn residue silage; DM = dry matter; OM = organic matter; CP = crude protein; ADIP = acid detergent insoluble protein; NDF = neutral detergent fiber; ADF = acid detergent fiber; TDN = total digestible nutrients; DMR = dry matter recovery. Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) Use of agro-industrial by-products and inoculant in Tifton 85 silages 209 Table 4. In vitro gas production (mL/100 mg fermented dry matter), degradation rates and in vitro digestibility (g/kg dry matter) of Tifton 85 silages as affected by addition of agro-industrial by-product and bacterial inoculant. Variable Inoculant By-product s.e.m. Inoculant. By-product Inoc. × TS TSSH TSCR By-pr. A (mL) Without 4.68b1 8.56a 6.84ab 0.62 0.73 0.03 0.45 With 5.63b 7.23a 6.46ab B (DR of A Without 0.16A2 0.11A 0.15A 0.016 0.08 075 0.04 mL/h) With 0.09B 0.13A 0.11A C (lag Without 5.05aA 2.14bA 0.19bB 0.57 0.54 <0.01 <0.01 time/h) With 2.17aB 2.30aA 2.03aA D (mL) Without 11.9b 17.8a 14.6ab 0.87 0.83 <0.01 0.85 With 12.0b 16.7a 14.9ab E (DR of D Without 0.05aA 0.04bA 0.04bA 0.002 0.02 <0.01 <0.01 mL/h) With 0.04aB 0.04aA 0.04aA A+D (mL) Without 16.6c 26.3a 21.4b 0.89 0.65 <0.01 0.42 With 17.6c 24.0a 21.4b IVDMD Without 606b 702a 714a 14.8 0.62 <0.01 0.60 With 619b 730a 699a IVOMD Without 629b 736a 747a 16.1 0.68 <0.01 0.58 With 642b 766a 728a IVCWD Without 632b 711a 588c 8.04 0.45 <0.01 0.36 With 627b 685a 596c 1Means within rows followed by different lower-case letters are different (P<0.05). 2Means within columns and variables followed by different upper-case letters are different (P<0.05). TS = Tifton 85 silage; TSSH = TS + soybean hulls silage; TSCR = TS + corn residue silage; A = gas volume produced from rapidly degradable fraction; B = degradation rate of rapidly degradable fraction (Fraction A) in mL per hour; DR of A = degradation rate of fraction A; C = lag time for bacterial colonization; D = gas volume produced from slowly degradable fraction; E = degradation rate of slowly degradable fraction (fraction D) in mL per hour; DR of D = degradation rate of fraction D; IVDMD: in vitro dry matter digestibility; IVOMD = in vitro organic matter digestibility; IVCWD = in vitro cell wall digestibility. Table 5. Hydrogen potential (pH), ammonia nitrogen concentration (NH3-N, g/kg total N) and bacterial colonization (log CFU/g) of Tifton 85 silages as affected by addition of agro-industrial by-products and bacterial inoculant. Variable Inoculant By-product s.e.m. Inoculant. By-product Inoc. × TS TSSH TSCR By-pr. pH Without 5.10aA1 5.54aA2 3.56bB 0.12 0.96 <0.01 <0.01 With 5.11aA 4.85aB 4.23bA NH3-N Without 62.7aA 79.0aA 22.1bA 7.02 0.52 <0.01 0.02 With 72.2aA 48.9abB 31.5bA Lactic acid Without 7.00aA 7.19aA 5.84bB 0.18 <0.01 <0.01 <0.01 bacteria With 7.28aA 6.98aA 7.15aA Enterobacteria Without 4.23aA 4.17aA 0.50bA 0.64 0.29 <0.01 0.02 With 4.74aA 1.33bB 1.12bA Clostridium Without 7.06aA 7.17aA 5.94bB 0.21 0.04 <0.01 0.04 spp. With 7.18aA 7.14aA 6.98aA 1Means within rows followed by different lower-case letters are different (P<0.05); 2Means within columns and variables followed by different upper-case letters are different (P<0.05). TS = Tifton 85 silage; TSSH = TS + soybean hulls silage; TSCR = TS + corn residue silage; Inoc = inoculant; By-pr = by-product. Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) 210 A.S. de Avila, M.A. Zambom, A. Faccenda, M.A. Neres, L. Muxfeldt, C.R. Schneider, M.M. Stum, R. Dri and P.R. Schneider not of inoculant, adding inoculant lowered pH in TSSH Crude protein concentration was not affected markedly and raised it in TSCR (P<0.05). Concentration of NH3-N by addition of agro-industrial by-products. This is scarcely followed a similar pattern with TS and TSSH producing surprising as CP% of Tifton 85 silage was 11.0 %, while higher levels than TSCR in the absence of inoculant, CP% of soybean hulls was only 11.2 % and of corn- while only TS exceeded TSCR with inoculant (P<0.05). processing residue was 10.9 %. Despite this, CP% of the Adding inoculant had an effect only for TSSH, where silages ranged from 10.9 to 11.9 %, which is adequate for inoculant lowered NH3-N concentration (P<0.05). feeding to non-lactating animals, although lactating cows Populations of lactic acid bacteria, enterobacteria and would need protein supplements, if fed solely on these Clostridium spp. were lower for TSCR than for TS silages. Neres et al. (2014) obtained a similar result, when and TSSH in absence of inoculant (P<0.05). However, evaluating benefits of additives to Tifton 85 forage ensiled in treatments with inoculant addition, populations at 38 days of vegetative growth. These authors found no of enterobacteria were the only ones affected by by- differences in CP% between treatments with soybean hulls products with TS>TSSH and TSCR (P<0.05). added and Control (17.6 and 17.3 %, respectively). Silo temperatures showed an interaction between With regard to ADIP concentrations in the silages, inoculant, by-product and aerobic exposure period values for all treatments with inoculant plus TSCR without (P<0.05). All silages were still aerobically stable at 144 inoculant were lower than for TS and TSSH without hours after exposure to air (Figure 1), except for TSCR inoculant. Acid detergent insoluble protein is lignin- without inoculant, which broke stability by 120 hours. associated protein that occurs due to non-enzymatic reactions with heating of the ensiled mass (Maillard reactions) and reduces protein digestibility (Van Soest 1994). According to Kung Jr et al. (2018), when excessive amounts of air are retained in forage mass at ensiling, temperatures increase and can reach above 45‒60 °C. When this occurs for a prolonged period, it may lead to protein being heat- damaged with increase in ADIP. However, in the present study all treatments were ensiled at the same pressure with similar specific mass, so it is difficult to associate elevated ADIP levels in some treatments with heat damage due to presence of additional air in those silages. Neutral detergent fiber and ADF concentrations in Figure 1. Aerobic stability of Tifton 85 silage (TS), Tifton 85 + silage were influenced by type of by-products used, soybean hulls silage (TSSH) and Tifton 85 + corn- processing with lowest levels in TSCR treatments. Lower fiber residue silage (TSCR) with or without addition of bacterial concentrations in corn-processing residue would have inoculant during 144 hours of exposure to air. been a contributing factor, as portions of corn grain are included in this by-product. As a result of composition of Discussion silages with by-product additives, TDN concentrations in these silages were higher than in straight Tifton 85 silage. This study has produced valuable information on benefits Dry matter recovery was not influenced by treatment of adding soybean hulls and corn-processing residue to and values varied between 911 and 988 g/kg. In evaluating Tifton 85 forage prior to ensiling and any additional benefit Tifton 85 silages with different additives, Neres et al. (2014) of adding bacterial inoculant. Inclusion of corn-processing also found no benefit from additives, with mean values of residue and soybean hulls promoted an increase in both DM 813 g/kg, well below those obtained in this study. Santos and OM of the silages. According to Rotz and Muck (1994), et al. (2014) evaluated the use of bacterial inoculant in DM concentration of around 300 g/kg may reduce potential guinea grass silages (Megathyrsus maximus syn. Panicum for undesirable fermentation and effluent production, which maximum) and obtained higher DMR in treatments with would reduce losses during storage. Andrade et al. (2012) homofermentative inoculant and attributed this outcome to added corn meal and soybean hulls to elephant grass at better fermentation profile in these silages resulting from ensiling and showed that these additives were good options inhibition of undesirable microorganisms. Higher DMR for increasing DM percentage, improving fermentation values are desirable as they indicate lower losses during standard and reducing losses via effluent. the ensiling process (Quaresma et al. 2010). Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) Use of agro-industrial by-products and inoculant in Tifton 85 silages 211 In vitro gas production from feedstuffs can be used to 2018); however, in our study, all treatments remained below estimate their nutritional value (Silva et al. 2014), based on these values, indicating good silage quality. Formation total volume of gases produced by fermentation of nutrients. of NH3-N occurs due to a group of proteolytic clostridia In terms of gas production from rapidly degradable (Clostridium spp.), which develop when pH is above 5.0 fraction (A), highest values occurred when soybean hulls (Driehuis 2013). This relationship between pH level and were used, with intermediate values following addition of development of proteolytic clostridia can be evidenced in corn-processing residue, indicating that addition of agro- this study, since lowest NH3-N values occurred for TSCR industrial by-products at ensiling of tropical grasses can with and without inoculant and for TSSH with inoculant, increase concentration of non-fibrous carbohydrates in which also had pH below 5.0. However, it is important to resulting silage. note that populations of clostridia evaluated in the present In slowly degradable fraction (D), TSSH again had study were high for all treatments. This may have occurred higher gas production regardless of use or not of inoculant, because some species of non-proteolytic clostridia tolerate which is consistent with greater in vitro cell wall digestibility pH values down to 4.2 (Driehuis 2013), which would (IVCWD) for these treatments. Higher NDF degradability explain the high population of these microorganisms in allows better microbial fermentation, which in turn can silages with lesser NH3-N concentration. increase energy availability of the diet (Arroquy et al. Regarding lactic acid bacteria, TSCR without 2014). Higher proportion of gas production from fraction inoculant had the lowest population. Low pH values for D, relative to fraction A, is due to characteristics of the this treatment are indicative of more-intense fermentation forage, in which fibrous fractions predominate, coupled soon after ensiling, which may have caused reduction in with low levels of soluble carbohydrates characteristic of concentration of substrates and later in the population of tropical grasses (Adesogan et al. 2004). lactic acid bacteria. Use of bacterial inoculant containing In terms of total gas production (A+D), inclusion of Lactobacillus plantarum and Pediococcus acidilatici soybean hulls produced the highest values, followed increased the population of lactic acid bacteria in by corn-processing residue, with lowest values for TS TSCR, but did not show this effect for TS and TSSH. By treatments. These results are consistent with IVDMD comparison, Neres et al. (2013) evaluated Tifton 85 without and IVOMD, which also increased following inclusion and with addition of soybean hulls, corn grits or inoculant of soybean hulls and corn-processing residue. Agro- at ensiling and found no differences in populations of lactic industrial by-products improved nutritional value of Tifton acid bacteria in resulting silages. 85 silage, indicating their potential for improving animal Concomitant with lactic fermentation, enterobacteria performance when silages are fed. are also present during initial stages of fermentation Regarding IVCWD, best results were with TSSH (698 and compete with lactic acid bacteria for nutrients, g/kg DM), followed by TS (635 g/kg DM) and TSCR (592 reducing silage quality. Rapid pH decline in the ensiled g/kg DM). Soybean hulls have high NDF digestibility mass to values below 4.5 is desirable to inhibit these (Zambom et al. 2001) and are an appropriate additive when microorganisms (Driehuis 2013). In fact, in the present making tropical grass silages, since this characteristic of study, smallest population of enterobacteria was observed fodder is correlated with DM intake and milk production in treatments with lower pH (TSCR with and without (Oba and Allen 1999). inoculant and TSSH with inoculant), while treatments that Addition of corn-processing residue, regardless of use showed pH above 5.0 (TS with and without inoculant and or not of inoculant, promoted lower silage pH, which may TSSH with inoculant) had higher counts of enterobacteria be related to greater concentration of available substrates (>4 log CFU/g). for lactic acid bacteria. It is important to note that only There was no loss of aerobic stability in silages, except TSCR silages (with and without inoculant) remained for TSCR without inoculant, which warmed to 2 °C below pH of 4.7, described by Kung Jr et al. (2018) as the above room temperature by 120 hours of exposure to air. maximum acceptable limit for grass silages. Andrade Penetration of air into the silage mass results in growth et al. (2012) evaluated elephant grass silages containing of yeasts, which assimilate lactate causing an increase in additives at 100 g by-product per kg fresh forage and pH temperature and pH (Kung Jr et al. 2018). This increase of silages containing corn residue was lower than for those in pH allows growth of aerobic bacteria and fungi that containing soybean hulls (3.49 vs. 4.3, respectively). increases temperature further, causing deterioration in Concentration of NH3-N normally found in grass silage quality (Muck 2013). Thus, loss of aerobic stability silages varies from 80 to 120 g/kg total N (Kung Jr et al. in TSCR without inoculant may be related to higher Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) 212 A.S. de Avila, M.A. Zambom, A. Faccenda, M.A. Neres, L. Muxfeldt, C.R. Schneider, M.M. Stum, R. Dri and P.R. Schneider availability of substrate for aerobic microorganisms, since and Technology. American Society of Agronomy, Madison, lowest pH values were found in this treatment, indicating WI, USA. p. 141–198. doi: 10.2134/agronmonogr42.c4 higher production of lactic acid. Driehuis F. 2013. Silage and the safety and quality of dairy Results of the present study showed that inclusion of foods: A review. Agricultural and Food Science 22(1):16– soybean hulls and corn-processing residue with Tifton 85 34. doi: 10.23986/afsci.6699 forage at ensiling improved nutritional value of resulting Evangelista AR; Lima JA de; Bernardes TF. 2000. Evaluation of some characteristics of star grass (Cynodon nlemfuensis silage, but only corn-processing residue improved bacterial Vanderyst) silage. Revista Brasileira de Zootecnia colonization. Addition of these by-products to this forage at 29(4):941–946. (In Portuguese) doi: 10.1590/S1516- ensiling could be considered desirable but there seems little 35982000000400001 merit in including bacterial inoculant. Further studies on a Fenner H. 1965. Method for determining total volatile bases in larger scale should be conducted to confirm these initial rumen fluid by steam distillation. Journal of Dairy Science laboratory findings and feeding studies with the various 48(2):249–251. doi: 10.3168/jds.S0022-0302(65)88206-6 silages would determine if apparent improvement in silage Goering HK; Van Soest PJ. 1970. Forage fiber analyses quality was reflected in improved animal performance. (apparatus, reagents, procedures, and some applications). United States Department of Agriculture, Washington, DC, Acknowledgments USA. naldc.nal.usda.gov/catalog/CAT87209099 Holden LA. 1999. Comparison of methods of in vitro dry matter digestibility for ten feeds. Journal of Dairy Science The authors thank Coordenação de Aperfeiçoamento de 82(8):1791–1794. doi: 10.3168/jds.S0022-0302(99)75409-3 Pessoal de Nível Superior - Brazil (CAPES) - Finance Jobim CC; Nussio LG; Reis RA; Schmidt P. 2007. Code 001 for their financial support. Methodological advances in evaluation of preserved forage quality. Revista Brasileira de Zootecnia 36(suppl):101–119. References (In Portuguese) doi: 10.1590/S1516-35982007001000013 Kung Jr L; Shaver RD; Grant RJ; Schmidt RJ. 2018. Silage (Note of the editors: All hyperlinks were verified 29 July 2022). review: Interpretation of chemical, microbial, and organoleptic components of silages. Journal of Dairy Adesogan AT; Krueger N; Salawu MB; Dean DB; Staples CR. Science 101(5):4020–4033. doi: 10.3168/jds.2017-13909 2004. The influence of treatment with dual purpose bacterial Kunkle WE; Bates DB. 1998. Evaluating feed purchasing inoculants or soluble carbohydrates on the fermentation and options: Energy, protein, and mineral supplements. aerobic stability of bermudagrass. Journal of Dairy Science Proceedings of Florida Beef Cattle Short Course, University 87(10):3407–3416. doi: 10.3168/jds.S0022-0302(04)73476-1 of Florida, Gainesville, FL, USA. p. 59–70. bit.ly/3OSgNlJ Andrade AP; Quadros DG de; Bezerra ARG; Almeida Mandebvu P; West JW; Hill GM; Gates RN; Hatfield RD; JAR; Silva PHS; Araújo JAM. 2012. Qualitative aspects Mullinix BG; Parks AH; Caudle AB. 1999. Comparison of Elephant grass silage with corn meal and soybean of Tifton 85 and coastal bermudagrasses for yield, hulls . Semina: Ciências Agrárias 33(3):1209–1218. (In nutrient traits, intake, and digestion by growing beef Portuguese) doi: 10.5433/1679-0359.2012v33n3p1209 steers. Journal of Animal Science 77(6):1572–1586. doi: AOAC (Association of Official Analytical Chemists). 10.2527/1999.7761572x 1990. Official Methods of Analysis. 15th Edn. AOAC Mauricio RM; Mould FL; Dhanoa MS; Owen E; Channa KS; International, Arlington, VA, USA. Theodorou MKA. 1999. A semi-automated in vitro gas Arriola KG; Queiroz OCM; Romero JJ; Casper D; Muniz E; production technique for ruminant feedstuff evaluation. Hamie J; Adesogan AT. 2015. Effect of microbial inoculants Animal Feed Science and Technology 79(4):321–330. doi: on the quality and aerobic stability of bermudagrass round- 10.1016/S0377-8401(99)00033-4 bale haylage. Journal of Dairy Science 98(1):478–485. doi: Muck RE. 2013. Recent advances in silage microbiology. 10.3168/jds.2014-8411 Agricultural and Food Science 22(1):3–15. doi: 10.23986/ Arroquy JI; Cornacchione MV; Colombatto D; Kunst Jr C. 2014. afsci.6718 Chemical composition and in vitro ruminal degradation of Neres MA; Zambom MA; Fernandes T; Castagnara DD; hay and silage from tropical grasses. Canadian Journal of Rodrigues JFH; Taffarel LE; Javorski CR; Pozza MSS. Animal Science 94(4):705–715. doi: 10.4141/cjas-2014-014 2013. Microbiological profile and aerobic stability of Tifton Bernardes TF; Chizzotti FHM. 2012. Technological innovations 85 bermudagrass silage with different additives. Revista in silage production and utilization. Revista Brasileira de Brasileira de Zootecnia 42(6):381–387. doi: 10.1590/S1516- Saúde e Produção Animal 13(3):629–641. doi: 10.1590/ 35982013000600001 S1519-99402012000300004 Neres MA; Hermes PR; Ames JP; Zambom MA; Castagnara Cherney JH; Cherney DJR. 2003. Assessing silage quality. In: DD; Souza LC de. 2014. Use of additives and pre-wilting Buxton DR; Muck RE; Harrison JH, eds. Silage Science in Tifton 85 bermudagrass silage production. Ciência e Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) Use of agro-industrial by-products and inoculant in Tifton 85 silages 213 Agrotecnologia 38(1):85–93. doi: 10.1590/S1413-7054201400 from in vitro gas production. Journal of Animal Science 0100010 72(11):2980–2991. doi: 10.2527/1994.72112980x Oba M; Allen MS. 1999. Evaluation of the importance of the Silva N; Junqueira VCA; Silveira NFA. 1997. Manual de digestibility of neutral detergent fiber from forage: effects métodos de análise de alimentos. Livraria Varela, São on dry matter intake and milk yield of dairy cows. Journal Paulo, SP, Brazil. of Dairy Science 82(3):589–596. doi: 10.3168/jds.S0022- Silva TC; Santos EM; Macedo CHO; Lima MA; Azevedo 0302(99)75271-9 JAG; Pinho RMA; Perazzo AF; Oliveira JS. 2014. Ruminal O’Kiely PO; Clancy M; Doyle EM. 2001. Aerobic stability of fermentation kinetics of sorghum hybrid silages. Arquivo grass silage mixed with a range of concentrate feedstuffs at Brasileiro de Medicina Veterinária e Zootecnia 66(6):1865– feed-out. Proceedings of the XIX International Grassland 1873. (In Portuguese) doi: 10.1590/1678-6221 Congress, Fundação de Estudos Agrários Luiz de Queiroz, Souza VG; Pereira OG; Valadares Filho SC; Ribeiro KG; Piracicaba, SP, Brazil. p. 794–795. uknowledge.uky.edu/ Pereira DH; Cecon PR; Moraes SA de. 2006. Effect of igc/19/21/21 replacing Tifton 85 haylage with sorghum silage on intake Paziani SF; Nussio LG; Pires AV; Ribeiro JL; Zopollatto and digestibility of nutrients and performance of beef M; Schmidt P. 2006. Effect of wilting and inoculant on cattle. Revista Brasileira de Zootecnia 35(6):2479–2486. (In Tanzania grass silage quality and heifers performance. Portuguese) doi: 10.1590/S1516-35982006000800038 Acta Scientiarum. Animal Science 28(4):393–400. (In Strazzi S. 2015. Derivados do milho são usados em mais de 150 Portuguese) doi: 10.4025/actascianimsci.v28i4.600 diferentes produtos industriais. Visão Agricola 13:146–150. Quaresma JPS; Abreu JG de; Almeida RG de; Cabral LS; bit.ly/3uvmKx5 Oliveira MA de; Rodrigues RC. 2010. Recovery of Theodorou M; Williams BA; Dhanoa MS; McAllan AB; dry matter and chemical composition of Cynodon sp. France JA. 1994. A simple gas production method using a stargrass silage under periods of pre-wilting. Ciência e pressure transducer to determine the fermentation kinetics Agrotecnologia 34(5):1232–1237. (In Portuguese) doi: of ruminant feeds. Animal Feed Science and Technology 10.1590/S1413-70542010000500022 48(3–4):185–197. doi: 10.1016/0377-8401(94)90171-6 Rodrigues PHM; Borgatti LMO; Gomes RW; Passini R; Meyer Tilley JMA; Terry RA. 1963. A two-stage technique for the PM. 2005. Effect of increasing levels of citrus pulp on the in vitro digestion of forage crops. Journal of the British fermentation quality and nutritive value of elephantgrass Grassland Society 18(2):104–111. doi: 10.1111/j.1365- silage. Revista Brasileira de Zootecnia 34(4):1138–1145. (In 2494.1963.tb00335.x Portuguese) doi: 10.1590/S1516-35982005000400007 Van Soest PJ; Robertson JB; Lewis BA. 1991. Methods Rotz CA; Muck RE. 1994. Changes in forage quality during for dietary fiber, neutral detergent fiber, and nonstarch harvest and storage. In: Fahey Jr GC, ed. Forage quality, polysaccharides in relation to animal nutrition. Journal of evaluation, and utilization. American Society of Agronomy, Dairy Science 74(10):3583–3597. doi: 10.3168/jds.S0022- Madison, WI, United States. p. 828–868. doi: 10.2134/1994. 0302(91)78551-2 foragequality.c20 Van Soest PJ. 1994. Nutritional Ecology of Ruminants. Cornell Santos EM; Pereira OG; Garcia R; Ferreira CLLF; Oliveira JS; University Press, Ithaca, NY, USA. Silva TC. 2014. Effect of regrowth interval and a microbial Vieira PF. 1980. Efeito do formaldeído na proteção de proteínas inoculant on the fermentation profile and dry matter e lipídeos em rações para ruminantes. Ph.D. Thesis. recovery of Guinea grass silages. Journal of Dairy Science Universidade Federal de Viçosa, Viçosa, MG, Brazil. locus. 97(7):4423–4432. doi: 10.3168/jds.2013-7634 ufv.br/handle/123456789/24369 Santos HG dos; Jacomine PKT; Anjos LHC dos; Oliveira VA Zambom MA; Santos GT dos; Modesto EC; Alcalde CR; de; Lumbreras JF; Coelho MR; Almeida JA de; Araujo Gonçalves GD; Silva DC da; Silva KT da; Faustino JO. Filho JC de; Oliveira JB de; Cunha TJF. 2018. Brazilian soil 2001. Valor nutricional da casca do grão de soja, farelo classification system. 5th Edn. EMBRAPA, Brasília, DF, de soja, milho moído e farelo de trigo para bovinos. Acta Brazil. bit.ly/3ImqINZ Scientiarum. Animal Science 23:937–943. doi: 10.4025/ Schofield P; Pitt RE; Pell AN. 1994. Kinetics of fiber digestion actascianimsci.v23i0.2648 (Received for publication 1 September 2020; accepted 22 June 2022; published 30 September 2022) © 2022 Tropical Grasslands-Forrajes Tropicales is an open-access journal published by International Center for Tropical Agriculture (CIAT), in association with The Tropical Crops Genetic Resources Institute of The Chinese Academy of Tropical Agricultural Sciences (TCGRI-CATAS). This work is licensed under the Creative Commons Attribution 4.0 International (CC BY 4.0) license. Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) Tropical Grasslands-Forrajes Tropicales (2022) Vol. 10(3):214–221 214 doi: 10.17138/TGFT(10)214-221 Research Paper Influence of plant population density of Chamaecrista rotundifolia on its value for hay making in the Eastern Amazon, Brazil Efecto de la densidad de plantas de Chamaecrista rotundifolia en su potencial para henificación en la Amazonía Oriental, Brasil ANGÉLICA LUCELIA DA SILVA NASCIMENTO1, NATAN LIMA ABREU1, RAIMUNDO VAGNER DE LIMA PANTOJA1, INGRID STEFANIE QUEIROZ DE OLIVEIRA1, JOSILENE DO NASCIMENTO GOMES1, RENÉ JEAN MARIE POCCARD CHAPUIS2 AND LETÍCIA DE ABREU FARIA1 1Universidade Federal Rural da Amazônia, Paragominas, PA, Brazil. ufra.edu.br 2French Agricultural Research Centre for International Development (CIRAD), Paragominas, PA, Brazil. bresil.cirad.fr Abstract Chamaecrista rotundifolia is a forage legume little used with Brazilian livestock; however, it has been studied for this purpose for over 40 years in Australia. The aim of this study was to characterize the influence of plant densities of approximately 444,400, 111,100 and 27,800 plants/ha (equivalent to spacings of 0.15 × 0.15; 0.30 × 0.30 and 0.60 × 0.60 m) on quantitative and qualitative parameters of C. rotundifolia grown in pure stands as forage under exclusive cropping for hay. While leaf dry matter yields in the first 93 days after planting ranged from 1.48 to 9.32 t DM/ha, declining to 0.71–4.92 t DM/ha in the subsequent 83 days, crude protein concentration of the material was only 7–8 %. Since this species tends to lose leaf during periods of stress, larger paddock studies are needed to determine how well leaf material is retained under conventional hay-making conditions. Optimal stubble height following harvesting should be investigated in an endeavor to increase DM yields at second harvest along with improved survival of plants. Keywords: Animal nutrition; fodder conservation; forage legumes; plant morphology. Resumen Chamaecrista rotundifolia es una leguminosa forrajera poco utilizada con el ganado brasileño; sin embargo, se ha estudiado para este propósito durante más de 40 años en Australia. El objetivo de este estudio fue caracterizar la influencia de densidades de plantas de aproximadamente 444,400, 111,100 y 27,800 plantas/ha (equivalente a espaciados de 0.15 × 0.15; 0.30 × 0.30 y 0.60 × 0.60 m) sobre parámetros cuantitativos y cualitativos de C. rotundifolia cultivado en masas puras como forraje bajo cultivos exclusivos para heno. Mientras que el rendimiento de materia seca de las hojas en los primeros 93 días después de la siembra osciló entre 1.48 y 9.32 t MS/ha, posteriormente descendió a 0.71–4.92 t MS/ha en los 83 días posteriores. La concentración de proteína bruta del material fue solo del 7–8 %. Dado que esta especie tiende a perder hojas durante los períodos de estrés, se necesitan estudios más amplios en los potreros para determinar qué tan bien se retiene el material vegetativo en las condiciones convencionales de producción de heno. Se debe investigar la altura óptima del rastrojo después de la cosecha en un esfuerzo por aumentar los rendimientos de MS en la segunda cosecha junto con una mejor supervivencia de las plantas. Palabras clave: Conservación de forraje; leguminosa forrajera; morfología; nutrición animal. Correspondence: Leticia de Abreu Faria, Universidade Federal Rural da Amazônia, Campus Paragominas, Pará, Brazil, 256, km 06, s/n, Bairro Nova Conquista. Email: leticiadeabreufaria@gmail.com Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) Chamaecrista plant density for hay making 215 Introduction cropping for hay production. While growing species in pure stands facilitates hay production, plant density can Legume hay is primarily made from alfalfa, although affect forage yield and quality as a consequence of plant there are many alternative legumes, mainly in tropical competition. Density effects may vary among species; environments. Chamaecrista rotundifolia (Pers.) Greene for example, the biomass of individual plants of Arachis is a legume native to South America (Argentina, Bolivia, pintoi cultivar ‘Belmonte’ was reduced dramatically when Brazil, Colombia, Paraguay, Uruguay and Venezuela), planted at high density, but the total yield per unit area Central America (Costa Rica and Panama) and North increased (Mamédio et al. 2020). America (Mexico) and is naturalized in countries such as We hypothesized that a similar situation would the USA and in Africa, in addition to being studied and occur with C. rotundifolia and designed this study to used as forage for more than 40 years in Australia. It is a quantify the influence of plant density on quantitative weak perennial or self-regenerating annual (in areas with and qualitative parameters of this legume in pure stands. heavy frost or a long dry season) with prostrate growth when young; when older its floral branches tend to die. Materials and Methods The main stem is erect to about 1 m high (rarely to 2.5 m) and laterals are ascendant with stems of 0.45‒1.1 m long The study was carried out on the experimental area (Cook et al. 2020). According to Strickland et al. (1985), at Universidade Federal Rural da Amazônia (UFRA), Cruz (1996) and Abreu et al. (2020) this is an interesting Campus Paragominas, Pará state, Brazil (2°59′26″ S, species with forage potential, which has stimulated 47°24′24″ W), between December 2018 and June 2019. The research into cropping techniques and use for animal feed. climate of Koppen’s Aw zone is characterized by a distinct Under high-density pure stands (0.25 × 0.25 m) in rainy season and a 6-month dry season (Bastos et al. 1993). tropical areas, C. rotundifolia yields of up to 22.4 t DM/ Average temperature during the experimental period was ha with protein concentration of 8.0 % at 133 days after 30.8 ºC with total precipitation of 2,082 mm (Figure 1). planting have been obtained (Abreu et al. 2020). These data The soil is Typic Hapludox (uniform, deep, clayey) suggest that the species could be utilized for hay production and physical analysis of the soil indicated composition via mechanical harvesting. Lopes (2001), reporting on of 77, 12.6 and 10 % of clay, silt and sand, respectively. the species’s potential for green manure, recommended Chemical analyses showed: pH (CaCl2) – 5.1; P resin planting of C. rotundifolia under spacings of 0.5 and 1.0 m (phosphorus) – 19 mg/dm3; K (potassium) – 3.7 mmolc/ for biomass and seed production, respectively. dm3; Ca (calcium) – 14 mmolc/dm3; Mg (magnesium) – 5 High nutrient composition of legumes makes them mmolc/dm3; H+Al (potential acidity) – 20 mmolc/dm3; ideal for hay production, while characteristics such as and V (base saturation) – 54 %. While C. rotundifolia ease of dehydration and leaf retention are also highly is well adapted to soils with low pH and base saturation, desirable. Several factors which are intrinsic to forage 1.3 t limestone/ha with total relative neutralizing power plants, e.g. cuticle thickness, diameter and length of of 88 % was applied in September 2018, according to stem and leaf:stem ratio, can have impacts on the drying the base saturation method (V%) to elevate the base process (Neres et al. 2010). saturation of soil to 70 % (Hohnwald et al. 2005). There are information gaps in cropping Seeds from native Brazilian plants of Chaemacrista recommendations for legumes, mainly in terms of exclusive rotundifolia (Pers.) Greene var. rotundifolia were 1,000 40 800 30 Precipitation (mm) 600 Temperature (°C) 20 400 200 10 0 0 December January February March April May June Figure 1. Monthly precipitation and mean temperature during the experimental period (INMET 2019). Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) Precipitation (mm) Temperature (°C) 216 A.L.S. Nascimento, N.L. Abreu, R.V.L. Pantoja, I.S.Q de Oliveira, J.N. Gomes, R.J.M.P. Chapuis and L.A. Faria collected manually and treated with hot water at 80 ºC at 105 ºC; crude protein (CP) by microKjeldahl method for 30 sec followed by immersion in water at room according to AOAC (2011); and neutral detergent fiber temperature (30 °C) for 12 hours to break seed dormancy and acid detergent fiber using the autoclave method (Gomes et al. 2021). Seedlings of Chamaecrista were described in Detmann et al. (2012). then produced on trays with substrate in a greenhouse Data were organized using Excel® and were tested and the experiment was established by transplanting for normality by the Shapiro-Wilk technique before seedlings on 28 December 2018. statistical analysis and any variable that failed to Weeds were controlled manually during the experimental follow a normal distribution was transformed through period. Fertilizers were applied according to Hohnwald et the procedure of BOX-COX from package “fpp” on al. (2005) with 8.9 kg P/ha as simple superphosphate at software R. All data were analyzed by ANOVA using transplanting of seedlings and 7.5 kg N/ha as urea and 6.7 the packages “expdoes.pt”, “emmeans” and “agricolae” kg K/ha as potassium chloride after 7 days. from software R as a randomized complete block design A randomized complete block design was used. with 3 planting densities and 2 harvest dates as fixed Treatments comprised 3 planting densities (approx. factors. To test for significance, variables were compared 444,400, 111,100 and 27,800 plants/ha; treated as high, by Tukey test at P<0.05 by command “pwpm”. intermediate and low densities), equivalent to plant spacings of 0.15 × 0.15, 0.30 × 0.30 and 0.60 × 0.60 m, Results respectively, with 7 replications. Plot size was 1.8 × 1.8 m. In February and March (before the canopy was Significant interactions between plant density and harvest closed) leaf necrosis was observed in young plants, date for DM production of C. rotundifolia leaves were possibly caused by incidence of fungi favored by recorded (Table 1). Leaf yield per plant decreased and leaf high rainfall and low soil drainage, independently of yield/ha increased progressively as plant density increased treatment. We controlled this problem by spraying plants (P<0.05; Table 1). In general, leaf yields at the first harvest with a solution of 2.3 kg copper oxide in 300 L water per were superior to those at the second harvest (Table 1). hectare, while at the second harvest (in the dry season) no symptoms of necrosis were observed. Table 1. Effects of plant spacing and harvest on leaf dry Harvests were carried out based on visual assessment, matter production (LDM) of Chamaecrista rotundifolia. aiming at harvesting with a balance between forage LDM (g/plant) LDM (t/ha) production and quality, as well as allowing natural Plant density1 Harvest 1 Harvest 2 Harvest 1 Harvest 2 reseeding, i.e. plants had more than 50 % of open pods 0.60 × 0.60 m 53.5Aa 25.6Ba 1.48Ac 0.71Ab (Abreu et al. 2020). The first harvest was performed on 31 0.30 × 0.30 m 52.6Aa 22.2Bab 5.65Ab 2.39Bb March 2019, i.e. 93 days after transplanting of seedlings, 0.15 × 0.15 m 21.0Ab 11.1Bb 9.32Aa 4.92Ba while the second occurred after a further 83 days on 22 P value <0.001 0.0085 June 2019 at the beginning of the dry season. At sampling CV (%) 26.9 35.9 for each harvest 3 whole plants in the central area (1 m²) 1Identified as low, intermediate and high density, respectively. from each plot were selected, although the branches were Means within columns followed by same lower-case letters spread over a larger area. The 3 plants were harvested and within rows and parameters followed by same upper-case at 5 cm from ground level and harvested material was letters do not differ significantly (P>0.05) by Tukey test. separated manually into stem and leaf. Plant components were weighed and diameter (StD) and length (StL) of There was no significant interaction between plant stems were recorded. Average diameter was determined spacing and harvest date for production of C. rotundifolia by taking measurements at 3 points along the main stems stem, but plant density influenced yields (Table 2). Stem using a digital caliper, while average length was measured production per plant was greater at the 2 wider plant on main stems using a tape. spacings than at the narrow spacing, while production Plant components were dried in a forced-air per ha increased as density increased (P<0.05; Table 2). circulation oven at 65 ºC for 72 hours and weighed to Stem yields were independent of harvest date (P = 0.08). determine dry matter yield. Samples from the first There was no significant interaction between plant harvest were ground in a Willey mill prior to chemical density and harvest date for leaf:stem ratio of plants, but analysis. Chemical analyses of plant components were plant density (P = 0.0046) and harvest date (P<0.001) performed as follows: dry matter (DM) by oven-drying each had significant effects (Table 3). The highest Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) Chamaecrista plant density for hay making 217 leaf:stem ratios were found at the low and intermediate plant density. Highest yields per plant and per hectare plant densities, while leaf:stem ratio at the first harvest occurred at the first harvest (Table 4). was greater than at the second harvest (Table 3). There was no significant interaction between plant density and harvest date for stem length (StL) (P>0.05), Table 2. Effects of plant density and harvest date on stem dry but intermediate plant density showed greater stem length matter (SDM) production of Chamaecrista rotundifolia. than high plant density (P = 0.0065; Table 5). However, a SDM (g/plant) SDM (t/ha) significant interaction between plant density and harvest Plant density1 date occurred for stem diameter with no effect of plant 0.60 × 0.60 m 42.0a 1.18c density on stem diameter at the first harvest, while stem 0.30 × 0.30 m 39.2a 4.39b diameter at the low plant density exceeded that at the high density at the second harvest (P = 0.0046). 0.15 × 0.15 m 25.3b 11.29a Chemical analyses of forage from the first harvest P value <0.001 <0.001 showed that plant density affected all parameters, CV (%) 6.01 11.37 except for CP and H concentrations in stem (Table 6). Harvest In general, CP, NDF, ADF and H concentrations in leaf Harvest 1 0.68a 4.19a and NDF and ADF concentrations in stem were lowest Harvest 2 0.66a 3.57a (P<0.05) at the high plant density. P value 0.0800 0.0810 Table 4. Total DM production (leaf plus stem) per plant and CV (%) 6.01 11.37 per hectare. 1Identified as low, intermediate and high density, respectively. Means within columns followed by same letters do not differ Plant density1 Total production Total production significantly (P>0.05) by Tukey's test. (g/plant) (t/ha) 0.60 × 0.60 m 80.0a 2.32c Table 3. Effects of plant density and harvest date on leaf:stem 0.30 × 0.30 m 73.8a 8.68b ratio of Chamaecrista rotundifolia. 0.15 × 0.15 m 40.9b 19.56a Plant density1 Leaf:stem ratio P value <0.001 <0.001 0.60 × 0.60 m 0.90a CV (%) 5.9 39.7 0.30 × 0.30 m 0.89a Harvest 0.15 × 0.15 m 0.61b Harvest 1 76.6a 12.02a P value 0.0046 Harvest 2 50.6b 8.36b CV (%) 29.8 P value <0.001 0.0064 Harvest CV (%) 5.9 39.7 Harvest 1 1.04a 1Identified as low, intermediate and high density, respectively. Harvest 2 0.56b Means within columns and parameters followed by same P value <0.001 letters do not differ significantly (P>0.05) by Tukey's test. CV (%) 29.8 Table 5. Effects of plant density and harvest date on length 1Identified as low, intermediate and high density, respectively. (StL) and diameter (StD) of stems of C. rotundifolia. Means within columns and parameters followed by same letters do not differ significantly (P>0.05) by Tukey's test. Plant density1 StL (cm) StD (mm) Harvest 1 Harvest 2 There was no interaction between plant density 0.60 × 0.60 m 73.1ba 2.67Aa 2.31Ba and harvest date for total DM production per plant 0.30 × 0.30 m 76.9a 2.87Aa 2.04Bab or per hectare (Table 4). However, both density and 0.15 × 0.15 m 63.3b 2.91Aa 1.95Bb harvest date influenced total DM production. Low and P value 0.0065 P value 0.0046 intermediate plant densities presented higher yields per CV (%) 19.0 CV (%) 18.1 plant than high plant density (P<0.001); on the other 1Identified as low, intermediate and high density, respectively. hand, production per hectare was directly related to plant Means within columns followed by the same lower-case letter density (P<0.001), i.e. highest production occurred at the and within rows for StD followed by the same upper-case highest plant density and lowest production at the lowest letter do not differ significantly (P>0.05) by Tukey test. Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) 218 A.L.S. Nascimento, N.L. Abreu, R.V.L. Pantoja, I.S.Q de Oliveira, J.N. Gomes, R.J.M.P. Chapuis and L.A. Faria Table 6. Effects of plant density on crude protein (CP), neutral detergent fiber (NDF), acid detergent fiber (ADF) and hemicellulose (H) concentrations in leaf and stem of Chamaecrista rotundifolia. Plant density1 Leaf Stem CP NDF ADF H CP NDF ADF H 0.60 × 0.60 m 10.3a 49.4a 28.4ab 21.1a 6.5a 75.7a 58.6b 16.2a 0.30 × 0.30 m 9.6ab 51.9a 30.5a 22.6a 6.2a 77.5a 61.1a 16.4a 0.15 × 0.15 m 8.6b 42.2b 26.0b 16.2b 5.6a 68.9b 52.6c 16.3a P value 0.0298 <0.001 <0.001 <0.001 0.2439 <0.001 <0.001 0.9571 CV (%) 12.9 10.5 10.5 16.2 12.9 6.1 7.3 6.7 1Identified as low, intermediate and high density, respectively. Means within columns followed by same letters do not differ significantly at P<0.05 by the Tukey test. Discussion investigated, since Lopes (2001), in similar climatic conditions, but in sandy soil, obtained up to 6 harvests This study has produced some valuable information per year with intervals between harvests of 56 days and on growth of C. rotundifolia when planted at different cutting at 30 cm above the soil surface. Yields reached densities in pure stands in Brazil. These data provide 17 t DM/ha from a single harvest at 6 months of age or an indication of likely yield and quality of this species a total yield of 25 t DM/ha with 3 harvests at 4, 7 and 14 in the particular environment. C. rotundifolia showed months of age. high potential for leaf production at the first harvest, Although C. rotundifolia is characterized as a mainly at the high plant density (Table 1), with markedly prostrate subshrub or small shrub, plant density lower production (30 % lower) at the second harvest. The and harvesting close to ground level can modify the reduced yields at the second harvest could be a function morphological composition of plants, mainly due to of this growth period extending into the beginning of competition for light and nutrients or reduction in the dry season, combined with forage being harvested amount of photosynthetic tissue, respectively. Different at near ground level at the first harvest. Total production planting densities influenced diameter and length of (Table 4) of the intermediate treatment (0.3 × 0.3 m) stem, which can be related to height and density of was similar to that of Abreu et al. (2020) at similar plant shrubs, and the area occupied by each plant, although density (0.25 × 0.25 m), although only one harvest was these characteristics were not evaluated. performed at 133 days after transplanting seedlings. As a result of the drastic cut, regrowth rates were Virtually no rain was registered after 30 April. Cruz slow and intervals between harvests increased, as also et al. (1999) evaluated accession CIAT 7792 and found observed in Arachis pintoi by Alonzo et al. (2017). no difference in DM production during dry and wet Besides, in dry weather, leaves were smaller and periods, which are characteristic of the Amazon climate. presented a different shape (Figure 2). According to Cook These authors found annual DM yields of 3.55 and 3.39 et al. (2020) this species is reasonably drought-tolerant t/ha for leaf and stem, respectively, in plants spaced with plants forming rosettes under heavier grazing, but at 0.50 m. The magnitude of difference between the 2 leaves often turn red and drop, if plants are left ungrazed studies can be justified by the different soil types, since and tall during dry conditions. This phenomenon was we cultivated C. rotundifolia on a clay soil with N:P:K observed in forage produced during the regrowth period fertilizer application, while Cruz et al. (1999) evaluated prior to Harvest 2, probably in response to the drop in C. rotundifolia in sandy soil without fertilizer, and used soil moisture levels, consequently resulting in lower DM a different accession and harvest height (30 cm above production. This is corroborated by values for leaf:stem ground). ratio under the effect of harvest (P = 0.0046) (Table 3) Harvesting plants at 5 cm from ground level resulted with a higher ratio at the first harvest. Leaf:stem ratio in long recovery periods following harvesting. It is (Table 3) is a plant characteristic with a significant suggested that harvesting this species near ground level influence on forage drying rates (Neres et al. 2010) and should be avoided to prevent low yields at subsequent animal intake of forage, so is an important attribute to harvests. In addition, higher harvest heights should be be considered. Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) Chamaecrista plant density for hay making 219 Lopes (2001) recommended planting this species at spacings of 0.5 and 1.0 m for the formation of pure stands. Our study showed closer plant spacings can result in high DM yields, e.g. at plant spacings between 0.15 × 0.15 and 0.3 × 0.3 m (Tables 1 and 2). High DM yields of forage are particularly desirable when harvesting forage for silage or hay production or cut-and-carry feeding, mainly due to reduction in processing costs per tonne. Lopes (2001) and Camarão et al. (2008) described this species as being well adapted to sandy soils, and even infertile soils in the Amazon. Our results indicated that excellent yields can be reached on predominantly clay soils as well, corroborating the findings of Abreu et al. (2020) in the same edaphoclimatic environment. However, CP concentration varied from 8.6 to 10.3 % for Figure 2. Branch with common, rounded leaves found at first leaves and 5.6 to 6.5 % for stems, which are low values harvest (A) and branch with modified, also smaller leaves for a leguminous plant. These values are lower than those from the second harvest (B) of Chamaecrista rotundifolia found by Cruz et al. (1999) of 16.0 and 18.6 % for leaves (leaves in B were not diseased). and 5.5 and 9.1 % for stems in the dry and rainy seasons, respectively, in a sandy Yellow Latosol, although these Estimated DM production per unit area considering authors obtained forage yields of only 4.1 t DM/ha in the both leaves and stems (Tables 1, 2 and 4) increased wet season and 2.84 t DM/ha in the dry season. as plant density increased, i.e. highest at the narrow The dynamic of production versus quality in both plant spacing/high plant density at the first harvest. studies emphasizes the importance of considering both This dynamic is a function of the greater number of quantity and quality of a species when determining harvest plants being able to utilize the available resources, i.e. frequency. Overall CP concentration of forage compared water, light and nutrients. However, during the second favorably with the 7 % CP, below which animal intake growing season, water availability, in particular, may can be compromised (Berchielli et al. 2006). While higher have become a limiting factor for maximum growth, plant densities resulted in increasing DM yields, forage particularly of leaves, in addition to the limited amount of produced under wider spacing showed better nutritional photosynthetic material available (Table 1). It was evident quality as leaf:stem ratio was superior to those at narrower that high density resulted in a decrease in individual spacings (Table 2). Although plant spacings between 0.15 components (leaves and stems) of the plants (see × 0.15 m and 0.30 × 0.30 m seem to be viable for hay individual DM production in Tables 1 and 2) according or forage production in tropical areas to maximize DM to the well-known deleterious effect of intra-specific yield, the forage produced would be of lesser quality than competition. Interestingly, Vieira et al. (2016) described at wider spacing. However, the increased yields would far C. rotundifolia plants as extremely sensitive to drought outweigh any reduction in quality. conditions, which can significantly reduce biomass Climatic conditions of the Amazon of high rainfall production, sometimes even resulting in plant death. On and temperature along with poor soil drainage can the other hand, Partridge and Wright (1992) described provide a favorable environment for fungal growth. Cruz the Australian cultivar 'Wynn' of this species as having (1996) observed that 6 accessions of C. rotundifolia (BR excellent adaptive capacity and being competitive in 000183, 000205, 000191, 000264, 000272 and 000256) high plant densities, even when intercropped with other were susceptible to the fungi Phomopsis subcircinata species, such as grasses. However, the cultivar used in and Rhizoctonia solani, although with little resultant this study is native to Brazil at about 23° S according damage. In that study the author worked with a harvest to Cruz et al. (1999); in the studies of Lopes (2001) and height of 20 cm, but recommended harvesting above Abreu et al. (2020) it demonstrated a desirable adaptation 30 cm at 56-day intervals, as harvests at 20 cm at the at latitudes about 1‒3° S, similar to 'Wynn' also showing same frequency caused the death of plants. This is susceptibility to drought. corroborated by Cruz et al. (1999) and Lopes (2001). Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) 220 A.L.S. Nascimento, N.L. Abreu, R.V.L. Pantoja, I.S.Q de Oliveira, J.N. Gomes, R.J.M.P. Chapuis and L.A. Faria Certainly, C. rotundifolia is an option for improving Alonzo LAG; Ferreira OGL; Vaz RZ; Costa OAD; Motta JF; livestock production in the Amazon, corroborating the Brondani WC. 2017. Perennial peanut handled with low finding of Camarão et al. (2008), but feeding studies grazing waste by sheep. Arquivo Brasileiro de Medicina would be needed to confirm acceptance by animals. Veterinária e Zootecnia 69(1):173‒180. (In Portuguese). Partridge and Wright (1992) showed improved growth of doi: 10.1590/1678-4162-9157 steers under grazing in Queensland, Australia, when this AOAC International. 2011. Official methods of analysis. 18th Edn. AOAC International, Gaithersburg, MD, USA. legume was sown into native grass pastures. One would Bastos TX; Rocha AMA da; Pacheco NA; Sampaio SMN. 1993. expect a similar result under conditions in the Amazon. Efeito da remoção da floresta ombrófila sobre o regime Pure stands of C. rotundifolia certainly showed pluviométrico no município de Paragominas - PA. Boletim potential for production of forage in the Amazon under de Geografia Teorética 23(45‒46):85‒92. bit.ly/3nl8QcA high planting densities. However, attempting to harvest Berchielli TT; Pires AV; Oliveira SG. 2006. Nutrição de the legume for hay might not be so successful, since this Ruminantes. 1st Edn. ABDR, Jaboticabal, SP, Brazil. species is recognized as shedding leaves under stress, and Camarão AP; Souza Filho APS; Lopes OMN. 2008. poor leaf retention during the hay-making process may Limitações e potencialidades da leguminosa Chamaecrista be an issue. Additional studies on optimal management rotundifolia para alimentação de ruminantes no Pará. 1st strategies, including reducing fungal attacks, persistence Edn. Embrapa, Belém, PA, Brazil. bit.ly/3yiJZg5 Cook BG; Pengelly BC; Schultze-Kraft R; Taylor M; Burkart under repeated harvests (it is only a weak perennial), S; Cardoso Arango JA; González Guzmán JJ; Cox K; acceptance by stock, digestibility, economic factors, Jones C; Peters M. 2020. Tropical Forages: An interactive soil influences and mechanical methods for planting and selection tool. 2nd and Revised Edn. International harvesting seed and harvesting forage are needed. Center for Tropical Agriculture (CIAT), Cali, Colombia and International Livestock Research Institute (ILRI), Conclusions Nairobi, Kenya. tropicalforages.info Cruz ED. 1996. Avaliação agronômica de leguminosas do Chamaecrista rotundifolia showed potential for forage gênero Chamaecrista na Região Bragantina, Pará, Brasil. production in the soil of the experimental site in pure Pasturas Tropicales 18(3):60‒64. bit.ly/3zYqTgt stands at high plant densities, producing yields of up to Cruz ED; Camarão AP; Simão Neto M. 1999. Forage production and nutritive value of Chamaecrista 14 t leaf DM/ha in 6 months. However, further studies rotundifolia (Persoon) Greene in the eastern Amazon are needed to determine how resilient plants are under Brazil. Pasturas Tropicales 21(3):46‒48. bit.ly/3HXZtJG repeated harvests, how well forage retains leaf during the Detmann E; Souza MA; Valadares Filho SC; Queiroz AC; hay-making process, how severe are fungal attacks during Berchielli TT; Saliba EOS; Cabral LS; Pina DS; Ladeira the rainy season, how irrigation at the beginning of the dry MM; Azevêdo JAG. 2012. Métodos para análise de season may impede the decline in production at this time alimentos. INCT - Ciência Animal. Suprema, Visconde and how productive it can be on different soil textures. do Rio Branco, MG, Brazil. Gomes JN; Abreu NL; Nascimento ALS; Rocha SM; Quadros Acknowledgments BR de; Faria LA. 2021. Dormancy-breaking treatments and soil types on Chamaecrista rotundifolia emergence and initial development. Ciência Animal Brasileira 22(1):e- The second author received grants from the National 66677. (In Portuguese) doi: 10.1590/1809-6891v22e-66677 Council of Scientific and Technological Development Hohnwald S; Rischkowsky B; Schultze-Kraft R; Rodrigues- (CNPq) for the grant of scientific initiation (CNPq Filho JA: Camarão AP. 2005. Experiences with legumes PIBIC/UFRA 2019/2020). as part of a ley pasture in a low input farming system of North-Eastern Pará, Brazil. Pasturas Tropicales References 27(3):1‒12. bit.ly/3OID7hw INMET (Instituto Nacional de Meteorologia). 2019. Dados (Note of the editors: All hyperlinks were verified 21 June 2022). históricos nacionais, 2019. Ministério da Agricultura, Pecuária e Abastecimento. portal.inmet.gov.br/uploads/ Abreu NL; Nascimento ALS; Pantoja RVL; Oliveira ISQ dadoshistoricos/2019.zip de; Gomes JN; Faria LA. 2020. Cutting times in the Lopes OMN. 2001. Chamaecrista rotundifolia - Leguminosa forage productivity and proteinvalue ofthe Chamaecrista para controle de mato e adubação verde do solo. rotundifolia. Amazonian Journal of Agricultural and Recomendações técnicas N° 11/2000. Embrapa Amazônia Environmental Sciences 63:1‒6. (In Portuguese). bit.ly/ Oriental, Altamira PA, Brazil. bit.ly/3brY9lP 3U0kdpK Mamédio D; Andrade CMS de; Sampaio AF; Loures DRS. Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) Chamaecrista plant density for hay making 221 2020. Soil management and planting spacing effects with steers in south-east Queensland. Tropical Grasslands on establishment of mixed swards of purple stargrass 26:263‒269. https://bit.ly/3On1Cku (Cynodon nlemfuensis cv. BRS Lua) and forage peanut Strickland RW; Greenfield RG; Wilson GPM; Harvey GL. (Arachis pintoi cv. Belmonte) in an area of degraded 1985. Morphological and agronomic attributes of Cassia Brachiaria brizantha. Revista Mexicana de Ciencias rotundifolia Pers., C. pilosa L., and C. trichopoda Benth., Pecuarias 11(1):241‒254. doi: 10.22319/rmcp.v11i1.5004 potential forage legumes for northern Australia. Australian Neres MA; Castagnara DD; Mesquita EE; Zambom MA; Journal of Experimental Agriculture 25(1):100–108. doi: Souza LC de; Oliveira PSR de; Jobim CC. 2010. Production 10.1071/EA9850100 of alfalfa hay under different drying methods. Revista Vieira RF; Camillo J; Coradin L. 2016. Espécies nativas da Brasileira de Zootecnia 39(8):1676‒1683. doi: 10.1590/ flora brasileira de valor econômico atual ou potencial: S1516-35982010000800008 plantas para o futuro: Região Centro-Oeste. p. 513—516. Partridge IJ; Wright J. 1992. The value of round-leafed cassia Ministério de Meio Ambiente, Brazilia, DF, Brazil. bit. (Cassia rotundifolia cv. Wynn) in a native pasture grazed ly/2UMZUzo (Received for publication 18 November 2020; accepted 10 June 2022; published 30 September 2022) © 2022 Tropical Grasslands-Forrajes Tropicales is an open-access journal published by International Center for Tropical Agriculture (CIAT), in association with The Tropical Crops Genetic Resources Institute of The Chinese Academy of Tropical Agricultural Sciences (TCGRI-CATAS). This work is licensed under the Creative Commons Attribution 4.0 International (CC BY 4.0) license. Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) Tropical Grasslands-Forrajes Tropicales (2022) Vol. 10(3):222–236 222 doi: 10.17138/TGFT(10)222-236 Research Paper The role of leucaena in cattle fattening and breeding production systems in eastern Indonesia El rol de la leucaena en los sistemas de producción de cría y engorde de ganado en el este de Indonesia FAHRUL IRAWAN1,2,3, DAHLANUDDIN2, MICHAEL J. HALLIDAY1, ROGER S. HEGARTY1 AND FRANCES C. COWLEY1 1School of Environmental and Rural Science, The University of New England, NSW, Australia. une.edu.au 2University of Mataram, Lombok, Indonesia. unram.ac.id 3Al-Azhar Islamic University of Mataram, Lombok, Indonesia. unizar.ac.id Abstract Cattle farming in West Nusa Tenggara province, Indonesia (NTB) is essential to support the high demand for beef cattle in Indonesia. Leucaena was introduced to smallholders as a high-quality feed to increase cattle production in NTB. A survey was conducted with both leucaena-using and non-leucaena-using smallholder cattle farmers in Sumbawa and West Sumbawa districts, NTB to understand the role of leucaena in NTB smallholder cattle enterprises (mixed breeding/fattening) and the effects of leucaena toxicity on cattle performance, especially cow-calf production. We found that farmers using leucaena feeding systems were able to keep more cattle than farmers using a traditional feeding system (9.1 vs 6.1 head/household). Many leucaena-using farmers (50.1 %) use leucaena for fattening cattle only. Other cattle classes (growers, breeding cows and bulls) were fed leucaena strategically, such as during the dry season (59 % of leucaena-using farmers) and at specific stages of pregnancy and lactation (41 % of leucaena-using farmers). Leucaena- using farmers in rainfed areas planted more leucaena (4,500 vs 1,984 trees) and fattened more bulls (5.8 vs 3.5 head/ household) than farmers in high-rainfall areas. Transmigrant Balinese farmers planted significantly more leucaena trees (7,500 vs 2,354 trees) and raised more fattening bulls (7.8 vs 3.7 head/household) than the local Sumbawanese farmers. Most Balinese farmers had been practising leucaena feeding systems since they migrated to Sumbawa, for as long as 3 decades. Most leucaena-using farmers (74 %) had observed symptoms of illness associated with leucaena toxicity in their cattle such as hair loss and salivation. Few farmers feeding leucaena to breeding cows (5 %) reported instances of reproductive failure. Almost all non-leucaena-using farmers (93 %) reported symptoms of illnesses associated with plant toxicities (among other potential causes), most commonly skin lesions, diarrhoea, cataracts, and listlessness. It was concluded that the priority use of leucaena in Sumbawa was for fattening cattle rather than breeding cattle. Leucaena supports smallholder farmers in Sumbawa to have more intensive, productive and income-earning cattle enterprises, but questions remain over whether it should be used for feeding breeding cows. Keywords: Breeding enterprises, leucaena users, toxicity. Resumen La ganadería en la provincia de West Nusa Tenggara, Indonesia (NTB) es esencial para respaldar la alta demanda de ganado vacuno en Indonesia. La leucaena se presentó a los pequeños agricultores como un alimento de alta calidad para aumentar la producción de ganado en NTB. Se llevó a cabo una encuesta con pequeños ganaderos que usaban y no usaban leucaena en los distritos de Sumbawa y West Sumbawa, NTB para comprender el papel de la leucaena en las empresas ganaderas Correspondence: Fahrul Irawan, Faculty of Mathematics and Science, Al-Azhar Islamic University of Mataram, Jl. Unizar No. 20, Turida, Kec. Sandubaya, Kota Mataram, NTB 83237, Indonesia. Email: fahrul.irawan@gmail.com Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) Leucaena for breeding and fattening cattle 223 de NTB (cría/engorde mixtos) y el efecto de la toxicidad de la leucaena sobre el rendimiento del ganado, especialmente en la producción de vacas y terneros. Descubrimos que los agricultores que usaban sistemas de alimentación con leucaena podían mantener más ganado que los agricultores que usaban un sistema de alimentación tradicional (9.1 vs. 6.1 cabezas por unidad familiar). Muchos agricultores usuarios de leucaena (50.1 %) utilizan la leucaena solo para engorde de ganado. Otras clases de ganado (animales en crecimiento, vacas reproductoras y toros) fueron alimentados con leucaena estratégicamente, como durante la estación seca (59 % de los agricultores que usan leucaena) y en etapas específicas de gestación y lactancia (41 % de los agricultores que usan leucaena). Los agricultores que utilizan leucaena en áreas de secano plantaron más leucaena (4,500 vs. 1,984 árboles) y engordaron más toros (5.8 vs. 3.5 cabezas/hogar) que los agricultores en áreas de alta precipitación. Los agricultores Balineses transmigrantes plantaron significativamente más árboles de leucaena (7,500 vs. 2,354 árboles) y criaron más toros de engorde (7.8 vs. 3.7 cabezas/hogar) que los agricultores locales de Sumbawan. La mayoría de los agricultores Balineses habían estado practicando por tres décadas sistemas de alimentación con leucaena desde que emigraron a Sumbawa. La mayoría de los productores que utilizan leucaena (74 %) habían observado síntomas de enfermedades asociadas con la toxicidad de la leucaena en su ganado, como pérdida de pelaje y salivación. Pocos granjeros que alimentaban con leucaena a vacas reproductoras (5 %) reportaron casos de falla reproductiva. Casi todos los agricultores que no usaban leucaena (93 %) informaron síntomas de enfermedades asociadas con la toxicidad de las plantas (entre otras posibles causas), más comúnmente lesiones en la piel, diarrea, cataratas y apatía. Se concluyó que el uso prioritario de la leucaena en Sumbawa era para el engorde de ganado más que para la cría de ganado. Leucaena permite que los pequeños agricultores de Sunbawa tengan empresas ganaderas más intensivas, productivas y generadoras de ingresos, pero quedan dudas sobre si debería usarse para alimentar a las vacas reproductoras. Palabras clave: Empresas de cría, usuarios de leucaena, toxicidad. Introduction promoted programs to increase the number of breeding cattle so as to improve domestic self-sufficiency in beef Leucaena (Leucaena leucocephala) is an edible production. The West Nusa Tenggara (NTB) Province forage tree legume widely used as feed for cattle and (comprising Lombok and Sumbawa Islands) in eastern other ruminants (buffalo, sheep and goats) in tropical Indonesia has been identified as an area with high potential and subtropical areas (Dalzell et al. 2012), especially for increasing the production of beef cattle in Indonesia. Indonesia. It provides high nutritive value fodder for However, the majority of beef cattle in NTB are kept by cattle. It is highly palatable, fast growing, able to be smallholder farmers with small-scale cattle ownership harvested for up to 40 years (Shelton and Dalzell 2007) (5-10 head/household) and managed under a traditional and a potential source of timber and firewood (Shelton cut-and-carry feeding system with poor nutritional feed, and Brewbaker 1994). The use of this tree legume can such as crop residues (rice straw and corn stover) or free- increase the growth rate by up to 0.83 kg live weight/ grazing on low-quality native pastures. Consequently, day for Bali cattle (Panjaitan et al. 2014) and improve efficiency of cattle production is low, with low calving the meat quality (Dahlanuddin et al. 2019) of fattening rates (~65 %), high calf mortality (10–20 %) and low bulls. It also has the potential to increase body condition growth rates (0.15–0.25 kg live weight/d) (Dahlanuddin score (BCS) and milk production when fed to breeding et al. 2019). Therefore, improving the supply and quality cows (Dahlanuddin et al. 2016). However, leucaena is of feed with forage tree legumes, such as leucaena, is not currently recommended as a feed for breeding cattle a promising strategy to overcome these productivity due to its concentration of the toxin mimosine and its problems. The adoption of leucaena-based feeding metabolites in all parts of the plant (leaves, pods, seeds systems for cattle in Sumbawa Island is increasing, with and bark) (Hegarty et al. 1964). These recommendations recent reports of more than 2,500 smallholder farmers for breeders are based on a limited body of reported using leucaena for feeding cattle (Dahlanuddin et al. evidence from farmers in Australia and other countries, 2019). suggesting that leucaena toxins could reduce reproductive Most studies on leucaena-based cattle production performance (Holmes 1980; Holmes et al. 1981; Jones et systems in NTB (Panjaitan et al. 2014; Dahlanuddin et al. 1989). al. 2017; 2018) and globally (Buck et al. 2019; Pachas et The demand for beef in Indonesia cannot currently be al. 2019) focus on the use of leucaena for fattening cattle met by local cattle, therefore, cattle and beef are imported only. This emphasis has arisen because of the profit and (Deblitz et al. 2011). The Indonesian Government has income implications of increased growth rate and sale Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) 224 F. Irawan, Dahlanuddin, M.J. Halliday, R.S. Hegarty and F.C. Cowley weight of fattening cattle (Cowley et al. 2020). There made by letter and phone call to all cattle farmers from are knowledge gaps remaining as to whether leucaena these databases in the study districts. Farmers who should also be used as feed for breeding animals and responded positively to the invitation to participate whether leucaena affects reproductive performance (112 leucaena-using farmers and 54 non-leucaena-using when fed to breeding bulls and cows. There is little farmers) were visited at their homes for a face-to-face information regarding the utilisation of this legume for interview. Participating farmers were classified by water breeding cattle by leucaena-using farmers in Indonesia management, either irrigated lands (2 or 3 crops per or elsewhere. Providing this information is important to year in irrigated systems) or rainfed areas (1 crop per improve understanding regarding feeding leucaena to year in rainfed systems) and by ethnicity (either local breeding cattle safely. Sumbawanese or transmigrant Balinese). This study aimed to understand the role of leucaena in NTB smallholder mixed breeding/fattening cattle Questionnaire and survey implementation enterprises, the use and effects of leucaena in cow-calf production systems and incidence, knowledge and farmer Questionnaires were developed for leucaena-using management strategies for alleviating leucaena toxicity farmers and for non-leucaena-using farmers. The issues, with specific attention to effects on cattle breeding questionnaires consisted of multiple-choice and Yes/No and reproduction. As a comparison, non-leucaena-using questions where a particular response was expected, and farmers were surveyed to provide a control reference to open-ended questions where the farmer could respond determine whether problems experienced by leucaena- freely. Leucaena-using farmers were asked questions on using farmers are attributable to the use of leucaena. the following topics: 1. Farm scale (e.g. property size, number of cattle Materials and Methods owned); 2. Growing leucaena (e.g. area of leucaena planted, Farmer recruitment number and cultivar of leucaena trees planted on their property); A survey was conducted from December 2019 to 3. Cattle management and leucaena use for their March 2020 in Sumbawa and West Sumbawa districts, cattle (when, why and how they feed leucaena to Sumbawa Island, NTB Province, Indonesia (Figure 1), their herd); through face-to-face interviews of leucaena-using and 4. Farmer observations of symptoms of leucaena non-leucaena-using smallholder cattle farmers. The toxicity, e.g. hair loss, skin lesions (Jones et al. survey methodology and questionnaire were approved 1978) salivation (Megarrity and Jones 1983), and by the Human Research Ethics Committee, University reduced reproductive performance of breeding of New England (HE19-040). cow/bull (Holmes 1980; Holmes et al. 1981); Farmer participants and survey areas were selected 5. Farmer knowledge of leucaena toxicity and its with a purposive sampling method. The criteria for prevention; inviting leucaena-using farmers from these districts to 6. Supplementary feeding; and participate in the research were that they had been feeding 7. Cattle performance. leucaena to fattening and breeding cattle for more than 1 Non-leucaena-using farmers were asked questions 1, year. Non-leucaena-using farmers from the same districts 4, 5, 6 and 7 above, but were not asked questions specific were invited to participate in the research on the proviso to growing and feeding leucaena to their cattle. that they had been raising either fattening or breeding The interviews were conducted in Indonesian by cattle for more than 1 year, had never fed leucaena to Indonesian enumerators and each took approximately 15 their cattle, and there was no leucaena growing near their minutes to complete. An information sheet was presented property accessible to their grazing cattle. to target farmers and a consent form was signed before Farmer recruitment to the survey was facilitated the interview commenced. by the Consortium for Large Ruminant Research - University of Mataram (for leucaena-using farmers), and Statistical Analyses Dinas Peternakan (Department of Animal Husbandry) NTB (for non-leucaena-using farmers), using their own All statistical analyses were conducted in R (R Core farmer records. Initial contact and recruitment were Team 2020). Chi-square tests were used for testing Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) Leucaena for breeding and fattening cattle 225 Figure 1. Map of Sumbawa Island, NTB Province of Indonesia showing the survey locations. rainfed; irrigated land (Google 2020). equality of proportions between groups. T-tests were and riversides or bought from other farmers. They also used for comparison of the means between groups. provided other feedstuffs such as Gliricidia (Gliricidia Univariate linear regressions were used to analyze sepium), crop residues and by-products (rice straw, corn correlations between continuous variables. stover, peanut haulms and rice bran) to supplement the diet when leucaena was not fed as 100 % of the diet, Results especially for non-fattening cattle classes, and during the dry season. Conversely, all non-leucaena-using Farmers interviewed farmers managed their cattle extensively with a low- input system. One hundred and sixty-six farmers were interviewed in this survey. Sixty-seven percent (112 of 166) of farmers practiced leucaena feeding in their cattle production system, while the remaining farmers were not using leucaena, instead raising cattle under the traditional cut-and-carry feeding system, with grasses and crop residues, or free-grazing native pastures. Leucaena was fed as a cut-and-carry fodder (Figure 2). Fifty-one percent (57 of 112) of leucaena-using farmers and 27 % (15 out of 54) of non-leucaena-using farmers were categorized as wetland farmers, using irrigation for crop production. The majority of leucaena-using farmers a. Leucaena harvested by a farmer. interviewed (81 %) were local Sumbawanese farmers, and the remainder (19 %) identified as Balinese ethnicity. All non-leucaena-using farmers were Sumbawanese. Leucaena users vs non-leucaena users There were substantial differences in the systems of cattle management by the 2 farmer groups. All leucaena- using farmers managed their cattle intensively with a cut- and-carry feeding system (Figure 2). With this system, cattle were kept in a pen or cattle house (kandang) and hand-fed fresh leucaena (leaves, small branches and b. Leucaena being fed to fattening bulls. pods) harvested from the farmer’s own land, roadsides Figure 2. Leucaena feeding systems for cattle in Sumbawa. Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) 226 F. Irawan, Dahlanuddin, M.J. Halliday, R.S. Hegarty and F.C. Cowley Leucaena-using and non-leucaena-using farmers of bulls fattened. The majority of leucaena-using farmers had a similar (P>0.05) land size (Table 1). On average, (71 %) fattened bulls, with an average of 4.7 ± 0.8 fattening land ownership was 2 ha/household with cattle, rice and bulls/farmer at the time of interview (Table 1). On average, maize as the primary enterprises. Typically, about half of each leucaena farmer could fatten 8.9 ± 10.0 bulls/year leucaena-growing farmers’ land (0.9 ± 1.0 ha/household) (range 2–60 head) under the leucaena feeding system. In was planted to leucaena (Table 1). Most of these farmers contrast, only 12 % of non-leucaena-using farmers kept initially started planting leucaena between 2010 and 2019. bulls for fattening, with an average of 2.9 ± 0.3 bulls per The number of leucaena trees planted per farm ranged year for non-leucaena-using farmers (Table 1). from 100 to 50,000 trees (Table 1), mostly of cultivar ‘Tarramba’. The majority (81 %) of non-leucaena-using Leucaena feeding systems in different land types farmers were unaware that leucaena could be fed to cattle and had never obtained any information from local Total area of land of leucaena-using farmers in irrigated extension services about its benefits in cattle production areas was not different from that of leucaena-using systems. The remaining non-leucaena-using farmers farmers in the rainfed area (Table 2), means being 2.2 and said that the animals did not like to eat leucaena leaves. 2.0 ha/household, respectively. However, in the irrigated At the time of the interview, leucaena-using farmers area, leucaena-using farmers allocated 1.3 ± 0.2 ha had cattle herds 50 % larger than non-leucaena-using (65 %) of their land for leucaena planting, whereas in the farmers (9.1 ± 0.7 vs 6.1 ± 0.5 total cattle/household, rainfed area only 0.7 ± 0.1 ha (35 %) of land was allocated respectively) (Table 1). However, 92 % of leucaena-using for leucaena planting (P<0.05). However, farmers in the farmers and all non-leucaena-using farmers had similar rainfed area planted more (P<0.05) leucaena trees (4,500 numbers of breeding cows (3.8 ± 0.2 vs 3.3 ± 0.2 cows/ trees ± 1,151/household) than the farmers in the irrigated household) (Table 1). Thus, the main difference between area (1,984 ± 595 trees/household). Leucaena-using the groups in the cattle production system was the number farmers typically split their land between broadacre crop Table 1. Farm-scale comparison between leucaena users and non-leucaena users at research sites in Sumbawa Island, NTB, Indonesia, at the time of the interviews. Items Leucaena users (n=112) Non-leucaena users (n=54) P-value Mean ± s.e. n Mean ± s.e. n Land size (ha) 1.9 ± 0.2 110 2.0 ± 0.2 50 0.34 Total number of cattle currently raised (head) 9.1 ± 0.7 112 6.1 ± 0.5 54 0.0003 Cattle production systems Fattening bulls (head) 4.7 ± 0.8 80 2.9 ± 0.3 7 0.003 Breeding cows (head) 3.8 ± 0.2 104 3.3 ± 0.2 54 0.141 Breeding bulls (head) 1.0 ± 0.0 12 1.8 ± 0.4 5 0.04 Growers (head) 1.2 ± 0.1 38 1.3 ± 0.1 21 0.217 Table 2. The comparison of farming scale between leucaena-using farmers in irrigated and rainfed areas. Items Irrigated Rainfed P-value Mean ± s.e. n Mean ± s.e. n Land ownership (ha) 2.2 ± 0.2 55 2.0 ± 0.1 57 0.166 Land planted to leucaena (ha) 1.3 ± 0.2 50 0.7 ± 0.1 51 0.005 Number of leucaena trees planted 1,984 ± 595 51 4,500 ± 1,151 52 0.040 Total number of cattle raised currently (head) 7.7 ± 0.7 55 10.5 ± 0.9 57 0.012 Cattle production systems Fattening bulls (head) 3.5 ± 0.9 39 5.8 ± 0.5 41 0.013 Breeding cows (head) 4.0 ± 0.2 54 3.4 ± 0.3 56 0.120 Breeding bulls (head) 1.0 5 1.0 7 Growers (head) 1.5 ± 0.1 13 1.3 ± 0.1 23 0.245 Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) Leucaena for breeding and fattening cattle 227 planting areas and leucaena areas, but some farmers Rainfed farm also had more cattle (P<0.05) than integrated broadacre and leucaena crops in an alley irrigated farms with the difference lying in the number of cropping configuration (Figure 3). bulls fattened (Table 2). Meanwhile the number of cattle in other classes (breeding cows and bulls plus growers) did not differ (P>0.05) between land types. Leucaena feeding systems in different ethnic groups The use of leucaena cattle production systems differed slightly between local Sumbawanese farmers and transmigrant Balinese farmers, particularly in the number of fattening bulls and the number of leucaena trees planted (Table 3). On average, Balinese farmers held more fattening bulls at the time of the interview (7.8 ± 1.9 bulls/household) than local farmers (3.7 ± 0.4 bulls/ household). The maximum number of bulls able to be a. Farmer’s land planted with leucaena. fattened in a fattening period (the duration of bulls being fattened from purchase to sale) by a Balinese farmer was 36 head. The greater cattle fattening focus by Balinese farmers was supported by increased leucaena plantings (7,500 trees/household) compared with Sumbawanese farmers (2,354 trees/household). On average, Balinese farmers had 7.3 years (range 3–15 years) of experience using leucaena in cattle production, especially for fattening bulls. In contrast, most Sumbawanese farmers were new to the leucaena feeding system, with an average of 3.8 years (range 1–5 years) experience, where they were previously practicing b. Leucaena integration with peanuts. a traditional cut-and-carry feeding system using poor Figure 3. Leucaena planting systems for cattle in Sumbawa. nutritional quality feed sources. Table 3. Comparison of farming scale between leucaena-using local Sumbawanese farmers and transmigrant Balinese farmers in Sumbawa Island. Items Local Balinese P-value Mean ± s.e. n Mean ± s.e. n Farming area Land ownership (ha) 2.1 ± 0.1 91 2.0 ± 0.2 21 0.46 Land area of leucaena planted (ha) 1.0 ± 0.1 82 0.9 ± 0.3 19 0.45 Number of leucaena planted (tree) 2,354 82 7,500 19 0.04 Number of cattle currently held (head) 8.6 ± 0.6 91 11.5 ± 2.0 21 0.05 Number of cattle fed leucaena per year (head) 7.2 ± 1.3 87 16.3 ± 1.4 19 0.01 Experience with leucaena feeding system (years) 3.8 ± 0.1 91 7.3 ± 0.9 21 0.00 % Irrigated land farmer 60 55 0 0 0.00 Cattle production systems Fattening bulls (head) 3.7 ± 0.4 62 7.8 ± 1.9 18 0.02 Breeding cows 3.7 ± 0.3 90 3.6 ± 0.4 20 0.44 Breeding bulls 1.0 12 1.0 1 Growers 1.2 ± 0.2 35 1.0 ± 0.1 2 0.00 Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) 228 F. Irawan, Dahlanuddin, M.J. Halliday, R.S. Hegarty and F.C. Cowley Seasonal leucaena feeding and significantly (P<0.05) from 7 % to a peak of 80 % in September-October, and the percentage of farmers There was consistently an average of 50 ± 10 % of feeding leucaena to growers increased from 3 % to 45 % leucaena-using farmers feeding leucaena to fattening in September-October (P<0.05). bulls in most months, even in the wet season (Figure 4a). In comparison, there was no change (P=0.12) in the The exception was in the early wet season (December) percentage of farmers who fed leucaena to fattening when leucaena-using farmers supplemented fattening bulls between the wet and the dry seasons. In addition to cattle diets with crop residues and other tree legumes. filling seasonal feed gaps, some leucaena-using farmers During the dry season, farmers reported that supplies of fed leucaena to breeding cows during the lactation leucaena decreased because they had to feed leucaena period, expecting their cows would be able to produce to other cattle classes in addition to fattening cattle at more milk for their calves. the time when other forages were not available (Figure 4 b–d). During the wet season, leucaena-using farmers fed Incidence of leucaena toxicity breeding cows and grower bulls fed stock crop residues, native grasses and other forage tree legumes (Gliricidia There was a significant difference between leucaena sepium and Sesbania grandiflora). and non-leucaena-using farmers regarding knowledge For breeding and growing cattle, fewer leucaena- of leucaena toxicity (Table 4). The majority of both using farmers fed leucaena to these animals year-round leucaena-using farmers and non-leucaena-using (31 ± 34 % and 17 ± 19 %, for breeding cows and grower farmers (75 and 93 %, respectively) reported they had bulls, respectively (Figure 4). The majority (59 %) of no knowledge of leucaena toxicity, despite the majority leucaena-using farmers fed leucaena to breeding cows of leucaena-using farmers having had long experience and growers during the dry season (between July and with leucaena. These farmers were unfamiliar with the October), when other forages were not available (Figure terms “plant toxicity” and many expressed surprise 5). During this period the percentage of farmers feeding when asked questions about this topic, as their herds leucaena to breeding cows increased progressively had shown good performance when fed leucaena. Mean: 50.1 Mean: 31.2 SD: 10.3 SD: 34.3 Mean: 17.4 Mean: 5.9 SD: 19.3 SD: 3.8 Figure 4. Percentage of leucaena-using farmers reporting leucaena use for each class of cattle in each month; (a) Fattening bulls, (b) Breeding cows, (c) Breeding bulls, (d) Grower bulls. Bar ( ) Percent of farmers; line (▬) Average rainfall. Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) Leucaena for breeding and fattening cattle 229 Eighteen percent of leucaena-using farmers and the traditional system (8.0 ± 3.1 months); however, participant remaining 7 % of non-leucaena-using farmers reported response rates to this question were low. There was a that they knew a little about leucaena toxicity from positive relationship between the numbers of leucaena vets and extension staff. Very few farmers (7 % of trees planted and the number of fattening cattle raised leucaena-using farmers and no non-leucaena-using (Figure 6). Farmers reported that the inter-calving interval farmers) reported that they had good knowledge of of breeding cows fed leucaena was 11.8 ± 0.3 months, leucaena toxicity and its prevention. whereas the inter-calving interval of cows managed under a traditional feeding system was 12.1 ± 0.3 months. Table 4. Farmer knowledge of leucaena toxicity and its prevention. Table 5. Symptoms of plant toxicities observed by leucaena- Knowledge of leucaena LU1 NLU2 Chi-Square using and non-leucaena-using cattle farmers in Sumbawa. toxicity and its prevention (n=112) (n=54) P-value Item LU1 (n=112) NLU2 (n=54) Good knowledge 7 0 0.008 Farmers reporting symptoms 74 15 Little knowledge 18 7 0.028 P<0.05 No knowledge/ awareness 75 93 0.165 Toxicity symptoms reported (n = 115) (n = 14) Hair loss 43 7 1LU= Leucaena users (%); 2NLU= Non-leucaena users (%). Skin lesion 1 14 Salivation 37 0 Unexpected performance 2 0 Cataract 0 36 Reproductive failure 5 0 Unexplained cattle death 4 0 Chronic diarrhoea 5 14 Listless 2 29 1LU= Leucaena users (%); 2NLU= Non-leucaena users (%). Figure 5. Lack of forage during the dry season in Sumbawa. Grasses and herbaceous forages are limited during this time, but trees, including forage tree legumes such as leucaena, retain green growth. P < 0.05 Despite having no knowledge of toxicity, when prompted the majority (74 %) of leucaena-using farmers agreed that they had observed symptoms associated with leucaena toxicity in their herd. The most common signs observed were alopecia (43 %) and excess salivation P = 0.00 (37 %) (Table 5). These symptoms occurred more frequently in newly purchased cattle naïve to leucaena, but disappeared within 2–3 weeks. Some leucaena- using farmers reported abortions and stillbirths in breeding cows fed leucaena. There were no reports of Figure 6. Relationship between numbers of leucaena trees these reproductive problems among non-leucaena-using planted and fattening bull ( ) and breeding cow ( ) herd size. farmers (Table 5), indicating that leucaena may have a negative side effect on breeding cows. Discussion Cattle performance Leucaena users vs non-leucaena users Leucaena-using farmers reported that their bulls could In general, leucaena-using and non-leucaena-using be fattened to finishing weight in 5.9 ± 0.4 months on farmers had a comparable farm size with cattle, rice and average, which is much faster than bulls fattened under the maize being the primary enterprises. This finding is in Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) 230 F. Irawan, Dahlanuddin, M.J. Halliday, R.S. Hegarty and F.C. Cowley agreement with data previously reported by Hilmiati et systems. According to Hilmiati et al. (2019) farmers can al. (2017) and Hilmiati et al. (2017; 2019) regarding land potentially earn profits up to IDR 21 million per year from area and usage of farmers in Sumbawa. The differences in leucaena-based fattening systems, compared with around area planted to leucaena on irrigated and rainfed farms is IDR 3 to 4 million per year from cropping systems. partly attributed to integration of limited numbers of trees Research in East Java smallholder cattle production in strips within cropping land on irrigated farms while systems found that use of leucaena in weaned Bali cattle trees were planted at high density in the absence of crops diets increased farmer by more than 100 % compared on rainfed farms. Leucaena cultivar ‘Tarramba’ was the with income on farms without leucaena (IDR 1,914,336 common cultivar planted by the farmers, and shows that vs IDR 704,076 gaps/head/period) (Priyanti et al. 2010). the introduction and development of this cultivar has been With the leucaena feeding system, farmers were able highly successful in eastern Indonesia since it provides a to keep more cattle than the non-leucaena feeding system, large quantity of high quality forage for cattle (Nulik et al. with fattening bulls forming the majority of the household 2019) during both wet and dry seasons (Sutaryono et al. herd. The greater number of fattening bulls raised by 2019). Tarramba leucaena has the advantages of greater leucaena-using farmers aligns with cattle raising being biomass production and increased psyllid tolerance and a primary income source for these farmers. Waldron drought tolerance (Nulik et al. 2013) and is preferred by et al. (2016) found that increasing the number of bulls cattle in Indonesia over other varieties (Nulik et al. 2019). fattened increased household income significantly due to In contrast, non-leucaena-using farmers use their land an increase in the revenue from extra cattle sold, whilst only for cropping activities, which are their primary the marginal cost to labour and feed remained low. For income source, while planting leucaena was new to example, increasing the number of fattening bulls from them. Most of them had no knowledge of the benefits 4 to 5 head increased feed demand by 20 % and labour of leucaena for cattle production systems. This suggests costs by 10 %, while farmer income increased from IDR that, despite numerous local and internally-funded 61,463 to IDR 77,848/day (Waldron et al. 2016). These projects focussed on leucaena integration in NTB cattle results are supported by other research showing that bull production systems over several decades, there is still fattening is the most common enterprise in intensive work to be done in extending this information to farmers. cattle production systems in NTB, where cattle are put in Deficits in farmer knowledge are not limited to leucaena the pen year-round with a cut-and-carry feeding system, specifically. Dahlanuddin et al. (2019) reported that many while breeding and backgrounding are more common smallholders in Sumbawa had little knowledge of the enterprises for traditional cattle production systems nutritional needs of the animals and did not understand (Hilmiati et al. 2019). the nutritional differences of the various feed sources This means that the inclusion of leucaena into cattle available to them. For example, they were unaware of the feeding systems offers great potential to improve the nutritional superiority of leucaena over other feedstuffs, cattle productivity and income of smallholder farmers. such as native grass, rice straw and maize stover. Therefore, expanding current leucaena extension and The low-input farming practices of the non-leucaena- adoption programs may be of benefit for farmers who using farmers coincided with cattle being a secondary currently do not use leucaena. According to Dahlanuddin income source for these farmers, to be sold whenever et al. (2019), the most effective aspect in the NTB they required extra money. These farming systems leucaena extension strategy was the establishment of require significantly lower labour input than the cut-and- on-farm demonstration sites, which are used for farmer carry leucaena farming systems. Given the low levels of cross-visits to teach best management practices (such awareness of leucaena among non-leucaena-using farmers, as leucaena establishment, harvesting and feeding it is possible that cattle remained a secondary income strategies) to non-leucaena-using farmers. source for these farmers because they lacked awareness of Kana Hau et al. (2014) identified several barriers leucaena as a high-quality and quantity, locally available to adopting intensive cattle management systems with feed source. Therefore, their cattle productivity remained leucaena feeding in eastern Indonesia that need to be low because of reliance on low quality native pastures and overcome for further extension of the system. These crop residues. It is also possible that these farmers chose include: to keep their farming enterprise (and returns) focussed 1. The ready availability of communal grazing on crop production, and preferred not to commit the time areas for cattle, so that farmers do not need to required by intensive cut-and-carry leucaena feeding allocate resources, labour, pens and feed; Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) Leucaena for breeding and fattening cattle 231 2. Farmers preferring to let the cattle graze in transmigrant Balinese farmers, particularly in number communal grazing areas, roadside or hills with a of fattening bulls and number of leucaena trees planted. herder, perceiving that they are safer from thieves Hilmiati et al. (2019) reported that a hectare of planted than when tethered or put in a pen; leucaena was able to feed 8–10 head of fattening bulls 3. Farmer beliefs that newly planted leucaena will during the wet season and 4–5 head of fattening bulls be ruined by free-grazing animals (cattle, goats during the dry season. With these leucaena resources, and buffalo); and the transmigrant farmers are able to keep at least twice 4. Limitations of skills and knowledge of farmers as many cattle annually as local farmers, showing that regarding tree establishment and limited access leucaena plays an essential role in increasing the income to leucaena planting material. of Balinese farmers. Indeed, one Balinese farmer had a Our results demonstrate that lack of farmer knowledge herd size of 60 head. Further, all Balinese farmers being of animal nutrition and awareness of the benefits of interviewed in this survey were located in rainfed areas feeding leucaena are also significant barriers to adoption with 1 cropping season, so they were reliant on cattle fattening as the main income source to support their Leucaena feeding systems in different land types families. The Balinese transmigrant farmers also had long In general, the total land area of leucaena-using farmers experience with leucaena feeding systems, having been in the irrigated area was no different from those of practicing this feeding system since they migrated leucaena-using farmers in the rainfed area. Differences to Sumbawa utilizing their previous experience of in number of leucaena trees planted is because most feeding leucaena to cattle in Bali (Dahlanuddin et al. farmers in rainfed areas used the land allocated to plant 2019). Meanwhile, the local Sumbawanese farmers are only leucaena at a high density, while the irrigated land relatively new to the leucaena feeding system, having farmers use their land to plant leucaena integrated with previously practiced traditional cut-and-carry and crops, e.g. maize, peanuts and beans, in alley cropping free-grazing feeding systems with poor nutritional systems. The integration of forage legumes with crops feed resources. Balinese farmers were one of the main in the more secure cropping land is one of the best drivers of the adoption of leucaena feeding systems strategies to inspire farmers to grow and provide high- by more than 1,000 local farmers in Sumbawa, who quality feed for cattle (Supriyadi et al. 2014). With this observed and replicated the successes of Balinese system, farmers can earn greater profits from combined farmers in fattening cattle under the leucaena feeding fattening cattle and crops (maize grain and cassava), up system (Dahlanuddin et al. 2019). to IDR 107 million/yr, compared with cropping only (IDR 43 million/yr) (Supriyadi et al. 2014). Role of leucaena in cattle production systems in NTB These results suggest that leucaena-using farmers in the rainfed area are focused on increasing feed resources Sumbawa Island is a tropical area where the average for cattle production. Subsequently, they are able to annual rainfall is 1,466 mm, of which 85 % falls during keep more cattle and to increase their income. These the wet season from November to May (Figure 4) farmers have only 1 cropping season each year (BPS- (BPS-Sumbawa 2020). The seasonal rainfall affects Sumbawa 2020) and are therefore more reliant on other the capacity of traditional smallholder farmers to income sources than farmers in the irrigated area, who provide feed for their herds since the production and the are able to triple-crop. Overall, the intensity of leucaena availability of grass and herbaceous forages fluctuates plantings, total cattle herd size and number of fattening during the year (Sutaryono et al. 2019). In the dry season, cattle suggest that rainfed leucaena-using farmers were forages other than leucaena or other tree legumes are more focused on cattle production and more reliant on rare (Figure 5). The herd sizes reported in the present leucaena to support this than the irrigated land leucaena- research were recorded during the wet season, and using farmers. so possibly may represent an annual peak in cattle ownership, if the dry season prompts farmers to reduce Leucaena feeding systems in different ethnic groups herd size by selling off fattening cattle. However, our results suggest that the leucaena-using farmers were less The use of leucaena cattle production systems differed affected by the seasonality of rainfall, since leucaena slightly between local Sumbawanese farmers and was always available for cattle during the year. Most Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) 232 F. Irawan, Dahlanuddin, M.J. Halliday, R.S. Hegarty and F.C. Cowley leucaena was used for fattening cattle only, while other gross margins for smallholders (Cowley et al. 2020). This cattle classes were fed leucaena only during the dry current research did not capture data on growth rates, season. The present survey asked farmers only whether sale weights or prices of fattening cattle. Nevertheless, they fed leucaena or not each month; however, Panjaitan some evidence shows that fattening bulls under leucaena et al. (2014) recorded the proportion of leucaena fed diets can double the growth rate compared with bulls in the diet on a seasonal basis, and reported that the fattened under the traditional system. Panjaitan et al. greatest proportion of leucaena (up to 100 %) in the diet (2014) reported that the overall average daily gain for fattening cattle occurred at the end of the wet season, (ADG) recorded for Bali bulls fattened with leucaena in while the lowest proportion of leucaena in the diet of Jatisari Village in Sumbawa was 0.42 kg/d. Similarly, fattening cattle (50 %) occurred in the dry season. Dahlanuddin et al. (2014) with a controlled experiment Evidence from this survey that some farmers feed reported that the ADG of Bali cattle fed dried leucaena leucaena to breeding cows at specific stages of the was greater (0.47 kg/d) than that of Bali bulls fed native breeding cycle shows that some smallholder farmers in grass only (0.2 kg/d). As a comparison, in Australia, the Sumbawa have an appreciation of changing nutritional ADG of steers grazing leucaena pastures was higher requirements with breeding cycles, and of how leucaena than those of steers grazing grass pasture only (0.48 use can support these. Dahlanuddin et al. (2016) vs 0.06 kg/d) (Rolfe et al. 2019). By increasing growth reported that feeding leucaena to cows increased body rates and minimising the proportion of dietary energy condition score and milk yield compared with cows fed going to maintenance requirements, the gross margin King grass only (2.1 ± 0.1 vs. 1.0 ± 0.1 kg milk/day). of smallholder forage tree legume fattening systems is Feeding a high-quality forage at key points is essential to comparable with high-input concentrate feeding systems improve reproductive performance, such as improving (Cowley et al. 2020). conception rates and milk production. Improving cow Leucaena feeding has previously been reported to BCS around parturition and early lactation is crucial have a positive impact in shortening the inter-calving since it determines their reproductive performance and interval (Mayberry et al. 2015, Wirdahayati et al. 1998). overall productivity (Herd and Sprott 1986). In the present study, however, farmer reports of typical inter-calving interval did not differ between leucaena Cattle Performance users and non-leucaena users. Anecdotal farmer reports of inter-calving interval, such as those collected in Leucaena-using farmers preferred fattening rather the present survey, are likely to be less reliable than than breeding because fattening is more profitable than experimental observations of calving dates. breeding cattle. The provincial agencies of NTB also Although many farmers reported some symptoms encourage farmers to fatten cattle as a way to increase related to leucaena toxicity, performance of both smallholder incomes in rural areas. Fattening enterprises fattening and breeding enterprises was significantly under leucaena feeding systems reduced labour and improved by leucaena. This suggests that any toxicoses feed cost (Halliday 2018). The finding from the present are short-lived, and that cattle become not only adapted survey that leucaena-feeding accelerated fattening from to leucaena, but are able to be highly productive. 8.0 to 5.9 months agrees with Dahlanuddin et al. (2017) and Hilmiati et al. (2019) who reported that the mean Incidence and farmer knowledge of leucaena toxicity duration of fattening with leucaena was 5.5 months. However, Panjaitan et al. (2014) reported that Balinese Although it has many benefits for animals, leucaena farmers in Sumbawa experienced with using leucaena contains mimosine and its derivatives (3-hydroxy-4(1H)- were able to complete 3 fattening periods each year, by pyridone and 3-hydroxy-2(1H)-pyridone, commonly feeding leucaena at a rate of 80 % of the diet or more. mentioned as 3,4-DHP and 2,3-DHP, Halliday 2018) Together with the greater year-round consistency in feed which are toxic and harmful for animals if they are not supply afforded by leucaena, this means that leucaena- adapted to their consumption. In this study, the most using farmers can fatten more than twice as many common symptoms reported by leucaena-using farmers cattle per year as non-leucaena-using farmers, using the were alopecia and excess salivation, although farmers same infrastructure and labour resources. Reducing the mostly did not understand that this was caused by fattening period duration and thus fattening more bulls leucaena toxicosis, and so, the vast majority of leucaena- per year by feeding a higher quality diet results in greater using farmers reported no knowledge that leucaena was Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) Leucaena for breeding and fattening cattle 233 toxic at first questioning. Hegarty et al. (1964) reported A few farmers (11 %) reported that "saltwater" and that hair loss is one of the most common signs of toxicity "tamarind water” (the extracted water of the tamarind observed in animals fed 100 % leucaena diets. Previous fruit) were given to their herd when they observed studies have reported abortions and stillbirths in ewes any symptoms of toxicity, such as salivation. They and heifers fed a high level of leucaena (Hamilton et al. claimed this strategy was successful in overcoming the 1971; Holmes 1980), and a small number of leucaena- symptoms within 2–3 weeks, which is also the time using farmers (5 %) also reported these issues in the commonly reported for adaptation of naïve animals to present study. Jones et al. (1989) suggested that this leucaena due to up-regulation of microbial and hepatic could be caused by the anti-mitotic action of mimosine leucaena detoxification pathways (Halliday 2018). or goitrogenic action of DHP. However, most reports of Although farmers in this survey did not use any an effect in cattle stem from early, unreplicated research method of transferring rumen fluid, there is evidence on cattle with unclear leucaena inoculation or adaptation emerging that rumen microbe genera able to detoxify status (Klieve et al. 2002). leucaena are naturally endemic in many, if not most, Non-leucaena-using farmers in this study reported ruminant populations in Indonesia, and potentially symptoms of health disorders which were rare in worldwide (McSweeney et al. 2019). However, extensive leucaena-using farms such as cataracts, listlessness, detailed research on leucaena-fed bulls in Sumbawa chronic diarrhoea and skin lesions. All non-leucaena- reported high concentrations of DHP, suggesting that the using farmers relied on free-grazing of their cattle, which bulls were not degrading all DHP in the rumen (Halliday increases the risk of ingestion of toxic plants and weeds. et al. 2014). This phenomenon shows that S. jonesii alone The different suite of symptoms reported between the 2 is not able to totally protect the animals. Halliday (2018) groups suggests that if both were the result of toxicities, reported more than 97 % of DHP in such animals was these could be caused by different plant toxins. Although excreted in a conjugated form. The conjugation of many there are many possible causes of these symptoms, other xenobiotic compounds, including DHP, commonly possible diagnoses consistent with these symptoms involves the hepatic process of glucuronication and include Malignant Catarrhal Fever disease (MCF), sulfation (Lindsay et al. 1974) with much evidence of which commonly occurs in cattle and other ruminants in this occurring in leucaena-fed ruminants (Hegarty et al. Indonesia, including in NTB (Muthalib 1988; Damayanti 1979; Elliott et al. 1985; Halliday 2018). Conjugation of a 2016). There is a high incidence of MCF in NTB cattle (92 compound increases water-solubility enabling it to more cases reported per year) (Wiyono and Damayanti 2018). readily be excreted via the urine, and in the case of DHP, Regardless of the precise cause of these symptoms, it is binds to and reduces the acute toxicity of the compound. possible that pathogenic illness may be more common Recent work has demonstrated that hepatic conjugation in non-leucaena farms since those cattle are managed plays an important role in protecting cattle from DHP under a traditional system where the cattle spent most of toxicity (Halliday 2018). their time in communal grazing areas with other cattle, sheep and goats. Conclusions The recommended management strategies to prevent negative effects of leucaena toxicity on naïve Leucaena plays an important role in providing a high- ruminants’ health include the transfer of the mimosine/ quality diet for cattle in Sumbawa, eastern Indonesia, DHP-degrading bacteria (including Synergistes jonesii) to achieve better performance and to support more (Allison et al. 1992) to ruminants newly introduced intensive, productive and income-earning cattle to leucaena, either by passive transfer from adapted enterprises. However, the majority of leucaena use by ruminants (Jones 1994) or by direct inoculation of DHP- farmers was focused on fattening cattle only, while other degrading media (available as a commercial inoculum cattle classes were fed leucaena mostly as a strategic feed in Australia only) (Klieve et al. 2002) together with a resource, during the dry season and at specific pregnancy gradual increase in the amount of leucaena in diets stages. High levels of productivity in both fattening and to promote the natural upregulation of detoxification breeding cattle fed leucaena were reported (e.g. high pathways. However, none of the leucaena-using farmers growth rate, reduced fattening interval of fattening bulls with knowledge of toxicity in Sumbawa used these and inter-calving interval of breeding cows), even though recommendations and all were unfamiliar with using several symptoms of leucaena toxicity such as hair loss, feeding management to reduce the toxicity of leucaena. salivation and reproductive failure were reported by Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) 234 F. Irawan, Dahlanuddin, M.J. Halliday, R.S. Hegarty and F.C. Cowley those farmers. These findings confirm that leucaena has Dahlanuddin; Henderson B; Dizyee K; Hermansyah; Ash great potential to be used for fattening cattle. However, A. 2017. Assessing the sustainable development and several reports of the incidence of reproductive issues intensification potential of beef cattle production in among the cows of leucaena-using farmers highlight Sumbawa, Indonesia, using a system dynamics approach. that knowledge gaps remain regarding the safe feeding PLoS One 12(8):e0183365. doi: 10.1371/journal.pone.0183365 of leucaena to breeding cattle. The confirmation by Dahlanuddin; Panjaitan TS; Sofyan; Poppi DP; Quigley SP. 2018. Bali × Hissar cattle fed Leucaena leucocephala this survey of rare, but nevertheless present, reports supplemented with maize grain grew faster than Bali of abortion and stillbirth by leucaena-using farmers cattle. Proceedings of the International Symposium on compared with the absence of these symptoms on non- the Nutrition of Herbivores, Clermont-Ferrand, France, leucaena-using farms suggests larger-scale and empirical 2-6 September 2018. espace.library.uq.edu.au/view/ research is needed to determine and define the risks of UQ:267ebae feeding leucaena to breeding cattle. Dahlanuddin; Panjaitan TS; Waldron S; Halliday MJ; Ash A; Morris ST; Shelton M. 2019. Adoption of leucaena-based Acknowledgments feeding systems in Sumbawa, eastern Indonesia and its impact on cattle productivity and farm profitability. We thank Mr Dedi Supriyadi, Mr Edi Irawan, Mr Tropical Grasslands-Forrajes Tropicales 7(4):428–436. doi: 10.17138/tgft(7)428-436 Erzal, Mr Dedi Purwansyah and Ms Febri Ariyanti Dalzell SA; Burnett DJ; Dowsett JE; Forbes VE; Shelton HM. for conducting the interviews and getting the 2012. Prevalence of mimosine and DHP toxicity in cattle photographs. This work was funded by Australian grazing Leucaena leucocephala pastures in Queensland, Centre for International Agricultural Research project Australia. Animal Production Science 52(5):365–372. doi: LPS/2015/047. 10.1071/AN11236 Damayanti R. 2016. Malignant Catarrhal Fever in Indonesia References and its control strategy. WARTAZOA, Indonesian Bulletin of Animal and Veterinary Sciences 26(3):103–114. (In (Note of the editors: All hyperlinks were verified 26 August 2022). Indonesian) doi: 10.14334/wartazoa.v26i3.1391 Deblitz C; Kristedi T; Hadi PU; Triastono J; Puspadi K; Allison MJ; Mayberry WR; McSweeney CS; Stahl DA. Nasrullah. 2011. Benchmarking the beef supply chain in 1992. Synergistes jonesii, gen. nov., sp.nov.: A rumen eastern Indonesia. Final Report. Australian Centre for bacterium that degrades toxic pyridinediols. Systematic International Agricultural Research (ACIAR), Canberra, and Applied Microbiology 15(4):522–529. doi: 10.1016/ Australia. bit.ly/3ABzC8g S0723-2020(11)80111-6 Google. 2020. Map of Sumbawa Island, NTB, Indonesia. BPS-Sumbawa. 2020. Sumbawa regency in figures (Sumbawa goo.gl/maps/Nhr9kLHtNjHyNHht9 Dalam Angka). Sumbawa Besar, NTB, Indonesia. Elliott R; Norton BW; Milton JTB; Ford CW. 1985. Effects Buck S; Rolfe J; Lemin C; English B. 2019. Adoption, of molasses on mimosine metabolism in goats fed fresh profitability and future of leucaena feeding systems and dried Leucaena with barley straw. Australian Journal in Australia. Tropical Grasslands-Forrajes Tropicales of Agricultural Research 36(6):867–875. doi: 10.1071/ 7(4):303–314. doi: 10.17138/tgft(7)303-314 AR9850867 Cowley FC; Syahniar TM; Ratnawati D; Mayberry DE; Halliday MJ. 2018. Unravelling Leucaena leucocephala Marsetyo; Pamungkas D; Poppi DP. 2020. Greater toxicity: ruminant studies in eastern Indonesia and farmer investment in well-formulated diets can increase Australia. Ph.D Thesis. School of Agriculture and Food liveweight gain and smallholder gross margins from cattle Science, The University of Queensland, Australia. doi: fattening. Livestock Science 242:104297. doi: 10.1016/j. 10.14264/uql.2018.382 livsci.2020.104297 Halliday MJ; Panjaitan T; Nulik J; Dahlanuddin; Padmanabha Dahlanuddin; Yanuarianto O; Poppi DP; McLennan SR; J; McSweeney CS; Depamede S; Kana Hau D; Kurniawan; Quigley SP. 2014. Liveweight gain and feed intake of Fauzan M; Sutartha; Yuliana BT; Pakereng C; Ara P; weaned Bali cattle fed grass and tree legumes in West Liubana D; Edison RG; Shelton HM. 2014. Prevalence Nusa Tenggara, Indonesia. Animal Production Science of DHP toxicity and detection of Synergistes jonesii in 54(7):915–921. doi: 10.1071/AN13276 ruminants consuming Leucaena leucocephala in eastern Dahlanuddin; Supriyadi M; Panjaitan TS; Poppi DP; Quigley Indonesia. Tropical Grasslands-Forrajes Tropicales SP. 2016. Increased body condition during lactation 2(1):71–73. doi: 10.17138/TGFT(2)71-73 increases milk production and pre-weaning growth of Hamilton RI; Donaldson LE; Lambourne LJ. 1971. Leucaena Bali cattle. Journal of Animal Science 94(suppl 5):399– leucocephala as a feed for dairy cows. Direct effect on 400. doi: 10.2527/jam2016-0830 reproduction and residual effect on the calf and lactation. Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) Leucaena for breeding and fattening cattle 235 Australian Journal of Agricultural Research 22(4):681– culture containing Synergistes jonesii, for protecting cattle 692. doi: 10.1071/AR9710681 against mimosine and 3-hydroxy-4(1H)-pyridone toxicity Hegarty MP; Schinckel PG; Court RD. 1964. Reaction of from feeding on Leucaena leucocephala. Australian Journal sheep to the consumption of Leucaena glauca benth. of Agricultural Research 53(1):1–5. doi: 10.1071/AR00121 and to its toxic principle mimosine. Australian Journal Lindsay RH; Aboul-Enein HY; Morel D; Bowen S. 1974. of Agricultural Research 15(1):153–167. doi: 10.1071/ Synthesis and antiperoxidase activity of propylthiouracil AR9640153 derivatives and metabolites. Journal of Pharmaceutical Hegarty MP; Phong Lee C; Christie GS; Court RD; Haydock Sciences 63(9):1383–1386. doi: 10.1002/jps.2600630910 KP. 1979. The Goitrogen 3-Hydroxy-4(1H)-Pyridone, a Mayberry D; Quigley S; Cramb R; Cowley F; Poppi D; ruminaI metabolite from Leucaena leucocephala: effects McCosker K; Priyanti A. 2015. Improving the reproductive in mice and rats. Australian Journal of Biological Sciences performance of cows and performance of fattening cattle 32:27–40. doi: 10.1071/BI9790027 in low input systems of Indonesia and northern Australia. Herd DB; Sprott LR. 1986. Body condition, nutrition and Final Report LPS/2008/038. Australian Centre for reproduction of beef cows. Texas Agricultural Extension International Agricultural Research (ACIAR), Canberra, Service, College Station, TX, USA. bit.ly/3KQDBRX Australia. bit.ly/3BjZCoN Hilmiati N; Budiwiranto B; Van de Fliert E. 2017. Gender, McSweeney CS; Padmanabha J; Halliday MJ; Hubbard B; ethnicity and engagement: Uptake strategies for Dierens L; Denman SE; Shelton HM. 2019. Detection of smallholder cattle farming innovation in West Nusa Synergistes jonesii and genetic variants in ruminants from Tenggara, Indonesia. Rural Extension and Innovation different geographical locations. Tropical Grasslands- Systems Journal 13(2):61–69. bit.ly/3AFTRSo Forrajes Tropicales 7(2):154–163. doi: 10.17138/tgft(7)154-163 Hilmiati N; Dahlanuddin; Panjaitan T. 2019. Improving Megarrity RG; Jones RJ. 1983. Toxicity of Leucaena livelihood of smallholder farmer through cattle fattening leucocephala in ruminants: The effect of supplemental based on forage tree legume in the arid Sumbawa. thyroxine on goats fed on a sole diet of Leucaena. Proceedings of The 8th International Seminar on Tropical Australian Journal of Agricultural Research 34(6):791– Animal Production, Yogyakarta, Indonesia, 23–25 798. doi: 10.1071/AR9830791 September 2019. bit.ly/3Ke65od Muthalib A. 1988. Malignant catarrhal fever disease Holmes JHG. 1980. Toxicity of Leucaena leucocephala. situation in West Nusa Tenggara (NTB). In: Daniels PW; II. Reduced fertility of heifers grazing Leucaena Sudarisman; Ronohadjo P, eds. Malignant Catarrhal Fever leucocephala. Papua New Guinea Agricultural Journal in Asian Livestock. ACIAR Monograph No. 7:59–63. bit.ly/ 31(1–4):47–50. bit.ly/3cgdNSo 3ABVQqy Holmes JHG; Humphrey JD; Walton EA; O’Shea JD. 1981. Nulik J; Dahlanuddin; Kana Hau D; Pakereng C; Edison RG; Cataracts, goitre and infertility in cattle grazed on an Liubana D; Ara SP; Giles HE. 2013. Establishment of exclusive diet of Leucaena leucocephala. Australian Leucaena leucocephala cv. Tarramba in eastern Indonesia. Veterinary Journal 57(6):257–261. doi: 10.1111/j.1751- Tropical Grasslands-Forrajes Tropicales 1(1):111–113. doi: 0813.1981.tb05805.x 10.17138/TGFT(1)111-113 Jones RJ; Blunt CG; Nurnberg BI. 1978. Toxicity of Nulik J; Kana Hau D; Halliday MJ; Shelton HM. 2019. Tarramba Leucaena leucocephala. The effect of iodine and mineral leucaena: A success story for smallholder bull fattening in supplements on penned steers fed a sole diet of Leucaena. eastern Indonesia. Tropical Grasslands-Forrajes Tropicales Australian Veterinary Journal 54(8):387–392. doi: 10.1111/ 7(4):410–414. doi: 10.17138/tgft(7)410-414 j.1751-0813.1978.tb02510.x Pachas NA; Radrizzani A; Murgueitio E; Uribe F; Cadavid ÁZ; Jones RM; McLennan MW; Dowsett KF. 1989. The effect of Chará J; Ruiz TE; Escalante E; Mauricio RM; Ramírez- Leucaena leucocephala on the reproduction of beef cattle Avilés L. 2019. Establishment and management of leucaena grazing Leucaena/grass pasture. Tropical Grasslands in Latin America. Tropical Grasslands-Forrajes Tropicales 23:108–114. bit.ly/3Biyg2A 7(2):127–132. doi: 10.17138/tgft(7)127-132 Jones RJ. 1994. Management of anti-nutritive factors — with Panjaitan T; Fauzan M; Dahlanuddin; Halliday MJ; Shelton special reference to Leucaena. In: Gutteridge RC; Shelton HM. 2014. Growth of Bali bulls fattened with Leucaena HM, eds. Forage Tree Legumes in Tropical Agriculture. CAB leucocephala in Sumbawa, Eastern Indonesia. Tropical International, Wallingford, UK. p. 216–231 bit.ly/3KRvFzI Grasslands-Forrajes Tropicales 2(1):116–118. doi: Kana Hau D; Panjaitan T; Nulik J; Dahlanuddin; Van de Fliert 10.17138/TGFT(2)116-118 E. 2014. Barriers to and opportunities for the use of forage Priyanti A; Quigley SP; Marsetyo; Pamungkas D; Dahlanuddin; tree legumes in smallholder cattle fattening systems Budisantoso E; Poppi D. 2010. Economic analysis of on-farm in Eastern Indonesia. Tropical Grasslands-Forrajes feeding strategies to increase post-weaning live weight gain Tropicales 2(1):79–81. doi: 10.17138/TGFT(2)79-81 of Bali calves. Proceedings of The 5th International Seminar Klieve AV; Ouwerkerk D; Turner A; Roberton R. 2002. The on Tropical Animal Production, Yogyakarta, Indonesia, production and storage of a fermentor-grown bacterial October 19‐22, 2010. p. 702–708. bit.ly/3QWACsO Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) 236 F. Irawan, Dahlanuddin, M.J. Halliday, R.S. Hegarty and F.C. Cowley R Core Team. 2020. R: A Language and Environment for International Conference, Canberra, Australia, 8-12 Statistical Computing. R Foundation for Statistical September 2014. p. 146. Computing, Vienna, Austria. Sutaryono YA; Supriadi D; Imran; Putra RA. 2019. Seasonal Rolfe J; Keating M; Lemin C; English B; Caird R; Black growth of Leucaena leucocephala cv. Tarramba in dry land E; Perry L; Brown G; Saunders T; Saunders C; O'Brien of west Sumbawa, Indonesia. Tropical Grasslands-Forrajes D; O'Brien L. 2019. Evaluating leucaena in timbered Tropicales 7(4):465–468. doi: 10.17138/tgft(7)465-468 northern basalt country in Queensland, Australia. Waldron S; Halliday M; Shelton M; Ngongo J; Nulik J; Tropical Grasslands-Forrajes Tropicales 7(4):333–335. doi: Kusuma Putri Utami S; Panjaitan T; Dahlanuddin; Tutik 10.17138/tgft(7)333-335 Yuliana B. 2016. Economic analysis of cattle fattening Shelton HM; Brewbaker JL. 1994. Leucaena leucocephala − systems based on forage tree legume diets in Eastern the most widely used forage tree legume. In: Gutteridge Indonesia. Report for ACIAR Project Report LPS-2014- RC; Shelton HM, eds. Forage Tree Legumes in Tropical 034. Australian Centre for International Agricultural Agriculture. CAB International, Wallingford, UK. bit. Research (ACIAR), Canberra, Australia. bit.ly/3QFUtNa ly/3RqJjfK Wirdahayati RB; Femandez PT; Liem C; Bamualim A. 1998. Shelton M; Dalzell S. 2007. Production, economic and Strategies to improve beef cattle productivity in Nusa environmental benefits of Leucaena pastures. Tropical Tenggara Region, Indonesia. International Seminar on Grasslands 41:174–190. bit.ly/3KWBlby Tropical Animal Production (ISTAP). Bulletin of Animal Supriyadi M; Mullik M; Sutaryono Y; Varela C; Coimbra Science, Supplement Edition, 1998:316–322. bit.ly/3pz6Kai L; Deus P; Dahlanuddin. 2014. Tree legume - maize - Wiyono A; Damayanti R. 2018. Outbreak of malignant cassava integration to improve Bali cattle productivity catarrhal fever in Bali cattle in Lombok Island, West Nusa in Timor Leste. Harnessing the Ecology and Physiology Tenggara, Indonesia. Jurnal Veteriner 19(3):419–429. (In of Herbivores. Proceedings of the Joint ISNH/ISRP Indonesian) doi: 10.19087/jveteriner.2018.19.419. (Received for publication 25 March 2021; accepted 18 August 2022; published 30 September 2022) © 2022 Tropical Grasslands-Forrajes Tropicales is an open-access journal published by International Center for Tropical Agriculture (CIAT), in association with The Tropical Crops Genetic Resources Institute of The Chinese Academy of Tropical Agricultural Sciences (TCGRI-CATAS). This work is licensed under the Creative Commons Attribution 4.0 International (CC BY 4.0) license. Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) Tropical Grasslands-Forrajes Tropicales (2022) Vol. 10(3):237–248 237 doi: 10.17138/TGFT(10)237-248 Research Paper The impact of COVID-19 on the sustainable intensification of forage-based beef and dairy value chains in Colombia: a blessing and a curse Las bondades y condenas del COVID-19 en la intensificación sostenible de las cadenas de valor de carne y leche en Colombia STEFAN BURKART1, MANUEL DÍAZ1, KAREN ENCISO1, ANDRÉS CHARRY2, NATALIA TRIANA1, MARTÍN MENA3, JOSÉ LUIS URREA-BENÍTEZ1, IRIELETH GALLO CARO1 AND REIN VAN DER HOEK4 1Tropical Forages Program, Alliance Bioversity-CIAT, Cali, Colombia. alliancebioversityciat.org 2Food Environment & Consumer Behavior, Alliance Bioversity-CIAT, Cali, Colombia. alliancebioversityciat.org 3Tropical Forages Program, Alliance Bioversity-CIAT, Managua, Nicaragua. alliancebioversityciat.org 4Tropical Forages Program, Alliance Bioversity-CIAT, Turrialba, Costa Rica. alliancebioversityciat.org Abstract The COVID-19 pandemic has exacerbated the difficulties associated with the need to transition the cattle sector in Latin America towards achieving sustainability and created a “double crisis” of pandemic and climate change. The increasing demand for animal sourced foods and the need to address the negative environmental impacts of cattle production, including greenhouse gas emissions, biodiversity loss and deforestation, and the implications of climate change on cattle production (prolonged droughts, prolonged rainy seasons, heat stress), have placed strong emphasis on sustainable intensification of forage-based beef and dairy systems for climate change mitigation and adaptation. This is needed to meet the commitments made by many Latin American countries to reduce greenhouse gas emissions under the Paris Agreement. Through a qualitative approach, this perspective paper reviews the present and potential impacts of the COVID-19 pandemic on progress towards sustainable intensification of the Colombian cattle sector. It also outlines new opportunities for sustainable intensification in Colombia that may provide useful examples for other Latin American countries. Short-term impacts such as: (i) increased input prices, (ii) limited access to inputs, credit, and technical assistance, and (iii) reduced incomes, have limited investment in sustainable intensification along the value chains. Reduced resources for research and development funding, unavailability of skilled and experienced staff, restrictions to travel and person-to-person interactions, in tandem, have caused setbacks in the development and application of sustainable technologies and programs. This has been addressed by increased use of technology for communication but there are difficulties with the broad availability of such technologies, especially for farmers. A long-term shift of consumer demand towards more sustainable animal products is occurring and expected to continue, and this should lead to new opportunities for sustainable intensification. Keywords: Cattle, climate change, crisis management, pandemic, tropical forages. Resumen La pandemia de COVID-19 ha provocado una nueva crisis, interconectada con la ya existente crisis ambiental, que entorpece los esfuerzos del sector ganadero latinoamericano en su transición hacia la sostenibilidad. La creciente demanda de alimentos de origen animal, los impactos ambientales de la producción ganadera (emisiones de gases de efecto invernadero, pérdida de la biodiversidad, deforestación) y las consecuencias del cambio climático en la producción ganadera (períodos Correspondence: Stefan Burkart, Tropical Forages Program, Alliance of Bioversity International and CIAT, Cali, Colombia. Email: s.burkart@cgiar.org Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) 238 S. Burkart, M. Díaz, K. Enciso, A. Charry, N. Triana, M. Mena, J.L. Urrea-Benítez, I. Gallo Caro and R. Van Der Hoek prolongados de sequía o lluvias, estrés de calor) son algunos de los desafíos que enfrenta la intensificación sostenible de los sistemas de carne y leche basados en forrajes, para lograr una reducción significativa de las emisiones de gases de efecto invernadero, uno de los compromisos adquiridos por varios países latinoamericanos en los Acuerdos de París. Lo anterior también dificulta que los productores locales de dichos sistemas ganaderos accedan a las herramientas y asistencia necesaria para su adaptación al cambio climático. Mediante un enfoque cualitativo, este documento tiene como objetivo evaluar los impactos existentes y potenciales, así como las oportunidades provocadas por la pandemia del COVID-19 en la intensificación sostenible del sector ganadero en Colombia, y se espera que esta información sea útil para otros países de América Latina. Los principales hallazgos muestran que los impactos a corto plazo, como el aumento de los precios de los insumos, el acceso limitado a los mismos, al crédito, la asistencia técnica y la reducción de los ingresos limitan las inversiones en intensificación sostenible a lo largo de las cadenas de valor. La reducción de la financiación para la investigación y el desarrollo provocaron retrocesos en la aplicación de tecnologías y en los programas de sostenibilidad. En el largo plazo, se espera un cambio en la demanda de los consumidores hacia productos sostenibles, lo que generaría nuevas oportunidades para la intensificación sostenible. Palabras clave: Cambio climático, forrajes tropicales, ganadería, manejo de la pandemia. Introduction of Colombian food systems and improving resilience is one of the major challenges in the medium- to long- In 2020, COVID-19 was declared by the World Health term. This is also likely to create opportunities along Organization (WHO 2020) as a pandemic, affecting the beef and dairy value chains through, for example, humankind in an unprecedented way. Since then, formalization of processes and adaptation to changing measures have been taken, and are continuing, by consumer demands (Burkart et al. 2020). national governments to protect public health. These Sustainable intensification of the cattle sector has have included travel bans, temporary closures of public/ been part of the political agendas of Latin American private establishments, confinement of individuals and countries for at least a decade prior to the pandemic. nationwide lockdowns that have caused significant The most important drivers have been increasing economic downturns. Compared to other sectors where demand for animal sourced foods (OECD/FAO 2019), negative impacts often become visible very quickly, it negative environmental impacts of some aspects of appears that agriculture and livestock have been little cattle production (FAO 2006; FAO 2018) and effects affected (FAO 2020a; ECLAC 2020a). However, this of enteric methane emissions of ruminants on climate perception may not reflect reality in Latin America, change (FAO 2016). Sustainable intensification has since many of the effects have not been quantified due become a central element for governments to meet the to monitoring difficulties, slow updates of databases commitments made under the Paris Agreement of the and general data scarcity, or the effects have not yet United Nations Framework Convention on Climate been communicated (Burkart et al. 2020). Disruptions Change for greenhouse gas emission reductions. in agricultural value chains have been noticeable and Sustainable intensification of forage-based cattle are likely to grow over time (Burkart et al. 2020; FAO systems consists of 3 pillars (Rao et al. 2015; Rudel et 2020b). Although most of the impacts are not yet visible al. 2015): (i) genetic intensification (the development and or fully analyzed, and despite the increased cattle use of improved forages and improved animal breeds for slaughter and beef exports in some Latin American increased productivity); (ii) ecological intensification countries in 2020 (Urrego 2021; Villamil 2020; Garza (the development and application of improved farm 2020), there have also been negative impacts. These and resource management options for increased include increasing input prices and reduced consumer efficiency); and (iii) socio-economic intensification (the demand which, in the short-term, affects the livelihoods development and improvement of policies, institutions, of low-income cattle producers. Since the beef and dairy and markets to support and increase technology uptake). sector is key to food security, nutrition and livelihoods, The transition towards sustainable intensification comes it is crucial for the involved stakeholders to identify with numerous livelihood and environmental benefits, and understand present and potential negative impacts, including income generation and diversification, food and where possible implement actions to mitigate their security, climate change mitigation and adaptation, effects. The COVID-19 crisis will indeed shape the future restoration of degraded lands and biodiversity Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) COVID-19 and sustainable intensification in Colombia 239 conservation (Rao et al. 2015), which contribute in turn producers. In the discussion we provide our perspective to the achievement of several Sustainable Development on how it should be possible to mitigate negative impacts Goals (UN 2021). In Colombia, significant advances have to take advantage of opportunities, and to support the been made in the past decade, such as the establishment sector in building resilience as well as in climate change of improved forages, protein banks or silvo-pastoral adaptation and mitigation. systems (Ganadería Colombiana Sostenible 2018), market improvements (Charry-Camacho et al. 2019; Results Ruden et al. 2020) and public policies (Ministerio de Agricultura y Desarrollo Rural 2019; Ministerio de COVID-19 and sustainable intensification at the market Agricultura y Desarrollo Rural 2020a; Presidencia de and consumer level la Nación Argentina 2018; Ministerio de Agricultura y Ganadería 2011). Annex 1 provides an overview on Food consumers have been subjected to an economic the achievements for Colombia as an example for other downturn caused by COVID-19 with increasing Latin American countries, focusing on public and private unemployment rates, part-time work and salary sector initiatives, public policy, finance mechanisms and reductions. This is evident especially in countries with market developments. Despite these advances and the yet little or no social safety net and a large proportion of limited visible effects of COVID-19 on the cattle sector, employment being in informal jobs (FAO 2020a; ECLAC the pandemic is expected to have significant impacts on 2020b; OIT 2020), such as in Colombia (NielsenIQ ongoing sustainable intensification efforts, i.e., when it 2020). According to the National Department for comes to technology adoption, product differentiation, Statistics (DANE 2021a), the country’s unemployment information sharing and financing (Burkart et al. 2020). rate increased from an average of 10.5 % in 2019 to Our objective is to describe and discuss how the 15.1 % in April 2021 as a direct result of the crisis. In pandemic is affecting sustainable intensification of the May 2020, 79 % of Colombian households were already cattle sector in Colombia, putting emphasis on both experiencing financial difficulties (Kantar 2020) and it is potential negative impacts and future opportunities. projected that during 2020, the number of people living in poverty increased from 35.7 to 42.6 % and in extreme Materials and Methods poverty from 9.6 to 12.8 %, respectively (DANE 2021d). These factors have reduced purchasing power, led to a A qualitative approach that included reviewing the lower demand for beef and some dairy products, and trends in current literature (scientific and non-scientific), affected resource availability along the value chains examination of archives and databases (from national and (CONtextoganadero 2020a). international organizations, public sector, and primary In the prevailing uncertainty, consumers have lost sources such as newspaper coverage), and ensuring confidence and have been spending less income on less- personal communications with key representatives essential items (Sullivan and Amos 2020). Although food of the cattle sector (e.g. cattle producers, beef and demand is generally considered as not very elastic and to dairy companies, distributors, public sector actors, have little effect on overall food consumption, reduced researchers, input suppliers, credit providers) was applied purchasing power can often lead to substitution effects to understand the impact of the COVID-19 pandemic and changes in the food basket towards more affordable on the livestock sector in Colombia. This allowed us foods. The National Federation of Cattle Producers in to gain a general perspective on the implications of the Colombia (Fedegan) projected a partial substitution of crisis, reflecting on how the subject is addressed by key beef with more economic protein sources, such as legume actors, media, and relevant literature, providing us with grains or poultry, at least in the short- to medium-term a wider context on different perceptions, implications (CONtextoganadero 2020a). Beef consumption and cattle and responses. Data was obtained in 2020 and 2021, slaughter was estimated to have declined by ~2.5 % in the with personal communications taking place via email first quarter of 2020 (CONtextoganadero 2020b; DANE- or telephone due to the public health measures related ESAG 2020a; DANE-ESAG 2020b), and by 30–40 % in to the COVID-19 pandemic. We completed the analysis April 2020 (Manrique 2020), respectively. Although cattle to cover three topics to assess and discuss the impacts slaughter appears to have recovered slightly in the first at the levels of (i) markets and consumers; (ii) public quarter of 2021, with a growth of 2.4 % compared to the policies and the value chain framework; and (iii) primary same period in 2020, this increase was mainly related to Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) 240 S. Burkart, M. Díaz, K. Enciso, A. Charry, N. Triana, M. Mena, J.L. Urrea-Benítez, I. Gallo Caro and R. Van Der Hoek exports (DANE 2021b), suggesting a negative effect of the COVID-19 and sustainable intensification at the public pandemic on the national beef industry. The dairy industry policy and value chain level has not been affected to the same extent, likely because most alternatives (e.g., soy or almond milk) are more Financial credits are a fundamental requirement for expensive than cow milk. There has been some reduction investments in sustainable intensification but are difficult for processed dairy products, such cheese and yogurt, to access for the many already indebted smallholder which is largely related to school and restaurant shutdowns cattle producers with low-productivity farms. To mitigate (Morais 2020; Durán 2020; González Bell 2020). effects of the pandemic, the Colombian Government Before the COVID-19 pandemic, sustainable food (FINAGRO 2020) provided an emergency credit line production was rapidly developing in Colombia in in 2020 that helped cattle producers pay interest on response to a growing demand by more discerning and existing credits to reduce bankruptcies. However, this better off consumers (which still make up only a small credit line was not intended for new investments in minority of the rather price-oriented consumer base). sustainable production alternatives and thus did not These same consumers show a high willingness to pay for provide incentives to revive pre-pandemic advances. To sustainable beef products (Charry-Camacho et al. 2019) foster the adoption of sustainable production systems, that creates opportunities along the value chain (Burkart the Government launched another credit line in 2020 et al. 2021; Charry-Camacho et al. 2018). Several beef for establishing silvo-pastoral systems with emphasis and dairy companies, NGOs and research institutions on improved tree species and the implementation of have developed, or are in the process of developing, living fences. This is the first initiative of this type in differentiated beef and dairy products through certification the country (Ministerio de Agricultura y Desarrollo or technical assistance and sustainable beef labels have Rural 2020a) and it is expected to further advance efforts already been launched (Annex 1). The pandemic might regarding sustainable intensification. lead to 2 opposing effects (Trujillo 2020; Sullivan and Another important factor for the transition towards Amos 2020). First, the substitution effect could lead to a sustainable cattle systems is access to rural extension and declining demand for sustainable beef and dairy products technical assistance. Despite the ongoing transformation of in the short-term, since they are generally more expensive, the rural extension system in Colombia through the SNIA and this would negatively affect the development of more (Sistema Nacional de Innovación Agropecuaria - National sustainable value chains and the implementation of Agricultural Innovation System) law of 2017 (Gobierno de sustainable production practices at farm level. This effect Colombia 2017), the reach of public rural extension is still has already been noted by private beef and dairy industries very limited. The pandemic has aggravated this situation, and has caused delays in the development or refinement leading to several consequences: (i) implementation of of differentiation efforts. Since resources became scarce, departmental extension plans (Planes Departamentales projects have been stopped and employees terminated. A de Extensión Agropecuaria) had to be accelerated at recent study by Ramirez et al. (2021) confirms that beef the expense of effectively prioritizing land areas and consumer preferences in Colombia have been affected crops for interventions; (ii) disruptions in procedures for by the pandemic, with a decline in beef consumption, requesting extension caused a mismatch of demand and especially amongst surveyed consumers from lower- supply (Ministerio de Agricultura y Desarrollo Rural income households or those facing economic reductions 2020c); and (iii) limited on-farm extension combined due to the pandemic. Second, a more optimistic scenario with reduced access to productive inputs affected farm considers that, once incomes recover in the medium- to productivity, resulting in reduced farm incomes and long-term, consumer preferences will shift towards more credit repayment capacity. However, there have been sustainable products, improved traceability, food safety positive developments too. Virtual extension efforts and animal welfare. This second scenario would boost have increased with social media and local radio stations the development of more sustainable value chains and the (Fedegan 2020; CONtextoganadero 2020c). Also, some implementation of sustainable production practices, and Colombian dairy and beef companies, input suppliers, both encourage and require investments on the part of private agencies and supermarkets have developed their the beef and dairy industries (Sullivan and Amos 2020; own rural extension and technical assistance systems, Manrique 2020). However, for most of the Colombian and many of these are focused on sustainable production beef and dairy consumers, price will continue to play the (Annex 1). Although the impacts of the pandemic on most important role in consumption decision-making. their efforts are not yet fully evident, they include some Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) COVID-19 and sustainable intensification in Colombia 241 important setbacks. Nevertheless, on-farm visits, face-to- sustainable markets/consumption and rural extension face rural education programs (e.g., Alquería’s Heirs of (Mesa de Ganaderia Sostenible 2020), which reached Tradition Program), and farm planning and monitoring over 20,000 attendees (Burkart and Urrea-Benítez 2020). (e.g. AngusAzul, GANSO) can only partly be provided The increased use of virtuality for capacity building or are being offered virtually, at the expense of producers and rural extension has attracted an unexpected high with limited internet connectivity. number of participants, yet concerns about inclusiveness, Research institutions and networks that focus on the quality (versus quantity) and the applicability of such development of sustainable production technologies, such virtually spread information remain valid. Around 52 % as improved forages, have been facing budget reductions of the Colombian population have limited internet access (e.g. the Colombian Forages Network) and there have (Revista Semana 2020) without the required equipment been shifts of focus towards crisis mitigation rather than such as smartphones and computers. The proportion in gradual improvements (CGIAR 2020; The Alliance of rural areas, where the principal target group of cattle Bioversity International and CIAT 2020). This might producers live, is likely to be substantially higher. A lead to lesser progress in developing technologies and positive trend shows that internet use grew by 40 % reducing knowledge gaps in the mid- and long-term. during the pandemic (Morales 2020) and there were In addition, institutions in charge of implementing large increases in the number of both fixed (+300 %) and sustainable intensification piloting and scaling projects mobile access points (+1.5 million), compared to pre- are facing reductions in their budgets and staff that will pandemic levels of 2019 (Revista Semana 2021). Also, likely lead to delays in scaling-up processes. These the Colombian Government plans further increases setbacks can, however, also represent new opportunities to reach 70 % of the population by 2022 (Revista since new elements can be considered in the reshaping Semana 2020). Despite these promising trends, internet or formulation of projects, such as a stronger focus on a access is a major issue that constrains the sustainable potentially changing consumer demand. intensification of the sector. Colombia has established numerous multi-actor initiatives for supporting sustainable intensification of COVID-19 and sustainable intensification at the primary the cattle sector, such as the Colombian Roundtable for producer level Sustainable Cattle (Mesa de Ganaderia Sostenible) with its 15 regional roundtables, the Tropical Forest Alliance According to ECLAC (2020a), the pandemic-caused (TFA), which established the Zero-Deforestation downturn of the global economy, particularly in the Agreements (Acuerdos Cero Deforestación), and United States, China, and Europe, has had negative the Sustainable Colombian Cattle Project (Proyecto impacts in Latin America and the Caribbean, affecting Ganadería Colombiana Sostenible), which ended in 2019 the trade of raw materials (volumes, unpredictable price (Fedegan et al. 2020; Ganadería Colombiana Sostenible changes). In Colombia, the US Dollar-Colombian Peso 2018; Mesa de Ganadería Sostenible Colombia 2019; exchange rate increased by 12.6 % during the pandemic Burkart and Urrea-Benítez 2020; Rodriguez 2020; (DolarWeb 2021; Banco de la República 2021) (Annex Triana-Ángel and Burkart 2021). The pandemic is 2). Both the Colombian beef and dairy sectors depend on affecting these initiatives to different extents, including imported agricultural inputs, such as forage seeds, feed through reduction of budgets, limitations in reaching concentrates, vaccines, salt, minerals and machinery. target groups, extended virtuality, delays in achieving Some of these inputs, particularly forage seeds and agreements or in public policy development, forcing fertilizer, are essential for the sustainable intensification the involved actors to seek for alternatives. The of traditional cattle systems. Forage seed exports from implementation of the Zero-Deforestation Agreements, Brazil, the principal tropical forage seed producer, for example, has slowed down and the committed decreased by 27 % between March and April 2020, and companies cannot properly visit their producers to by 11 % compared to April 2019 (Legiscomex 2020). verify compliance with the established agreements. Data regarding forage seed imports into Colombia are To counteract these developments, the Colombian not yet available but it appears that there has been both Roundtable for Sustainable Cattle launched a virtual a reduced demand from the Colombian cattle sector and seminar series (from May to August 2020) on sustainable a reduced supply from Brazil. Urea fertilizer prices have cattle (Conversatorios sobre Ganadería Sostenible), on average increased by 9 % between March and April focusing on elements like climate change, biodiversity, 2020 (DANE 2020a). In 2020, urea was more expensive Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) 242 S. Burkart, M. Díaz, K. Enciso, A. Charry, N. Triana, M. Mena, J.L. Urrea-Benítez, I. Gallo Caro and R. Van Der Hoek in 35 % of the municipalities and in 2021 (until May), increase their support for sustainable intensification in 69.12 %, respectively (DANE 2020b; DANE 2021c). of the cattle sector despite the setbacks caused by the The combination of reduced input quantities and pandemic. The occasion represents a major opportunity increased prices results in additional burdens for cattle with various confluent demands of local populations, producers to access the required means for sustainable such as increasing food security and sovereignty, intensification, leading to disincentives. ensuring incomes and livelihoods of the most Owing to the effects of the pandemic, young cattle vulnerable and affected population, promoting rural producers, who are potential investors in sustainable justice and redistribution, and addressing the national production practices, perceive less incentives for and international commitments of environmental pursuing a future in the sector and thus might migrate protection, climate change mitigation and adaptation. to cities in larger numbers, aggravating the problems Positive examples for this are the new credit line for of generational transfer, already critical prior to the the establishment of silvo-pastoral systems by the pandemic (Parra Cortés and Magaña Magaña 2021). Colombian government, and the development of the Female cattle producers, who have crucial roles in the GANSO label for sustainable beef. Both initiatives were adoption of new technologies (Triana-Ángel and Burkart launched during the peak of the pandemic in 2020. 2019) and often focus on deriving specialty products Although this study focuses on Colombia, there are such as ‘branded’ cheese or yogurt, are likely to lose some similarities with other Latin American countries, market access and be discouraged if dairy products are such as Costa Rica and Argentina, not only in the being substituted with non-dairy products, leading to similarity of their cattle systems but also their advances lower household incomes and reducing the capacities for towards achieving sustainability. Other studies, although investing in sustainable intensification. not directly related to sustainable intensification, have come to similar conclusions regarding the impacts of Discussion the pandemic on the agricultural sector. For example, Hammond et al. (2022) found that in several African During the last decade, Colombia has made significant and Asian countries, the measures employed during the advances in the transition towards a more sustainable pandemic have affected food purchase, product sales beef and dairy cattle sector. The COVID-19 pandemic (volumes and prices), and access to productive inputs and has caused adverse impacts on both existing and planned markets, with a tendency towards more severe effects in initiatives but also provides some opportunities for the those countries with stricter measures. Siche (2020), in future. The critical impacts are certainly the reduction a more general analysis, found that the pandemic had of budgets in many of the involved research and strong impacts on food supply chains, mostly regarding development institutions, a temporary shift in priorities, food demand, ultimately affecting the most vulnerable reduced consumer incomes, substitution of beef and actors. Sridhar et al. (2022) conisder the adoption of dairy products with other protein sources, disruptions in sustainable agricultural practices and use of digital tools rural extension and technical assistance programs, and as suitable options for building a more resilient food reduced ongoing monitoring of on-farm interventions and system that can more easily absorb disruptions. commitments. All of these disturb the implementation of sustainable production practices. Positive effects Acknowledgments of the pandemic on sustainable intensification are an apparently increased consumer consciousness towards This work was undertaken as part of the CGIAR Research more environmentally friendly products, animal welfare, Program (CRP) on Livestock. In addition, it was supported and traceability, which can be harnessed through product by the LivestockPlus project funded by the CRP on Climate differentiation. While the shift towards virtual capacity Change, Agriculture and Food Security (CCAFS), which building and rural extension can reach larger numbers is a strategic partnership of CGIAR and Future Earth of participants, internet accessibility and connectivity and by the OneCGIAR Initiative on Livestock, Climate in rural areas remain challenging and lead to questions and System Resilience (LCSR). We thank all donors that about inclusiveness. globally support the work of the CRP programs through To preserve the achieved progresses and to take their contributions to the CGIAR system. We gratefully advantage of new opportunities, actors at different levels, acknowledge funding from the Biotechnology and e.g. public policy, market, primary producer, should Biological Sciences Research Council Project Advancing Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) COVID-19 and sustainable intensification in Colombia 243 sustainable forage-based livestock production systems por WhatsApp. Federación Nacional de Ganaderos in Colombia - CoForLife (BB/S01893X/1) and the UK (Fedegán), Bogotá, Colombia. bit.ly/2ZIic7L Research and Innovation (UKRI) Global Challenges DANE. 2020a. Informe especial-Componente de insumos y Research Fund (GCRF) GROW Colombia grant via the factores asociados a la producción agropecuaria de SIPSA. UK’s BBSRC (BB/P028098/1). Departamento Administrativo Nacional de Estadística (DANE), Bogotá, Colombia. bit.ly/3gnLb6K References DANE. 2020b. Boletín mensual insumos y factores asociados a la producción agropecuaria. Departamento Administrativo Nacional de Estadística (DANE), Bogotá, (Note of the editors: All hyperlinks were verified 1 September 2022). Colombia. bit.ly/3cuPPzp Alquería. 2020. Programa para el desarrollo de proveedores: DANE. 2021a. Tasa Global de Participación, Ocupación y Herederos de tradición [PowerPoint slide]. Alquería, Desempleo. Departamento Administrativo Nacional de Bogotá, Colombia. bit.ly/3Kzy6XK Estadística (DANE), Bogotá, Colombia. Banco de la República (Banrep). 2021. Serie Histórica de DANE. 2021b. Encuesta de Sacrificio de Ganado (ESAG). la Tasa representativa del Mercado 1991–2021. Banrep, Departamento Administrativo Nacional de Estadística Bogotá, Colombia. bit.ly/3TuIxzR (DANE), Bogotá, Colombia. bit.ly/3vaKeoF Burkart S; Díaz MF; Enciso-Valencia K; Urrea-Benítez JL; DANE. 2021c. Boletín mensual N°104. Insumos y factores de Charry-Camacho A; Triana-Ángel N. 2020. COVID-19 la producción agropecuaria. Departamento Administrativo and the bovine livestock sector in Colombia: Current and Nacional de Estadística (DANE), Bogotá, Colombia. bit.ly/ potential developments, impacts and mitigation options. 3v6PVny CIAT Publication No. 498. International Center for DANE. 2021d. Pobreza Monetaria En Colombia-Resultados Tropical Agriculture (CIAT). Cali, Colombia. hdl.handle. 2020 [PowerPoint slide]. Departamento Administrativo net/10568/108354 Nacional de Estadística (DANE), Bogotá, Colombia. bit.ly/ Burkart S; Urrea-Benítez JL. 2020. The Colombian Roundtable 3wnf1Qo for Sustainable Beef and Dairy: 2020 activities. Alliance DANE-ESAG. 2020a. Encuesta de Sacrificio de Ganado of Bioversity International and CIAT, Cali, Colombia. hdl. (ESAG) [Database]. Departamento Administrativo handle.net/10568/111099 Nacional de Estadística (DANE), Bogotá, Colombia. bit.ly/ Burkart S; Arango J; Enciso-Valencia K; Ruden A; Charry- 2yI5CdK Camacho A; Díaz MF; Gutierrez JF; Castro JP. 2021. DANE-ESAG. 2020b. Censo Sacrificio de ganado total y Business opportunities along sustainably-intensified beef departamental: vacunos, porcinos y otras especies. Marzo and dairy value chains. Business Model Brief. Alliance of 2020 [Database]. Departamento Administrativo Nacional Bioversity International and CIAT, Nairobi, Kenya. hdl. de Estadística (DANE), Bogotá, Colombia. bit.ly/2XQbOcj handle.net/10568/116579 DolarWeb. 2021. Dólar histórico en Colombia. bit.ly/3iw7NWj CGIAR. 2020. A food systems response to COVID-19. Durán C. (April 20, 2020). Ganaderos “en jaque” por el CGIAR. bit.ly/3cDGyZi Covid-19. eDairyNews. bit.ly/3AIBuL0 Charry-Camacho A; Jäger M; Enciso-Valencia K; Romero M; ECLAC. 2020a. Dimensionar los efectos del COVID-19 para Sierra L; Quintero M; Hurtado JJ; Burkart S. 2018. Cadenas pensar en la reactivación. Informe Especial COVID-19 de valor con enfoque ambiental y cero deforestación en No. 2. Economic Commission for Latin America and the la Amazonía colombiana – Oportunidades y retos para el Caribbean (ECLAC), Santiago, Chile. hdl.handle.net/11362/ mejoramiento sostenible de la competitividad regional. 45445 CIAT Políticas en Síntesis No. 41. Centro Internacional de ECLAC. 2020b. América Latina y el Caribe ante la pandemia Agricultura Tropical (CIAT), Cali, Colombia. hdl.handle. del COVID-19: efectos económicos y sociales, Informe net/10568/97203 Especial COVID-19 No. 1. Economic Commission for Charry-Camacho A; Narjes M; Enciso-Valencia K; Peters Latin America and the Caribbean (ECLAC), Santiago, M; Burkart S. 2019. Sustainable intensification of Chile. hdl.handle.net/11362/45337 beef production in Colombia - Chances for product FAO. 2006. Livestock’s long shadow. Environmental issues differentiation and price premiums. Agricultural and Food and options. Food and Agriculture Organization of the Economics 7:22. doi: 10.1186/s40100-019-0143-7 United Nations, Rome, Italy. bit.ly/3Q5A8Qb CONtextoganadero. 2020a. Así sería el efecto del COVID-19 FAO. 2016. Livestock & Climate Change. Food and para el agro. Federación Nacional de Ganaderos (Fedegán), Agriculture Organization of the United Nations, Rome, Bogotá, Colombia. bit.ly/2XBFuJN Italy. bit.ly/3eg39vh CONtextoganadero. 2020b. Cambios en los horarios de FAO. 2018. FAO's role in livestock and the environment. algunas plantas de beneficio. Federación Nacional de Food and Agriculture Organization of the United Nations, Ganaderos (Fedegán), Bogotá, Colombia. bit.ly/2Mi2OHz Rome, Italy. bit.ly/3pZwnBt CONtextoganadero. 2020c. Coregán capacita a sus afiliados FAO. 2020a. Mitigating the impacts of COVID-19 on the Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) 244 S. Burkart, M. Díaz, K. Enciso, A. Charry, N. Triana, M. Mena, J.L. Urrea-Benítez, I. Gallo Caro and R. Van Der Hoek livestock sector. Food and Agriculture Organization of the bovina sostenible. n9.cl/3ziy3 United Nations. Rome, Italy. doi: 10.4060/ca8799en Mesa de Ganaderia Sostenible. 2020. Conversatorios sobre FAO. 2020b. Nueva enfermedad por coronavirus (COVID-19). Ganadería Sostenible. Bogotá, Colombia. bit.ly/2Ad9pzY Preguntas frecuentes: pandemia del COVID-19, su Ministerio de Agricultura y Desarrollo Rural. 2019. Política impacto en la alimentación y la agricultura. Food and agropecuaria y de desarrollo rural 2018-2022: un campo Agriculture Organization of the United Nations, Rome, para la equidad. n9.cl/yjgle Italy. shar.es/aHRYip Ministerio de Agricultura y Desarrollo Rural (MADR). Fedegan. 2020. Redes sociales y radio comunitaria, alternativas 2020a. MinAgricultura Rodolfo Zea lanza nueva línea para capacitar productores (May 20, 2020). Federación para financiar los sistemas silvopastoriles MADR, Bogotá, Nacional de Ganaderos (Fedegán), Bogotá, Colombia. bit.ly/ Colombia. bit.ly/3sNo6ze 2X6euTI Ministerio de Agricultura y Desarrollo Rural (MADR). Fedegan; CIPAV; TNC; Fondo Acción. 2020. Resultados 2020b. Actualización de la Contribución Determinada Proyecto Ganadería Colombiana Sostenible [PowerPoint a Nivel Nacional de Colombia (NDC). MADR, Bogotá, slide]. Federación Nacional de Ganaderos (Fedegán), Bogotá, Colombia. bit.ly/3R66CuW Colombia. bit.ly/3Q5NUCm Ministerio de Agricultura y Desarrollo Rural (MADR). FINAGRO. 2020. Línea Especial de Crédito Colombia Agro- 2020c. Circular 002 de 2020. MADR, Bogotá, Colombia. Produce. Circular reglamentaria P-11 2020. FINAGRO. , bit.ly/3ddTkZp Bogotá, Colombia. bit.ly/2ZHdPtK Ministerio de Agricultura y Ganadería (MAG). 2011. Política Fundación Alpina. 2019. Informe de Sostenibilidad Alpina de Estado para el sector agroalimentario y el Desarrollo 2019. Fundación Alpina, Bogotá, Colombia. bit.ly/3czoek1 Rural Costarricense 2010-2021. MAG, San José, Costa Ganadería Colombiana Sostenible. 2018. ¿En qué consiste el Rica bit.ly/2QlzRjr proyecto Ganadería Colombiana Sostenible?. bit.ly/32IkHat Morais D. 2020. Cuarentena redujo el acopio de la leche. GANSO. 2020. GANSO: Creando Paisajes Productivos y eDairyNews. bit.ly/3KA3ZPU Sostenibles. Ganadería Sostenible (GANSO), Villavicencio, Morales C. 2020. Acceso a Internet está garantizado en Colombia Colombia. ganso.com.co durante pandemia: MinTic. RCN Radio bit.ly/2ZNpuXN Garza J. 2020. Ganadería es inmune a la crisis. La República. NielsenIQ. 2020. In Latin america COVID-19 will hit low bit.ly/2SuM6uV income consumers the most. bit.ly/3B0L3Xj Gobierno de Colombia. 2017. Por medio de la cual se crea el OECD/FAO. 2018. OECD-FAO Agricultural Outlook Sistema Nacional de Innovación Agropecuaria y se crean (Edition 2018). OECD Agriculture Statistics [Database]. otras disposiciones. Ley 1876 del 29 de diciembre de 2017. doi: 10.1787/d4bae583-en Diario Oficial No. 50.461. Gobierno de Colombia, Bogotá, OIT (Organización Mundial del Trabajo). 2020. Observatorio Colombia. bit.ly/2M4HKUz de la OIT: La COVID-19 y el mundo del trabajo. Tercera González Bell J. 2020. La venta de yogures ha disminuido edición. Estimaciones actualizadas y análisis. bit.ly/3Ttfnky 20% por la falta de estudiantes y loncheras. Diario La Parra Cortés RI; Magaña-Magaña MA. 2021. Características República. bit.ly/36zz8ik técnico-económicas de sistemas de producción bovina de Hammond J; Siegal K; Milner D; Elimu E; Vail T; Cathala P; las razas criollas colombianas Romosinuano y Hartón del Gatera A; Karim A; Lee J-E; Douxchamps S; Thanh Tu M; Valle. Revista MVZ Córdoba 26(2):e2079. doi: 10.21897/ Ouma E; Lukuyu B; Lutakome P; Leitner S; Wanyama I; rmvz.2079 Pham Thi T; Hong Phuc PT; Herrero M; van Wijk M. 2022. Presidencia de la Nación Argentina. 2018. Objetivos de Perceived effects of COVID-19 restrictions on smallholder desarrollo sostenible. Buenos Aires, Argentina. bit.ly/ farmers: Evidence from seven lower- and middle-income 3tlDzbb countries. Agricultural Systems 198:103367. doi: 10.1016/j. Ramirez O; Charry-Camacho A; Diaz MF; Enciso-Valencia agsy.2022.103367 K; Mejía D; Burkart S. 2021. The effects of COVID-19 Kantar. 2020. Consumer Thermometer Colombia 3° Edición, on beef consumer preferences and beliefs in Colombia: Mayo 2020. Kantar, Barcelona, España. A logit model approach. Frontiers in Sustainable Food Legiscomex. 2020. Detalle de las importaciones y Systems 5:725875. doi: 10.3389/fsufs.2021.725875 exportaciones por país [Database]. bit.ly/3Twf88n Rao IM; Peters M; Castro A; Schultze-Kraft R; White D; Londoño M. 2019. AGROSAVIA y el CIAT articulados en pro Fisher M; Miles J; Lascano C; Blümmel M; Bungenstab D; del desarrollo de una ganadería sostenible. AGROSAVIA. Tapasco J; Hyman G; Bolliger A; Paul BK; Van der Hoek bit.ly/3pWd7Vo R; Maass BL; Tiemann TT; Cuchillo M; Douxchamps Manrique F. 2020. Estas son las acciones que propone Fedegán S; Villanueva C; Rincon A; Ayarza M; Rosenstock T; para enfrentar la pospandemia. CONtextoganadero. bit.ly/ Subbarao GV; Arango J; Cardoso J; Worthington M; 2X6aeUe Chirinda N; Notenbaert A; Jenet A; Schmidt A; Vivas N; Mesa de Ganadería Sostenible Colombia. 2019. Bases técnicas Lefroy R; Fahrney K; Guimaraes EP; Tohme JM; Cook para la formulación de la política nacional de ganadería S; Herrero M; Chacón M; Searchinger T; Rudel T. 2015. Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) 245 S. Burkart, M. Díaz, K. Enciso, A. Charry, N. Triana, M. Mena, J.L. Urrea-Benítez, I. Gallo Caro and R. Van Der Hoek LivestockPlus-The sustainable intensification of forage engine for change: strategic partnerships with a private based agricultural systems to improve livelihoods and company (CIAT/CCAFS-Alqueria): The beginnings, ecosystem services in the tropics. Tropical Grasslands- incentives, and objectives of the Herederos de Tradición Forrajes Tropicales 3(2):59–82. doi: 10.17138/tgft(3)59-82 (Heirs of Tradition) initiative carried out by Alquería’s Revista Semana. 2020. Día internacional del internet: ¿Cómo farmer training programs. CCAFS Info Note. CGIAR está Colombia en conectividad? Revista Semana, Bogotá, Research Program on Climate Change, Agriculture and Colombia. bit.ly/3gzkImM Food Security (CCAFS), Cali, Colombia. hdl.handle. Revista Semana. 2021. Acceso a internet en Colombia se net/10568/106800 aceleró durante la pandemia. Revista Semana, Bogotá, Triana-Ángel N; Burkart S. 2019. Between silences and Colombia. bit.ly/3vplwRk opportunities: Gender and bovine livestock in Latin Rodriguez R. 2020. Acuerdos de Cero Deforestación con America, a state of the question. Info note. International Cadenas Productivas [PowerPoint slide]. MinAmbiente- Center for Tropical Agriculture (CIAT). Cali, Colombia. Alliance Bioversity CIAT-BID. bit.ly/3AYJcSG hdl.handle.net/10568/101292 Rudel TK; Paul BK; White D; Rao IM; Van der Hoek R; Triana-Ángel N; Ariza Aya M; Burkart S. 2020. Youth in Castro A; Boval M; Lerner A; Schneider L; Peters M. livestock and the transformative power of rural education: 2015. LivestockPlus: Forages, sustainable intensification, the case of Heirs of Tradition, 2012–2020. CCAFS Info and food security in the tropics. Ambio: a journal of the Note. CGIAR Research Program on Climate Change, human environment 44(7):685–693. doi: 10.1007/s13280- Agriculture and Food Security (CCAFS), Cali, Colombia. 015-0676-2 hdl.handle.net/10568/111012 Ruden A; Castro JP; Gutiérrez JF; Koenig S; Arango J 2020. Triana-Ángel N; Burkart S. 2021. Managing changes, GANSO: New business model and technical assistance transforming the sector: the case of the Colombian for the professionalization of sustainable livestock Roundtable for Sustainable Cattle. CCAFS Info farming in the Colombian Orinoquia region. CCAFS Info Note. CGIAR Research Program on Climate Change, Note. CGIAR Research Program on Climate Change, Agriculture and Food Security (CCAFS), Wageningen, Agriculture and Food Security (CCAFS), Cali, Colombia. The Netherlands. hdl.handle.net/10568/116074 hdl.handle.net/10568/110456 Trujillo CA. 2020. ¿Puede el confinamiento por COVID 19 Siche R. 2020. What is the impact of COVID-19 disease propiciar un estilo de vida y un consumo más sostenible? on agriculture? Scientia Agropecuaria 11(1):3–6. doi: bit.ly/2B4JCue 10.17268/sci.agropecu.2020.01.00 UN. 2021. The 17 Goals. United Nations – Department of Sridhar A; Balakrishnan A; Jacob MM; Sillanpää M; Economic and Social Affairs, Sustainable Development. Dayanandan N. 2022. Global impact of COVID-19 on New York City, NY, USA. sdgs.un.org/goals agriculture: role of sustainable agriculture and digital Urrea JL. 2018. Segundo taller de la Red Colombiana de farming. Enviromental Science and Pollution Research Forrajes CIAT-AGROSAVIA. CIAT. bit.ly/3B4cGP2 8:1–17. doi: 10.1007/s11356-022-19358-w Urrego A. 2021. Los sectores de ganadería bovina y de Sullivan R; Amos N. (May 15, 2020). How consumer producción lechera crecieron el año pasado. Agronegocios. concerns impact farm animal welfare. Innovation Forum. bit.ly/3CHYYT9 bit.ly/36CmMWI Villamil L. 2020. El coronavirus no pudo frenar el boom The Alliance of Bioversity International and CIAT. 2020. La exportador de la carne argentina. Clarín Rural. bit.ly/3vH4m2d respuesta de la Alianza frente al COVID-19. ciat.cgiar.org/ WHO (World Health Organization). 2020. WHO Director- covid-19 General’s opening remarks at the media briefing on Triana-Ángel N; Ariza Aya M. 2019. Youth in livestock, the COVID-19. https://bit.ly/2B6xj0w Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) COVID-19 and sustainable intensification in Colombia 246 Annex 1. Sustainable intensification of the Colombian cattle sector: initiatives and achievements Initiative/project and duration Level of intervention Interventions and achievements Sources Sustainable Colombian Cattle Primary producer • 4,100 cattle farms from 5 regions participated Fedegan et Project1,d (2010-2019) • 18,300 ha under conservation al. (2020); • 38,400 ha under sustainable land-use Ganadería • 4,800 ha under intensive silvo-pastoral systems Colombiana • (effect of payments for ecosystem services) Sostenible (2018) • 25% increase in milk production • 10% increase in forage digestibility and quality • 3% increase in milk quality • 25% increase in forage biomass Colombian Forages Primary producer • Development, evaluation, and release of forage materials for different Colombian agro- Urrea (2018); Network2,a (2018-present) Forage seed sector ecosystems Londoño (2019) Research institutions • Development of forage niche products • Release of various new forage materials e.g., Megathyrsus maximus cultivar ‘Sabanera’, Urochloa brizantha CIAT 26124 cultivar ‘Caporal’, Avena sativa AV25T cultivar ‘Altoandina’ • In release process: Cenchrus ciliaris Alquería-SENA rural Primary producer • 2-year technology level formal training program for young dairy producers Alquería (2020); education program Heirs of (i.e., youth) • Capacity building through farmer field schools and farm planning Triana Ángel et Tradition3,b (2012-present) • Agreement with SENA to facilitate technical skills for integrated cattle management al. (2020); Triana • Agreement with the Alliance of Bioversity International-CIAT (since 2020): trainings on Ángel and Ariza improved pasture management, sustainable dairy production, value chains and the role of Aya (2019) women in agriculture • 188 young graduates: 60 women and 129 men (graduates of the program) • Training of 3,350 farmers (mainly women and young farmers) provided on good management practices, milk quality and milking routines Fundación Alpina4,c Primary producer • 2,374 participants in productive projects aimed at strengthening technical capacities to Fundación Dairy value chain improve income and livelihoods and support the reconversion of cattle systems Alpina (2019) • Participation in the national roundtable for rural development • 400 women increased their milk sales by 16% due to training and assistance in their economic and social empowerment • 18% increase in the creation of new cooperatives Zero-Deforestation Public policy • Identification of strategies for monitoring deforestation in the cattle sector Rodriguez Agreements for Beef and Monitoring • 35 public and private sector actors already signed their commitment (2020); Burkart Dairy1,b (2018-present) Communication • Development of tools for monitoring, reporting and verification (MRV) of zero-deforestation and Urrea • Development of governance mechanisms for the execution of programs for the reduction of Benítez (2020) deforestation Continued Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) 247 S. Burkart, M. Díaz, K. Enciso, A. Charry, N. Triana, M. Mena, J.L. Urrea-Benítez, I.G. Caro and R. Van Der Hoek Initiative/project and duration Level of intervention Interventions and achievements Sources Colombian Roundtable Primary producer • >60 members from the public (30%) and private sectors, academia, NGOs, among others (70%) Personal comm. for Sustainable Cattle1a,b Dairy and beef value • Contribution to the development of public policy frameworks on sustainable cattle technical (2014-present) chains • 13 established regional roundtables in the main cattle regions of Colombia secretary (May Public policy • Face-to-face and virtual capacity building events on sustainable cattle production and value 10, 2021); MGS- chains Col (2019); • Seminar series on sustainable cattle in 2020 with >20,000 participants from >15 countries Burkart and • Member of: Global Roundtable for Sustainable Beef Urrea Benítez • Draft public policy on sustainable cattle under review by the Ministry of Agriculture (2020) Sustainability Program Primary producer • Establishment of a protocol for the evaluation of sustainable practices in cattle farms Personal comm. AngusAzul4,b (2012-present) • 12% of the total area of cattle farms that sell to AngusAzul are protected forests (>4,000 ha) AngusAzul (May • Member: Zero-Deforestation Agreements and the Colombian Roundtable for Sustainable Cattle 12, 2021) Sustainable Cattle Primary producer • Technical assistance on sustainable beef farming to primary producers in the Meta and GANSO (2020); (GANSO)3,b (2018-present) Beef value chain Casanare Departments Ruden et al. • Development of sustainable and scalable business models (2020) • Development of the sustainability seal Aval GANSO together with a major supermarket chain • Establishment of 7 pilot and 1 demonstration farms Nationally Appropriate Primary producer In the planning phase, but with tentative commitments: Ministerio de Mitigation Actions (NAMA) Financing • 69,000 ha made available for restoration purposes Agricultura y for the cattle sector1,e • 2.2 million ha under improved pastures Desarrollo Rural (2014-present) • 601,000 ha planted with shattered trees (MADR 2020b) • 3,800 ha planted with mixed-forage banks • 660 ha with integrated living fences • 61,000 ha planted with forage hedges • 61,000 ha under intensive silvo-pastoral systems Credit line program for Primary producer • Implemented at a regional level (mainly in 82 municipalities) MADR (2020a) silvo-pastoral systems5,c • Promotion of sustainable practices in the different cattle regions of the country (2020-present) • Credits are directed to the purchase and planting of tree species • First initiative in this regard Departmental Agricultural Primary producer • Prioritizes crops and regions that receive public rural extension MADR (2020c) Extension Plans5,b Extensionists SNIA law 18765,b Primary producer • Establishes the restructuring of the national agricultural innovation system Ley 1876 de 2017 (2017-present) Extensionists • Purpose of improving the productivity and sustainability of the national agricultural sector (2017) • Focus on public rural extension Type of initiative = 1Multi-actor initiative, 2Research agreement, 3Public-private partnership, 4Private sector initiative, 5Public sector initiative. Effects of the COVID-19 pandemic = abudget reductions/constraints, bdifficulties in reaching target groups, ceffects still unknown, dinitiative ended before the pandemic started, einitiative has not yet started. Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) COVID-19 and sustainable intensification in Colombia 248 Annex 2. US$-COP exchange rate 2019-2021, with reference to the start of the COVID-19 pandemic. Source: own elaboration based on Banco de la República (2021). (Received for publication 6 December 2021; accepted 29 August 2022; published 30 September 2022) © 2022 Tropical Grasslands-Forrajes Tropicales is an open-access journal published by International Center for Tropical Agriculture (CIAT), in association with The Tropical Crops Genetic Resources Institute of The Chinese Academy of Tropical Agricultural Sciences (TCGRI-CATAS). This work is licensed under the Creative Commons Attribution 4.0 International (CC BY 4.0) license. Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) Tropical Grasslands-Forrajes Tropicales (2022) Vol. 10(3):249–260 249 doi: 10.17138/TGFT(10)249-260 Artículo científico Predicción del valor nutricional de sorgo para forraje mediante espectroscopia de reflectancia en el infrarrojo cercano (NIRS) y ecuaciones empíricas Prediction of the nutritional value of sorghum for forage using near-infrared reflectance spectroscopy (NIRS) and empirical equations SONIA PEREIRA-CRESPO1, ADRIÁN BOTANA1, MARCOS VEIGA1, CÉSAR RESCH1, LAURA GONZÁLEZ1, ROBERTO LORENZANA2, VALENTÍN GARCÍA-SOUTO1, MARÍA DEL PILAR MARTÍNEZ-DIZ1 Y GONZALO FLORES-CALVETE1 1Centro de Investigacións Agrarias de Mabegondo (CIAM), Mabegondo, Abegondo, A Coruña, España. ciam.gal 2Laboratorio Interprofesional Galego de Análise do Leite (LIGAL), Mabegondo, Abegondo, A Coruña, España. ligal.gal Resumen En este trabajo se evalúa la capacidad de la tecnología NIRS para estimar la composición química (n=171) y la digestibilidad de la materia orgánica (n=71) de la planta entera de sorgo aprovechado para forraje y sus componentes morfológicos, y se desarrollan ecuaciones empíricas basadas en parámetros químicos para la estimación de la digestibilidad de la materia orgánica (DMO), comparando su capacidad predictiva con la obtenida mediante NIRS. Las calibraciones NIRS obtenidas para la predicción de todos los parámetros mostraron valores del coeficiente de determinación en validación externa (r2) igual o superior a 0.90, indicando una alta confiabilidad. Asimismo, los valores de la relación entre la desviación estándar y el error de predicción (RPD) en validación externa fueron superiores a 3. La tecnología NIRS mostró una superioridad predictiva de la DMO, comparada con la mejor ecuación empírica, permitiendo reducir el error de predicción de validación externa, desde ±3.9 % a ±1.9 %. Se concluye que las ecuaciones NIRS desarrolladas son una herramienta útil y precisa para la evaluación nutricional rápida y precisa de la planta entera de sorgo y sus componentes morfológicos. Palabras clave: Composición química, digestibilidad, espectroscopia del infrarrojo cercano, modelos empíricos, Sorghum bicolor. Abstract In the present work studied the predictive ability of NIRS for the estimation of chemical composition (n = 171) and organic matter digestibility (n=71) of whole plant forage sorghum and morphological components, developed empirical equations based on chemical parameters to estimate the organic matter digestibility (OMD) values and compared the predictive ability of empirical models vs. NIRS equations. The predictive ability of NIRS models for estimating the OMD and chemical composition showed high reliability, according to the coefficient of determination in external validation (r2≥0.90), whilst the ratio of the standard deviation of the original data to standard error of external validation (RPD) values were higher than 3.0 for all parameters studied. Applying NIRS models to the prediction of OMD of whole plants and morphological components of forage sorghum led to the reduction in the standard error of external validation, in comparison with the best empirical model based on the chemical composition of samples (from ±3.9 to ±1.9 %). It is concluded that the NIRS equations developed in the present work are valuable tools for the fast and accurate assessment of the nutritive value of the whole plant and components of forage sorghum. Keywords: Chemical composition, digestibility, empirical models, near infrared spectroscopy, Sorghum bicolor. Correspondencia: Sonia Pereira-Crespo, CIAM, 15318 Abegondo, A Coruña, España. Email: soniapereira@ciam.gal Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) 250 S. Pereira-Crespo et al. Introducción Materiales y Métodos El ensilaje de maíz es el principal forraje conservado Material vegetal consumido en las explotaciones lecheras de Galicia (España), pues se estima que alrededor del 75 % El colectivo de muestras de sorgo en estado fresco de la producción de leche de la región procede de proviene de diferentes trabajos realizados por el Centro de explotaciones donde el ensilaje de maíz constituye una Investigaciones Agrarias de Mabegondo (CIAM) desde parte importante o mayoritaria de la ración de las vacas 2012 hasta 2018 en las localidades gallegas de Mabegondo lecheras (Flores et al. 2017). El clima predominante en (A Coruña) y Pobra do Brollón (Lugo), que incluían la zona de producción de leche de Galicia es atlántico diferentes variedades (de grano y forrajeras, así como bmr) húmedo, aunque existen episodios relativamente y distintos estados fenológicos, desde estado vegetativo frecuentes de sequías estacionales durante el verano, hasta grano pastoso. La colección está constituida por 171 que pueden comprometer el rendimiento del maíz muestras (120, 28 y 23 muestras de planta entera, parte forrajero, especialmente en zonas con suelos arenosos vegetativa e inflorescencia, respectivamente). o con escasa profundidad. El sorgo (Sorghum bicolor (L.) Moench) es un cultivo que se caracteriza por Métodos analíticos de referencia la tolerancia a condiciones de escasez de agua y nutrientes (Farré y Faci 2006), por lo que se considera La preparación de las muestras para las posteriores cada vez más una alternativa al maíz en esas áreas. determinaciones por vía húmeda se realizó mediante La utilización de forrajes producidos en la propia secado en estufa de aire forzado (Unitherm, Russell- explotación y la mejora de la eficiencia de su uso en Lindsey Engineering, Birmingham, England) a 80 la alimentación de las vacas es una de las principales °C durante 16 h. Las muestras fueron posteriormente claves del sector lácteo gallego para mantener su molidas a 1 mm en un molino de martillos (Christy posición dentro de las 10 principales regiones lecheras y Norris Ltd., Chelmsford, England). Los resultados de la Unión Europea (López-Iglesias 2015). Con se expresaron en porcentaje sobre MS corregida por este propósito, se requiere un sistema avanzado de humedad residual. La determinación del contenido en análisis de la calidad del forraje que proporcione una materia orgánica (MO), proteína bruta (PB) y almidón evaluación rápida, económica y precisa. (ALM) se realizó según Horwitz y Latimer (2010), y La estimación del valor energético (energía los contenidos en fibra neutro detergente (FND), fibra metabolizable) se realiza habitualmente a partir ácido detergente (FAD) y hemicelulosa (HCEL) se del valor de la digestibilidad de la materia orgánica realizaron siguiendo la metodología de Goering y (DMO) (Dijkstra et al. 2005). Los estudios in vivo Van Soest (1970). Los contenidos en carbohidratos proporcionan los valores de referencia de la DMO de solubles en agua (CSA) y no estructurales (CNET) se los alimentos para rumiantes, pero no son prácticos determinaron según Castro (2000). como análisis de rutina debido a su complejidad, duración y costo (Gosselink et al. 2004). Los métodos Determinación de la digestibilidad alternativos para evaluar en la práctica el valor energético de los forrajes incluyen el uso de ecuaciones La determinación de la digestibilidad in vivo se realizó empíricas basadas en parámetros químicos, técnicas mediante evaluaciones con un mínimo de cinco ovinos de digestión in vitro y espectroscopia de reflectancia machos castrados de más de dos años de edad, de raza del infrarrojo cercano (NIRS) (Gosselink et al. 2004; Gallega, alojados en jaulas metabólicas dotadas de Pereira-Crespo et al. 2022a, 2022b). separadores de heces y orina. El nivel de alimentación El objetivo de este estudio fue evaluar la capacidad fue ad libitum, ajustándose la oferta para permitir un de la tecnología NIRS para estimar la composición 10 % de rechazos. El procedimiento de las evaluaciones química y DMO de la planta entera y los componentes in vivo se realizó siguiendo la metodología indicada en morfológicos de sorgo forrajero cultivado en Galicia, y Pereira-Crespo et al. (2022a, 2022b). La digestibilidad in desarrollar ecuaciones empíricas basadas en parámetros vitro de la materia orgánica fue determinada siguiendo químicos para la estimación de la DMO, comparando su el método descrito por Tilley y Terry (1963) modificado capacidad predictiva con la obtenida por NIRS. por Alexander y McGowan (1966), utilizando como Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) Modelos de predicción del valor nutricional de sorgo para forraje 251 animales donantes de inóculo ruminal dos vacas secas basadas en parámetros químicos, persiguiendo el objetivo canuladas en el rumen alimentadas según lo indicado de establecer una base de comparación de la capacidad en Pereira-Crespo et al. (2022b). Con la finalidad predictiva de este modelo y el obtenido mediante la de controlar la variabilidad entre series de análisis tecnología NIRS, aplicada a la estimación del valor sucesivas, los resultados de digestibilidad in vitro se nutricional de planta entera de sorgo y sus componentes corrigieron utilizando un conjunto de cuatro muestras morfológicos (hojas + tallos e inflorescencias). de referencia cuya digestibilidad in vivo era conocida, Se examinaron las relaciones entre los valores de colocados por duplicado al inicio, mitad y al final de DMO y las determinaciones de laboratorio mediante cada serie. La corrección se ajustó según lo indicado por análisis de correlación y de regresión linear simple y Pereira-Crespo et al. (2022a, 2022b). múltiple, acompañada de un proceso de regresión paso Del total de muestras utilizadas con DMO conocida, a paso, siendo seleccionados los mejores modelos uni cinco (n=5) provenían de ensayos in vivo y las restantes y multivariables en cuanto a porcentaje de varianza (n=66) de ensayos in vitro. explicada de la variable dependiente. Los procedimientos utilizados fueron CORR, STEPWISE, GLM y REG del Error estándar de los métodos de referencia paquete estadístico SAS v. 9.4 (SAS Institute 2012). En la construcción de los modelos predictivos se utilizó el Para evaluar los resultados de las calibraciones NIRS en método “leave one out” (Massart et al. 1997), un proceso relación al error del método de referencia, se calculó el secuencial de validación cruzada, en el cual del conjunto error estándar del método de referencia (SEL) para cada de calibración era excluida secuencialmente una de parámetro, realizado a partir de la varianza entre replicados las muestras, sobre la que se realizó la validación de la de la misma muestra. Los valores de SEL de los métodos predicción a partir del modelo de calibración generado, de referencia del presente trabajo figuran en la Tabla 1. repitiendo el proceso un número de veces igual al número total de muestras. Se computó el error estándar Tabla 1. Error estándar del método de referencia para la de validación cruzada (SECV) a partir de los errores de determinación de la composición química y digestibilidad. estimación (ei= diferencia entre el valor observado y el Parámetro Media SEL CV predicho por el modelo) de las muestras del conjunto de Composición química (%MS) validación como MO 87.63 0.57 0.65 PB 11.30 0.50 4.48 , FND 44.34 1.74 3.92 siendo n el número total de muestras. Los modelos FAD 36.22 1.18 3.26 con un porcentaje de varianza explicada en validación CSA 13.09 0.92 7.08 cruzada inferior al 20 % fueron excluidos, así como CNET 17.43 0.88 5.06 aquellos que incluían como predictores variables con un ALM 17.37 1.57 9.04 elevado grado de colinealidad, dado su efecto negativo Digestibilidad (%) en la estabilidad de los coeficientes de los modelos de regresión, siendo aquel estimado a través de las DMOTT 64.77 1.94 3.00 herramientas de diagnóstico del procedimiento REG DMO in vivo 66.83 2.33 3.49 de SAS. Los modelos de predicción de DMO fueron SEL = error estándar de laboratorio; CV = coeficiente de objeto de validación externa. A tal efecto, ordenadas las variación; MS = materia seca (%); MO = materia orgánica; PB = proteína bruta; FND = fibra neutro detergente; FAD = muestras de la colección por su valor de digestibilidad, fibra ácido detergente; CSA = carbohidratos solubles en agua; cada quinta muestra se asignó al grupo de validación CNET = carbohidratos no estructurales; ALM = almidón; independiente, siendo construido el modelo predictivo DMOTT = digestibilidad de la MO in vitro método Tilley- sobre los cuatro quintos restantes. El valor del error de Terry; DMO = digestibilidad de la MO. predicción (SEP) se calculó como el error estándar de la diferencia entre los valores observados y predichos, y el Análisis estadístico en la construcción de las ecuaciones valor del bias como el valor medio de la diferencia entre de regresión basadas en parámetros químicos los valores observados y predichos, siendo obtenido el coeficiente de determinación en validación externa (r2) a Para la colección de muestras de las que se disponía valores partir de la regresión lineal entre los valores observados de DMO (n=71) se desarrollaron ecuaciones de regresión y los predichos. Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) 252 S. Pereira-Crespo et al. Análisis e instrumentación NIRS que contenía el 25 % de las muestras totales, estando representados en ambos grupos los 3 tipos de muestras La información espectral de las muestras se registró (planta entera, parte vegetativa e inflorescencia). en un espectrofotómetro monocromador de espectro Las ecuaciones de calibración se realizaron a través continuo Foss NIRSystem 6500 (Silver Spring, de regresión de mínimos cuadrados parciales modificada Washington, USA), situado en una sala con temperatura (MPLS, Martens y Næs 1987) entre los datos espectrales controlada (24 ± 1 °C). El instrumento estaba dotado y los valores de referencia, aplicándose el procedimiento de un módulo de giro y realiza medidas de reflectancia de validación cruzada para determinar el número óptimo (R) en la región espectral comprendida entre 1100 nm y de términos de la ecuación. Se incluyeron cuatro grupos 2500 nm. Las muestras secas y molidas se analizaron de validación cruzada para prevenir el sobreajuste, por duplicado como log(1⁄R). La adquisición de la que fueron secuencialmente utilizados para efectuar información espectral y el análisis quimiométrico de la validación de las ecuaciones generadas. Además, los datos se realizó mediante el software WinISI II v.1.5 durante el proceso de desarrollo de cada ecuación de (Infrasoft International 2000, Port Matilda, PA, USA). calibración se aplicaron como máximo 2 etapas de eliminación de anómalos químicos (T > 2.5; Mark y Tratamiento quimiométrico de los datos espectrales Workman 2003). La selección de las mejores ecuaciones de calibración NIRS se realizó en base al menor valor Mediante el algoritmo CENTER (Shenk y Westerhaus, del error estándar validación cruzada (SECV) y al mayor 1991) del software WinISI, se realizó un Análisis valor del coeficiente de determinación en validación de Componentes Principales, seguido del cálculo de cruzada (1 − VR; Shenk y Westerhaus 1996). distancias entre espectros en un espacio n-dimensional a través de la distancia de Mahalanobis, la cual Resultados permitió estudiar la estructura y variabilidad espectral de la población y detectar muestras anómalas, Ecuaciones empíricas para la predicción de DMO siendo identificadas como muestras atípicas (outliers espectrales) aquellas con valores GH>3 (Shenk y Valores medios de la colección. En la Tabla 2 se puede Westerhaus, 1991). Aquellos espectros anómalos observar la media, desviación típica, coeficiente de fueron eliminados, repitiendo el proceso de detección variación y rango de valores para la composición química y eliminación de anómalos espectrales hasta conseguir y DMO del colectivo total de muestras (n = 71) de las que que todas las muestras presentaran un valor de GH se disponía el valor de referencia de DMO. Las muestras inferior al valor máximo recomendado. de planta entera constituían el 70 % del total, repartiéndose Los pretratamientos espectrales evaluados fueron: el resto entre muestras de inflorescencia (panícula) y parte SNV (Standard normal variate); DT (Detrend), la vegetativa (plantas desprovistas de inflorescencia). combinación de ambos (SNV+D) (Barnes et al., 1989) Los contenidos medios (y rango de variación) en % y W-MSC (Weighted multiplicative scatter correction; MS fueron, MO: 94.2 (88.6 a 97.4), PB: 7.6 (3.5 a 16.2), Shenk y Westerhaus 1993), combinados con derivadas FND: 60.3 (37.2 a 79.2), FAD: 31.1 (16.2 a 39.1), HCEL: como: 1,4,4,1; 1,5,5,1; 1,6,4,1; 1,8,4,1; 1,10,5,1; 1,10,10,1; 29.1 (19.9 a 40.2), ALM: 8.9 (0.0 a 45.4), CNET: 23.8 (9.1 2,4,4,1; 2,5,5,1; 2,6,4,1; 2,8,4,1; 2,10,5,1; 2,10,10,1; donde el a 55.6) y CSA: 12.7 (2.7 a 27.4). El valor medio de DMO primer dígito expresa el orden de la derivada (1=primera fue 63.2 %, oscilando los valores mínimo y máximo entre derivada, 2=segunda derivada), el segundo dígito 53.8 y 77.6 %. Los contenidos en ALM, CSA, CNET indica el tamaño del segmento sobre el cual se realiza y PB fueron los parámetros de mayor variabilidad, con la derivación (intervalo expresado en nanómetros) y el valores del coeficiente de variación de 111.1 %, 55.4 %, tercero y cuarto dígito indican el tamaño de los intervalos, 37.6 %, y 32.1 %, respectivamente, mientras que la expresados en nanómetros, utilizados para el cálculo de concentración de MO fue el menos variable (CV= 1.8 %), suavización de la derivada (Shenk y Westerhaus 1994). siendo intermedia la de las distintas fracciones de la fibra, El colectivo de espectros se dividió de forma aleatoria con valores de CV de 13.8 % para FND, 14.0 % para FAD (Shenk y Westerhaus 1991) en un grupo de calibración al y 15.3 % para HCEL. La variabilidad de la DMO con un que se destinaron el 75 % de las muestras seleccionadas valor del CV de 10.3 % se situó a un nivel inferior de la aleatoriamente, siendo el resto de las muestras asignando observada para los parámetros de composición química, a un segundo grupo de validación independiente (externa) a excepción del contenido en materia orgánica. Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) Modelos de predicción del valor nutricional de sorgo para forraje 253 Tabla 2. Composición química y digestibilidad de la colección Correlaciones entre composición química y digestibilidad. de muestras de planta entera de sorgo y sus componentes La matriz de correlaciones entre los parámetros de morfológicos, utilizadas para la obtención de las ecuaciones composición química y la digestibilidad de la materia empíricas. orgánica se muestra en la Tabla 3. Fue observada una Parámetro n Media SD CV Min Max correlación negativa moderada pero significativa (p<0.001) MO 71 94.2 1.7 1.8 88.6 97.4 entre los valores de DMO y las distintas fracciones de fibra, PB 71 7.6 2.4 32.1 3.5 16.2 con valores del coeficiente de correlación r de -0.47 para FND 71 60.3 8.3 13.8 37.2 79.2 FND, de -0.49 para FAD, y de -0.39 para HCEL. También FAD 71 31.1 4.3 14.0 16.2 39.1 se observó una correlación aproximadamente de la misma HCEL 71 29.1 4.5 15.3 19.9 40.2 magnitud, pero de signo positivo, entre el valor de DMO y ALM 71 8.9 9.9 111.1 0.0 45.4 el de materia orgánica y la concentración de carbohidratos no estructurales, con valores de r de 0.47 para MO, de CNET 71 23.8 8.9 37.6 9.1 55.6 0.46 para CNET y de 0.48 para CSA. El coeficiente de CSA 71 12.7 7.0 55.4 2.7 27.4 correlación entre DMO y los contenidos en proteína y DMO 71 63.2 6.5 10.3 53.8 77.6 almidón fue muy bajo y no alcanzó significación. MS = materia seca; MO = materia orgánica (%MS); PB = Destaca la alta correlación negativa entre los valores de proteína bruta; FND = fibra neutro detergente (%MS); FAD carbohidratos estructurales y no estructurales (valores de r = fibra ácido detergente (%MS); HCEL = hemicelulosa (%MS); ALM = almidón (%MS); CNET = carbohidratos no de -0.94, -0.89 y -0.88 entre CNET con FND, FAD y HCEL, estructurales (%MS); CSA = carbohidratos solubles (%MS); respectivamente) reflejando parcialmente el resultado del DMO = digestibilidad de la MO; n = número de muestras; SD proceso de maduración de la planta, donde la proporción de = desviación estándar; CV = coeficiente de variación; Max = la fracción fibra en la MS tiende a reducirse con el incremento valor máximo; Min = valor mínimo. la acumulación de azúcares y almidón en la planta. Tabla 3. Matriz de correlaciones entre parámetros de composición química y digestibilidad de la materia orgánica de la colección de planta entera de sorgo y sus componentes morfológicos. PB FND FAD HCEL ALM CNET CSA DMO MO r -0.61 -0.55 -0.51 -0.53 0.28 0.69 0.53 0.47 p *** *** *** *** * *** *** *** PB r - 0.22 -0.02 0.43 0.17 -0.37 -0.71 -0.17 p NS NS *** NS ** *** NS FND r - - 0.94 0.94 -0.78 -0.94 -0.10 -0.47 p *** *** *** *** NS *** FAD r - - - 0.77 -0.82 -0.89 0.06 -0.49 p *** *** *** NS *** HCEL r - - - - -0.65 -0.88 -0.25 -0.39 p *** *** * *** ALM r - - - - - 0.72 -0.49 0.09 p *** *** NS CNET r - - - - - - 0.20 0.46 p NS *** CSA r - - - - - - - 0.48 p *** MS = materia seca; MO = materia orgánica (%MS); PB = proteína bruta (%MS); FND = fibra neutro detergente (%MS); FAD = fibra ácido detergente (%MS); HCEL = hemicelulosa (%MS); ALM = almidón; CNET = carbohidratos no estructurales; CSA = carbohidratos solubles; DMO = digestibilidad de la materia orgánica (%); r = coeficiente de correlación de Pearson; p = significación (*** p<0.001; ** p<0.01; * p<0.05; NS p>0.05: no significativo). Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) 254 S. Pereira-Crespo et al. Ecuaciones de predicción de DMO basadas en parámetros la predicción de muestras independientes, el que químicos. Las ecuaciones de regresión seleccionadas, sería coincidente con el criterio de Shenk et al. (2001) así como los estadísticos de calibración y de validación que sitúan un umbral de 0.60 en el valor mínimo del cruzada y externa se muestran en la Tabla 4, estando coeficiente de determinación en validación externa para aquellas ordenadas de mayor a menor por el porcentaje que el modelo pueda ser usado de forma fiable en la de la varianza explicada en calibración. Las ecuaciones estimación de muestras problema. de predicción de la digestibilidad de la materia orgánica basadas en análisis químicos de la población de muestras Ecuaciones NIRS para predicción de la composición de planta entera de sorgo y sus fracciones mostraron química y DMO una moderada capacidad predictiva. Los dos mejores modelos de dos variables, que incluyeron FAD y CSA o Las características descriptivas de los parámetros de FAD y ALM como predictores, explicaron en validación composición química y DMO del colectivo de calibración cruzada el 56 % y el 47 % de la varianza con un error (n = 126) y validación externa (n = 45) para el desarrollo estándar de validación cruzada de ±4.29 % y ±4.78 %, de modelos de predicción NIRS se muestran en la Tabla respectivamente. La eliminación de CSA o ALM de la 5. La composición química media (y rango de variación) ecuación redujo el porcentaje de varianza explicada hasta de las muestras de calibración y validación expresadas el 34 % y aumentó el error hasta ±5.01 % para el modelo en %MS, fueron, respectivamente, MO: 94.5 (83.7 a univariable con FAD como único predictor. 97.9) y 94.1 (89.1 a 97.9), PB: 8.9 (3.5 a 21.1) y 9.4 (3.8 En validación externa, el ajuste de las ecuaciones a 20.2), FND: 57.5 (28.4 a 79.2) y 58.4 (30.0 a 78.7), empeoró notablemente, con valores de r2 y de SEP de FAD: 29.4 (12.0 a 39.5) y 29.5 (13.8 a 38.5), CNET: 25.2 0.44 y ±5.74 % y de 0.40 y ±5.51 % para las ecuaciones (4.7 a 65.3) y 24.1 (4.3 a 62.6), CSA: 11.8 (1.5 a 27.4) y con dos variables y de 0.21 y ±6.14 % para la ecuación 10.9 (1.3 a 24.5) y ALM: 11.0 (0.0 a 56.2) y 10.5 (0.0 a univariable con FAD como único predictor. Los valores 54.4). Los valores medios de digestibilidad de la materia de los estadísticos RPDve y RERve en validación externa orgánica fueron 63.3 % y 62.4 % para la colección de oscilaron entre 1.17 y 1.05 para el primero y entre 4.3 calibración y validación, respectivamente, con rangos y 3.8 para el segundo. Atendiendo a los criterios que de variación de 55.1 a 77.6 % para el primer grupo y juzgan la calidad de las ecuaciones en base a estos de 53.8 a 73.5 % para el segundo. Se observa un amplio valores (Williams 2014), los valores de RPD inferiores intervalo de variabilidad para todos los parámetros, tanto a 1.5 indican que el modelo no sería de utilidad para en el colectivo de calibración como en el de validación Tabla 4. Ecuaciones de predicción de la digestibilidad de la materia orgánica en base a parámetros químicos de la colección total (planta entera y sus componentes morfológicos) de sorgo. Ecuación Calibración Validación cruzada Validación externa R2 SEC 1-VR SECV r2 SEP SEPc Bias Pend DMO = 57.41 +3.54 CSA -0.099 FAD×CSA 0.59 4.18 0.56 4.29 0.44 5.74 5.53 -1.508 0.79 s.e ±1.03 ±0.398 ±0.0125 P *** *** *** *** *** *** DMO = 128.53 -1.92 FAD -0.634 ALM 0.55 4.42 0.47 4.74 0.40 5.51 5.49 0.415 0.75 s.e ±7.28 ±0.211 ±0.093 P *** *** *** *** *** *** DMO = 86.20 -0.74 FAD 0.38 4.88 0.34 5.01 0.21 6.14 6.09 0.795 0.80 s.e ±4.92 ±0.156 P *** *** *** *** NS DMO = digestibilidad de la materia orgánica (%); CSA = carbohidratos solubles en agua (%MS); FAD = fibra ácido detergente (%MS); ALM = almidón (%MS); s.e = error estándar; P = significación (*** P<0.001; ** P<0.01; * P<0.05; NS no significativo P>0.05); R2 = coeficiente de determinación del modelo de calibración; SEC = error estándar de calibración; 1-VR = coeficiente de determinación en validación cruzada; SECV = error estándar de validación cruzada; r2 = coeficiente de determinación en validación externa; SEP = error estándar de predicción; SEPc = error estándar de predicción corregido por bias; Bias = diferencia media entre los valores observados y los predichos por la ecuación; Pend = pendiente de la regresión entre valores observados y predichos por la ecuación. Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) Modelos de predicción del valor nutricional de sorgo para forraje 255 Tabla 5. Estadísticos de calibración y validación cruzada de las ecuaciones NIRS para estimación de la composición química y de la digestibilidad de la materia orgánica de muestras de planta entera de sorgo y sus componentes morfológicos. Calibración Validación cruzada Parámetro Scatter MT n Outliers T Media SD Min Max R2 SEC 1-VR SECV RPDvc RERvc MO SNV+D 2,6,4,1 126 9 5 94.5 2.3 83.7 97.9 0.97 0.31 0.95 0.38 6.2 37.5 PB SNV+D 2,4,4,1 126 6 8 8.9 3.4 3.5 21.1 0.99 0.30 0.99 0.31 10.9 56.0 FND SNV+D 2,6,4,1 126 6 8 57.5 9.9 28.4 79.2 0.99 1.20 0.99 1.44 6.8 35.2 FAD SNV+D 2,4,4,1 126 7 3 29.4 5.9 12.0 39.5 0.98 0.83 0.98 0.90 6.5 30.7 CNET SNV+D 1,4,4,1 125 5 4 25.2 12.8 4.7 65.3 0.99 1.13 0.99 1.21 10.6 50.2 CSA SNV+D 1,10,5,1 125 5 9 11.8 6.4 1.5 27.4 0.99 0.63 0.99 0.83 7.8 31.4 ALM SNV+D 1,10,5,1 126 5 9 11.0 14.2 0.0 56.2 0.99 1.13 0.99 1.38 10.3 40.8 DMO SNV+D 2,10,5,1 51 1 6 63.3 6.9 55.1 77.6 0.94 1.64 0.92 1.93 3.6 11.7 Validación externa Parámetro n Media SD Min Max r2 SEP SEPc Bias Pendiente RPDve RERve MO 45 94.1 2.1 89.1 97.9 0.97 0.36 0.37 0.04 0.98 5.9 24.2 PB 45 9.4 4.0 3.8 20.2 0.99 0.31 0.31 -0.06 0.98 12.7 52.3 FND 45 58.4 9.9 30.0 78.7 0.98 1.35 1.35 -0.19 0.99 7.4 36.1 FAD 45 29.5 5.7 13.8 38.5 0.98 0.85 0.86 -0.07 1.02 6.6 28.9 CNET 45 24.1 13.3 4.3 62.6 0.99 1.18 1.00 0.08 1.00 11.3 49.4 CSA 45 10.9 6.4 1.3 24.5 0.99 0.72 0.67 0.25 0.98 8.9 32.5 ALM 45 10.5 14.5 0.0 54.4 0.99 1.36 1.37 -0.27 1.00 10.7 39.9 DMO 20 62.4 5.5 53.8 73.5 0.90 1.73 1.77 0.08 1.00 3.2 11.4 MT = tratamiento matemático; SNV+D = standard normal variate+detrend; n = número de observaciones; Outliers = número de muestras anómalas eliminadas; T = número de factores de la regresión; SD = desviación estándar; Min = valor mínimo; Max = valor máximo; SEC = error estándar de calibración; R2 = coeficiente de determinación del modelo de calibración; 1-VR = coeficiente de determinación en validación cruzada; SECV = error estándar de validación cruzada; RPD = ratio entre la desviación estándar y el error de predicción en validación cruzada o externa; RER = ratio entre rango de valores y el error de predicción en validación cruzada o externa; r2 = coeficiente de determinación en validación externa; SEP = error estándar de predicción; SEPc = error estándar de predicción corregido por bias; Bias = diferencia media entre los valores observados y los predichos por la ecuación; MS = materia seca; MO = materia orgánica (%MS); PB = proteína bruta (%MS); FND = fibra neutro detergente (%MS); FAD = fibra ácido detergente (%MS); CNET = carbohidratos no estructurales (%MS); CSA = carbohidratos solubles en agua (%MS); ALM = almidón (%MS); DMO = digestibilidad de la materia orgánica (%). externa, confirmando la elevada diversidad de muestras eliminadas del conjunto de calibración respecto del incluidas en este producto, tanto muestras de planta número inicial de muestras osciló entre 5.1 y 3.9 % para entera de sorgo como de los diferentes componentes las calibraciones NIRS de los parámetros químicos, morfológicos que la conforman. siendo del 1.9 % para DMO. Las ecuaciones de predicción para todos los Las calibraciones NIRS obtenidas para la predicción de parámetros se obtuvieron con SNV+D como todos los parámetros mostraron valores de los coeficientes pretratamiento espectral para corregir el efecto de la de determinación iguales o superiores a 0.90. tanto en dispersión de la luz o scatter debido a diferencias en el la etapa de validación cruzada (1-VR) como en la de tamaño de partícula o de compresión de las muestras en validación externa (r2), indicando una calidad excelente, la cápsula. Los mejores modelos para la predicción de según lo indicado por Shenk et al. (2001). Las ecuaciones la concentración de los carbohidratos no estructurales NIRS para la predicción de PB, CNET, CSA y ALM (CNET, CSA y ALM) se obtuvieron utilizando la primera presentaron el mayor porcentaje de varianza explicada derivada del espectro, mientras que para la estimación de (99 %) en la etapa de validación externa, seguidos de las los restantes parámetros (MO, PB, FND, FAD y DMO) ecuaciones para estimar los contenidos en FND y FAD las ecuaciones se vieron optimizadas mediante el uso de (98 %) y MO (97 %) así como el valor de DMO (90 %). la segunda derivada. El número de muestras anómalas Los valores de SECV y SEP fueron, respectivamente, de Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) 256 S. Pereira-Crespo et al. ±0.38 y ±0.36 % para MO, de ±0.31 y ±0.31 % para PB, en la Figura 1, donde se puede observar el buen ajuste de ±1.44 y ±1.35 % para FND, de ±0.90 y ±0.85 % para obtenido por los modelos de predicción, los cuales FAD, de ±1.21 y ±1.18 % para CNET, de ±0.83 y ±0.72 % cumplieron con los criterios indicados por Windham et para CSA, de ±1.38 y ±1.36 % para ALM, siendo de ±1.93 al. (1989) y Shenk et al. (2001), relativos a los límites y ±1.73 % para DMO. recomendados para que las ecuaciones se consideren Los valores de los estadísticos RPD y RER, tanto en el fiables. A este respecto, los valores de la pendiente de proceso de validación cruzada como en el de validación regresión entre los valores observados y predichos, los externa, superan ampliamente los valores mínimos del bias y los del SEPc estuvieron dentro de los intervalos recomendados (3 para RPD y 10 para RER) por Williams recomendados por los citados autores (pendiente y Sobering (1996) y Williams (2014), confirmando entre 0.9 y 1.1; bias= ±0.60 SEC y SEPc < 1.30 SEC), la buena capacidad de predicción cuantitativa de las mientras que los valores de r2 de las ecuaciones para la ecuaciones y asegurando la validez de las mismas en la estimación de los parámetros de composición química aplicación a futuras muestras independientes. y DMO superaron ampliamente el valor límite de 0.60 La representación gráfica de los valores de referencia mostrando el buen comportamiento predictivo de las y de los valores predichos por las ecuaciones NIRS ecuaciones NIRS para la predicción del valor nutricional obtenidos en la etapa de validación externa se muestra de muestras de sorgo. Figura 1. Plot de valores de composición química y digestibilidad determinados por los métodos de referencia y de valores predichos mediante las ecuaciones NIRS para muestras de planta entera de sorgo y sus componentes morfológicos. Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) Modelos de predicción del valor nutricional de sorgo para forraje 257 Discusión Ecuaciones empíricas Los valores medios de composición química de la planta Marten et al. (1975) estudiaron diversos métodos para la de sorgo forrajero indicados en las Tablas INRA (2010) predicción de la DMO in vivo de ensilajes de maíz y de entre los estados vegetativo y de grano lechoso son (en sorgo en base a una colección de 51 muestras de ambos %MS) de 91.7 (88.4 a 94.1) para MO, de 11.8 (6.9 a 19.0) forrajes. El mejor predictor de la DMO fue el contenido para PB, de 60.6 (55.7 a 60.7) para FND y de 35.3 (31.2 en FAD, con valores de R2 de 0.80 y 0.61 para los ensilajes a 35.4) para FAD, siendo el valor medio de DMO (%) de de sorgo y de maíz, respectivamente, no siendo útil la 63.7, oscilando entre 71.0% para el estado vegetativo y inclusión de PB para mejorar la capacidad predictiva de 60.0% para el estado de grano lechoso. las ecuaciones. El método de la ecuación sumativa de Van En las Tablas FEDNA (Calsamiglia et al. 2016) se Soest no proporcionó estimaciones satisfactorias de la recoge un número limitado de valores para ensilados de digestibilidad del sorgo ni del maíz, si bien la concentración sorgo, siendo los valores medios y rango de variación de de lignina-permanganato explicó el 74% de la variación la composición química (en %MS) de 9.0 (8.3 a 11.2) para en DMO del sorgo pero solo el 23% en el caso del maíz. PB, de 59.4 (57.1 a 61.3) para FND, y de 6.3 (3.5 a 9.7) para Esta observación coincide con lo indicado por Dickerson ALM. El valor medio de DMO indicado en las citadas (1986), quien reporta un coeficiente de correlación de -0.88 Tablas es de 59.8%, con una oscilación entre 58.5 y 62.4%. para la planta de sorgo y de -0.51 para la planta de maíz El rango de variación de la composición química entre el grado de lignificación de la FAD y la digestibilidad, de las muestras de planta de sorgo y sus componentes señalando que el mayor descenso de digestibilidad de los morfológicos utilizadas en el presente trabajo es más sorgos forrajeros comparados con el maíz se atribuye al amplio que el indicado en las publicaciones anteriores, con hecho de que en el sorgo se incrementa la lignina en la valores (en %MS) de MO entre 88.6 a 94.7, de PB entre FAD con el avance de la madurez, mientras que en el maíz 3.5 a 16.2, de FND entre 37.2 a 79.2, de FAD entre 16.2 la relación LAD/FAD no cambia substancialmente con a 39.1 y de ALM entre 0 a 45.4, mientras que el de DMO el tiempo, pasando del 10% en plantas a comienzo de la varió entre 53.8 a 77.6%. La inclusión de muestras en la floración al 12% en plantas en estado maduro. colección procedentes de diferentes genotipos de plantas En el presente trabajo, las correlaciones observadas de sorgo (incluidos los tipos bmr), diferentes localidades, entre DMO y la composición química de las muestras años y estados de cosecha en la colección utilizada, unido son de escasa magnitud, con valores del coeficiente de a la presencia de muestras de fracciones de la planta (parte correlación r de 0.47 con MO, de 0.48 con CSA y de -0.47 vegetativa e inflorescencia) junto con el de planta entera y -0.49 con FND y FAD, respectivamente. Por lo tanto, explican la mayor variabilidad observada en la población las ecuaciones empíricas obtenidas para la estimación de de muestras utilizada en el presente trabajo. la DMO basadas en los contenidos en FAD y CSA o en Heuzé et al. (2015) indican que a finales del estado FAD y ALM, son de un poder predictivo mediocre, con vegetativo la DMO del sorgo en climas templados es de bajos valores del coeficiente de correlación en validación 70–71%, descendiendo hasta el 61% cuando la floración externa r2 entre 0.44 y 0.40 y altos errores de predicción está completa y permaneciendo más o menos estable SEP de ±5.74 y ±5.51%, respectivamente. tras este estado. Los valores de DMO de planta entera Una limitación adicional de las ecuaciones basadas en de sorgo de este trabajo poseen un rango mayor (estado FAD para predecir la digestibilidad del sorgo es que pueden vegetativo: 75.3% - estado grano pastoso: 53.8%). En subestimar el valor de las variedades bmr como indican un trabajo realizado en el CIAM (Pereira-Crespo et Brouk y Bean (2011). En el citado trabajo de Pereira-Crespo al. 2018), para una variedad de sorgo forrajero y otra et al. (2018) se indican valores medios de DMO de 61.5% y de maíz cultivados en un estudio en gran parcela y de 67.6% para variedades bmr y no bmr, respectivamente, recogidos a los 125 días tras la siembra, que fueron sin diferencias apreciables en los contenidos en fibra, con evaluados in vivo en estado fresco se reportan valores valores medios de 32.2 y 33.1%MS para FAD y de 59.6 y muy semejantes de MO entre ambas especies, (97.0 y 59.7 %MS para FND, respectivamente. 97.6 %MS) siendo el sorgo algo más proteico y sobre todo más fibroso que el maíz (PB: 9.8 y 7.4 %MS; FAD: Ecuaciones NIRS para la estimación de la composición 30.7 y 19.7 %MS, FND: 56.2 y 38.3 %MS) con un química y DMO contenido muy inferior de almidón (10.4 y 34.9 %MS) y mucho menos digestible (DMO in vivo: 58.3 y 70.1%). Los resultados de las calibraciones NIRS obtenidas Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) 258 S. Pereira-Crespo et al. en este trabajo para la predicción de la composición de los ensilajes de sorgo, si bien en este trabajo no se química y el valor de DMO de las muestras de sorgo realizó la prueba de validación externa. En un trabajo más muestran una excelente calidad de las estimaciones, reciente realizado en Argentina con una amplia colección con valores de r2 ≥ 0.90 y RPDve ≥ 3.0 en el proceso de muestras de planta entera de sorgo y de maíz y de las de validación externa de las ecuaciones sobre muestras diferentes fracciones morfológicas que las componen, independientes. Por otro lado, la capacidad predictiva Juan et al. (2016) reportaron una excelente capacidad de la DMO mediante calibraciones NIRS mostró una predictiva de las calibraciones NIRS para la estimación elevada superioridad en comparación con las ecuaciones de los contenidos en PB y FND y de la digestibilidad de empíricas. En la estimación de la digestibilidad de la materia seca de las muestras con un alto valor de la muestras independientes, la técnica NIRS permitió varianza explicada en calibración (96–97%) y con valores duplicar el porcentaje de varianza explicada por la mejor del error estándar de predicción SECV y de RPDvc en ecuación empírica, desde el 44% al 90% y reducir el validación cruzada de ±0.54% y 5.6 para PB, de ±2.51% y error de predicción casi en un 70%, desde ±5.74% a 5.0 para FND y de ±1.4 y 5.1 para DMS. ±1.73% unidades porcentuales de DMO. En un trabajo con muestras de ensilajes de sorgo de Conclusiones Kansas (USA), Dubois et al. (1989) reportaron que la tecnología NIRS podía ser usada satisfactoriamente para Las calibraciones NIRS obtenidas en el presente trabajo predecir el contenido de PB, FAD y FND, reportando para la estimación de la composición química y el valor valores de r2 y SEP, en validación externa, de 0.86 y de DMO de planta entera de sorgo y sus componentes ±0.41% para PB, de 0.86 y ±1.75% para FAD y de 0.92 morfológicos mostraron una predicción cuantitativa y ±1.83% para FND. Sin embargo, los correspondientes satisfactoria y, por tanto, una alternativa rápida y precisa valores de RPDve en este trabajo fueron de 1.9, 1.9 y 1.2 para determinar estos parámetros en relación a los métodos para las ecuaciones de predicción de PB, FAD y FND, analíticos convencionales. La técnica NIRS mostró una respectivamente, lo que sugiere una escasa robustez de capacidad predictiva de la DMO superior a las ecuaciones las ecuaciones, mostrando un poder predictivo bajo para empíricas basadas en parámetros químicos. los citados parámetros. Es posible encontrar trabajos en los que la estimación Agradecimientos del valor de DMO de la planta de sorgo mediante NIRS no es satisfactoria. Por ejemplo, Hicks et al. (2002) Trabajo financiado por el proyecto FEADER 2013/22 de estudiaron la utilidad de la técnica para estimar el valor la Xunta de Galicia. nutricional de líneas puras de sorgo granífero en el marco de trabajos de mejora genética, reportando que Referencias mientras la predicción del contenido en PB era realizada de forma precisa por NIRS, no era satisfactoria la (Nota de los editores: Enlaces verificados el 8 de septiembre de 2022). predicción del valor de DMO in vitro de las muestras, Alexander RH; McGowan M. 1966. The routine determination con valores de RPDvc en validación cruzada de 2.3 y 1.0, of in vitro digestibility of organic matter in forages. Grass respectivamente. Sin embargo, dado el bajo valor medio and Forage Science 21(2):140–147. doi: 10.1111/j.1365- reportado para DMO (38.9%), la escasa fiabilidad de la 2494.1966.tb00462.x ecuación NIRS cabría atribuirla a la baja variabilidad de Barnes RJ; Dhanoa MS; Lister SJ. 1989. Standard normal la colección para este parámetro. variate transformation and detrending of near infrared Kański et al. (2013) estudiaron la utilidad de NIRS para diffuse reflectance spectra. Applied Spectroscopy estimar la digestibilidad in vivo de la materia orgánica de 43(5):772–777. doi: 10.1366/0003702894202201 ensilajes de sorgo a partir de una colección de 58 muestras Brouk MJ; Bean B. 2011. Sorghum in dairy production. que incluían ensilajes de planta entera de sorgo y mezclas Feeding guide. United Sorghum Checkoff Program, con otras materias primas. El valor del coeficiente de Lubbock, Texas,USA. bit.ly/3cOHXw7 Castro P. 2000. Determinación de carbohidratos no determinación en calibración para la estimación de estructurales en forrajes. En: III Reunión Ibérica de DMO fue de 0.83 y los de SEC y SECV, respectivamente, Pastos y Forrajes. Consellería de Agricultura, Gandería e fueron de ±4.4 y ±4.8%, en tanto que el valor RPDvc en Política Agroalimentaria, Xunta de Galicia, ed. Santiago validación cruzada fue de 2.4, concluyendo los autores de Compostela, España, p. 447–453. que NIRS permite la predicción directa del valor DMO Dickerson JT. 1986. Yield, composition and nutritive value Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) Modelos de predicción del valor nutricional de sorgo para forraje 259 of forage sorghum silages: hybrid and stage of maturity Kański J; Pyś JB; Górka P. 2013. Prediction of the nutritive effects. M.Sc. Thesis. Kansas State University, Manhattan, value of wet whole-crop sorghum silage according to the Kansas, USA. bit.ly/3RG6XVf INRA feeding system by near-infrared spectroscopy. Dijkstra J; Forbes JM; France J. 2005. Quantitative aspects of Journal of Animal and Feed Sciences 22(4):360–365. doi: ruminant digestion and metabolism. CAB International, 10.22358/jafs/65928/2013 Wallingford, UK. doi: 10.1079/9780851998145.0000 López-Iglesias E. 2015. O complexo lácteo galego nun Dubois PC; Garcia G; Bolsen KK; Harbers LH. 1989. Near- mercado liberalizado. Estratexias e políticas ante a infrared reflectance spectroscopy calibrations for sorghum desaparición do sistema de cotas na UE. Documento silage. Kansas Agricultural Experiment Station Research 13/2015. Foro Económico de Galicia, Ourense, España. Reports 0(1):90–93. doi: 10.4148/2378-5977.2296 Mark H; Workman J. 2003. Statistics in Spectroscopy. Farré I; Faci JM. 2006. Comparative response of maize (Zea Academic Press, Amsterdam, The Netherlands. doi: mays L.) and sorghum (Sorghum bicolor L. Moench) 10.1016/B978-0-12-472531-7.X5040-9 to deficit irrigation in a Mediterranean environment. Marten GC; Goodrich RD; Schmid AR; Meiske JC; Jordan Agricultural Water Management 83(1–2):135–143. doi: RM; Linn JG. 1975. Evaluation of laboratory methods 10.1016/j.agwat.2005.11.001 for determining quality of corn and sorghum silages: II. Calsamiglia S; Ferret A; Bach A. 2016. Tablas FEDNA de valor Chemical methods for predicting in vivo digestibility. nutritivo de forrajes y subproductos fibrosos húmedos. Agronomy Journal 67(2):247–251. doi: 10.2134/agronj197 Fundación Española para el Desarrollo de la Nutrición 5.00021962006700020019x Animal (FEDNA). Madrid, España. bit.ly/3BiPfSk Martens H; Næs T. 1987. Multivariate calibration by data Flores-Calvete G; Martínez-Fernández A; Doltra J; García- compression. En: Williams P; Norris K, eds. Near- Rodríguez A; Eguinoa-Ancha P. 2017. Estructura y infrared technology in the agricultural and food industries sistemas de alimentación de las explotaciones lecheras American Association of Cereal Chemists, Inc. St. Paul, de Galicia, Cornisa Cantábrica y Navarra. INIA, Madrid, MN, USA España. bit.ly/3d2H8A8 Massart DL, Vandeginste BGM, Buydens LMC, De Jong Goering HK; Van Soest PJ. 1970. Forage fiber analysis S, Lewi PJ, Smeyers-Verbeke J 1997. Handbook of (apparatus, reagents, procedures and some applications). Chemometrics and Qualimetrics, Part A. Data Handling Agriculture Handbook No. 379, USDA, Washington, DC, in Science and Technology, volume 20, part A. Elsevier: USA. bit.ly/3x3jMB3 Amsterdam. bit.ly/3cOB8L2 Gosselink JMJ; Dulphy JP; Poncet C; Jailler M; Tamminga Pereira-Crespo S; Díaz N; Piñeiro J; Plata-Reyes D; Gómez- S; Cone JW. 2004. Prediction of forage digestibility in Miranda A; González-Alcántara F; Botana A; Veiga M; ruminants using in situ and in vitro techniques. Animal Valladares J; González L; Resch C; Flores-Calvete G. Feed Science and Technology 115(3–4):227–246. 10.1016/j. 2018. El sorgo como alternativa forrajera al cultivo del anifeedsci.2004.01.008 maíz. Vaca Pinta 5:116–124. bit.ly/3cUSYMn Heuzé V; Tran G; Giger-Reverdin S; Lebas F. 2015. Sorghum Pereira-Crespo S; Botana A; Veiga M; González L; Resch forage. Feedipedia, a programme by INRA, CIRAD, AFZ C; García-Souto V; Martínez-Diz MP; Lorenzana and FAO. feedipedia.org R; Flores-Calvete G. 2022a. Prediction of quality of Hicks C; Tuinstra MR; Pedersen JF; Dowell FE; Kofoid KD. total mixed ration for dairy cows by near infrared 2002. Genetic analysis of feed quality and seed weight reflectance spectroscopy and empirical equations. of sorghum inbred lines and hybrids using analytical Journal of Applied Animal Research 50(1):69–79. doi: methods and NIRS. Euphytica 127(1):31–40. doi: 10.1080/09712119.2021.2022488 10.1023/A:1019943805514 Pereira-Crespo S; Botana A; Veiga M; González L; Resch Horwitz W; Latimer GW 2010. Official methods of analysis C; García-Souto V; Lorenzana R; Martínez-Diz MP; of AOAC International. En: Horwitz W, ed. 1997. Volume Flores-Calvete G. 2022b. Modelos de predicción del valor I, agricultural chemicals, contaminants, drugs. AOAC nutricional de henos de hierba. Información Técnica International, Gaithersburg, MD, USA. Económica Agraria (ITEA) 118(3):377–395. doi: 10.12706/ INRA. 2010. Alimentation des bovins, ovins et caprins. itea.2021.031 Besoins des animaux-Valeurs des aliments. Tables INRA Shenk JS; Westerhaus MO. 1991. Population definition, sample 2007. Institut National de la Recherche Agronomique, selection, and calibration procedures for near infrared (INRA) Éditions Quae, Versailles, France. bit.ly/3Bg3ck2 reflectance spectroscopy. Crop Science 31(2):469–474. Juan NA; Ortiz DA; Pordomingo AB; Funaro DO. 2016. doi: 10.2135/cropsci1991.0011183X003100020049x Tecnología NIRS para estimar el valor nutritivo de Shenk JS; Westerhaus MO. 1993. Near infrared planta entera de maíz y sorgo para silaje y sus fracciones reflectance analysis with single and multiproduct (tallo, hoja, panoja/espiga). Revista de la Asociación calibrations. Crop Science 33(3):582–584. doi: 10.2135/ Argentina de Producción Animal 36:230. hdl.handle.net/ cropsci1993.0011183X003300030032x 20.500.12123/771 Shenk JS; Westerhaus MO. 1994. The application of near Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) 260 S. Pereira-Crespo et al. infrared reflectance Spectroscopy (NIRS) to forage vitro digestion of forage crops. Grass and Forage Science analysis. En: Fahey GC, ed. Forage Quality, Evaluation, 18(2):104–111. doi: 10.1111/j.1365-2494.1963.tb00335.x and Utilization. ASA, CSSA & SSSA, Madison, WI, Williams PC; Sobering DC. 1996. How do we do it: a brief USA. p. 406–449. doi: 10.2134/1994.foragequality.c10 summary of the methods we use in developing near Shenk JS; Westerhaus MO. 1996. Calibration the ISI way. infrared calibrations. En: Davies AMC; Williams PC, En: Davies AMC; Williams PC, eds. Near Infrared eds. Near Infrared Spectroscopy: The Future Waves. NIR Spectroscopy: the future waves. Proceedings of the 7th Publications, Chichester, United Kingdom. p. 185–188. International Conference on Near Infrared Spectroscopy. Williams PC. 2014. The RPD statistic: a tutorial note. NIR NIR Publications, Chichester, West Sussex, United News 25(1):22–26. doi: 10.1255/nirn.1419 Kingdom, p. 198–202. Windham WR; Mertens DR Barton II FE. 1989. Protocol Shenk JS; Workman JJ; Westerhaus MO. 2001. Application of for NIRS calibration: sample selection and equation NIR spectroscopy to agricultural products. En: Burns DA; development and validation. En: Marten GC; Shenk JS; Ciurczak EW, eds. Handbook of Near Infrared Analysis Barton FE, eds. Near Infrared Reflectance Spectroscopy Marcel Dekker, New York, USA. p. 419–474. (NIRS): Analysis of forage quality USDA Handbook Nº Tilley JMA; Terry RA. 1963. A two-stage technique for the in 643, Washington, DC, USA. p. 96–103. bit.ly/3exbJpW (Recibido para publicación 11 enero 2022; aceptado 26 agosto 2022; publicado 30 septiembre 2022) © 2022 Tropical Grasslands-Forrajes Tropicales is an open-access journal published by International Center for Tropical Agriculture (CIAT), in association with The Tropical Crops Genetic Resources Institute of The Chinese Academy of Tropical Agricultural Sciences (TCGRI-CATAS). This work is licensed under the Creative Commons Attribution 4.0 International (CC BY 4.0) license. Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) Tropical Grasslands-Forrajes Tropicales (2022) Vol. 10(3):261–270 261 doi: 10.17138/TGFT(10)261-270 Research Paper Ingestive behavior and dry matter intake of dairy cattle grazing Kikuyu grass (Cenchrus clandestinus) pastures Comportamiento ingestivo y consumo de materia seca de vacas lactantes en pastoreo de pasto kikuyo (Cenchrus clandestinus) YESID AVELLANEDA-AVELLANEDA, EDGAR MANCIPE-MUÑOZ AND JUAN VARGAS-MARTÍNEZ Corporación Colombiana de Investigación Agropecuaria, Agrosavia, Cundinamarca, Colombia. agrosavia.co Abstract The objective of this study was to evaluate the effect of animal characteristics, grazing management, and supplementation on ingestive behavior and dry matter intake (DMI) of Kikuyu grass in lactating cows. Four trials were conducted with multiparous Holstein dairy cows in non-limiting forage conditions using 9 cows in each trial, 1 cow per paddock. Individual DMI was estimated through forage mass difference (pre- and post-grazing mass), ingestive behavior, and using markers [chromium oxide and undegradable acid detergent fibre (uADF)]. DMI was also estimated using 3 nutritional models (CSIRO, NRC and AFRC). Grazing time and bite mass were positively related to the cow body weight, while bite rate showed a negative relationship with forage mass. The grazing time on a pasture of 42 d regrowth was less than the time spent grazing on a pasture of 28 or 56 d regrowth. DMI estimated by forage mass difference showed a positive relation with forage mass, supplement intake and liveweight. DMI estimated using markers showed a positive relation with milk production and liveweight and a negative relationship with forage height. Forage mass difference and ingestive behavior measurements provided good estimates (R2>0.8) of DMI associated with forage mass, liveweight and supplement intake in cows grazing Kikuyu grass. Keywords: Bite mass, bite rate, external and internal markers, grassland systems, grazing time. Resumen El objetivo de este estudio fue evaluar el efecto de las características de los animales, el manejo del pastoreo y la suplementación sobre el comportamiento ingestivo y el consumo de materia seca (CMS) del pasto kikuyo en vacas lactantes. Se realizaron cuatro ensayos con vacas lecheras Holstein multíparas en condiciones en que la provisión de forraje no fue una limitante utilizando 9 vacas en cada ensayo, 1 vaca por potrero. El CMS individual se estimó a través de la diferencia de la disponibilidad de forraje (antes y después del pastoreo), el comportamiento ingestivo y el uso de marcadores [óxido crómico y fibra detergente ácida no degradable (uADF)]. El CMS también se estimó utilizando 3 modelos nutricionales (CSIRO, NRC y AFRC). El tiempo de pastoreo y tamaño de bocado se relacionaron positivamente con el peso corporal de la vaca, mientras que el tamaño de bocado mostró una relación negativa con la disponibilidad de forraje. El tiempo de pastoreo fue menor en una pradera de 42 días de rebrote que en las de 28 o 56 días. El CMS estimado por diferencia en la disponibilidad de forraje mostró una relación positiva con la masa de forraje, el consumo de suplementos y el peso vivo. El CMS estimado mediante marcadores mostró una relación positiva con la producción de leche y el peso vivo y una relación negativa con la altura del forraje. Las mediciones de la diferencia en la disponibilidad de forraje y el comportamiento ingestivo proporcionaron buenas estimaciones (R2>0.8) del CMS asociado con la disponibilidad de forraje, el peso vivo y el consumo de suplementos en vacas que consumen pasto kikuyo. Palabras clave: Masa de bocado, marcadores externos e internos, sistemas de pastizales, tasa de bocado, tiempo de pastoreo. Correspondence: Yesid Avellaneda-Avellaneda, Centro de Investigación Tibaitatá, Corporación Colombiana de Investigación Agropecuaria, Agrosavia. Km 14 vía Bogotá-Mosquera, Cundinamarca, Colombia. Email: yavellaneda@agrosavia.co Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) 262 Y. Avellaneda-Avellaneda, E.A. Mancipe-Muñoz and J.J. Vargas-Martínez Introduction measure in the field, such as forage mass, plant height, supplement supply, cow liveweight and grazing time, can The main factor that defines animal performance in be used to make more accurate predictions of Kikuyu ruminants is dry matter intake (DMI) (Sollenberger and grass intake in lactating cows. Vanzant 2011). Physiological and physical constraints, optimization of oxygen consumption and animal behavior Materials and Methods have been used to explain DMI by ruminants in different contexts (NASEM 2016). However, physical rumen gut Animal management and procedures were approved by fill and animal behavior are more related to DMI of the bioethics committee of the Corporación Colombiana ruminants in grassland conditions (Boval et al. 2015; de Investigación Agropecuaria (Agrosavia) act number Sollenberger et al. 2020a). Also, supplementation has an 029. Four experiments were conducted in the dairy unit associative effect on DMI in ruminants because it may at Tibaitatá research center, Agrosavia, at 2516 masl maintain or increase forage intake and increase total DMI (latitude 4°35´56´́ N, longitude 74°04´51́ ´ W) and a (additive effect) or reduce forage intake but increase total mean temperature of 16 °C in Mosquera, Colombia. Two DMI (substitutive effect) (Bargo et al. 2003). hectares of pre-established Kikuyu grass were used. Dairy production feeding systems in the Colombian The area was mowed at 10 cm and lime (2 t lime/ha), highlands consist of forages, especially Kikuyu urea (100 kg urea/ha) and DAP (50 kg DAP/ha) were grass (Cenchrus clandestinus), plus concentrate applied following the recommendation of ICA (1992). supplementation (Carulla and Ortega 2016). Kikuyu The area was divided into 18 separately fenced paddocks grass is a C4 species that tolerates acid soils, drought (approximately 1,100 m2 each), with 9 paddocks used in conditions and poor management, resulting in low animal each of the 4 experiments conducted. productivity (Vargas et al. 2018). Literature suggests that good management of Kikuyu grasslands and appropriate Cow management and experimental design supplementation may promote high milk production and farm profitability (Fariña et al. 2011). There is interest Multiparous Holstein dairy cows were used in each of in understanding the environmental and management the 4 experiments. Kikuyu grass was offered at 3 kg factors that modify Kikuyu grass productivity and forage dry matter/100 kg liveweight to ensure forage nutritive value to define management recommendations mass was not limiting (Correa et al. 2008). Each cow for increasing ruminant performance (Fonseca et al. was assigned to an individual paddock with water 2016; Escobar et al. 2020; Avellaneda et al. 2020). ad libitum. Supplementation was supplied at the milking In Colombia, DMI of dairy cows in Kikuyu grass parlor twice per day. Each trial was implemented for pastures has been evaluated using external and internal 15 days. The first 10 days were an adaptation period to markers (Aguilar et al. 2009; Correa et al. 2009; Mojica management and supplement intake and the last 5 days et al. 2009; Parales et al. 2016) or by calculating the were the measurement period. difference between the forage mass on offer and the Experiment 1: Effects of cow liveweight and level of milk forage mass remaining following a grazing event (Gómez- production. Nine cows with different liveweight (low: Vega et al. 2019). Studies have evaluated and modelled 441±14 kg; medium: 502±21 kg; high: 676±38 kg) and the effect of different animal characteristics such as milk milk production (low: 9.0±0.6 L/d; medium: 11.9±1.7 production or liveweight (NRC 2001; CSIRO 2007), L/d; high: 17.3±0.9 L/d) were allocated, 1 cow per supplementation level (Alderman and Cottrill 1993), paddock, to 9 paddocks after 43 days of Kikuyu grass forage management such as grazing frequency or time regrowth with 3 fence movements throughout the day (Abrahamse et al. 2008) or ingestive behavior (Boval and (06.00, 10.00 and 15.00 h). In addition to grazing, cows Sauvant 2019) on the DMI. This approach has not been received 1 kg supplement per 4.25 kg of milk produced. thoroughly evaluated in milk production systems of the Measurements were taken per cow per paddock for each Colombian highland tropics or used for development of combination of liveweight and milk production. models specific to the production system and conditions Experiment 2: Effect of different lengths of regrowth of the region. This research aimed to evaluate the relation period. Nine cows with similar milk production between animal characteristics, grazing management and (13.5±2.7 L/d) but different liveweight (low: 435±6 kg; supplementation amount on DMI and animal behavior. medium: 502±27 kg; high: 657±75 kg) were allocated, 1 We hypothesized that using variables that are easy to cow per paddock, to each of 9 paddocks. Three different Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) Dry matter intake of dairy cattle grazing Kikuyu 263 regrowth periods (28, 42 or 56 d) of Kikuyu grass were analyzed using the near-infrared spectroscopy (NIRS) used and cows received 1 kg supplement per 4 kg of methodology (Ariza-Nieto et al. 2017). The agronomic milk produced. The experimental unit was a cow in and chemical composition of Kikuyu grass, and the an individual paddock with 3 replicates per treatment. chemical composition of supplements of each experiment Regardless of treatment, there were 3 fence movements are presented in tables 1 and 2, respectively. throughout the day (06.00, 10.00 and 15.00 h). Experiment 3: Effect of number of times cows were Table 1. Agronomic characteristics and chemical composition moved to a new, ungrazed area in the paddock per day. of Kikuyu grass. Nine cows with similar body weight (500±33 kg) and Agronomic Trial milk production (14.2±1.9 L/d) were allocated, 1 cow per characteristics 1 2 3 4 paddock, to each of 9 paddocks. Daily forage availability Regrowth period (d) 43 28 42 56 43 43 was varied by using electric-fence movements at 2 (6.00 Plant height (cm) 17.0 10.2 18.3 25.0 19.1 16.9 and 14.30 h), 4 (6.00, 10.00, 12.00 and 14.30 h) or 6 (6.00, Pre-grazing mass 9.00, 10.00, 11.00, 12.00 and 14.30 h) times throughout (kg DM/m2) 0.17 0.11 0.18 0.21 0.12 0.11 the day with cows also receiving 1 kg supplement per 4 Chemical composition % DM kg milk produced. The experimental unit was a cow in Dry matter1 24.6 25.8 23.4 19.6 15.0 16.6 an individual paddock with 3 replicates per treatment. Crude protein 12.8 16.6 14.2 13.8 19.5 18.1 Experiment 4: Effect of rate of supplementation and Neutral detergent milk production of cow. Nine cows with different fiber 61.4 56.6 59.4 60.0 55.8 56.4 milk production (low: 11.9±0.4 L/d; medium: 15.4±1.0 Acid detergent fiber 34.4 32.7 33.0 33.4 32.4 33.2 L/d; high: 19.1±1.8L/d) but similar liveweight (578+53 Calcium 0.30 0.27 0.32 0.27 0.20 0.24 kg) were allocated, 1 cow per paddock, to each of 9 Phosphorus 0.34 0.32 0.36 0.34 0.40 0.39 individual paddocks with a regrowth period of Kikuyu NEL (Mcal/kgDM) 1.20 1.28 1.24 1.22 1.33 1.30 grass of 43 days and 3 fence movements throughout the 1 (% as fed) day (6.00, 10.00 and 15.00 h). Cows with similar milk production and lactating days were randomly assigned to Table 2. Ingredients and chemical composition of supplements. 1 of the 3 supplementation rates (1 kg of the supplement per 2, 3 or 4 kg of milk produced). Measurements were Item Trial taken per cow per paddock for each combination of milk 1 2 3 4 production and supplementation rate. Corn grain meal 26.1 23.0 23.0 23.0 Bakery residues 15.0 10.0 12.0 12.0 Forage management, supplement composition and Glycerin 20.0 25.0 25.0 25.0 chemical analysis Cottonseed, whole 20.0 23.0 23.0 23.0 Distiller's dried grains with Pre-grazing and post-grazing forage mass were measured solubles (%) 18.9 19.0 17.0 17.0 in each paddock during the last 5 days of each experimental Chemical composition % DM period. Pre-grazing forage mass was measured using the Dry matter 91.0 91.8 92.6 89.9 plate-meter (EC-10, Jenquip®), while quantification of Crude protein 12.2 13.2 11.0 12.8 post-grazing forage mass was done using a metric ruler Neutral detergent fiber 20.7 25.5 22.2 21.9 following the methodology of Avellaneda et al. (2020) because the resting cows crushed the grass, affecting the Acid detergent fiber 12.0 17.2 11.5 15.0 measurement with the forage plate-meter. Pre-grazing Ether extract 5.55 5.62 5.31 5.26 forage samples for each paddock were collected, dried Calcium 0.11 0.11 0.10 0.11 and conserved for subsequent analysis. Supplements were Phosphorus 0.51 0.48 0.50 0.48 manufactured for each experiment to supply the animal NE L (Mcal/kgDM) 1.87 1.74 1.88 1.79 requirements (NRC 2001) and offered individually at the milking parlor. A sample of each supplement was retained Variables evaluated for subsequent analysis. During the measurement period, orts of each supplement were weighed to calculate Individual DMI was estimated using different the supplement intake. Forages and supplements were methodologies. Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) 264 Y. Avellaneda-Avellaneda, E.A. Mancipe-Muñoz and J.J. Vargas-Martínez a. Forage mass difference: Forage intake was Equation 5: Feces=(Chromium supplied + concentration calculated individually as the difference between pre- of chromium in the supplement × supplement intake) / and post-grazing forage mass. Total DMI was defined as (Concentration of chromium in feces/0.79) forage intake plus supplement intake. d. Model estimation: Intake was estimated using Equation 1: Intake (kg/d)=(Pre-grazing biomass – post- equations described in NRC (2001), CSIRO (2007) and grazing biomass) + supplement AFRC (Alderman and Cottrill 1993) for dairy cows, b. Ingestive behavior: Forage intake was estimated respectively. as the product between grazing time, bite rate, and bite Equation 6 (NRC): Intake=(fat-corrected milk × 0.372 mass. Total DMI was defined as the addition of forage + 0.0968 × body weight0.75) × (1-e (-0.192*(Days in milk) +3.67)) and supplement intake. Fat-corrected milk is milk adjusted on a 4 % fat basis Equation 2: Intake (kg/d)=(grazing time × bite rate × ((0.4+(0.15*milk fat(%)))*milk production). bite mass) + supplement Equation 7 (CSIRO): Intake=Potential intake × intake Animal behavior was classified as grazing, level ruminating and resting. The grazing time was defined Equation 8 (AFRC): Intake=0.076 + 0.404 × following the animal behavior during each experimental concentrate intake + 0.013 × Body weight + 4.12 × log period. Each animal was observed every 10 min for 24 10 (days in milk) – 0.129*n + 0.14 × milk production h during the measurement period of each trial. Grazing Milk production was measured for each cow at 1, 3 and time was calculated as the time that animals spent in 5 days of each measurement period. A milk sub-sample grazing activity. Bite rate was calculated as the number of each animal per measurement day was analyzed for of bites during 5 min, observed every 15 min during the protein and fat (MilkoScanTM FT120, AOAC 2016). grazing period. Mouth movements during rumination (rumination rate) were calculated for 5 min observed Statistical analysis every 15 min during the rumination period. Bite mass was defined through 2 different approaches. Initially, Data on feeding behavior of different trials were a hand-picked sample was determined considering the evaluated with regression analysis. The independent width and depth of the bites of each cow, mimicking the variables were days in milk, liveweight, metabolic ingestive behavior. Also, bite mass was estimated using liveweight, pre-grazing forage mass, milk production, the relation between bite mass and liveweight (equation corrected milk production, milk fat concentration, milk 3, Boval and Sauvant 2019). The methodologies to protein concentration and supplement intake. The REG estimate bite mass were applied each day during the procedure was used for linear regression. The stepwise measurement period of each trial. selection method, assessing contributions of effects as Equation 3: Log10 Bite mass=0.20 + 0.97 × Log10 Body they were added to or removed from the model, was used Weight to select the explicative variables (P<0.05, SAS 2017). c. Markers: Internal and external markers were used Cow behavior of experiments 2 and 3 were analyzed to estimate forage intake (Correa et al. 2009). Cows as a completely randomized design using a GLM received 10 g of chromium oxide (Cr2O3), divided into procedure (SAS 2017), where the fixed effect was the 2 doses daily, to estimate fecal production, assuming regrowth period or the movements of the electric fence, 79 % of chromium-marker recovery rate (Lippke 2002; respectively, and the error was the variation of each cow Correa et al. 2009). Feces were collected twice a day between measurements. Differences were considered during the measurement period of each trial. Feces with an alpha value lower than 5 %. The linear and were dried and mixed by cow per period. Undegradable quadratic responses of fixed effects were determined. acid detergent fibre (uADF) at 144 h of incubation and The individual DMI using forage mass difference, cow chromium concentration were calculated for forage behavior and markers were calculated through regression samples, supplements and feces. The recovery of uADF analysis. The independent variables were pre-grazing was assumed as 0.8 (Sunvold and Cochran 1991). forage mass, forage height, animal activity, bite rate, Equation 4: Intake=((Feces × concentration of uADF rumination rate, bite mass, supplement intake, liveweight, in feces/0.8) - (supplement intake × concentration of metabolic liveweight, milk production, corrected milk uADF in supplement))/ concentration of uADF in production, milk fat and protein concentration. The REG forage) + supplement procedure and the stepwise option were used to select the Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) Dry matter intake of dairy cattle grazing Kikuyu 265 explicative variables (P<0.1, SAS 2017). Similarly, the Grazing time showed a positive relationship with cow DMI of experiments 2 and 3 were analyzed as a completely liveweight (Table 4). While rumination was negatively randomized design using a GLM procedure (SAS 2017). related to pre-grazing forage mass and supplement intake, DMI using different approaches was evaluated through it positively correlated with corrected milk production. the Pearson correlation. The percentage and absolute Inversely, time resting showed a positive relationship mean bias error were defined to evaluate the relationship with pre-grazing forage mass and supplement intake between different methodologies. and a negative relationship with the fat-corrected milk production (Table 4). The bite rate was negatively related Results to pre-grazing forage mass, while the rumination rate had a positive relationship with supplement intake (Table 4). Behavior and intake characteristics in dairy cows Only the bite mass, using the hand-picked methodology, was positively related to the animal's liveweight (Table 4). Dairy cows spent 18, 30 and 39 % of time resting, grazing Regrowth period of the Kikuyu grass affected the and ruminating throughout the day, respectively (Table proportion of time spent in grazing (P<0.05) but not the 3). The average bite and rumination rate were 0.55 bite/ duration of resting or rumination (P>0.05). Regrowth sec and 1.03 bite/sec, respectively (Table 3). Regardless period did not affect the bite rate, rumination rate, or of the methodology, the average bite mass was 0.71g bite mass (P>0.05). Conversely, more fence movements DM/bite or 0.72 g DM/bite (Table 3). increased resting and decreased rumination times (P<0.05) but did not affect grazing time (P>0.05). Fence Table 3. Description of animal behavior and intake traits of movement did not change the bite rate, rumination rate or dairy cows in grassland systems. bite mass (P>0.05) (Table 5). Variable Unit Mean Range Grazing h/d 7.2 5.1 - 8.8 Estimation of DMI using different methodologies Ruminating h/d 9.3 6.4 - 11.4 The average DMI in dairy cows was estimated between Resting h/d 4.2 0.7 - 6.8 13.7 and 14.2 kg/d using the different methodologies Bite rate times/s 0.55 0.41 - 0.70 (Table 6). The linear regression of DMI, calculated by Rumination rate times/s 1.03 0.94 - 1.12 different methodologies according to the variables of Bite mass1 g DM/bite 0.71 0.55 - 0.99 forage, ingestive behavior, and animal performance, Bite mass2 g DM/bite 0.72 0.30 - 1.13 is presented in Table 7. Pre-grazing forage mass, 1Hand-picked sample simulating animal bite. supplement intake and metabolic body weight variables 2Log10 Bite mass = 0.20 + 0.97 × Log10 Body Weight (Boval proved suitable for estimating DMI, calculated as the and Sauvant 2019). difference between pre- and post-grazing forage mass. Table 4. Regression analysis of cow behavior or intake traits with forage mass, supplement intake and animal characteristics in dairy cows (mean from all experiments). Variable Intercept FM IS CMP BW R2 P Circadian behavior (h/d) Grazing 5.27*** 0.0035* 0.12 * Rumination 9.94*** -0.98+ -0.34* 0.16* 0.20 + Rest 3.51* 1.23+ 0.37+ -0.18+ 0.15 ns Intake behavior (times/s) Bite rate 0.61*** -0.06* 0.11 * Rumination rate 0.95*** 0.014* 0.17 * Bite mass (g DM/bite) Hand-picked 0.32+ 0.00074* 0.15 * FM = pre-grazing forage mass (t/ha); IS = intake of supplement (kg/d); CMP = fat-corrected milk production; BW = liveweight; ns = non-significant; + = P< 0.1; * = P< 0.05; *** = P< 0.001. Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) 266 Y. Avellaneda-Avellaneda, E.A. Mancipe-Muñoz and J.J. Vargas-Martínez Table 5. Behavior of lactating cows under different grass-regrowth periods and electric fence movement schemes in Kikuyu pastures. Variable Regrowth period (d) Fence movement MSE Effect MSE3 Effect4 28 42 56 2 4 6 Circadian behavior (h/d) Grazing 7.3a 5.9b 7.8a 0.34 C 7.1 7.6 6.4 0.61 ns Rumination 9.3 9.2 9.0 0.74 ns 10.8a 9.4ab 7.9b 0.60 L Rest 3.9 5.4 3.7 0.67 ns 2.6b 3.5b 6.1a 0.87 L Intake behavior (times/s) Bite rate 0.55 0.50 0.49 0.04 ns 0.5 0.5 0.5 0.02 ns Rumination rate 1.03 1.03 1.09 0.04 ns 1.1 1.0 1.0 0.03 ns Bite mass (g DM/bite) Hand-picked1 0.63 0.94 0.87 0.21 ns 0.7 0.6 0.7 0.09 ns Estimated2 0.67 0.66 0.74 0.08 ns 0.7 0.7 0.6 0.05 ns 1Hand-picked sample simulating animal bite; 2Log10 Bite mass = 0.20 + 0.97 × Log10 Body Weight (Boval and Sauvant 2019); 3MSE = mean square error; 4L = lineal effect; ns = not significant. Different letters in the same row mean significant differences (P<0.05). Table 6. DMI (kg/d) of dairy cows estimated using different methodologies in Kikuyu grassland systems. Variable Mean Range Forage mass difference 13.7 8.7 – 19.1 Ingestive behavior1 13.8 8.7 – 19.9 Ingestive behavior2 14.2 7.7 – 21.0 Markers 13.8 10.2 – 18.9 1Hand-picked sample simulating animal bite; 2Log10 Bite mass=0.20 + 0.97 × Log10 Body Weight (Boval and Sauvant 2019). Table 7. Regression analysis of DMI (kg/d) estimation using different methodologies in dairy cattle. Variable Intercept FM FH GT BR BS IS MP BW MBW R Forage mass difference -1.18ns 3.72*** 0.55** 0.078*** 0.81 Ingestive behavior1 -18.51* 1.54* 16.98* 11.30* 0.02* 0.84 Markers 8.94** -0.11* 0.27* 0.01* 0.38 FM = pre-grazing forage mass (t/ha); FH = forage height (cm); GT = grazing time (h/d); BR = bite rate (times/sec); IS = intake of supplement (kg/d); MP = milk production (kg/d); BW = liveweight (kg); MBW = metabolic live weight (kg); 1Hand-picked sample simulating animal bite; *** = P<0.0001; * = P<0.05; ns = not significant. The estimation of DMI using ingestive behavior had a number of fence movements did not affect the estimation positive relationship with grazing time, bite rate, bite of DMI using other methodologies (Table 8). mass, and supplement intake. The estimation of DMI DMI estimated as the difference between pre- and using markers showed a positive relationship between post-grazing showed a positive correlation (0.62) with the milk production and body weight and a negative NRC model. Estimation of DMI using ingestive behavior relationship with forage height. The coefficients of and calculating the bite mass (Boval and Sauvant 2019) determination to estimate DMI through forage mass had a positive correlation (0.64 and 0.70) with the AFRC difference or ingestive behavior were greater than and NRC models, respectively. The estimation of DMI internal and external markers (Table 7). using the 2 methodologies of ingestive behavior showed a The regrowth period did not affect DMI, regardless positive correlation (0.68) between them. DMI estimated of the methodology used to determine intake (P>0.05). with the AFRC model had a positive correlation (0.68 Greater fence movements increased the DMI only when and 0.78) with NRC and CSIRO models, respectively. calculated as the difference between pre- and post- Estimation of DMI with internal and external markers did grazing. DMI increased 36 % when the electric fence not show a significant relationship with other estimation was moved 6 vs 2 times throughout the day. However, the methodologies or the CSIRO model (Table 9). Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) Dry matter intake of dairy cattle grazing Kikuyu 267 Table 8. DMI (kg/d) of lactating cows under different grass-regrowth periods and electric fence movements schemes in Kikuyu pastures. Regrowth period (d) Fence movement (times/day) Variable MSE3 Effect4 MSE Effect 28 42 56 2 4 6 Forage mass difference 11.9 12.6 14.7 1.31 ns 12.0b 12.3b 16.5a 1.0 L Ingestive Behavior1 13.4 13.0 15.5 2.64 ns 13.0 13.0 12.0 1.4 ns Ingestive Behavior2 13.6 10.7 13.8 1.56 ns 12.4 13.3 11.3 0.9 ns Markers 15.4 12.7 14.8 3.94 ns 12.1 12.7 12.5 1.1 ns 1Bite mass calculated as a hand-picked sample simulating an animal bite; 2Bite mass calculated using Log10 Bite mass = 0.20 + 0.97 × Log10 Body Weight (Boval and Sauvant 2019); 3MSE = mean square error; 4L = lineal effect; ns = not significant. Different letters in the same row mean significant differences (P<0.05). Table 9. Pearson correlations between methodologies and models to estimate DMI in dairy cows. Methodology Behavior1 Behavior2 Markers NRC CSIRO AFRC Mass difference 0.50** 0.51** 0.18ns 0.62*** 0.11ns 0.41* Ingestive Behavior1 0.68*** 0.16ns 0.46** 0.32+ 0.47** Ingestive Behavior2 0.25ns 0.70*** 0.49** 0.64*** Markers 0.47** 0.21ns 0.31+ NRC 0.44** 0.68*** CSIRO 0.78*** 1Bite mass calculated as a hand-picked sample simulating an animal bite; 2Bite mass calculated using Log10 Bite mass=0.20 + 0.97 × Log10 Body Weight (Boval and Sauvant 2019). Table 10. Percentage difference between estimates of DMI in dairy cows (absolute means above the diagonal while percentage of change under the diagonal) using different methodologies. Methodology Forage mass difference Behavior1 Behavior2 Markers NRC CSIRO AFRC Forage mass difference 2.28 1.85 2.21 2.11 2.30 1.89 Ingestive Behavior1 3.33 % 1.92 2.79 2.50 2.45 2.56 Ingestive Behavior2 0.63 % 2.73 % 2.56 2.20 2.28 1.92 Markers -0.07 % 3.26 % 0.55 % 2.24 2.31 2.18 NRC -12.14 % -8.39 % -11.44% -12.05 % 1.67 2.86 CSIRO -6.84 % -3.27 % -6.18 % -6.76 % 4.71 % 2.11 AFRC 8.41 % 11.46 % 8.98 % 8.47 % 18.31 % 14.27 % 1Bite mass calculated as a hand-picked sample simulating an animal bite; 2Bite mass calculated using Log10 Bite mass=0.20 + 0.97 × Log10 Body Weight (Boval and Sauvant 2019). The NRC and CSIRO models overestimated (i.e. different levels of intake and production. Dairy cows on negative percentage bias), while the AFRC model ryegrass and clover pastures spent 38 % of their time underestimated (i.e. positive percentage bias) DMI grazing (Rombach et al. 2019). The shorter grazing time calculated through different forage mass approaches. on Kikuyu grass in this experiment may be explained by Also, the AFRC model showed closer estimations of a greater concentration of neutral detergent fiber relative DMI (i.e. lower absolute bias) with respect to the other to ryegrass (Vargas et al. 2014; Aguilar et al. 2009), models. Ultimately, the estimation of DMI through constraining the total daily intake due to a lower passage forage mass difference had the lowest absolute bias rate and physical restriction (Allen 2000; NASEM 2016). relative to other methodologies (Table 10). Ruminants can increase DMI in diets with a lower concentration of structural carbohydrates (Mertens Discussion 1987). However, a similar concentration of structural carbohydrates in Kikuyu grass across regrowth periods Grazing behavior is affected by internal and external precluded reaching any conclusions on their effect on factors that modify the animal response, resulting in DMI in lactating cows in the current study. Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) 268 Y. Avellaneda-Avellaneda, E.A. Mancipe-Muñoz and J.J. Vargas-Martínez Forage traits may explain animal grazing behavior. mass difference methodology. However, there were Rombach et al. (2019) reported that bite rate and bite no differences in the DMI using other methodologies mass were 1.21 bite/s and 0.47 g DM/bite, respectively, when increasing the frequency of fence movements. in dairy cows grazing ryegrass and clover pastures. The forage mass difference methodology may have an Those values suggested lower DMI per bite relative to implicit methodological bias that limits the accuracy of the current experiments, requiring more grazing time DMI estimation. to supply nutrient requirements. It is recognized that There was a positive but not strong relationship cattle can modulate grazing time, bite mass, or bite between measurements of DMI and those calculated rate according to the forage characteristics (Boval and using nutritional models. Correa et al. (2009) suggested Sauvant 2019). However, the biological ranges across a strong relationship between the DMI estimate with which ruminants can modify these responses under external and internal markers and the NRC (2001) or grazing are not well defined (Sollenberger et al. 2020b). CNCPS (Fox et al.1992) models. However, NRC (2001) Younger forages have greater nutritive value but less and CSIRO (2007) models tended to overestimate, while mass than older ones, requiring more grazing time to Alderman and Cottrill (1993) tended to underestimate acquire the nutrient requirements due to the small bite DMI relative to the measurement methodologies mass. Mature forages show greater forage mass but evaluated. Therefore, it is necessary to recognize the lesser forage quality, increasing the grass selection main factors that influence DMI to determine the most and requiring more grazing time to supply the energy appropriate methodology to define DMI in grazing requirements (Galyean and Gunter 2016). conditions of Kikuyu pastures. The bite mass is associated with the capability of the animal to access forage and is associated with the animal's Conclusions liveweight and forage characteristics (Gordon et al. 1996; Boval and Sauvant 2019; Sollenberger et al. 2020a). Bite DMI is a cornerstone variable, and it is necessary to mass increases in taller forages (Gregorini et al. 2008). identify methodologies that provide more accurate However, long stems reduce bite mass, especially in estimations under grazing conditions. Cow behavior pastures with low bulk density (Galyean and Gunter was related to forage mass, supplement intake and 2016). In the current experiment, there were no bite mass animal traits. Frequency of fence movements affected differences among the regrowth periods. However, there cow behavior, while grazing Kikuyu pastures at an was a positive correlation between the grazing time, bite intermediate regrowth period of 42 d reduced the mass, and bite rate with DMI, suggesting that the animal grazing time. Conversely, average DMI was related response to forage characteristics may modify forage to forage traits, cow behavior and milk production. intake (Holecheck et al. 1995; Sollenberger et al. 2020a). There was a positive but weak relationship between Determining DMI in grazing conditions presents methodologies used to measure intake and the different challenges due to the difficulty of accurately defining models used to predict intake. Ultimately, NRC (2001) the animal response for forage selection, especially in and CSIRO (2007) models overestimated DMI, while diverse pastures or rangeland conditions (Boval and Alderman and Cottrill (1993) underestimated DMI Sauvant 2019). DMI showed different relationships with using the measurement methodologies in the study. forage traits and animal characteristics according to Based on these data, we conclude that measurement of the methodology used to estimate intake with greater forage mass, nutritional quality and cow liveweight are cow liveweight, grazing time, bite mass, bite rate, relatively easy to measure and can be used to estimate supplementation intake and forage mass positively DMI in field conditions. The measurement of the DMI associated with greater DMI. through the other methodologies tested was laborious Forage management may promote or reduce DMI and and required high investment with no consistent results. modify animal behavior and performance (Holecheck et al. 1995). Abrahamse et al. (2008) reported that cows Acknowledgments grazing in a small paddock with frequent rotation showed greater intake than those grazing in bigger The authors thank the Ministry of Agriculture and Rural ones with less rotation. In this experiment, increasing Development, Colombia, for funding and the Colombian the frequency at which new grass was offered increased Agricultural Research Corporation, Agrosavia, for the DMI in dairy cows as calculated using the forage supporting this research. Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) Dry matter intake of dairy cattle grazing Kikuyu 269 References Chiov.) for milk production in Colombia: A review. II. Energy value, intake, production and nutritional efficiency. (Note of the editors: All hyperlinks were verified 13 September 2022). Livestock Research for Rural Development 20(4):61. (In Spanish). lrrd.org/lrrd20/4/corr20061.htm Abrahamse PA; Dijkstra J; Vlaeminck B; Tamminga S. 2008. Correa HJ; Pabón ML; Carulla JE. 2009. Estimation of dry Frequent allocation of rotationally grazed dairy cows matter intake of lactating Holstein cows under grazing changes grazing behavior and improves productivity. in Antioquia. Livestock Research for Rural Development Journal of Dairy Science 91(5):2033–2045. doi: 10.3168/ 21(4):59. (In Spanish). lrrd.org/lrrd21/4/corr21059.htm jds.2007-0579 CSIRO. 2007. Nutrient requirements of domesticated Aguilar OX; Moreno BM; Pabón ML; Carulla JE. 2009. ruminants. Commonwealth Scientific and Industrial Effect of kikuyo (Pennisetum clandestinum) or ryegrass Research Organization. CSIRO Publishing, Collingwood, (Lolium hibridum) intake on conjugated linoleic acid VIC, Australia. concentration and fatty acids composition of milk fat. Escobar MA; Cárdenas EA; Carulla JE. 2020. Effect of Livestock Research for Rural Development 21(4):49. (In altitude and defoliation frequency in the quality and Spanish) lrrd.org/lrrd21/4/agui21049.htm growth of Kikuyu grass (Cenchrus clandestinus). Revista Alderman G; Cottrill BR. 1993. Energy and protein Facultad Nacional Agronomía Medellín 73(1):9121–9130. requirements of ruminants. An advisory manual prepared doi: 10.15446/rfnam.v73n1.77330 by the AFRC Technical Committee on Responses to Fariña SR; Garcia SC; Fulkerson WJ; Barchia IM. 2011. Nutrients. CAB International, Wallingford, UK. Pasture-based dairy farm systems increasing milk Allen MS. 2000. Effects of diet on short-term regulation of feed production through stocking rate or milk yield per cow: intake by lactating dairy cattle. Journal of Dairy Science Pasture and animal responses. Grass and Forage Science 83(7):1598–1624. doi: 10.3168/jds.S0022-0302(00)75030-2 66(3):316–332. doi: 10.1111/j.1365-2494.2011.00795.x AOAC (Official Methods Analysis of the Association of Fonseca C; Balocchi O; Keim JP; Rodríguez C. 2016. Effect of Analytical Communities). 2016. Fat, Lactose, Protein defoliation frequency on yield and nutritional composition and solids in Milk. Mid-Infrared Spectroscopic Method of Pennisetum clandestinum Hochst.ex Chiov. Agro Sur 972.16. AOAC International, Arlington, VA, USA. 44(3):67–76. (In Spanish). doi: 10.4206/agrosur.2016. Ariza-Nieto CJ; Mayorga OL; Mojica B; Parra DM; Afanador- v44n3-07 Tellez G. 2017. Use of LOCAL algorithm with near-infrared Fox DG; Sniffen CJ; O´Connor JD; Russell JB; Van Soest spectroscopy in forage resources for grazing systems PJ. 1992. A net carbohydrate and protein system for in Colombia. Journal of Near Infrared Spectroscopy evaluating cattle diets: III. Cattle requirements and diet 26(1):44–52. doi: 10.1177/0967033517746900 adequacy. Journal of Animal Science 70(11):3578–3596. Avellaneda Y; Mancipe EA; Vargas JJ. 2020. Effect of regrowth doi: 10.2527/1992.70113578x period on morphological development and chemical Galyean MY; Gunter SA. 2016. Predicting forage intake in composition of kikuyu grass (Cenchrus clandestinus) extensive grazing systems. Journal of Animal Science in Colombian’s highlands. Revista CES de Medicina 94(Suppl 6):26–43. doi:10.2527/jas.2016-0523 Veterinaria y Zootecnia 15(2):23–37. (In Spanish). doi: Gómez-Vega S; Caicedo-Pinzón R; Vargas-Martínez JJ. 10.21615/cesmvz.15.2.2 2019. Strategic supplementation effect in a dairy system Bargo F; Muller LD; Kolver ES; Delahoy JE. 2003. Production in Cundinamarca, Colombia. Revista de Investigaciones and digestion of supplemented dairy cows on pasture. Veterinarias del Perú 30(3):1109–1116. (In Spanish). doi: Journal of Dairy Science 86(1):1–42. doi: 10.3168/jds. 10.15381/rivep.v30i3.15302 S0022-0302(03)73581-4 Gordon IJ; Illius AW; Milne JD. 1996. Sources of variation in Boval M; Edouard N; Sauvant D. 2015. A meta-analysis of the foraging efficiency of grazing ruminants. Functional nutrient intake, feed efficiency and performance in cattle Ecology 10(2): 219–226. doi: 10.2307/2389846 grazing on tropical grasslands. Animal 9(6):973–982. doi: Gregorini P; Gunter SA; Beck PA. 2008. Matching plant 10.1017/S1751731114003279 and animal processes to alter nutrient supply in strip- Boval M; Sauvant D. 2019. Ingestive behaviour of grazing grazed cattle: Timing of herbage and fasting allocation. ruminants: meta-analysis of the components of bite mass. Journal of Animal Science 86(4):1006–1020. doi: 10.2527/ Animal Feed Science and Technology 251:96–111. doi: jas.2007-0432 10.1016/j.anifeedsci.2019.03.002 Holecheck JL; Pieper RD; Herbel CH. 1995. Range Carulla JE; Ortega E. 2016. Dairy production systems management: Principles and practices (2nd ed.). Prentice of Colombia: challenges and opportunities. Archivos Hall, Englewood Cliffs, New Jersey, USA. Latinoamericanos de Producción Animal 24(2):83–87. (In ICA (Instituto Colombiano Agropecuario). 1992. Fertilización Spanish). bit.ly/3RJjdEX en diversos cultivos, quinta aproximación. Manual Correa HJ; Pabón ML; Carulla JE. 2008. Nutritional value de Asistencia Técnica No.25. Produmedios. Bogotá, of kikuyu grass (Pennisetum clandestinum Hoechst Ex Colombia. hdl.handle.net/20.500.12324/14124 Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) 270 Y. Avellaneda-Avellaneda, E.A. Mancipe-Muñoz and J.J. Vargas-Martínez Lippke H. 2002. Estimation of forage intake by ruminants Sollenberger LE; Vanzant ES. 2011. Interrelationships among on pasture. Crop Science 42(3):869–872. doi: 10.2135/ forage nutritive value and quantity and individual animal cropsci2002.8690 performance. Crop Science 51(2):420–432. doi: 10.2135/ Mertens DR. 1987. Predicting intake and digestibility using cropsci2010.07.0408 mathematical models of ruminal function. Journal Sollenberger LE; Aiken GE; Wallau MO. 2020a. Managing of Animal Science 64(5):1548–1558. doi: 10.2527/ grazing in forage-livestock systems. In: Rouquette M; jas1987.6451548x Aiken G, eds. Management Strategies for Sustainable Mojica JE; Castro E; León J; Cárdenas EA; Pabón ML; Cattle Production in Southern Pastures. Academic Press, Carulla JE. 2009. Effect of the offer of kikuyu grass Cambridge, MA, USA. 77–100. doi: 10.1016/B978-0-12- and oat silage on milk bovine production and quality 814474-9.00005-0 composition. Revista Corpoica–Ciencia y Tecnología Sollenberger LE; Wallau MO. 2020b. Plant-Herbivore Agropecuaria 10(1):81–90. (In Spanish). doi: 10.21930/ interactions. In: Moore K; Collins M; Jerry C; Redfearn rcta.vol10_num1_art:132 D, eds. Forages: The Science of Grassland Agriculture II, NASEM (National Academies of Sciences, Engineering and 7th Edition. John Wiley & Sons, New York, NY, USA. Medicine). 2016. Nutrient requirements of Beef Cattle: 201–214. doi: 10.1002/9781119436669.ch10 8th Revised Edition. The National Academies Press, Sunvold GD; Cochran RC. 1991. Technical note: Evaluation Washington, DC, USA. doi: 10.17226/19014 of acid detergent lignin, alkaline peroxide lignin, acid NRC (National Research Council). 2001. Nutrient insoluble ash, and indigestible acid detergent fiber as Requirements of Dairy Cattle: 7th Revised Edition. The internal markers for prediction of alfalfa, bromegrass, and National Academies Press, Washington, DC, USA. doi: prairie hay digestibility by beef steers. Journal of Animal 10.17226/9825 Science 69(12):4951–4955. doi: 10.2527/1991.69124951x Parales JE; Pabón ML; Carulla JE. 2016. Supplementation Vargas J; Pabón M; Carulla J. 2014. In vitro methane with corn oil to grazing cows: ruminal fermentation, milk production in the incubation of mixtures of grasses and yield, and fatty acid profile. Revista Brasilera de Zootecnia legumes from colombian high-altitude areas. Archivos 45(11):693–703. doi: 10.1590/S1806-92902016001100008 de Zootecnia 63(243):397–407. (In Spanish). doi: 10.4321/ Rombach M; Südekum KH; Münger A; Schori F. 2019. S0004-05922014000300001 Herbage dry matter intake estimation of grazing dairy Vargas JJ; Sierra AM; Mancipe EA; Avellaneda Y. 2018. cows based on animal, behavior, environmental, and feed Kikuyu, present grass in ruminant production systems variables. Journal of Dairy Science 102(4):2985–2999. intropic Colombian highlands. Revista CES de Medicina doi: 10.3168/jds.2018-14834 Veterinaria y Zootecnia 13(2):137–156. (In Spanish). doi: SAS. 2017. Online Doc ver 9.4. SAS Institute Inc., NC, USA. 10.21615/cesmvz.13.2.4 (Received for publication 5 May 2021; accepted 30 August 2022; published 30 September 2022) © 2022 Tropical Grasslands-Forrajes Tropicales is an open-access journal published by International Center for Tropical Agriculture (CIAT), in association with The Tropical Crops Genetic Resources Institute of The Chinese Academy of Tropical Agricultural Sciences (TCGRI-CATAS). This work is licensed under the Creative Commons Attribution 4.0 International (CC BY 4.0) license. Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) Tropical Grasslands-Forrajes Tropicales (2022) Vol. 10(3):271–279 271 doi: 10.17138/TGFT(10)271-279 Research Paper Digital imaging outperforms traditional scoring methods for spittlebug tolerance in Urochloa humidicola hybrids Las imágenes digitales superan los métodos de evaluación tradicionales para la tolerancia al salivazo en los híbridos de Urochloa humidicola LUIS HERNÁNDEZ, PAULA ESPITIA AND JUAN ANDRÉS CARDOSO Tropical Forages Program, Alliance of Bioversity International-CIAT, Cali, Colombia. alliancebioversityciat.org Abstract American spittlebug species (Hemiptera: Cercopidae) are major pests in Urochloa humidicola (syn. Brachiaria humidicola) cultivars in the neotropics. The U. humidicola breeding program of the Alliance Bioversity-CIAT aims to increase tolerance to spittlebugs. To develop tolerant U. humidicola genotypes, adequate screening methods are needed. Currently, visual scores of plant damage by spittlebugs is the standard method to screen for variation in plant tolerance. However, visual scoring is prone to human bias, is of medium throughput and relies on the expertise of well-trained personnel. In this study, estimations of plant damage from SPAD chlorophyll meter measurements and digital images with visual scoring from an inexpert evaluator and visual scoring from an expert were compared. This information should inform if different methods could be implemented in the U. humidicola breeding program. Time needed to evaluate damage was recorded for each method. Lin’s correlation coefficient, Pearson’s correlation coefficient and broad sense heritability values were calculated. Damage estimated from digital images showed the highest throughput (twice as fast as visual scoring from an expert), high correlations with visual scoring (r>0.80, P<0.0001) and heritability values for plant damage as good or better (>0.7) than those obtained by visual scoring from an expert. Results indicate that digital imaging could improve the efficiency of phenotyping in breeding for increased tolerance to spittlebugs in U. humidicola. Keywords: Aeneolamia varia, Brachiaria, high-throughput phenotyping, host-plant resistance, sensors, tropical forage grasses. Resumen Las especies de salivazo (Hemiptera: Cercopidae) constituyen una importante plaga en cultivos de Urochloa humidicola (sinónimo de Brachiaria humidicola) en el neotrópico. El Programa de Mejoramiento de U. humidicola de la Alianza Bioversity- CIAT tiene como objetivo incrementar la tolerancia al salivazo. Para desarrollar genotipos de U. humidicola tolerantes a estas especies, se necesitan métodos de detección adecuados. Actualmente, las evaluaciones visuales del daño causado por salivazo sobre las plantas es el método estándar para detectar variaciones en la tolerancia de las plantas. Sin embargo, la calificación visual es propensa al sesgo humano, tiene un rendimiento medio y depende de la experiencia de personal bien capacitado. En este estudio, se compararon las estimaciones de daños en las plantas a partir de mediciones del medidor de clorofila SPAD, análisis de imágenes digitales y puntuación visual de un evaluador inexperto y otro experto. Esta investigación debe confirmar si se pueden implementar métodos alternativos de evaluación en el programa de mejoramiento de U. humidicola. Se registró el tiempo necesario para evaluar el daño con cada método. También se calcularon el coeficiente de correlación de Lin, el coeficiente de correlación de Pearson y los valores de heredabilidad en sentido amplio. El daño estimado a partir de imágenes digitales mostró el rendimiento más alto (dos veces más rápido que la puntuación visual de un experto), altas correlaciones con la puntuación visual (r > 0.80, p < 0.0001) y valores de heredabilidad para el daño de la planta tan buenos o mejores (> 0.7) Correspondence: Juan Andres Cardoso, Alliance Bioversity-CIAT, Km 17 Recta Cali – Palmira, Valle del Cauca, Colombia. Email: j.a.cardoso@cgiar.org Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) 272 L. Hernández, P. Espitia and J.A. Cardoso que los obtenidos por puntuación visual de un experto. Los resultados indican que las imágenes digitales podrían mejorar la eficiencia del fenotipado en trabajos de mejoramiento para una mayor tolerancia a los salivazos en U. humidicola. Palabras clave: Aeneolamia varia, Brachiaria, fenotipado de alto rendimiento, gramíneas forrajeras tropicales, sensores, resistencia varietal. Introduction Increasing tolerance to spittlebugs in U. humidicola is a major target for the Urochloa breeding program of Urochloa humidicola is an important forage grass in the Alliance Bioversity-CIAT and adequate screening the tropical savannas of America (Berchembrock et methods are needed to increase the accuracy of the al. 2020). The productivity of current cultivars of selection process for tolerance. Currently, visual scoring U. humidicola is challenged by several American of plant damage is the standard phenotyping method spittlebug species (Hemiptera: Cercopidae) (Cardona to evaluate plant tolerance to the spittlebug complex et al. 2004). The damage in Urochloa grasses is caused in Urochloa grasses. Visual scores rely on estimates of when nymphs and adults feed from the xylem sap of percentages of dead leaf tissue (Parsa et al. 2011). Overall, roots in their immature stage (5 instars) and from the visual scoring is a low cost and medium throughput shoot in their adult stage (Valério et al. 2001). Thus, phenotyping method that has proven successful in the visual damage depends on the insect stage. In the Urochloa breeding program of the Alliance Bioversity- first 4 instars the damage is imperceptible, but when CIAT (Cardona et al. 1999; Miles et al. 2006). nymphs reach stage 5 an ascendant acropetal chlorosis Visual scoring is prone to subjectivity of the evaluator is observed and, under a severe attack, the entire and may not be accurate enough for use for selection in plant above-ground portion of the plant appears dry and dead breeding programs (Walter et al. 2012). Factors that can (Valério et al. 2001). When adults suck xylem sap the affect scoring of plants include expertise of the evaluator damage is observed in young leaves, where whitish- (different scores from different evaluators) and fatigue chlorotic spots appear around suction points due to over working hours. To overcome these, sensor-based parenchyma tissue dilution from the caustic substances measurements are gaining momentum in the Urochloa present in saliva (Valério et al. 2001). The spots tend to breeding program (Cardoso and Rao 2019). Hand-held coalesce in chlorotic lesions from the tip to the base of devices such as the SPAD series meters are used to non- the leaf and, when there is heavy infestation, the leaves destructively record greenness of leaves. These devices appear entirely yellow or necrotic (Figure 1) (Sotelo and Cardona 2001; Thompson and León-González measure the difference between the leaf transmittance in 2005). 650 nm (red) and 950 nm (infrared) regions using 2 light- emitting diodes and a photodiode receptor, delivering a relative SPAD meter value proportional to the amount of chlorophyll of the sample (Ling et al. 2011; Yuan et al. 2016). Measurements using SPAD meters have been shown to be positively and linearly correlated with percentages of dead tissue in Urochloa grasses (Cardoso et al. 2013). Another method used to record percentages of dead leaf tissue in Urochloa grasses is digital imaging (Jiménez et al. 2020). Sensor-based measurements are currently used in the Urochloa breeding program, but not to measure tolerance to spittlebugs (Cardoso et al. 2019; Jiménez et al. 2017; Jiménez et al. 2020; Mazabel et al. 2020). Therefore, the main objective of the present work was to compare alternative phenotyping methods (SPAD measurements and digital images) and a visual scoring from an inexperienced evaluator with evaluation of visual Figure 1. Symptoms of the damage caused by spittlebug scoring of damage from an expert. This information nymphs (Aeneolamia varia) on Urochloa species. should inform which screening methodology is the most Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) Digital images for scoring Urochloa tolerance 273 appropriate in terms of ease, accuracy and throughput, scoring from an expert and an inexpert evaluator; 2) and identify refinements needed. Improved screening SPAD measurements; and, 3) digital images. Plant methods should allow more accurate and intense damage, observed as chlorotic leaf area, was estimated selection, and hence, greater genetic gain for tolerance and expressed in percentage as described below. Time to spittlebugs in U. humidicola hybrids. spent for plant damage evaluation using the different methods was recorded. Materials and Methods Visual scoring Thirty-one U. humidicola genotypes were used in the present study, which was conducted at CIAT (Palmira, Visual scoring for plant damage was made as an Colombia, 3°31′ N, 76°19′ W; 965 masl.). Genotypes assessment of the proportion of green to senescing leaf with unknown tolerance included 24 hybrids originating tissue (yellow to brown) of the whole plant. Visual from the U. humidicola breeding Program of the Alliance scoring used a 11-point scale as follows: Bioversity-CIAT and 6 checks with known tolerance 0 = all leaves are green; to spittlebugs. Checks included 3 tolerant genotypes 1 = 10 % of senescent leaves; (cultivars ‘Llanero’ and ‘Tully’ and 1 germplasm 2 = 20 % of senescent leaves; accession, CIAT/16888) and 3 susceptible ones (2 3 = 30 % of senescent leaves; germplasm accessions, CIAT/26146 and CIAT/26375, 4 = 40 % of senescent leaves; and a hybrid, Bh13/2768). The germplasm accessions 5 = 50 % of senescent leaves; CIAT/16888 and CIAT/26146 are the foundation parents 6 = 60 % of senescent leaves; of the U. humidicola breeding program. All genotypes 7 = 70 % of senescent leaves; were planted as root splits from vegetative material. For 8 = 80 % of senescent leaves; each genotype, 10 root splits with 1 single tiller were 9 = 90 % of senescent leaves; harvested from plants maintained under greenhouse 10 = 100 % of senescent leaves. conditions at 28 °C and 80 % RH and then immersed for To test whether the visual scoring was affected 5 minutes in a 1 % sodium hypochlorite solution. Root by a given person during an evaluation, an expert splits were rinsed in water and planted in cylindrical and an inexpert evaluator carried out visual scorings polyvinyl chloride (PVC) pots (5.3 cm wide × 6.5 cm independently. deep) that contained 40 g of sterilized soil (3:1 weight soil: weight sand). Plants were watered daily and fertilized SPAD measurements with 30 mL of nutrient solution prepared with a 15 % N-15 % P-15 % K soluble fertilizer at 3 g/L two weeks SPAD meters (SPAD-502, Konica Minolta, Japan) were after planting. One month after planting, when sufficient used to estimate greenness of different leaves. SPAD superficial roots were available to serve as feeding sites units were recorded on 3 fully expanded leaves for each for the nymphs, 5 plants/genotype were infested with 6 plant and the mean taken. Plant damage was estimated mature eggs of Aeneolamia varia as previously described from the difference in SPAD measurements between by Cardona et al. (1999). The other 5 plants/genotype consecutive weeks as follows: were not infested and used as controls. The eggs were Damage = [(SPADn- SPADn+1)/SPADn]*100. previously obtained from the Alliance Bioversity-CIAT where: spittlebug mass rearing colony, selected for viability SPADn is a SPAD recording at any given week; by visual inspection and incubated under controlled SPADn+1 is the SPAD recording the week after. conditions (28 ⁰C, 85 % RH) (Parsa et al. 2011). Plants were organized in a randomized complete block with 2 Digital imaging treatments (infested with A. varia and un-infested) and 5 replicates. For image acquisition, individual plants were placed within a closed chamber (dimensions: 2×1.5×1 m) and Plant damage evaluation illuminated from above with a 120 cm long, 32 W, T8 LED tube producing 2,500 lumens. Images were taken Three phenotyping methods were used to assess plant with a digital color camera (Nikon Coolpix P6000, damage at weekly intervals for 5 weeks: 1) visual Nikon, Japan) with the following set up: F-stop: f/2.7, Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) 274 L. Hernández, P. Espitia and J.A. Cardoso Exposure time: 1/60, and ISO speed ISO-89 and from a clusters of colors in the image: background, green tissue Nadir, i.e. vertical, view of the plant. Images were saved and senescing tissue (Figure 2). The number of pixels for in a 4,224 x 3,168 pixel JPEG format. To account for each cluster was then quantified and plant damage was difference in illumination and color tones in images, calculated as: images were pre-processed with GIMP software (GIMP Damage = [SP/(SP+GP)]*100 2.10) to apply a pre-saved color tone matching curve where: to all JPEG files. Images were then processed and SP is number of pixels clustered as senescing tissue; analyzed using ImageJ (ImageJ 1.51). Image processing GP is number of pixels clustered as green tissue. consisted of splitting the images into their color channels (red, green and blue), and then normalizing the blue Statistical analysis channel (blue channel / red channel + green channel + blue channel). The normalized blue channel was used Mean values and standard deviations were calculated for image segmentation using the default threshold for estimations of plant damage for different dates and method of ImageJ. Image segmentation consisted of evaluation methods. Two-way analyses of variance the separation of shoot (white pixels) from background were calculated. Analyses were performed only for (black pixels). Once the image was segmented, a mask infested plants and conducted in R (R Development was laid onto the original unsegmented image using CoreTeam 2015). Calculations of agreement, Lin’s the AND logic operation. The masked image was then concordance index (Lin 1989) and Pearson correlation used to calculate the difference between green and red coefficient, were performed between estimates of channels (G-R), which enhances contrast between plant damage from alternative methods. Broad sense green tissue and senescing tissues. Once the G-R was heritability (H2) was calculated for each of the different calculated, K-means clustering was used to create 3 evaluation methods (Piepho and Möhring 2007). Figure 2. Summary of the image processing pipeline. *Green-Red is the result of subtracting green channel minus red channel. Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) Digital images for scoring Urochloa tolerance 275 Results 2 methods (SPAD measurements and digital images) were found from the first week of evaluation (Table 2). Comparison of throughput and estimated damage Throughout the experiment, estimates of damage were greater in visual scores compared to those obtained from The digital images method was significantly faster than SPAD meters (about 1.5-fold greater) and digital images the other methods (Table 1). There were no significant (about 1.3-fold greater). differences between the expert and inexpert evaluation in time needed for visual scoring. Concordances, correlations and heritability Table 1. Mean values of 5 evaluations showing the time Values of Lin’s concordance coefficient (CCC) and required to perform evaluations. Pearson correlations (r) increase with the time for all Evaluation method Number of plants scored/h1 the methods compared with the visual scoring from the Visual scoring (expert) 58 ± 13c expert, obtaining values over 0.7 for CCC and over 0.8 Visual scoring (inexpert) 71 ± 25bc for r (Table 3). Highest concordances and correlations SPAD measurements 80 ± 15b were observed between visual scoring from the expert Digital images 113 ± 15a and inexpert evaluators (Table 3). 1Values denote means ± standard deviations. Different letters Table 4 shows the weekly broad sense heritability (H2) next to standard deviation values denote significant differences values according to evaluation method. All the H2 values at P=0.05. increased through time for the 4 evaluation methods. Greater values of H2 (values closer to 1) were obtained using The visual scoring methodology generally had higher the digital images method, indicating that a large portion of values of damage for all the assessments, followed by the variation is due to genetic factors and a smaller portion digital images and SPAD measurements (Figure 3). due to environment and genotype-environment interaction. Differences in estimates of damage between visual scores Conversely, lowest H2 values were obtained for the visual (from expert and inexpert evaluators) and the other scoring from an inexpert evaluator. 100 c Visual Scoring From an Expert d c b Visual Scoring From an Inexpert c 80 SPAD Measurements d a Digital Images c b 60 a c c b 40 a b b b a 20 a a 0 7 14 21 28 35 Days After Infestation Figure 3. Comparison of damage percentage over time. *Column bars represent means and error bars indicate the standard deviation. Letters over bars indicate the differences by evaluation method at 7, 14, 21, 28 or 35 days after infestation. Columns with different letter are significantly different (P<0.05). Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) Damage Percentage (%) 276 L. Hernández, P. Espitia and J.A. Cardoso Table 2. Analysis of variance (2-way-ANOVA) made at different days after infestation (DAI) with Aeneolamia varia. Genotypes evaluated correspond to 27 U. humidicola hybrids. Method corresponds to evaluation techniques used (Visual scoring (expert), Visual scoring (inexpert), SPAD measurements and Digital images). DAI* Source Df Sum of Squares Mean of Square F value P value 7 Genotype 29 13869 478.2 4.849 7.06E-14 *** Method 3 2471 823.7 8.352 0.000021 *** Genotype × Method 87 9659 111 1.126 0.225 Residuals 416 41026 98.6 14 Genotype 29 33021 1139 5.712 <2e-16 *** Method 3 24089 8030 40.276 <2e-16 *** Genotype × Method 87 15286 176 0.881 0.761 Residuals 416 82935 199 21 Genotype 29 48623 1677 8.695 <2e-16 *** Method 3 103661 34554 179.185 <2e-16 *** Genotype × Method 87 12537 144 0.747 0.951 Residuals 416 80220 193 28 Genotype 29 64296 2217 11.771 <2e-16 *** Method 3 124900 41633 221.037 <2e-16 *** Genotype × Method 87 9217 106 0.562 0.999 Residuals 416 78355 188 35 Genotype 29 121290 4182 12.627 <2e-16 *** Method 3 76836 25612 77.323 <2e-16 *** Genotype × Method 87 20665 238 0.717 0.97 Residuals 416 137794 331 ***Significant at the 0.001 probability level; *DAI = days after infestation Table 3. Lin’s concordance coefficient and Pearson correlation analysis between damage percentages obtained from the evaluation methods. DAI* Index** Visual scoring from an expert vs. Visual scoring from an expert vs. Visual scoring from an expert vs. Visual scoring from an inexpert SPAD measurements digital images 7 CCC 0.89 0.16 0.29 CI 0.86 - 0.91 0.09 - 0.24 0.22 - 0.36 r 0.9*** 0.25*** 0.44*** 14 CCC 0.89 0.24 0.42 CI 0.86 - 0.91 0.17 - 0.32 0.36 - 0.49 r 0.91*** 0.36*** 0.59*** 21 CCC 0.89 0.34 0.58 CI 0.87 - 0.91 0.28 - 0.4 0.52 - 0.64 r 0.92*** 0.6*** 0.76*** 28 CCC 0.93 0.56 0.75 CI 0.91 - 0.94 0.5 - 0.61 0.71 - 0.79 r 0.94*** 0.82*** 0.88*** 35 CCC 0.94 0.70 0.86 CI 0.93 - 0.95 0.64 - 0.75 0.83 - 0.89 r 0.95*** 0.83*** 0.9*** *DAI = days after infestation; **CCC = Lin’s concordance correlation coefficient; CI = confidence interval (95 %); r = Pearson correlation coefficient; *** = correlation significance (P<0.001). Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) Digital images for scoring Urochloa tolerance 277 Table 4. Broad sense heritability according to treatment, evaluation method and days after infestation for damage percentage. Method 7 DAI* 14 DAI 21 DAI 28 DAI 35 DAI H2 P-value H2 P-value H2 P-value H2 P-value H2 P-value Digital images 0.6 0.007 0.7 0.001 0.8 0.001 0.8 0.001 0.8 0.001 SPAD Measurements 0.4 0.13 0.5 0.04 0.7 0.001 0.7 0.001 0.8 0.001 Visual Scoring from expert 0.6 0.0024 0.6 0.0077 0.7 0.001 0.8 0.001 0.8 0.001 Visual Scoring from inexpert 0.5 0.0083 0.5 0.0077 0.5 0.0079 0.5 0.0212 0.5 0.0088 *DAI = days after infestation. Discussion and time of detection of damage were greater using visual scoring (for both expert and inexpert evaluators), The present study indicated that capture of digital it is likely that visual scores over-estimated damage as images was the fastest method to record plant damage, previously identified by Bock et al. (2010). as previously shown for other traits (Jiménez et al. Results indicate the inexpert evaluator got better with 2020; Büchi et al. 2018). Reduction of time is among time in the visual scoring of plant damage, as shown in the improvements sought by most phenotyping methods other studies (Bock et al. 2016; Bock et al. 2020). Despite (Shakoor et al. 2017; Araus et al. 2018) to allow more the improvement gained by the inexpert evaluator, they plants to be evaluated for plant damage, reduce the time were unable to distinguish percentages of damage below needed or allow more intensive phenotyping (recording 20 %, whereas the expert evaluator could distinguish at of additional traits that might be of interest). 10 % (data not shown). Similar results were found when Results showed that there were no significant experienced and inexperienced evaluators assessed severity differences between estimates of damage from visual of Phomopsis leaf blight of strawberry (Nita et al. 2003). scoring from an expert and an inexpert evaluator, Also, the level of agreement between estimates of damage suggesting the inexpert evaluator followed carefully from visual scoring and digital images is considered low the instructions given by the expert evaluator. However, (McBride 2005). This is not surprising as estimates of this might not always be the case for new evaluators. damage from visual scoring were discrete values being Successful training of a new evaluator is dependent compared to continuous values of plant damage estimated on the inherent characteristics of the individual and from digital images (McBride 2005) and with a likely previous knowledge of the plants, which likely affects overestimation of damage from visual scoring. the accuracy of any evaluation. Clear instruction and All the methodologies except for visual scoring training increase the accuracy of visual estimates of plant from an inexpert evaluator showed a high accuracy with damage minimizing errors (Bock et al. 2020). Despite heritability values over 0.7. Similar results for heritability estimates of damage from the expert and inexpert were obtained by other authors when comparing image- evaluators being similar, measures of data variability based phenotyping methods to visual evaluations (i.e., standard deviation) from the inexpert evaluator (Makanza et al. 2018; Singh et al. 2019). A phenotyping were greater than those from the expert evaluator. Similar procedure, such as digital imaging, that detects high results were found by El Jarroudi et al. (2015) when heritability of any given trait allows a broader selection comparing estimates of septoria leaf blotch severity process, hence, the genetic advance through the breeding (and measures of data variability) in winter wheat from cycles is faster (Holland et al. 2002). Different methods different evaluators. require different equipment and skills and have different Development of damage could be detected earlier costs and advantages/disadvantages that also have to be under the visual scoring method. Since the magnitude taken into account together with accuracy (Table 5). Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) 278 L. Hernández, P. Espitia and J.A. Cardoso Table 5. Comparison among plant damage estimation methodologies. Digital images SPAD measurements Visual scoring Equipment • Digital camera. • SPAD meter. • Software for image correction and processing. • Photobox with constant light. Labor and • Semi-skilled labor for image capture. • Unskilled labor. • Highly skilled evaluators. skill level • Skilled labor to process and analyze the images. Advantages • Higher accuracy through time for • Only needs one equipment and • Earlier detection of symptoms. plant damage quantification. Fastest does not need trained personal. methodology -allows to collect higher numbers of data in less time. Disadvantages • Requires qualified personal to • Time consuming. • Over estimation of plant damage. automatize the process. • Low correlation to the standard • Costly because of continued visual scoring assessment. rigorous training of new • Depends on the evaluator expertise. evaluators. Conclusions Berchembrock YV; de Figueiredo UJ; Nunes JAR; Valle CB do; Barrios SCL. 2020. Comparison of selection methods among The present work showed that estimation of plant damage and within full-sibling progenies in Urochloa humidicola. from digital images yielded similar results to those Grass and Forage Science 75:145–152. doi: 10.1111/gfs.12468 obtained by the standard method of visual scoring by an Bock CH; Poole GH; Parker PE; Gottwald TR. 2010. Plant expert evaluator. One of the major drawbacks of visual disease severity estimated visually, by digital photography scoring is the dependence on an expert evaluator. Training and image analysis, and by hyperspectral imaging. Critical Reviews in Plant Sciences 29(2):59–107. doi: of new evaluators for visual scoring of plant damage might 10.1080/07352681003617285 be a straightforward mechanism to ensure continuity over Bock CH; Hotchkiss MW; Wood BW. 2016. Assessing disease time. However, inter-rater variation represents a major severity: accuracy and reliability of rater estimates in drawback for this method. Overall, SPAD measurements relation to number of diagrams in a standard area diagram were more time consuming and showed a low correlation set. Plant Pathology 65(2):261–272. doi: 10.1111/ppa.12403 with the standard evaluation of visual scoring from an Bock CH; Barbedo JGA; Del Ponte EM; Bohnenkamp D; expert, which makes this method less suitable to assess Mahlein AK. 2020. From visual estimates to fully automated large numbers of hybrids in the U. humidicola breeding sensor-based measurements of plant disease severity: status program. Higher values of broad sense heritability and and challenges for improving accuracy. Phytopathology faster recording of plant damage from digital images Research 2:9. doi: 10.1186/s42483-020-00049-8 suggests that this phenotyping method could be used to Büchi L; Wendling M; Mouly P; Charles R. 2018. Comparison improve the efficiency of breeding for increased tolerance of visual assessment and digital image analysis for canopy cover estimation. Agronomy Journal 110(4):1289–1295. to spittlebugs in U. humidicola. doi: 10.2134/agronj2017.11.0679 Cardona C; Miles JW; Sotelo G. 1999. An improved Acknowledgments methodology for massive screening of Brachiaria spp. genotypes for resistance to Aeneolamia varia (Homoptera: This work was undertaken as part of the OneCGIAR cercopidae). Journal of Economic Entomology 92(2):490– Initiative on Accelerated Breeding. 496. doi: 10.1093/jee/92.2.490 Cardona C; Fory P; Sotelo G; Pabon A; Diaz G; Miles JW. References 2004. Antibiosis and tolerance to five species of spittlebug (Homoptera: Cercopidae) in Brachiaria spp.: implications (Note of the editors: All hyperlinks were verified 12 September 2022). for breeding for resistance. Journal of Economic Entomology 97(2):635–645. doi: 10.1093/jee/97.2.635 Araus JL; Kefauver SC; Zaman-Allah M; Olsen MS; Cairns Cardoso JA; Rao IM. 2019. Drought resistance of tropical forage JE. 2018. Translating high-throughput phenotyping into grasses: Opening a fertile ground for innovative research. In: genetic gain. Trends in Plant Science 23(5):451–466. doi: Pessarakli M, ed. Handbook of Plant and Crop Stress. CRC 10.1016/j.tplants.2018.02.001 Press, Boca Raton, FL, USA. hdl.handle.net/10568/105437 Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) Digital images for scoring Urochloa tolerance 279 Cardoso JA; Rincon J; Jiménez JC; Noguera D; Rao IM. Nita M; Ellis MA; Madden LV. 2003. Reliability and 2013. Morpho-anatomical adaptations to waterlogging accuracy of visual estimation of phomopsis leaf blight of by germplasm accessions in a tropical forage grass. AoB strawberry. Phytopathology 93(8):995–1005. doi: 10.1094/ PLANTS 5:plt047. doi: 10.1093/aobpla/plt047 PHYTO.2003.93.8.995 El Jarroudi M; Kouadio AL; Mackels C; Tychon B; Delfosse Parsa S; Sotelo G; Cardona C. 2011. Characterizing herbivore P; Bock CH. 2015. A comparison between visual estimates resistance mechanisms: spittlebugs on Brachiaria spp. as and image analysis measurements to determine septoria an example. Journal of Visualized Experiments 52:e3047. leaf blotch severity in winter wheat. Plant Pathology doi: 10.3791/3047 64(2):355–364. doi: 10.1111/ppa.12252 Piepho HP; Möhring J. 2007. Computing heritability and selection Holland JB; Nyquist WE; Cervantes-Martínez CT. 2002. response from unbalanced plant breeding trials. Genetics Estimating and interpreting heritability for plant breeding: 177(3):1881–1888. doi: 10.1534/genetics.107.074229 an update. In: Janick J, ed. Plant Breeding Reviews, Vol. R Development CoreTeam. 2015. A language and environment 22 p. 9–112. John Wiley & Sons, Inc., Hoboken, NJ, USA. for statistical computing. R Foundation for Statistical doi: 10.1002/9780470650202.ch2 Computing. Vienna, Austria. Jiménez JC; Cardoso JA; Leiva LF; Gil J; Forero MG; Shakoor N; Lee S; Mockler TC. 2017. High throughput Worthington ML; Miles JW; Rao IM. 2017. Non-destructive phenotyping to accelerate crop breeding and monitoring phenotyping to identify Brachiaria hybrids tolerant to of diseases in the field. Current Opinion in Plant Biology waterlogging stress under field conditions. Frontiers in 38:184–192. doi: 10.1016/j.pbi.2017.05.006 Plant Science 8:167. doi: 10.3389/fpls.2017.00167 Singh D; Wang X; Kumar U; Gao L; Noor M; Imtiaz M; Singh Jiménez JC; Leiva L; Cardoso JA; French AN; Thorp KR. RP; Poland J. 2019. High-throughput phenotyping enabled 2020. Proximal sensing of Urochloa grasses increases genetic dissection of crop lodging in wheat. Frontiers in selection accuracy. Crop and Pasture Science 71(4):401– Plant Science 10:394. doi: 10.3389/fpls.2019.00394 409. doi: 10.1071/CP19324 Sotelo G; Cardona C. 2005. Manejo integrado del salivazo de Lin LK. 1989. A concordance correlation coefficient to evaluate los pastos con énfasis en resistencia varietal. In: Herrero M; reproducibility. Biometrics 45:255–268. doi: 10.2307/2532051 Ramírez A; Joaquín N, eds. Manejo y evaluación de pasturas Ling Q; Huang W; Jarvis P. 2011. Use of a SPAD-502 tropicales. Centro Internacional de Agricultura Tropical meter to measure leaf chlorophyll concentration (CIAT), Cali, Colombia p. 140–150. bit.ly/3BTE7eT in Arabidopsis thaliana. Photosynthesis Research Thompson V; León-González R. 2005. The identity and 107(2):209–214. doi: 10.1007/s11120-010-9606-0 distribution of sugar cane and pasture spittlebugs Makanza R; Zaman-Allah M; Cairns JE; Magorokosho (Homoptera: Cercopidae) in Costa Rica. Manejo Integrado C; Tarekegne A; Olsen M; Prasanna BM. 2018. High- de Plagas y Agroecologia. 75:43–51. (In Spanish). throughput phenotyping of canopy cover and senescence repositorio.catie.ac.cr/handle/11554/6456 in maize field trials using aerial digital canopy imaging. Valério JR; Cardona C; Peck DC; Sotelo G. 2001. Spittlebugs: Remote Sensing 10(2):330. doi: 10.3390/rs10020330 bioecology, host plant resistance and advances in IPM. Mazabel J; Worthington M; Castiblanco V; Peters M; Arango Proceedings of the XIX International Grassland Congress, J. 2020. Using near infrared reflectance spectroscopy for São Pedro, SP, Brazil, 11–21 February, 2001. uknowledge. estimating nutritional quality of Brachiaria humidicola uky.edu/igc/19/5/6 in breeding selections. Agrosystems, Geosciences & Walter A; Studer B; Kölliker R. 2012. Advanced phenotyping Environment 3(1):e20070. doi: 10.1002/agg2.20070 offers opportunities for improved breeding of forage and McBride GB. 2005. A proposal for strength-of-agreement criteria turf species. Annals of Botany 110(6):1271–1279. doi: for Lin's Concordance Correlation Coefficient. NIWA client 10.1093/aob/mcs026 report. NIWA, Hamilton, New Zealand. bit.ly/3esrFcK Yuan Z; Cao Q; Zhang K; Ata-Ul-Karim ST; Tian Y; Zhu Y; Miles JW; Cardona C; Sotelo G. 2006. Recurrent selection in Cao W; Liu X. 2016. Optimal leaf positions for SPAD a synthetic brachiariagrass population improves resistance meter measurement in rice. Frontiers in Plant Science to three spittlebug species. Crop Science 46(3):1088–1093. 7:719. doi: 10.3389/fpls.2016.00719 doi: 10.2135/cropsci2005.06-0101 (Received for publication 23 June 2021; accepted 30 August 2022; published 30 September 2022) © 2022 Tropical Grasslands-Forrajes Tropicales is an open-access journal published by International Center for Tropical Agriculture (CIAT), in association with The Tropical Crops Genetic Resources Institute of The Chinese Academy of Tropical Agricultural Sciences (TCGRI-CATAS). This work is licensed under the Creative Commons Attribution 4.0 International (CC BY 4.0) license. Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) Tropical Grasslands-Forrajes Tropicales (2022) Vol. 10(3):280–287 280 doi: 10.17138/TGFT(10)280-287 Regional communication Nitrogen and phosphorus fertilizer application to Elephant grass (Cenchrus purpureus syn. Pennisetum purpureum) cultivar ‘Cameroon’ in an arenosol in Rio Grande do Norte, Brazil Fertilización con nitrógeno y fósforo en pasto elefante cv. Cameroon en un arenosol en Rio Grande do Norte, Brasil LUIZ E.C. OLIVEIRA1, FÁBIO H.T. OLIVEIRA1, GUALTER G.C. SILVA2, MARCIO G. SILVA BEZERRA1, ÉRIC G. MORAIS1, GABRIEL F.R. BEZERRA2, GIOVANA S. DANINO2, ERMELINDA M.M. OLIVEIRA2 AND FRANCISCO V.S. SÁ1 1Center of Agricultural Sciences, Universidade Federal Rural do Semi-Árido, Mossoró, RN, Brazil. ufersa.edu.br 2Agricultural Sciences Unit, Universidade Federal do Rio Grande do Norte, Macaíba, RN, Brazil. ufrn.br Abstract Elephant grass (Cenchrus purpureus syn. Pennisetum purpureum) stands out for its high dry matter production per unit area and good nutritional value and is cultivated throughout Brazil. This study aimed to evaluate the performance of Elephant grass cultivar ‘Cameroon’ fertilized with nitrogen (N) and phosphorus (P) at different rates. The experimental design was in randomized blocks with 10 treatments and 4 replicates. The treatments consisted of 5 doses of N (0, 200, 400, 600 and 800 kg N/ha) all with 66 kg P/ha and 5 doses of P (0, 22, 44, 66 and 88 kg P/ha) all with 600 kg N/ha. The variables evaluated during 3 harvests were: shoot dry matter (DM) yield, N and P concentrations in shoots, and uptakes of N and P in forage. Results showed that, in the arenosol of the experimental area, high doses of N and P could produce high yields of the grass (40‒41 t DM/ha) over 260 days. The grass extracted large amounts of N (on average, 800 kg N/ha over 260 days) and P concentrations were significantly affected by P fertilization only in the last harvest, where it increased from 0.27 to 0.78 g P/kg DM. However, application of only 200 kg N/ha will produce more than 60 % of the DM yield response achieved with 800 kg N/ha. Similarly, there seems little merit in applying more than 22 kg P/ha with the N. Longer-term studies are needed to test these hypotheses along with economic assessments to determine the financial soundness of such decisions. Keywords: Export, extraction, nutrient, yield. Resumen El pasto elefante destaca por su alta producción de materia seca por unidad de área y por su buen valor nutricional, se cultiva en todo Brasil. Este trabajo tuvo como objetivo evaluar la productividad del pasto elefante cultivar 'Cameroon' fertilizado con diferentes dosis de nitrógeno (N) y fósforo (P). El diseño experimental fue en bloques al azar con diez tratamientos y cuatro repeticiones. Los tratamientos consistieron en cinco dosis de nitrógeno (0, 200, 400, 600 y 800 kg/ha de N) todo con 66 kg/ha de P y cinco dosis de fósforo (0, 22, 44, 66 y 88 kg/ha de P) todo con 600 kg/ha de N. Las variables evaluadas fueron: rendimiento de materia seca (MS) de los brotes, concentraciones de N y P en los brotes y absorción de N y P en el forraje. Los resultados mostraron que, en el arenosol del área experimental, se puede producir altos rendimientos de pasto (40–41 t MS/ha) en 260 días, cuando se aplica altas dosis de N y P. El pasto extrajo grandes cantidades de N (en promedio, 800 kg N/ha durante 260 días) y las concentraciones de P se vieron significativamente afectadas por la fertilización con fósforo solo en la última cosecha, donde aumentó de 0.27 a 0.78 g P/kg MS. Sin embargo, la aplicación de solo 200 kg Correspondence: Francisco Vanies da Silva Sá, Center of Agricultural Sciences, Universidade Federal Rural do Semi-Árido, Mossoró, CEP 59625-900, RN, Brazil. Email: vanies_agronomia@hotmail.com Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) Nitrogen and phosphorus doses in elephant grass 281 N/ha producirá más del 60 % de la respuesta de rendimiento de MS lograda con 800 kg N/ha. De manera similar, no se justifica aplicar más de 22 kg P/ha con el N. Se necesitan estudios a más largo plazo para probar estas hipótesis junto con evaluaciones económicas para determinar la solidez financiera de tales decisiones. Palabras clave: Exportación, extracción, nutrientes, producción. Introduction component of genes and chromosomes, as well as many coenzymes, phosphoproteins and phospholipids. Elephant grass (Cenchrus purpureus Schum.) Morrone Considering the importance of both nutrients in (syn. Pennisetum purpureum Schum.) stands out for its maintaining the productive potential of Elephant grass, as high dry matter (DM) production per unit area and good well as the lack of tables of recommendations calibrated nutritional value, being cultivated throughout Brazil for the arenosols in the region, it is important to conduct and tolerating unfavorable climatic conditions, such studies evaluating the responses of agricultural crops as drought and cold temperatures (Queiroz Filho et al. as a function of varying doses of essential nutrients. 2000). Historically, this plant has been used as a forage Thus, this study aimed to evaluate the performance of reserve, and is an important complement to bulky feed Elephant grass cultivar ‘Cameroon’ fertilized with N for animals on rural properties. In addition to this use, and P at a range of doses in an endeavor to develop in recent years research has shown that Elephant grass an understanding of optimal levels of these nutrients has the potential to be used as an alternative biomass to apply. A serious concern for soil management is the for energy production, mainly due to its high dry matter amount of N and P which is taken up by the pasture yields (Mello et al. 2002; Quesada et al. 2003). grass. The amount of N and P that is absorbed by the Elephant grass has high nutritional requirements, due pasture is crucial for the extrapolation of the fertilizer to its high production potential. Thus, any nutritional dose. The correct amount of N and P promotes optimal deficiency during its development will limit production plant growth and it makes fertilization economically (Avalhaes et al. 2009). Arenosols are sandy soils with low viable. In addition, the correct dose of N can avoid levels of organic matter and low water retention (Costa contamination of water courses by leaching of excess et al. 2013). In Brazil, arenosols are common in coastal of N. We also evaluated N and P concentration and regions and in regions with high precipitation rates, uptake to understand Elephant grass growth in response where leaching of basic cations can cause low fertility to N and P doses. (Speratti et al. 2018). However, there is little information on the effects of fertilizer application on production of Materials and Methods Elephant grass on arenosols. Nitrogen (N) is directly linked to processes involved The field experiment was installed in an area of the in plant growth and development (Porto et al. 2012), Agricultural School of Jundiaí, belonging to the Federal being an essential constituent of proteins and important University of Rio Grande do Norte, in Macaíba-RN, in the photosynthetic process, due to its participation Brazil (5°53' S, 35°21' W; 15 masl). in the chlorophyll molecule (Martuscello et al. 2016). The local climate is a transition between the types As Optimal fertilizer requirements must be carefully defined and BSw of Köppen’s classification system, with high as cost of N fertilizers is high (Primavesi et al. 2004). temperatures throughout the year (annual average 27 °C, After N, phosphorus (P) is the nutrient that most limits maximum 32 °C and minimum 21 °C). Average annual forage production when limited in supply (Oliveira et rainfall is 1,071 mm, with rainy season from March to al. 2007; Foloni et al. 2008). Although the total amount July (IDEMA 2013). of P in soil is relatively high, it is generally not found The soils of the experimental area are classified as in its plant-available form (Santos et al. 2010). P has arenosols (quartzipsamments), with sandy texture and important functions at the initial stage of development gently sloping topography (Beltrão et al. 1975). Before of forage plants, when there is intense meristematic the experiment, 20 individual soil samples were collected activity due to root system development and tillering. In from the 0‒20 cm horizon of the experimental area and addition, P is essential for cell division, due to its role in homogenized to obtain a composite sample, which the structure of nucleic acids (Cantarutti et al. 2002). It was sent to the laboratory for chemical and physical is also crucial for plant metabolism in energy transfer in characterization (Table 1), following the methodology cells, respiration and photosynthesis, being a structural proposed by EMBRAPA (1997). Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) 282 L.E.C. Oliveira et al Table 1. Chemical and physical characteristics in the 0‒20 cm horizon of the soil of the experimental area before the experiment pH1 OM Ntotal P K+ Na+ Ca2+ Mg2+ Al3+ (H+Al) Sand Silt Clay (H2O) (g/kg) (mg/dm3) (cmolc/dm3) (g/kg) 5.9 2.64 0.70 1 32.8 5.3 0.4 1.2 0.05 0.83 940 40 20 OM = organic matter. The experimental design was randomized blocks with sulfate and potassium chloride used in these top-dressings 11 treatments and 4 replicates. The treatments consisted were applied in furrows 0.15 m from the grass rows. of 5 levels of nitrogen (N) (0, 200, 400, 600 and 800 The Elephant grass (Cenchrus purpureus) cultivar kg N/ha) in the presence of 66 kg P/ha and 5 levels of ‘Cameroon’ was planted by placing whole stems of the phosphorus (P) (0, 22, 44, 66 and 88 kg P/ha) in the grass at the bottom of the furrows according to the end- presence of 600 kg N/ha. All treatments were fertilized to-end system, and then cutting them into approximately with 125 kg K/ha, 30 kg S/ha, 1.0 kg B/ha, 3.0 kg Zn/ha 70-cm-long pieces. During the study, whenever necessary, and 0.5 kg Cu/ha (Table 2). The commercial fertilizers the grass was irrigated using a conventional sprinkler. used in applying the nutrients were urea, ammonium Following planting, the Elephant grass was studied for sulfate, triple superphosphate, potassium chloride and 260 days between December 2016 and August 2017. The FTE BR-12 as source of the micronutrients. first harvest was performed at 110 days after planting at an average height of 288 cm, the second harvest at 75 days Table 2. Doses (kg/ha) of nutrients applied in the various later at an average height of 245 cm and the third harvest treatments to evaluate effects on DM yield of Elephant grass at 75 days later at an average height of 227 cm, i.e. at and concentrations and uptake of N and P. intervals of 110, 75 and 75 days. Treatment N P K S B Zn Cu At each harvest, plants from the middle 2 rows (usable 1 0* 66 125 30 1 3 0.5 area) of each plot were cut at soil level to determine the 2 200 66 125 30 1 3 0.5 amount of dry matter produced in each plot. After weighing 3 400 66 125 30 1 3 0.5 of the fresh forage, all harvested plants from each plot were 4 600 66 125 30 1 3 0.5 chopped with a forage chopper machine (M-600 Chopper 5 800 66 125 30 1 3 0.5 with a 3 hp engine) and then homogenized. Samples of this chopped material were then collected and dried in 6 600 0* 125 30 1 3 0.5 a forced-air circulation oven at 65 ºC. After reaching a 7 600 22 125 30 1 3 0.5 constant weight, the samples were weighed and ground in a 8 600 44 125 30 1 3 0.5 Wiley-type mill and mineralized by sulfuric acid digestion, 9 600 66 125 30 1 3 0.5 following the methodology of Tedesco et al. (1995). N was 10 600 88 125 30 1 3 0.5 quantified by the Kjeldahl method and P was quantified by *Control colorimetric method. N and P uptakes were calculated by multiplying DM yield by the N and P concentrations. Each experimental plot was 2.8 m wide by 3.0 m long After data collection and tabulation, results were and contained 4 rows of Elephant grass, with inter-row subjected to analyses of variance and regression, using the spacing of 0.7 m. The usable area of the plot measured 4.2 software program SISVAR v. 5.6 (Ferreira 2011). For each m2, consisting of the 2 central rows. characteristic evaluated, the mean square of residuals of Initially, soil tillage was performed as light harrowing the analysis of variance was used as an experimental error using a tractor, the area was demarcated by delimiting the to test the significance of the coefficients of the regression total space of the experiment, blocks and plots, and then models to be fitted. We also compared the means by the furrows were opened manually with a hoe. All doses of P, LSD test at the 5 % probability level. S, B, Zn and Cu were applied at the bottom of the furrow before planting, together with 10 % of the N dose and Results 10 % of the K dose. The remaining 90 % of N and K fertilizers was split into Analyses of variance and regression revealed a significant equal amounts and applied on 6 occasions: 30 and 60 days positive effect of increasing N fertilizer level on forage dry after planting, 15 and 45 days after the first harvest and 15 matter (DM) yield at all harvests and, consequently, on the and 45 days after the second harvest. Urea, ammonium accumulated DM yield over the 3 harvests (Table 3). Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) Nitrogen and phosphorus doses in elephant grass 283 Table 3. Effects of N fertilizer level on dry matter yield of most suitable for representing DM yield at the second Elephant grass at 1st, 2nd and 3rd harvests (intervals of 110, harvest (Table 3). Highest DM yield over the 260 days of 75 and 75 days) and overall, and estimates of the parameters of occurred at a dose of 800 kg N/ha (41.3 t DM/ha). Mean the fitted regression models (square root, linear and quadratic). DM yield at the first harvest exceeded those at the second N dose Dry matter yield (t DM/ha) Total and third harvests. (kg/ha) 1st harvest 2nd harvest 3rd harvest DM yield increase with the first 200 kg N/ha was 0 (Control) 11.2 7.23 6.64 25.1 greater than increases with subsequent increments of 200 200 16.8 8.33 9.64 34.8 kg/ha (Table 3). 400 18.0 8.68 9.80 36.5 There was a positive effect of N fertilizer level on 600 19.1 10.25 12.17 39.8 N concentration in harvested forage at the first and 800 18.1 12.48 10.60 41.3 third harvests (P<0.01), but differences failed to reach significance at the second harvest, despite absolute Average 16.7 9.5 9.8 35.5 values being similar for all harvests (Table 4). At the first LSD 4.7 2.3 2.2 7.2 harvest, N applications of 600 and 800 kg N/ha increased ANOVA ** ** ** ** N concentration in forage relative to Control, while at b0 11.18 6.91 6.68 25.18 the third harvest applications of 400 kg N/ha and above b1: X -0.0108 0.0062 -0.0041 -0.0051 increased N concentration in forage (P<0.01). b2: X 0.5 0.5629 - 0.2518 0.7097 N uptake by Elephant grass increased significantly b (P<0.01) with applications of 200 kg N/ha and above at 3: X 2 - - - - R2 0.98 0.93 0.99 0.99 first and third harvests and overall (Table 4). However, at the second harvest, only N doses of 600 and 800 kg N/ha Application of all fertilizer levels increased DM yield had greater N uptake than Control (Table 4). relative to Control at first and third harvests and overall N uptake by Elephant grass at the first harvest (208‒480 (total production) (Table 3). However, at the second kg N/ha) exceeded those at the second and third harvests harvest, only N doses higher than 600 kg N/ha produced (150‒260 and 124‒290 kg N/ha, respectively) (Table 4). more DM than the Control (Table 3). DM yields of Elephant grass at the first and second The square root model was the most suitable for harvests were not affected by P application (P>0.05; Table representing the effects of N fertilizer level on DM yields 5). However, at the third harvest and overall, application at the first and third harvests plus total production of of 44 kg P/ha and above significantly increased DM yields Elephant grass, while the simple linear model was the (P<0.01; Table 5). Table 4. Effects of N fertilizer level on N concentration in forage and N uptake by Elephant grass at 1st, 2nd and 3rd harvests (intervals of 110, 75 and 75 days, respectively) and overall, and estimates of the parameters of regression models (square root, linear and quadratic). N dose (kg/ha) N concentration (g/kg) N uptake (kg/ha) Total 1st harvest 2nd harvest 3rd harvest 1st harvest 2nd harvest 3rd harvest 0 (Control) 18.7 20.8 18.7 208.2 149.8 123.6 482 200 19.3 19.3 21.6 320.6 160.7 207.9 689 400 21.0 17.7 23.6 380.6 156.8 231.0 768 600 26.7 24.5 26.7 509.1 251.1 277.8 1,038 800 25.4 21.9 25.4 451.0 268.2 301.6 1,021 Average 22.2 20.9 23.2 373.9 197.3 228.4 800 LSD 4.4 5.7 3.1 89.1 69.8 52.1 141.3 ANOVA ** ns ** ** ** ** ** b0 18.53 19.36 18.4 200.2 147.7 124.3 477.2 b1: X -0.1244 - - - - 4.7208 - b2: X 0.5 0.0146 0.0037 0.0199 0.7264 0.0056 0.055 1.0838 b3: X 2 - - 0.00001 0.00049 0.0002 - 0.00046 R2 0.83 0.20 0.95 0.93 0.88 0.99 0.95 Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) 284 L.E.C. Oliveira et al Table 5. Effects of P fertilizer level on dry matter yield of trend as data for P concentration in forage, with no Elephant grass at 1st, 2nd and 3rd harvests (intervals of 110, significant effect of P fertilizer application (P>0.05) at 75 and 75 days, respectively) and overall, and estimates of the first and second harvests and a positive linear effect the parameters of regression models (square root, linear and at both the third harvest and overall (P<0.01) (Table 6). quadratic). P dose Dry matter yield (t DM/ha) Total Discussion (kg/ha) 1st harvest 2nd harvest 3rd harvest 0 14.76 8.19 6.98 29.93 While this study has shown that Elephant grass (Control) cultivar ‘Cameroon’ showed good growth responses to 200 19.02 10.24 9.28 38.54 application of N fertilizer, the greatest response per unit 400 17.92 10.63 11.02 39.57 of N applied occurred with the first 200 kg N/ha. Sixty percent of the total N response in DM yield occurred 600 16.61 11.56 11.28 39.45 with the application of only 200 kg N/ha. In fact, the 800 19.30 11.22 12.08 42.58 increase in DM yield with the first 200 kg N/ha was 50 % Average 17.79 10.37 10.13 38.29 greater than the total response to the next 600 kg N/ha, LSD 3.49 2.52 2.45 5.97 a fine example of the law of diminishing returns. While ANOVA ns ns ** ** average DM yield decreased by 42 % from the first to b the second harvest and remained at a similar level for 0 14.98 8.29 7.05 30.10 b1: X -0.0623 0.0892 0.1144 -0.0849 the third harvest, the longer growth period prior to the b first harvest would have played a major part in the higher 2: X 0.5 0.9173 - - 2.0232 yields at that stage, along with the higher application of N b3: X 2 - -0.0006 -0.0007 - to all treatments during that growth period (40 vs. 30 % R2 0.59 0.96 0.99 0.96 of total N applied). Release of nutrients in soil following P concentrations in Elephant grass forage at the first cultivation prior to planting may have also contributed. and second harvests were not significantly affected by Higher DM yields at the first harvest after planting, P fertilizer level applied (P>0.05; Table 6), while at the compared with those at later harvests, have also been third harvest P concentration increased linearly with reported by other authors (Morais et al. 2009; Santos increasing level of P fertilizer application (P<0.01). et al. 2014). In those studies, growth period between Uptake of P by Elephant grass followed the same planting and the first harvest was longer than growth Table 6. Effects of P fertilizer level on P concentration in forage and P uptake by Elephant grass forage at 1st, 2nd and 3rd harvests (intervals of 110, 75 and 75 days, respectively) and overall, and estimates of the parameters of regression models (square root, linear and quadratic). P dose (kg/ha) P concentration (g/kg) P uptake (kg/ha) Total 1st harvest 2nd harvest 3rd harvest 1st harvest 2nd harvest 3rd harvest 0 (Control) 0.83 0.20 0.27 12.35 1.67 1.33 15.35 200 0.83 0.51 0.34 15.96 4.88 3.23 24.07 400 0.89 0.49 0.44 16.12 5.19 4.86 26.17 600 0.82 0.56 0.55 13.04 5.54 7.36 25.94 800 0.88 0.54 0.78 17.09 6.92 7.77 31.78 Average 0.85 0.46 0.48 14.91 4.84 4.91 24.66 LSD 0.48 0.37 0.15 9.00 3.77 1.15 8.36 ANOVA ns ns ** ns ns ** ** b0 0.832 0.20 0.23 12.55 1.75 1.51 17.72 b1: X - 0.0835 - 0.7973 0.6523 - - b2: X 0.5 0.0004 -0.0051 0.0056 -0.0507 -0.0150 0.0770 0.1589 b3: X 2 - - - - - - - R2 0.19 0.97 0.95 0.41 0.96 0.97 0.85 Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) Nitrogen and phosphorus doses in elephant grass 285 periods prior to subsequent harvests. The average matter and breakdown of plant residues incorporated accumulated DM yields for the 3 harvests obtained in into the soil before planting plus biological nitrogen our study, despite being high, were still lower than the fixation. Morais et al. (2011), using 15N natural abundance yields of 61.9‒85.3 t DM/ha observed in the studies technique, reported contributions of biological N fixation of Santos et al. (2014) and Morais et al. (2009). Those above 50 % of the N requirement in Elephant grass. The authors cultivated Elephant grass in soils with higher association of Elephant grass plants with some N2-fixing fertility than that of the arenosol used in the present bacteria, such as Gluconacetobacter diazotrophicus, study (Table 1). In addition, total growth periods in those Herbaspirillum seropedicae and H. frisingense, has been studies (660 days) were much longer than the 260 days reported in the literature (Kirchhof et al. 2001; Reis et al. in the present study. 2001; Camelo et al. 2021; Santos et al. 2021). In another Despite limited differences in N concentration in study, Morais et al. (2009) observed N accumulation of harvested forage with increasing N application rates, N 217 kg N/ha in Elephant grass cultivar ‘Cameroon’, when uptake by the forage generally increased with increase harvest was delayed until 12 months after planting. in N application. It is worth pointing out that both N Martuscello et al. (2009), studying critical levels concentration and N uptake are highly subject to the of P in soil and in the shoots of Elephant grass during effects of dilution and DM yields of forage, because of establishment, obtained lower DM yields than those the increase in DM accumulation in shoots following found in this study. The highest yields were 9.3 and 12.7 application of N fertilizer. Thus, the limited variation in t DM/ha, combined yields from first and second harvests N concentration in forage merely masked the increased at 50 and 110 days after planting, following applications uptake of N by plants as level of N fertilizer increased. of 52 and 131 kg P/ha, respectively. P concentration However, N concentration in forage at harvest ranged in forage at the second and third harvests in our study from 1.87 to 2.67 % for individual treatments at increased with increase in P doses applied to the soil. individual harvests. The N content exceed the level the Martuscello et al. (2009) also observed an increase in P N content normally found in grasses (1.5 to 2.0 %), but concentration in Elephant grass as a function of increase the N contents are adequate for Elephant grass (Avalhaes in P doses applied. At both first and second harvests, et al. 2009; Morais et al. 2011). Although converting these authors recorded average P concentrations in plant leaf nitrogen content to leaf protein by multiplying by material of 1.0 g/kg DM for the control and 1.4 g/kg 6.25 may overestimate the true protein content of raw DM when 131 kg P/ha was applied, higher than those materials, it is widely accepted as an industry standard obtained in the present study for similar conditions of (Walsh et al. 2018). The higher levels of N, equivalent low initial P availability (0.6 mg/dm3). On the other to 16.7 % protein, are adequate even for high-producing hand, Moreira et al. (2006) found P concentrations in dairy cows (Busanello et al. 2017; Alessio et al. 2020). plants at the first harvest of 0.7 g/kg DM for control and A serious concern for soil management is the amount 1.7 g/kg DM, when 131 kg P/ha was applied. In general, of N which is taken up by the pasture grass. Total uptake P concentrations in herbage >0.3 % are adequate for of N by the various treatments during the 260 days ranged ruminants (Suttle 2010). In Elephant grass, application from 480 to 1,040 kg N/ha. Under a cut-and-carry system of 22 kg P/ha fulfilled this objective in our study. of feeding, removal of these amounts of N from the soil- pasture system could lead to a depletion of soil N levels Conclusions over time unless the manure from animals is returned to the pasture or N fertilizer is applied. Flores et al. (2012), In the arenosol of the experimental site, while high doses testing the effects of N fertilizer and harvest age on DM of nitrogen and phosphorus produced the highest yields production of Elephant grass in the Brazilian Cerrado, of elephant grass cultivar ‘Cameroon’, i.e. 37.4‒40.0 t observed that applying 150 kg N/ha produced an uptake DM/ha over 3 harvests (260 days) with doses of 600 kg of 471 kg N/ha at 180 days after planting. Fagundes et al. N/ha, satisfactory DM yields were obtained at 200 kg (2007) evaluated the influence of N fertilizer application N/ha. Further longer-term studies (over at least 2 years) on DM production of Elephant grass cultivar ‘Guaçu’ and are needed to confirm this finding along with economic observed an increase in N extraction by the grass as level assessments to determine the financial soundness of these of N applied increased. In the Control treatment in our strategies. It appears that there is little merit in applying study, 482 kg N/ha was taken up by Elephant grass, which more than 22 kg P/ha to the grass pastures, but longer- would have come from the mineralization of soil organic term studies would determine if this hypothesis is correct. Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) 286 L.E.C. Oliveira et al Acknowledgments EMBRAPA (Empresa Brasileira de Pesquisa Agropecuária). 1997. Manual de Métodos de Análises de Solo. 2nd Edn. The authors thank the Federal Rural University of the Embrapa Solos, Rio de Janeiro, Brazil. Semi-Arid Region (UFERSA) and all professors who Fagundes JL; Fonseca DM da; Mistura C; Salgado LT; Queiroz are part of the Graduate Program in Soil and Water DS; Morais RV de; Vitor CMT; Moreira LM. 2007. Nitrogen Management (PPGMSA), as well as the Coordination and potassium fertilization on Elephant grass cv. napier under rotational grazing. Boletim da Indústria Animal for the Improvement of Higher Education Personnel 64(2):149–158. (In Portuguese). bit.ly/3NNsY29 (CAPES) for all financial assistance, and the Agricultural Ferreira DF. 2011. Sisvar: a computer statistical analysis School of Jundiaí - UFRN for their support and for system. Ciência e Agrotecnologia 35(6):1039–1042. doi: providing the study area. 10.1590/S1413-70542011000600001 Flores RA; Urquiaga SS; Alves BJR; Collier LS; Morais References RF de; Prado RM. 2012. Effect of nitrogen fertilizer and cutting age on the dry matter production of Elephant grass (Note of the editors: All hyperlinks were verified 8 July 2022). in Savana. Revista Brasileira de Engenharia Agrícola e Ambiental 16(12):1282–1288. (In Portuguese). doi: Avalhaes CC; Prado RM; Rozane DE; Romualdo LM; Correia 10.1590/S1415-43662012001200004 MAR. 2009. Macronutrients omission and the growth and Foloni JSS; Tiritan CS; Calonego JC; Alves Junior J. 2008. nutritional status of Elephant grass (cv. Mott) growing Rock phosphate fertilization and phosphorus recycling by in nutrient solution. Scientia Agraria 10(3):215–222. (In pearl millet, Brachiaria sp., corn and soybean. Revista Portuguese). doi: 10.5380/rsa.v10i3.14528 Brasileira de Ciência do Solo 32(3):1147–1155. (In Alessio DRM; Velho JP; Tambara AAC; Gomes IPO; Knob Portuguese) doi: 10.1590/S0100-06832008000300023 DA; Haygert-Velho IMP; Busanello M; Thaler-Neto A. IDEMA (Instituto de Desenvolvimento Sustentável e Meio 2020. Dietary roughage sources affect lactating Holstein Ambiente). 2013. Macaíba, Perfil do seu Município, x Zebu cows under experimental conditions in Brazil: a Brazil. idema.rn.gov.br meta-analysis. Tropical Animal Health and Production Kirchhof G; Eckert B; Stoffels M; Baldani JI; Reis VM; 52:185–193. doi: 10.1007/s11250-019-02005-x Hartmann A. 2001. Herbaspirillum frisingense sp. nov., Beltrão VA; Freire LCM; Santos MF. 1975. Levantamento a new nitrogen-fixing bacterial species that occurs in Semidetalhado da Área do Colégio Agrícola de Jundiaí C4-fibre plants. International Journal of Systematic - Macaíba/RN. SUDENE - Recursos de Solos, Recife, and Evolutionary Microbiology 51(1):157–168. doi: Brazil. 10.1099/00207713-51-1-157 Busanello M; Velho JP; Alessio DRM; Haygert-Velho IMP; Martuscello JA; Fonseca DM da; Moreira LM; Ruppin RF; Tambara AAC; Thaler-Neto A. 2017. In situ ruminal Cunha DNFV da. 2009. Critical soil and shoot phosphorus degradability of protein feeds with distinct physical forms: levels for the establishment of Elephant grass. Revista a meta-analysis. South African Journal of Animal Science Brasileira de Zootecnia 38(10):1878–1885. doi: 10.1590/ 47(1):91-95. doi: 10.4314/sajas.v47i1.13 S1516-35982009001000004 Camelo A; Barreto CP; Vidal, MS; Rouws JRC; Lédo FJS; Martuscello JA; Majerowicz N; Cunha DNFV; Amorim Schwab S; Baldani JI. 2021. Field response of two seed PL; Braz TGS. 2016. Productive and physiological propagated elephant grass genotypes to diazotrophic characteristics of Elephant grass under nitrogen bacterial inoculation and in situ confocal microscopy fertilization. Archivos de Zootecnia 65(252):565–570. (In colonization analyses. Symbiosis 83:41–53. doi: 10.1007/ Portuguese) bit.ly/3aqp55m s13199-020-00730-8 Mello ACL de; Lira MA; Dubeux Júnior JCB; Santos MVF Cantarutti RB; Tarré RM; Macedo R; Cadisch G; Resende CP; dos; Freitas EV de. 2002. Characterization and selection of Pereira JM; Braga JM; Gomide JA; Ferreira E; Alves BJR; Elephant grass clones (Pennisetum purpureum Schum.) at Urquiaga S; Boddey RM. 2002. The effect of grazing intensity the Pernambuco forest zone. Revista Brasileira Zootecnia and the presence of a forage legume on nitrogen dynamics 31(1):30–42. (In Portuguese). doi: 10.1590/S1516- in Brachiaria pastures in the Atlantic forest region of the 35982002000100004 South of Bahia, Brazil. Nutrient Cycling in Agroecosystem Morais RF de; Souza BJ de; Leite JM; Soares LHB; Alves BJR; 64:257–271. doi: 10.1023/A:1021415915804 Boddey RM; Urquiaga S. 2009. Elephant grass genotypes Costa A da; Albuquerque JA; Costa A da; Pertile P; Silva for bioenergy production by direct biomass combustion. FR da. 2013. Water retention and availability in soils Pesquisa Agropecuária Brasileira 44(2):133–140. doi: of the state of Santa Catarina-Brazil: effect of textural 10.1590/S0100-204X2009000200004 classes, soil classes and lithology. Revista Brasileira de Morais RF de; Quesada DM; Reis VM; Urquiaga S; Alves Ciência do Solo 37(6):1535–1548. doi: 10.1590/S0100- BJR; Boddey RM. 2011. Contribution of biological 06832013000600010. nitrogen fixation to Elephant grass (Pennisetum purpureum Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) Nitrogen and phosphorus doses in elephant grass 287 Schum.). Plant and Soil 356(1–12):23–24. doi: 10.1007/ nitrogen fixation associated with tropical pasture grasses. s11104-011-0944-2 Australian Journal of Plant Physiology 28(9):837–844. Moreira LM; Fonseca DM da; Martuscello JA; Nóbrega EB doi: 10.1071/PP01079 da. 2006. Phosphorus absorption and critical levels in the Santos CM dos; Nascimento WBA; Nascimento BP do; Schwab shoot for the Elephant grass (Pennisetum purpureum cv. S; Baldani JI; Vidal MS. 2021. Temporal assessment of Napier) production maintenance. Ciência e Agrotecnologia root and shoot colonization of Elephant grass (Pennisetum 30(6):1170–1176. (In Portuguese). doi: 10.1590/S1413- purpureum Schum.) host seedlings by Gluconacetobacter 70542006000600020 diazotrophicus strain LP343. Microbiological Research Oliveira PPA; Oliveira WS de; Corsi M. 2007. Residual effect 244:126651. doi: 10.1016/j.micres.2020.126651 of soluble phosphorus fertilizers on pasture recovery Santos DH; Tiritan CS; Foloni JSS; Fabris LB. 2010. (Brachiaria brizantha cv. Marandu) in sandy soil. Revista Sugarcane yield under fertilization with filter cake Brasileira de Zootecnia 36(6):1715–1728. (In Portuguese) enriched with soluble phosphate. Pesquisa Agropecuária doi: 10.1590/S1516-35982007000800003 Tropical 40(4):454-461. (In Portuguese) doi: 10.1590/ Porto MLA; Puiatti M; Fontes PCR; Cecon PR; Alves S1983-40632010000400002 JC; Arruda JA de. 2012. Zucchini yield and nitrate Santos MMP; Daher RF; Ponciano NJ; Gravina GA; Pereira accumulation in fruits as a function of the nitrogen AV; Santos CL. 2014. Elephant grass responses under fertilization. Bragantia 71(2):190–195. (In Portuguese) doses of nitrogen fertilization coverage for energy doi: 10.1590/S0006-87052012005000020 purposes. Revista de Ciências Agrárias 37(1):100–108. (In Primavesi AC; Primavesi O; Corrêa LA; Cantarella H; Silva Portuguese) doi: 10.19084/rca.16804 AG; Freitas AR; Vivaldi LJ. 2004. Nitrogen fertilization Speratti AB; Johnson MS; Sousa HM; Dalmagro HJ; Couto in coastcross grass: effects on nutrient extraction and EG. 2018. Biochar feedstock and pyrolysis temperature apparent nitrogen recovery. Revista Brasileira de effects on leachate: DOC characteristics and nitrate losses Zootecnia 33(1):68–78. (In Portuguese). doi: 10.1590/ from a Brazilian Cerrado Arenosol mixed with agricultural S1516-35982004000100010 waste biochars. Journal of Environmental Management Queiroz Filho JL; Silva DS da; Nascimento IS do. 2000. 211(1):256–268. doi: 10.1016/j.jenvman.2017.12.052 Dry matter production and quality of Elephant grass Suttle NF. 2010. Mineral nutrition of livestock. 4th Edn. CABI (Pennisetum purpureum Schum.) cultivar Roxo at different Publishing, Cambridge, UK. cutting ages. Revista Brasileira de Zootecnia 29(1):69–74. Tedesco MJ; Gianello C; Bissani CA; Bohnen H; Volkweiss (In Portuguese) doi: 10.1590/S1516-35982000000100010 SJ. 1995. Análise de solo, plantas e outros materiais. 2nd Quesada DM; Coelho CHM; Boddey R; Reis VM; Alves B; Edn. Departamento de Solos da Universidade Federal do Urquiaga S. 2003. Effect of green manure and N-fertilizer Rio Grande do Sul, Porto Alegre, Brazil. on biomass accumulation and biological nitrogen fixation in Walsh JJ; Jones DL; Chadwick DR; Williams AP. 2018. different genotypes of Elephant grass (Pennisetum purpureum Repeated application of anaerobic digestate, undigested Schum.). Agronomia 37(2):54–59. bit.ly/3NWOEsu cattle slurry and inorganic fertilizer N: Impacts on pasture Reis VM; Reis Junior FB do; Quesada DM; Oliveira OCA de; yield and quality. Grass and Forage Science 73(3):758– Alves BJR; Urquiaga S; Boddey RM. 2001. Biological 763. doi: 10.1111/gfs.12354 (Received for publication 15 June 2020; accepted 21 June 2022; published 30 September 2022) © 2022 Tropical Grasslands-Forrajes Tropicales is an open-access journal published by International Center for Tropical Agriculture (CIAT), in association with The Tropical Crops Genetic Resources Institute of The Chinese Academy of Tropical Agricultural Sciences (TCGRI-CATAS). This work is licensed under the Creative Commons Attribution 4.0 International (CC BY 4.0) license. Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) Tropical Grasslands-Forrajes Tropicales (2022) Vol. 10(3):288–296 288 doi: 10.17138/TGFT(10)288-296 Regional Communication Forage production and quality of Urochloa decumbens cultivar ‘Basilisk’ in Okinawa, Japan Producción y calidad del forraje de Urochloa decumbens cultivar ‘Basilisk’ en Okinawa, Japón TAKASHI HANAGASAKI Okinawa Livestock Research Center, Nakijin, Okinawa, Japan. pref.okinawa.jp/site/norin/chikuken Abstract Two studies were conducted to assess forage growth and nutritive value of Urochloa decumbens (syn. Brachiaria decumbens) cultivar ‘Basilisk’ in comparison with other grass species grown in Okinawa during 2002–2005 and 2006–2008. Harvests were performed every 40 days from April to October and every 55 days from November to March. In Experiment 1, from 2002 to 2005 total dry matter (DM) yield of ‘Basilisk’ (119.5 t/ha) was significantly higher than that of Digitaria eriantha cultivar ‘Transvala’ (87.4 t/ha; P = 0.01), one of the most popular recommended grass varieties in Okinawa Prefecture. Mean DM digestibility of ‘Basilisk’ was 56.7 %, significantly higher than that of other recommended grass varieties (54.5–51.4 %). In addition, total digestible DM yield (64.8 t/ha) and crude protein (CP) yield (13.7 t/ha) of ‘Basilisk’ were significantly higher than those of other varieties including ‘Transvala’ (P<0.01). In Experiment 2, total DM yield of ‘Basilisk’ during 2006–2008 was 93.0 t/ha and significantly higher only than that of Urochloa mutica (syn. Brachiaria mutica) (78.6 t/ha; P<0.01), whereas mean DM digestibility (54.8 %) was significantly higher only than that of Chloris gayana cultivar ‘Katambora’ (52.8 %; P<0.05). Total digestible DM yield (48.8 t/ha) of ‘Basilisk’ was significantly higher only than that of U. mutica (40.3 t/ha; P<0.01) while its total CP yield (10.4 t/ha) was similar to those of other Urochloa cultivars (P>0.05). As a result, in 2016 ‘Basilisk’ was approved to be added to the list of grasses recommended for sowing in Okinawa Prefecture for improving beef production in the area. A suitable supply of seed to allow this cultivar to be sown widely is essential if its potential for improving beef production in the region is to be realized. Keywords: Brachiaria decumbens, crude protein, dry matter digestibility, seasonal production, tropical forage. Resumen Se realizaron dos estudios para evaluar el crecimiento y el valor nutritivo del cultivar ‘Basilisk’ de Urochloa decumbens (syn. Brachiaria decumbens) en comparación con otras especies de gramíneas cultivadas en Okinawa durante 2002‒2005 y 2006‒2008. Las cosechas se realizaban cada 40 días desde abril a octubre y cada 55 días desde noviembre a marzo. En el Experimento 1, de 2002 a 2005 el rendimiento total de materia seca (MS) de ‘Basilisk’ (119.5 t/ha) fue significativamente mayor que el del cultivar ‘Transvala’ de Digitaria eriantha (87.4 t/ha; P = 0.01), una de las variedades de pasto más popularmente recomendadas en la prefectura de Okinawa. La digestibilidad media de la MS de ‘Basilisk’ fue del 56.7 %, significativamente más alta que la de otras variedades de gramíneas recomendadas (54.5- 51.4 %). Además, el rendimiento total de MS digestible (64.8 t/ha) y el rendimiento de proteína cruda (PC) (13.7 t/ ha) de ‘Basilisk’ fueron significativamente más altos que los de otras variedades, incluida ‘Transvala’ (P<0.01). En el Experimento 2, el rendimiento total de MS de ‘Basilisk’ durante 2006‒2008 fue de 93.0 t/ha y solo significativamente mayor que el de Urochloa mutica (sin. Brachiaria mutica) (78.6 t/ha; P<0.01), mientras que la digestibilidad media de la MS (54.8 %) fue solo significativamente mayor que la del cultivar de ‘Katambora’ de Chloris gayana (52.8 %; Correspondence: Takashi Hanagasaki, Okinawa Agricultural Research Center, 906-0012, Miyakojima, Okinawa, Japan . Email: hangskit@yahoo.co.jp Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) Growth of Urochloa decumbens in Okinawa 289 P<0.05). El rendimiento total de MS digestible (48.8 t/ ha) de ‘Basilisk’ fue solo significativamente mayor que el de U. mutica (40.3 t/ha; P<0.01) mientras que su rendimiento total de PC (10.4 t/ha) fue similar al de otros cultivares de Urochloa (P>0.05). Como resultado, en 2016 se aprobó que ‘Basilisk’ se agregara a la lista de pastos recomendados para la siembra en la prefectura de Okinawa para mejorar la producción de carne en el área. Un suministro adecuado de semillas que permita sembrar ampliamente este cultivar es esencial si se quiere aprovechar su potencial para mejorar la producción de carne en la región. Palabras clave: Brachiaria decumbens, digestibilidad, forraje tropical, producción estacional, proteína cruda. Introduction Materials and Methods In Okinawa, the southernmost part of Japan, the beef Research location and period industry accounted for 24.5 % of gross agricultural production in 2019 and sales of calves during the past 10 The research was conducted during 2 periods (2002– years ranked fourth throughout Japan. According to the 2005 and 2006–2008) at Okinawa Livestock Research official figures of the Okinawa prefectural government, Center (Nakijin, Okinawa, Japan) (26º68' N, 127º94' E; more than 3 million foreign tourists visited Okinawa in 90 masl). Soils of the experimental area were Kunigami 2018, the highest number recorded in any year at that Merge, a red acidic Acrisol (Miyagi and Kondo 1990) stage, and 10 million of domestic and international and are composed of 32.8 % clay (<0.002 mm particle tourists were recorded in 2019. There is an urgent need size), 63.7 % silt (0.002–0.2 mm) and 4.6 % sand (>0.2 to boost the production of Okinawa’s famous unique mm) (Oshiro 2007). Chemical characteristics are: pH beef brands such as ‘IshigakiGyu’, ‘YamashiroGyu’ 4.7, total carbon, 1.33 %; total nitrogen, 0.12 %; organic and ‘MiyakoGyu’ to cope with this increase. To ensure matter, 2.2 %; cation exchange capacity, 13.5 meq/100 breeding cows are healthy with high reproductive rates g soil; Ca, 74.4 meq/100 g soil; Mg, 40.6 meq/100 g and high growth rates are achieved in growing animals, soil; and K, 14.0 meq/100 g soil. Specific gravity is feeding of high-quality grass plays an important role. relatively high. Main clay mineral is kaolinite. Climatic The Prefecture of Okinawa consists of many small conditions during the 2 experimental periods (2002– islands where cattle are raised on pasture, and droughts 2005 and 2006–2008) are illustrated in Figures 1 and 2, are experienced on some islands. Okinawa experiences indicating a cold and dry winter season from November a subtropical climate so tropical perennial grasses can to March. be grown successfully. Growing highly productive In Experiment 1, between June 2002 to April 2005 grasses would remove the need for farmers to depend performance of ‘Basilisk’ was compared with that on imported forage. Introduction of new species suitable of the other Urochloa species and cultivars as well as for grazing by cattle would be a significant advance and cultivars from other genera, while in Experiment 2 from drought tolerance would be an added benefit. Urochloa December 2006 to December 2008 comparisons were decumbens (Stapf) R.D. Webster (syn. Brachiaria made with other species within the genus Urochloa as decumbens Stapf) cultivar ‘Basilisk’, originally from well as Chloris gayana cultivar ‘Katambora’. Uganda, is known for its drought tolerance (Oram 1990; Miles et al. 1996). Actually, ‘Basilisk’ is sown Experiment 1 in Queensland, Australia and since the 1970s has been grown in Brazil, well-adapted to infertile acid soils in the The design was a complete randomized block with 7 Brazilian savanna, while showing high productivity and grasses and 3 replications. Seeds of selected grasses persistence (Kissmann 1977). were sown at 27.8 kg/ha on 11 October 2001, except It was considered that ‘Basilisk’ could be possibly more for Digitaria eriantha cultivar ‘Transvala’ and productive than grasses currently grown in Okinawa so Cynodon nlemfuensis (Table 1), which were planted we investigated yield of dry matter, digestible dry matter vegetatively using stolons at 4 stolons/m2 on 29 October and crude protein of Basilisk in comparison with those 2001. Seed of Ch. gayana cultivar ‘Katambora’ was sown of recommended grass varieties of Okinawa Prefecture on 22 April 2002 to increase the number of cultivars and other species of the genus Urochloa. involved. Plot size was 2 × 3 m = 6 m2. Basal fertilizer Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) 290 T. Hanagasaki. Monthly rainfall Mean monthly rainfall Mean temperature Maximum temperature Minimum temperature 35 500 30 400 25 300 20 200 15 100 10 0 5 Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar AprMay Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar AprMay Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr 2002 2003 2004 2005 Figure 1. Average, maximum and minimum temperatures, monthly rainfall from June 2002 to April 2005 and mean monthly rainfall. Monthly rainfall Mean monthly rainfall Mean temperature Maximum temperature Minimum temperature 35 500 30 400 25 300 20 200 15 100 10 0 5 Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 2006 2007 2008 Figure 2. Average, maximum and minimum temperatures, monthly rainfall from December 2006 to December 2008 and mean monthly rainfall of N:P:K at 50:13:25 kg/ha was applied at planting April 2002, while ‘Katambora’ was harvested for the and further N:P:K at 70:17:45 kg/ha was applied as first time in June 2002. Harvesting occurred at locally maintenance fertilizer after each harvest. Measurement used intervals, i.e. approximately every 40 days from of species performance commenced in June 2002, when April to October (summer season) and approximately original plantings had become established following a every 55 days from November to March (winter season) few preliminary harvests. The previous harvest was in in each year, ceasing in April 2005. Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) Rainfall (mm) Rainfall (mm) Temperature(℃) Temperature(℃) Growth of Urochloa decumbens in Okinawa 291 Table 1. List of grass varieties investigated in Experiments 1 In vitro dry matter digestibility and crude protein and 2. analysis Experiment 1 Experiment 2 U. decumbens 'Basilisk' U. decumbens 'Basilisk' Grass samples were analyzed for in vitro DM digestibility U. humidicola U. brizantha 'Marandu' according to the pepsin-cellulase assay (Goto and Minson Ch. gayana 'Callide' U. humidicola 1977). Crude protein (CP) concentration was determined by the microKjeldahl method using an Auto Analyzer. Ch. gayana 'Katambora' U. brizantha 'MG5' Cynodon nlemfuensis U. ruziziensis Statistical analysis D. eriantha 'Transvala' U. mutica Megathyrsus maximus Ch. gayana 'Katambora' One-way analysis of variance (ANOVA) was used for the 'Gatton ' statistical analysis of DM yield, in vitro DM digestibility, digestible DM yield, CP concentration and CP yield in Experiment 2 each experiment using the RStudio Version 1.4.1717. Tukey’s test was used to identify the differences between The design was again a complete randomized block with Basilisk and other grass varieties. 7 grasses and 3 replications. On 18 October 2005 seeds of grasses were sown at 41.7 kg/ha, while Urochloa mutica Results was planted with 4 stolons/m2 (Table 1). Basal fertilizer of N:P:K at 80:14:40 kg/ha was applied at planting and Experiment 1 the same quantities were applied as maintenance fertilizer after each harvest. Again plants were allowed to establish Dry matter. In Experiment 1 from 2002 to 2005, there before observations commenced on 18 December 2006. was no significant difference in DM yield between The previous harvest was on 7 November 2006. Harvests ‘Basilisk’ and other varieties in 2002 (Table 2). occurred approximately every 55 days from November However, in 2003 DM yield of ‘Basilisk’ (46.7 t/ha) to March and every 40 days from April to October as was significantly greater than that of U. humidicola described above until December 2008. (33.1 t/ha; P = 0.02), C. nlemfuensis (33.8 t/ha; P = 0.03) and ‘Transvala’ (30.2 t/ha; P = 0.004). As a result, Dry matter yield total DM yield of ‘Basilisk’ (119.5 t/ha) from 2002 to 2005 was significantly greater than that of ‘Transvala’ In both studies, forage on each plot was harvested (6 m2) (87.4 t/ha; P = 0.01). at approximately 10 cm from ground level and forage samples (500 g) for each plot were collected and dried Digestibility. Mean DM digestibility of ‘Basilisk’ at 70 ℃ for 48 h to determine DM yield. The dried (56.7 %) from 2002 to 2005 was significantly greater samples were ground with a mill and the powder was than that of U. humidicola (54.3 %; P = 0.02), ‘Callide’ sieved through a 1 mm mesh for analyzing for nitrogen (54.5 %; P = 0.03), ‘Katambora’ (51.4 %; P<0.001) and concentration for crude protein concentration and in C. nlemfuensis (52.9 %; P<0.001) (Table 3). vitro DM digestibility. For digestible DM yield, there was no significant Table 2. Dry matter yield (t/ha) of forage of a range of tropical grasses in Okinawa (Experiment 1). Grass variety 2002 S4h1 2003 S7h 2004 S8h 2005 S2h Total Urochloa decumbens 'Basilisk' 21.6 ± 1.1 46.7 ± 1.4 45.7 ± 0.7 5.5 ± 0.2 119.5 ± 0.5 Urochloa humidicola 17.1 ± 1.6 33.1 ± 3.5* 43.4 ± 1.4 5.1 ± 0.5 98.7 ± 6.9 Chloris gayana 'Callide' 20.2 ± 0.7 36.9 ± 0.7 38.5 ± 1.8 5.3 ± 0.4 100.9 ± 3.3 Chloris gayana 'Katambora' 18.5 ± 0.5 35.4 ± 3.1 37.0 ± 3.6 4.4 ± 0.8 95.3 ± 7.6 Cynodon nlemfuensis 19.6 ± 0.8 33.8 ± 0.9* 38.6 ± 2.2 3.2 ± 0.3 95.1 ± 2.7 Digitaria eriantha 'Transvala' 16.7 ± 0.5 30.2 ± 1.7** 36.5 ± 1.5 4.1 ± 0.2 87.4 ± 3.0* Megathyrsus maximus 'Gatton' 21.4 ± 0.9 42.2 ± 0.9 43.7 ± 0.6 4.9 ± 0.4 112.1 ± 1.0 Within columns, * and ** indicate significant differences (P<0.05) and (P<0.01) compared with ‘Basilisk’; 1Sum of n harvest. Source: Mochizuki et al. 2005. Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) 292 T. Hanagasaki. Table 3. Mean dry matter digestibility (%) of forage of a range of tropical grasses in Okinawa (Experiment 1). Grass variety 2002 M4h1 2003 M7h 2004 M8h 2005 M2h Overall mean U. decumbens 'Basilisk' 54.4 ± 0.2 56.5 ± 0.5 54.3 ± 0.2 72.8 ± 1.0 56.7 ± 0.1 U. humidicola 53.2 ± 0.1 53.8 ± 0.7* 51.3 ± 0.2* 71.6 ± 0.2 54.3 ± 0.3* Ch. gayana 'Callide' 52.8 ± 0.2 54.0 ± 0.3 50.9 ± 0.2** 74.7 ± 0.1 54.5 ± 0.2* Ch. gayana 'Katambora' 48.3 ± 0.3*** 50.7 ± 0.1*** 48.1 ± 0.5*** 74.4 ± 0.4 51.4 ± 0.3*** C. nlemfuensis 50.7 ± 0.7* 53.0 ± 0.5** 48.8 ± 0.8*** 74.6 ± 0.4 52.9 ± 0.7*** D. eriantha 'Transvala' 51.3 ± 0.4 56.5 ± 0.5 53.5 ± 0.3 76.5 ± 0.6* 56.1 ± 0.1 M. maximus 'Gatton' 51.0 ± 1.0* 54.8 ± 0.2 53.7 ± 0.5 76.2 ± 0.6* 55.5 ± 0.5 Within columns, * and *** indicate significant differences (P<0.05), (P<0.01) and (P<0.001) compared with ‘Basilisk’; 1Mean of n harvest. Source: Hanagasaki et al. 2006. difference between ‘Basilisk’ and other varieties in that of ‘Katambora’ (10.1 t/ha; P = 0.02) and ‘Transvala’ 2002 and 2005, while in 2003 that of ‘Basilisk’ (25.5 t/ (9.5 t/ha; P = 0.008). ha) was significantly greater than those of U humidicola (17.4 t/ha; P = 0.006), ‘Katambora’ (17.7 t/ha; P = Experiment 2 0.008), C. nlemfuensis (17.0 t/ha; P = 0.004) and ‘Transvala’ (16.1 t/ha; P = 0.001) (Table 4), and in 2004 Dry matter. In Experiment 2 from 2005 to 2008, DM that of ‘Basilisk’ (23.8 t/ha) was significantly greater yield of ‘Basilisk’ (2.6 t/ha) was significantly lower than those of ‘Katambora’ (17.1 t/ha; P = 0.01) and than that of U. brizantha ‘MG5’ (3.9 t/ha; P = 0.007) in C. nlemfuensis (18.1 t/ha; P = 0.03). Over the complete 2006 but was significantly higher than that of U. mutica study, total digestible DM yield of ‘Basilisk’ (64.8 t/ (46.3 vs. 37.9 t/ha; P = 0.01) in 2007 and overall ha) was significantly higher than that of U. humidicola (93.0 vs.78.6 t/ha; P = 0.008) (Table 7). (51.2 t/ha; P = 0.04), ‘Katambora’ (46.8 t/ha; P = 0.006), Digestibility. Mean DM digestibility of ‘Basilisk’ C. nlemfuensis (47.2t/ha; P = 0.007) and ‘Transvala’ (54.8 %) from 2006 to 2008 was significantly higher (46.3 t/ha; P = 0.005). than that of ‘Katambora’ (52.8 %; P = 0.02) (Table 8). Crude protein. Mean CP concentration (13.1 %) of Digestible DM yield of Basilisk in 2007 and 2008 ‘Basilisk’ was significantly higher than that of ‘Callide’ (23.6 and 23.5 t/ha) was significantly greater than those and ‘Katambora’ (P<0.05) (Table 5). of U. mutica (19.4 t/ha; P = 0.03 and 19.6 t/ha; P = 0.01) Regarding CP yield, that of ‘Basilisk’ (5.1 t/ha) was (Table 9). Total digestible DM yield of ‘Basilisk’ significantly higher than that of U. humidicola (3.6 t/ha; (48.8 t/ha) was also significantly greater than that of U. P = 0.03) and ‘Transvala’ (3.3 t/ha; P = 0.01) in 2003, mutica (40.3 t/ha; P = 0.004). while in 2005 that of ‘Basilisk’ (1.0 t/ha) was significantly Crude protein. Mean CP concentration (12.7 %) higher than that of ‘Katambora’ (0.6 t/ha; P = 0.04) and and total CP yield (10.4 t/ha) of ‘Basilisk’ were not C. nlemfuensis (0.5 t/ha; P = 0.02) (Table 6). Total CP significantly different from those of other Urochloa yield of ‘Basilisk’ (13.7 t/ha) was significantly higher than cultivars (Tables 10 and 11). Table 4. Digestible dry matter yield (t/ha) of forage of a range of tropical grasses in Okinawa (Experiment 1). Grass variety 2002 S4h1 2003 S7h 2004 S8h 2005 S2h Total U. decumbens 'Basilisk' 11.4 ± 0.6 25.5 ± 0.4 23.8 ± 0.1 4.2 ± 0.1 64.8 ± 0.3 U. humidicola 8.9 ± 0.8 17.4 ± 1.9** 21.2 ± 0.5 3.8 ± 0.3 51.2 ± 3.5* Ch. gayana 'Callide' 10.3 ± 0.5 19.8 ± 0.3 18.9 ± 1.0 4.2 ± 0.3 53.2 ± 1.9 Ch. gayana 'Katambora' 8.6 ± 0.1 17.7 ± 1.6** 17.1 ± 1.8* 3.5 ± 0.6 46.8 ± 4.0** C. nlemfuensis 9.6 ± 0.5 17.0 ± 0.6** 18.1 ± 0.7* 2.5 ± 0.3 47.2 ± 1.2** D. eriantha 'Transvala' 8.3 ± 0.2 16.1 ± 0.8** 18.7 ± 0.7 3.2 ± 0.2 46.3 ± 1.5** M. maximus 'Gatton' 10.6 ± 0.6 21.7 ± 0.5 22.2 ± 0.4 3.8 ± 0.3 58.3 ± 0.9 Within columns, * and ** indicate significant differences (P<0.05) and (P<0.01) compared with ‘Basilisk’. 1Sum of n harvest. Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) Growth of Urochloa decumbens in Okinawa 293 Table 5. Mean crude protein concentration (%) of forage of a range of tropical grasses in Okinawa (Experiment 1). Grass variety 2002 M4h1 2003 M7h 2004 M8h 2005 M2h Overall mean U. decumbens 'Basilisk' 13.0 ± 0.2 12.8 ± 0.1 12.2 ± 0.1 17.9 ± 0.7 13.1 ± 0.1 U. humidicola 13.3 ± 0.4 12.8 ± 0.4 11.2 ± 0.2 16.6 ± 0.6 12.7 ± 0.3 Ch. gayana 'Callide' 11.9 ± 0.3 11.5 ± 0.2 10.7 ± 0.0 13.9 ± 0.2** 11.5 ± 0.1* Ch. gayana 'Katambora' 11.6 ± 0.3 11.6 ± 0.3 10.6 ± 0.5* 13.5 ± 0.5*** 11.4 ± 0.3* C. nlemfuensis 14.0 ± 0.4 13.6 ± 0.4 12.0 ± 0.3 17.1 ± 0.1 13.5 ± 0.3 D. eriantha 'Transvala' 13.4 ± 0.3 12.6 ± 0.3 10.9 ± 0.2 15.2 ± 0.2* 12.4 ± 0.1 M. maximus 'Gatton' 13.6 ± 0.4 13.7 ± 0.1 11.4 ± 0.2 16.5 ± 0.5 13.1 ± 0.1 Within columns, *, ** and *** indicate significant differences (P<0.05), (P<0.01) and (P<0.001) compared with ‘Basilisk’. 1Mean of n harvest. Table 6. Crude protein yield (t/ha) of forage of a range of tropical grasses in Okinawa (Experiment 1). Grass variety 2002 S4h1 2003 S7h 2004 S8h 2005 S2h Total U. decumbens 'Basilisk' 2.6 ± 0.2 5.1 ± 0.1 4.9 ± 0.1 1.0 ± 0.0 13.7 ± 0.2 U. humidicola 2.2 ± 0.1 3.6 ± 0.4* 4.3 ± 0.1 0.8 ± 0.1 10.9 ± 0.7 Ch. gayana 'Callide' 2.2 ± 0.2 3.8 ± 0.1 3.8 ± 0.2 0.7 ± 0.1 10.5 ± 0.4 Ch. gayana 'Katambora' 2.0 ± 0.1 3.8 ± 0.4 3.7 ± 0.5 0.6 ± 0.1* 10.1 ± 1.1* C. nlemfuensis 2.5 ± 0.2 4.1 ± 0.2 4.3 ± 0.1 0.5 ± 0.0* 11.5 ± 0.4 D. eriantha 'Transvala' 2.0 ± 0.1 3.3 ± 0.2* 3.6 ± 0.1 0.6 ± 0.0 9.5 ± 0.4** M. maximus 'Gatton' 2.6 ± 0.2 5.0 ± 0.1 4.5 ± 0.1 0.8 ± 0.1 12.9 ± 0.2 Within columns, * and ** indicate significant differences (P<0.05) and (P<0.01) compared with ‘Basilisk’; 1Sum of n harvest. Table 7. Dry matter yield (t/ha) of forage of Urochloa cultivars and Chloris gayana cultivar ‘Katambora’ in Okinawa (Experiment 2). Grass variety 2006 1 harvest 2007 S8h1 2008 S8h Total U. decumbens 'Basilisk' 2.6 ± 0.1 46.3 ± 2.4 44.1 ± 1.4 93.0 ± 3.7 U. brizantha 'Marandu' 3.1 ± 0.2 42.1 ± 0.9 40.3 ± 0.2 85.5 ± 1.3 U. humidicola 2.9 ± 0.0 44.7 ± 0.4 43.0 ± 0.3 90.5 ± 0.4 U. brizantha 'MG5' 3.9 ± 0.2** 46.8 ± 0.6 46.6 ± 1.3 97.2 ± 0.9 U. ruziziensis 2.1 ± 0.2 44.4 ± 1.4 47.4 ± 0.9 94.0 ± 2.2 U. mutica 2.0 ± 0.2 37.9 ± 0.6* 38.7 ± 0.6 78.6 ± 0.9** Ch. gayana 'Katambora' 2.8 ± 0.1 44.1 ± 0.6 48.3 ± 1.2 95.3 ± 1.7 Within columns, * and ** indicate significant differences (P<0.05) and (P<0.01) compared with ‘Basilisk’; 1Sum of n harvest. Source: Kudaka et al. 2020. Table 8. Mean dry matter digestibility (%) of forage of Urochloa cultivars and Chloris gayana cultivar ‘Katambora’ in Okinawa (Experiment 2). Grass variety 2006 1 harvest 2007 M8h1 2008 M8h Overall mean U. decumbens 'Basilisk' 62.7 ± 1.0 54.4 ± 0.6 54.2 ± 0.1 54.8 ± 0.4 U. brizantha 'Marandu' 61.2 ± 0.9 55.3 ± 0.2 53.5 ± 0.8 54.8 ± 0.4 U. humidicola 61.2 ± 0.5 53.7 ± 0.3 54.2 ± 1.0 54.4 ± 0.5 U. brizantha 'MG5' 54.9 ± 0.5*** 53.5 ± 0.1 52.4 ± 0.1 53.1 ± 0.1 U. ruziziensis 63.2 ± 0.2 56.6 ± 0.2* 55.2 ± 0.3 56.4 ± 0.1 U. mutica 63.7 ± 0.4 54.2 ± 0.2 51.5 ± 0.3 53.5 ± 0.2 Ch. gayana 'Katambora' 60.9 ± 0.8 53.3 ± 0.3 51.3 ± 0.1 52.8 ± 0.1* Within columns, * and ** indicate significant differences (P<0.05) and (P<0.01) compared with ‘Basilisk’; 1Mean of n harvest. Source: Kudaka et al. 2020. Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) 294 T. Hanagasaki. Table 9. Digestible dry matter yield (t/ha) of forage of Urochloa cultivars and Chloris gayana cultivar ‘Katambora’ in Okinawa (Experiment 2). Grass variety 2006 1 harvest 2007 S8h1 2008 S8h Total U. decumbens 'Basilisk' 1.6 ± 0.1 23.6 ± 1.4 23.5 ± 0.8 48.8 ± 2.2 U. brizantha 'Marandu' 1.9 ± 0.1 22.2 ± 0.5 21.2 ± 0.2 45.3 ± 0.4 U. humidicola 1.7 ± 0.0 22.2 ± 0.2 22.9 ± 0.4 46.8 ± 0.2 U. brizantha 'MG5' 2.1 ± 0.1 23.9 ± 0.4 24.2 ± 0.7 50.3 ± 0.4 U. ruziziensis 1.4 ± 0.1 23.8 ± 0.8 25.3 ± 0.4 50.5 ± 1.1 U. mutica 1.3 ± 0.1 19.4 ± 0.4* 19.6 ± 0.3* 40.3 ± 0.6** Ch. gayana 'Katambora' 1.7 ± 0.0 21.5 ± 0.2 24.0 ± 0.6 47.3 ± 0.7 Within columns, * and ** indicate significant differences (P<0.05) and (P<0.01) compared with ‘Basilisk’; 1Sum of n harvest. Source: Kudaka et al. 2020. Table 10. Mean crude protein concentration (%) of forage of Urochloa cultivars and Chloris gayana cultivar ‘Katambora’ in Okinawa (Experiment 2). Grass variety 2006 1 harvest 2007 M8h1 2008 M8h Overall mean U. decumbens 'Basilisk' 18.2 ± 0.5 12.5 ± 0.3 12.2 ± 0.2 12.7 ± 0.1 U. brizantha 'Marandu' 18.0 ± 0.7 13.9 ± 0.4 13.0 ± 0.3 13.7 ± 0.4 U. humidicola 17.1 ± 0.2 12.4 ± 0.3 11.7 ± 0.1 12.4 ± 0.2 U. brizantha 'MG5' 16.1 ± 0.3 13.2 ± 0.2 12.3 ± 0.2 12.9 ± 0.2 U. ruziziensis 18.9 ± 0.4 13.5 ± 0.3 12.3 ± 0.3 13.2 ± 0.1 U. mutica 20.6 ± 0.4 13.9 ± 0.1 11.1 ± 0.2 12.9 ± 0.2 Ch. gayana 'Katambora' 17.0 ± 0.3 12.3 ± 0.2 11.3 ± 0.2 12.1 ± 0.2 1Mean of n harvest. Source: Kudaka et al. 2020 Table 11. Crude protein yield (t/ha) of forage of Brachiaria cultivars and Chloris gayana cultivar ‘Katambora’ in Okinawa (Experiment 2). Grass variety 2006 1 harvest 2007 S8h1 2008 S8h Total U. decumbens 'Basilisk' 0.5 ± 0.0 5.1 ± 0.3 4.8 ± 0.2 10.4 ± 0.5 U. brizantha 'Marandu' 0.6 ± 0.0 5.2 ± 0.1 4.8 ± 0.0 10.6 ± 0.1 U. humidicola 0.5 ± 0.0 4.9 ± 0.1 4.4 ± 0.0 9.8 ± 0.1 U. brizantha 'MG5' 0.6 ± 0.0 5.4 ± 0.2 4.9 ± 0.2 10.9 ± 0.2 U. ruziziensis 0.4 ± 0.0 5.1 ± 0.3 4.5 ± 0.1 10.0 ± 1.4 U. mutica 0.4 ± 0.0 4.6 ± 0.0 3.9 ± 0.1 8.9 ± 0.1 Ch. gayana 'Katambora' 0.5 ± 0.0 4.7 ± 0.1 4.9 ± 0.2 10.1 ± 0.2 1Sum of n harvest. Discussion supporting claims that this cultivar has a good level of tolerance of drier conditions. However, even in this ‘dry’ This study has provided valuable information on the year, rainfall received was 1,530 mm, which should be relative performance of a range of tropical grasses adequate to support good DM yields given that >100 mm including U. decumbens cultivar ‘Basilisk’ over a was registered in 7 of the 12 months. number of years in Okinawa. ‘Basilisk’ performed as As well as having good DM production, ‘Basilisk’ well as all other cultivars evaluated in most years and showed CP concentration as high as all other cultivars outperformed some cultivars in some years. While tested with mean values of 13.1 and 12.7 % in the rainfall in 2003 was well below the long-term mean for 2 experiments. The level of U. decumbens (14 %) the area (Figure 1), ‘Basilisk’ maintained a high level of was within the range expected for immature leaves in production and had higher DM yield than U. humidicola, Costa Rica (Lascano and Euclides 1996). ‘Basilisk’ C. nlemfuensis and D. eriantha cultivar ‘Transvala’, would provide an excellent diet for grazing ruminants, Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) Growth of Urochloa decumbens in Okinawa 295 especially given their ability to select a better quality diet Before it can be sown widely a reliable source of seed than the mean of total feed on offer. DM digestibility of will be needed. However, ‘Basilisk’ seed production in 56.7 and 54.8 % in the 2 experiments would ensure that Okinawa has been insufficient to meet demand (Kouki there was an adequate supply of available energy. et al. 2009), despite the fact ‘Basilisk’ seed production is In Okinawa, area of grass pasture covers more a success at 17‒22 °S and elevations of 600‒1,000 masl than 95 % of the total with 4 major grass varieties in Australia and Brazil (Hare et al. 2005). Unless seed recommended, i.e. Ch. gayana, D. eriantha, M. maximus production can be increased successfully in Okinawa, and C. nlemfuensis in the past 15 years. D. eriantha efforts will be needed to source supplies of seed from cultivar ‘Transvala’ has become the major grass sown countries where seed is already grown successfully if its in Okinawa and represented 27.9 % of total area sown potential is to be realized in the area. to grass in Okinawa in 2020 because of its perceived excellent characteristics and its suitability for local conditions (Hanagasaki 2022). However, ‘Basilisk’ was obviously superior to ‘Transvala’ in Experiment 1, in terms of yields of DM, digestible DM and CP, suggesting that it could be an acceptable substitute. In addition, ‘Basilisk’ performed significantly better than U. mutica, which is also a recommended grass variety for sowing in Okinawa Prefecture. In fact, ‘Basilisk’ is a high yielding species, particularly if nitrogen fertilizer is provided. Up to 30 t DM/ha/yr can be obtained on fertile soils in Vanuatu, and the same biomass production is possible under coconut plantations in the Solomon Islands (Cook et al. 2020). The average yield is, however, generally lower at about 10 t DM/ha/yr. ‘Basilisk’ yielded 4 t DM/ha/yr without fertilizer in Colombia (FAO 2016). In studies in northeast Brazil (Rodrigues et al. 2014), total forage production of ‘Basilisk’ (8.94t DM/ha) over 2 years was significantly higher than Koronivia grass (Brachiaria humidicola) and Gamba grass (Andropogon gayanus). In addition, milk yield of cows grazing ‘Basilisk’ (8 kg/cow/d) was greater than on Brachiaria Figure 3. Appearance of Urochloa decumbens cultivar dictyoneura (now: Urochloa humidicola) (6 kg/cow/d) ‘Basilisk’ pasture in Okinawa. in Colombia (Lascano and Euclides 1996). U. decumbens has an allelopathic effect, inhibiting Acknowledgments germination of seeds of other plants (Barbosa et al. 2008). In our study it prevented invasion by other grasses I am grateful to Messrs Yokota, Morikawa and Ebina throughout the 10 years since ‘Basilisk’ was planted (see for providing technical advice; Messrs Nagatoshi, bottom photo in Figure 3). Kudaka, Majikina and Touma and Mrs Mochizuki for Consequently, if ‘Basilisk’ is introduced and sown data collection; Messrs Kohama and Teruya and Mrs widely in Okinawa, it should result in production of Takeuchi for supporting this research; Mr Ahagon for providing the photographs; and Mr Nagayama for kindly high-quality forage, which could reduce annual cost showing how to use Rstudio. of keeping breeding cows (Kouki and Ebina 2009). Based on results of the current study it seems that U. References decumbens cultivar ‘Basilisk’ could be added to the list of suitable species for sowing on these islands. In fact, (Note of the editors: All hyperlinks were verified 1 July 2022). in 2016 ‘Basilisk’ was approved to be added to the list of grasses recommended for sowing in Okinawa Prefecture Barbosa EG; Pivello VR; Meirelles ST. 2008. Allelopathic for improving beef production in the area and significant evidence in Brachiaria decumbens and its potential to demand for seed has emerged. invade the Brazilian Cerrados. Brazilian Archives of Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) 296 T. Hanagasaki. Biology and Technology 51(4):825‒831. doi: 10.1590/ Kudaka M; Shioyama A; Nagatoshi M; Hanagasaki T; Nitta M. S1516-89132008000400021 2010. The test of choosing tropical forage grasses. Bulletin Cook BG; Pengelly BC; Schultze-Kraft R; Taylor M; Burkart of The Okinawa Prefectural Livestock and Grassland S; Cardoso Arango JA; González Guzmán JJ; Cox K; Research Center 48:63-70. (in Japanese). agriknowledge. Jones C; Peters M. 2020. Tropical Forages: An interactive affrc.go.jp/RN/2010814813.pdf selection tool. 2nd and Revised Edn. International Center Lascano CE; Euclides VPB. 1996. Nutritional quality and animal for Tropical Agriculture (CIAT), Cali, Colombia and production of Brachiaria pastures. In: Miles JW; Maass International Livestock Research Institute (ILRI), Nairobi, BL; Valle CB do, eds. Brachiaria: Biology, Agronomy, and Kenya. www.tropicalforages.info Improvement. CIAT publication 259. CIAT, Cali, Colombia FAO (Food and Agriculture Organization of the United and CNPGC/EMBRAPA, Campo Grande, MS, Brazil. p. Nations). 2016. Grassland Index. A searchable catalogue 106‒123. hdl.handle.net/10568/82028 of grass and forage legumes. FAO, Rome, Italy. Miles JW; Maass BL; Valle CB do, eds. 1996. Brachiaria: Goto I; Minson DJ. 1977. Prediction of the dry matter Biology, agronomy, and improvement. CIAT, Cali, digestibility of tropical grasses using a pepsin-cellulase Colombia and CNPGC/EMBRAPA, Campo Grande, MS, assay. Animal Feed Science and Technology 2(3):247‒253. Brazil. hdl.handle.net/10568/54362 doi: 10.1016/0377-8401(77)90028-1 Miyagi N; Kondo T. 1990. On the physical properties of Hanagasaki T; Mochizuki T; Morikawa N; Nagatoshi M; Kunigami Maaji soil – Physical and engineering properties Majikina M. 2006. The test of choosing tropical forage of local soils in Okinawa area (I). Transactions of the grasses. Bulletin of The Okinawa Prefectural Livestock Japanese Society of Irrigation, Drainage and Reclamation and Grassland Research Center 44:79-84. (in Japanese) Engineering 149:39–44. (In Japanese). doi: 10.11408/ agriknowledge.affrc.go.jp/RN/2010813245.pdf jsidre1965.1990.149_39 Hanagasaki T. 2022. Characteristics of Digitaria eriantha cv. Mochizuki T; Morikawa N; Nagatoshi M; Majikina M. 2005. Transvala grown in Okinawa, Japan. Taiwanese Journal of The test of choosing tropical forage grasses. Bulletin of The Agricultural Chemistry and Food Science. (In press). Okinawa Prefectural Livestock and Grassland Research Hare MD; Tatsapong P; Lunpha A; Wongpichet K. 2005. Center 43:30-36. (in Japanese). agriknowledge.affrc.go.jp/ Brachiaria species in north-east Thailand: Dry matter RN/2010813222.pdf yields and seed production. Tropical Grasslands 39:99‒106. Oram RN. 1990. Brachiaria. In: Register of Australian herbage bit.ly/3NuXQnX plant cultivars. 3rd Edn. CSIRO Division of Plant Industry, Kissmann KG. 1997. Plantas infestantes e nocivas. BASF, São Melbourne, Victoria, Australia. Bernardo do Campo, SP, Brazil. Oshiro K. 2007. Okinawan agriculture and soil – the aim of Kouki K; Ebina M. 2009. Breeding and forage utilization sustainable and developmental agriculture. Mitsuwa Social of Brachiaria grass (Brachiaria spp.). Japanese Journal Welfare Corporation, Okinawa, Japan. of Grassland Science 55(2):179‒187. (In Japanese). doi: Rodrigues RC; Sousa TVR; Melo MAA; Araújo JS; Lana 10.14941/grass.55.179 RP; Costa CS; Oliveira ME; Parente MOM; Sampaio Kouki K; Ebina M; Suenaga K; Nakanishi Y; Yokota M; Inafuku IBM. 2014. Agronomic, morphogenic and structural M; Hanagasaki T. 2009. Seed production properties of the characteristics of tropical forage grasses in Northeast genus Brachiaria. Bulletin of the Okinawa Prefectural Brazil. Tropical Grasslands-Forrajes Tropicales 2:214‒222. Livestock and Grassland Research Center 46:75‒79. (In doi: 10.17138/TGFT(2)214-222 Japanese). (Received for publication 20 October 2020; accepted 3 June 2022; published 30 September 2022) © 2022 Tropical Grasslands-Forrajes Tropicales is an open-access journal published by International Center for Tropical Agriculture (CIAT), in association with The Tropical Crops Genetic Resources Institute of The Chinese Academy of Tropical Agricultural Sciences (TCGRI-CATAS). This work is licensed under the Creative Commons Attribution 4.0 International (CC BY 4.0) license. Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) Tropical Grasslands-Forrajes Tropicales (2022) Vol. 10(3):297–301 297 doi: 10.17138/TGFT(10)297-301 Short Communication Effects of feeding dried olive (Olea europaea) leaves with wheat straw- concentrate rations on feed conversion efficiency in Awassi rams Efectos de la alimentación de hojas secas de olivo (Olea europaea) con raciones de concentrado de paja de trigo sobre la eficiencia de conversión alimenticia en carneros Awassi MAZEN ALOMAR, M. RATEB AL-MASRI AND MOUTAZ ZARKAWI Division of Animal Production, Department of Agriculture, Atomic Energy Commission, Damascus, Syria. aec.org.sy Abstract Three groups of Awassi rams were fed for 6 weeks either a conventional wheat straw-concentrate ration (Control) or 2 experimental rations, where 30 (G1) and 60 % (G2) of the wheat straw was replaced with dried olive leaves following oil extraction. All rations were isocaloric and isonitrogenous. Feed intake (FI), bodyweight gain (BWG) and feed conversion efficiency (FCE) were measured. BWG and FI during the 6-week period were not significantly (P>0.05) different for the different rations, averaging 4.75 and 116 kg/animal, respectively. In addition, there were no significant (P>0.05) differences in FCE values between the Control and experimental groups fed wheat straw + olive leaves, averaging 24.6 kg feed/kg gain. These results suggest that dried olive leaves can replace wheat straw in wheat straw-concentrate rations at levels up to 60 % without affecting performance. Further studies are needed to determine optimal combinations of straw, olive leaves and concentrate to achieve different goals as well as intakes and performance when offered rations ad lib. Economic assessments would determine if including olive leaves would reduce the costs of feeding. Keywords: Bodyweight gain, feeding, growth performance, tree forage. Resumen Tres grupos de carneros Awassi fueron alimentados durante 6 semanas con una ración convencional de concentrado-paja de trigo (control) o 2 raciones experimentales, donde el 30 (G1) y el 60 % (G2) de la paja de trigo se reemplazó con hojas secas de olivo después de la extracción del aceite. Todas las raciones fueron isocalóricas e isonitrogenadas. Se midieron el consumo de alimento (CA), la ganancia de peso (GDP) y la eficiencia de conversión alimenticia (ECA). GDP y CA durante el período de 6 semanas no fueron significativamente diferentes (P>0.05) para las distintas raciones, con un promedio de 4.75 y 116 kg/animal, respectivamente. Además, no hubo diferencias significativas (P>0.05) en los valores de ECA entre los grupos control y experimentales alimentados con paja de trigo + hojas de olivo, con un promedio de 24.6 kg de alimento/kg de ganancia. Estos resultados sugieren que las hojas secas de olivo pueden reemplazar la paja de trigo en las raciones de concentrado-paja de trigo en niveles de hasta el 60 % sin afectar el rendimiento. Se necesitan más estudios para determinar las combinaciones óptimas de paja, hojas de olivo y concentrado para lograr diferentes objetivos, así como consumos y rendimiento cuando se ofrecen raciones ad libitum. Las evaluaciones económicas determinarían si la inclusión de hojas de olivo reduciría los costos de alimentación. Palabras clave: Alimentación, forraje de árboles, ganancia de peso corporal, rendimiento del crecimiento. Correspondence: Mazen Alomar, Division of Animal Production, Department of Agriculture, Atomic Energy Commission, Damascus, Syria. Email: ascientific15@aec.org.sy Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) 298 M. Alomar, M.R. Al-Masri and M. Zarkawi Introduction (ME). For Groups G1 and G2, portion (30 and 60 %, respectively) of the wheat straw was replaced by dried Recycling of olive (Olea europaea L.) by-products olive leaves. Level of concentrate in the experimental and their utilization in animal nutrition could improve rations was adjusted, while the level of concentrate was the economics and efficiency of animal production in reduced and barley added to make all rations (Control Mediterranean regions, as well as having environmental and experimental) isocaloric and isonitrogenous as benefits (Molina-Alcaide and Nefzaoui 1996). Large recommended by Kirchgessner (1982) and Friesecke quantities of olive leaves are available as by-products of (1984). Animals were offered the same amounts of CP the olive-growing industry in winter in these regions and (166 g) and ME (15.3 MJ) daily. The balanced rations have potential for alleviating some of the feed shortages for the 3 groups are shown in Table 1. The concentrate and nutritional deficiencies experienced. mixture contained (fresh weight basis) 400 g barley/kg, Olive leaves represent around 10 % of the total 400 g wheat bran/kg, 180 g cotton cake/kg, 16 g salt/kg weight of olives retrieved at harvesting (Delgado- and 4 g premix of minerals and vitamins/kg. Nutritional Pertíňez et al. 1998). Air-drying for 7 days in the shade components of the feed ingredients are shown in Table 2. at room temperature had no detrimental effects on the The present study lasted for 8 weeks (July‒August) nutritive value of olive leaves and could represent a under natural photoperiod and temperature conditions. simple and low-cost procedure for their preservation Rams were fed on the same rations for 2 weeks to allow (Martín-García and Molina-Alcaide 2008). However, them to adapt to the rations before the beginning of the phenolic components (particularly tannins) in some tree 6-week experimental period. They were individually leaves may bind to protein, thus rendering the protein housed in metal pens with all feeds presented in separate undegradable by rumen microbes. troughs. Rams were weighed at the beginning of the We considered that performance of rams fed wheat experiment and weekly during the study in the morning straw-concentrate rations would not be affected by before feeding. The different components of each ration replacing some of the wheat straw with dried olive were thoroughly mixed prior to feeding. Animals were leaves. Therefore, the objective of the present work fed twice daily, at 08:30 and 14:30 h, and all feeds was to study effects on feed intake (FI), bodyweight offered were consumed. Water was available ad libitum. gain (BWG) and feed conversion efficiency (FCE) of FI (kg/animal), BWG (kg/animal) and FCE (kg feed/kg partial substitution of dried olive leaves for wheat straw gain) were determined during the 6-week period. in wheat straw-concentrate rations for Awassi rams plus determine concentrations of total phenols, total tannins Estimation of nutrient components and condensed tannins in the leaves. Standard methods as described in AOAC (1990) were Materials and Methods used for determination of dry matter (DM), ash, ether extract (EE) and crude protein (CP) concentrations. Materials tested, experimental animals and managment Nitrogen (N) concentration was measured by the Kjeldahl method, and CP concentration was calculated as N% × Olive (Olea europaea) leaves, as a by-product of olive 6.25. Crude fiber (CF) was determined by the method oil extraction, were collected from 4 factories and dried of Naumann and Bassler (1976) and in vitro digestible in the shade for 7 days. The dried leaves were thoroughly organic matter (IVDOM) and metabolizable energy mixed to be used later as feed supplement for Awassi (ME) according to the methods of Menke et al. (1979) rams. using a gas-production technique. Details of methods of Twelve adult Syrian Awassi rams aged between 2 and incubation and estimation of gas production, IVDOM 3 years were randomly allocated to 3 similar groups in and ME have been described previously (Al-Masri terms of age and initial body weight: Control group (n 2010). Total phenols were quantified by Folin Ciocalteu = 4; 71.3 ± 9.6 kg); G1 group (n = 4; 71.8 ± 4.7 kg); and reagent and total tannins were calculated as the difference G2 group (n = 4; 68.3 ± 6.1 kg). The Control group was between phenolics before and after tannin removal from fed a conventional wheat straw-concentrate ration (51 % the extract using insoluble polyvinyl pyrolidone (Makkar wheat straw and 49 % concentrate), containing 90.1 g/kg et al. 1993). Condensed tannins were determined by the crude protein (CP) and 8.37 MJ/kg metabolizable energy butanol-HCl method (Porter et al. 1986). Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) Dried olive leaves for sheep 299 Table 1. Daily intake of nutrients and feed ingredients by Awassi rams in Control and experimental groups (G1 and G2). Feed ingredient Intake (g/d) Crude Protein (g/d) Metabolizable Energy (MJ/d) Control G1 G2 Control G1 G2 Control G1 G2 Wheat straw 1,000 700 400 26.0 18.2 10.4 6.18 4.33 2.47 Olive leaves - 300 600 - 20.2 40.4 - 1.79 3.58 Concentrate mix 1,000 730 400 139.6 101.9 55.8 9.20 6.71 3.68 Barley - 220 500 - 25.8 58.7 - 2.47 5.61 Total 2,000 1,950 1,900 165.6 166.1 165.3 15.38 15.30 15.34 G1, G2 = 30 and 60 % of the straw replaced by olive leaves, respectively. Table 2. Nutritional components (g/kg DM) and metabolizable Table 3. Effects on BWG, FI and FCE of partial substitution of energy (ME; MJ/kg DM) of the feed ingredients. dried olive leaves (DOL) for wheat straw in straw-concentrate CP A CF EE IVDOM ME rations for Awassi rams over 6 weeks. Wheat straw 27.7 115 319 11.8 505 6.6 BWG FI FCE Dried olive leaves 72.9 99.1 163 50 490 6.5 Control 4.35a 119a 27.4a Concentrate 155 53.9 66 36.2 734 10.2 G1 (30 % DOL) 5.00a 116a 23.2a Barley 126 25.0 54 24 816 12.0 G2 (60 % DOL) 4.89a 113a 23.1a CP = crude protein s.e.m. 0.20 1.73 1.42 A = ash P-value 0.893 0.689 0.827 CF = crude fiber Means in the same column followed by different letters are EE = ether extract different at P<0.05. BWG= (kg/animal); FI= (kg/animal); FCE= IVDOM = in vitro digestible organic matter. (kg feed/kg gain). Statistical analyses Discussion Data were subjected to analysis of variance (ANOVA) using a Statview-IV program (Abacus Concepts, The absence of significant differences between the 2 Berkeley, CA, USA) to test the effects of ration type on experimental groups and Controls in terms of BWG, FI FI, BWG and FCE. Means were separated using Fisher’s and FCE indicates that there was no adverse effect of least significant difference test at the 95 % confidence including olive leaves in straw-concentrate rations for level. Awassi rams during a period of 6 weeks. One would expect similar performance since FI was controlled and Results rations were adjusted to be isocaloric and isonitrogenous, provided no anti-nutritional components were present. BWG and FI during the 6-week feeding period were not Since leaves of some shrubs can contain significant affected by ration type (P = 0.893 and 0.689, respectively) amounts of tannins, which can affect utilization of protein (Table 2) and averaged 4.75 and 116 kg/animal, in the rumen, it might have been expected that utilization respectively. As a result, FCE was not significantly (P of N from the olive leaves might be lower than that of N = 0.827) affected by ration type fed, averaging 24.6 in the concentrate. Getachew et al. (2002) reported that kg feed/kg gain (Table 3). Average amounts of total plant samples containing total phenols and tannins at phenols, total tannins and condensed tannins in the olive concentrations (g tannic acid equivalent/kg DM) up to 40 leaves used in this experiment were 77.3, 17.0 and 5.4 g/ and 20 g/kg, respectively, are not expected to precipitate kg DM, respectively. protein and, therefore, are unlikely to adversely affect Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) 300 M. Alomar, M.R. Al-Masri and M. Zarkawi ruminant production. However, total condensed tannin Delgado-Pertíñez M; Chesson A; Provan GJ; Garrido A; concentrations exceeding 50 g/kg DM can inhibit ruminal Gómez-Cabrera A. 1998. Effect of different drying systems microbial activity, depress dry matter digestibility (Kumar for the conservation of olive leaves on their nutritive value and Vaithiyanathan 1990) and reduce voluntary FI for ruminants. Annales De Zootechnie 47:141–150. hal. (Waghorn et al.1990). The anti-nutritional components archives-ouvertes.fr/hal-00889721 (total phenols, total tannins and condensed tannins) Delgado-Pertíñez M; Gόmez-Cabrera A; Garrido A. 2000. Predicting the nutritive value of the olive leaf (Olea measured in the olive leaves used in this experiment were europaea): digestibility and chemical composition and in low (77.3, 17.0 and 5.4 g/kg DM, respectively), which vitro studies. Animal Feed Science and Technology 87(3– should not depress protein utilization. The levels in our 4):187–201. doi: 10.1016/s0377-8401(00)00195-4 sample of leaves were not abnormal as Delgado-Pertíňez Friesecke H. 1984. Handbuch der praktischen Fütterung. BLV- et al. (2000) failed to detect hydrolyzable or condensed Verlagsgesellschaft mbH, München, Germany. tannins in olive leaves, and others (Yáňez-Ruiz et al. Getachew G; Makkar HPS; Becker K. 2002. Tropical 2004; Martín-García and Molina-Alcaide 2008) found browses: contents of phenolic compounds, in vitro gas low amounts of total condensed tannins in olive leaves production and stoichiometric relationship between short (range 9.6‒11.1 g/kg DM). chain fatty acid and in vitro gas production. Journal of Agricultural Science 139(3):341–352. doi: 10.1017/ S0021859602002393 Conclusions Kirchgessner M. 1982. Tierernährung, 5. Auflage. DLG- Verlag, Frankfurt, Germany. The data suggest that replacing 30 or 60 % of wheat Kumar R; Vaithiyanathan S. 1990. Occurrence, nutritional straw in a 50:50 straw-concentrate mixture with dried significance and effect on animal productivity of tannins olive leaves and adjusting the overall rations to be in tree leaves. Animal Feed Science and Technology 30(1– isocaloric and isonitrogenous had no significant effect on 2):21–38. doi: 10.1016/0377-8401(90)90049-E performance of rams. Therefore, olive leaves could play Makkar HPS; Blümmel M; Borowy NK; Becker K. 1993. an important role as a source of feed in extensive animal Gravimetric determination of tannins and their correlations production systems in arid and semi-arid zones, making with chemical and protein precipitation methods. Journal effective use of a by-product of olive production. More of the Science of Food and Agriculture 61(2):161–165. doi: 10.1002/jsfa.2740610205 studies are needed to determine the optimal combinations Martín-García I; Molina-Alcaide E. 2008. Effect of different of straw, olive leaves and concentrates to achieve drying procedures on the nutritive value of olive (Olea different goals as well as others to measure voluntary FI europaea var. europaea) leaves for ruminants. Animal and performance under ad libitum feeding. Economic Feed Science and Technology 142(3–4):317–329. doi: analyses would determine to what extent including olive 10.1016/j.anifeedsci.2007.09.005 leaves in rations would alter costs of feeding rams. Menke KH; Raab L; Salewski A; Steingass H; Fritz D; Schneider W. 1979. The estimation of the digestibility Acknowledgments and metabolizable energy content of ruminant feedstuffs from the gas production when they are incubated with The authors thank the Director General and Head of rumen liquor in vitro. The Journal of Agricultural Science 93(1):217–222. doi: 10.1017/S0021859600086305 Agriculture Department, Atomic Energy Commission of Molina-Alcaide E; Nefzaoui A. 1996. Recycling of olive oil Syria, for their encouragement. by-products: possibilities of utilization in animal nutrition. International Biodeterioration and Biodegradation 38(3– References 4):227–235. doi: 10.1016/S0964-8305(96)00055-8 Naumann C; Bassler R. 1976. Die chemische Untersuchung (Note of the editors: All hyperlinks were verified 7 July 2022). von Futtermitteln, Methodenbuch Band III. Neumann- Neudamm, Berlin, Germany. Al-Masri MR. 2010. In vitro rumen fermentation kinetics Porter IJ; Hrstich LN; Chan BG. 1986. The conversion and nutritional evaluation of Kochia indica as affected of procyanidin and prodelphinidins. Phytochemistry by harvest time and cutting regimen. Animal Feed 25(1):223–230. doi: 10.1016/S0031-9422(00)94533-3 Science and Technology 157(1–2):55–63. doi: 10.1016/j. Waghorn GC; Jones WT; Shelton ID; McNabb WC. 1990. anifeedsci.2010.01.013 Condensed tannins and the nutritive value of herbage. AOAC (Association of Official Analytical Chemists). 1990. Proceedings of the New Zealand Grassland Association Official Methods of Analysis. 15th Edn. Association of 51:171–176. doi: 10.33584/jnzg.1990.51.1894 Official Analytical Chemists, Arlington, VA, USA. Yáňez-Ruiz DR; Martín-García AI; Moumen A; Molina- Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) Dried olive leaves for sheep 301 Alcaide E. 2004. Ruminal fermentation and degradation leaves. Journal of Animal Science 82(10):3006–3014. doi: patterns, protozoa population and urinary purine derivatives 10.2527/2004.82103006x excretion in goats and wethers fed diets based on olive (Received for publication 17 November 2020; accepted 16 June 2022; published 30 September 2022) © 2022 Tropical Grasslands-Forrajes Tropicales is an open-access journal published by International Center for Tropical Agriculture (CIAT), in association with The Tropical Crops Genetic Resources Institute of The Chinese Academy of Tropical Agricultural Sciences (TCGRI-CATAS). This work is licensed under the Creative Commons Attribution 4.0 International (CC BY 4.0) license. Tropical Grasslands-Forrajes Tropicales (ISSN: 2346-3775) TGFT Editorial Team A.A. 6713, Km 17 Recta Cali-Palmira. Cali, Valle del Cauca, Colombia. Phone: +57 2 4450100 Ext. 3084 Email: CIAT-TGFT-Journal@cgiar.org