' ' ·- ANNUAL -REPORT 1997 Project PE-5 Sustainable Systems for Smallholders: Integrating improved germplasm and resource management for enhanced crop and livestock production systems ~Ag3 5 5 5 o o 4 Mean 1.2 1.7 1.5 2.7 1.2 1.5 Table 4. Percent respondents (n=93) clearing given forest area (ha) by main agricultura! systems, Pucallpa, Peru, 1996 S lash-and-B urn Are a Forest Riverine Total Cattle Oil Palm TOTAL < 1.0 9 5 7 12 17 8 1.0-1.9 41 39 40 12 42 38 2.0-2.9 29 28 29 26 33 29 >= 3.0 21 28 24 50 8 25 Mean 1.8 2.0 1.9 2.6 1.5 1.9 Table 5. Percent of respondents (n=83) reporting criteria for choice of location of forest field to clear and crop by main agricultura! system, Pucallpa, Peru S lash-and-B urn Are a Forest Riverine Total Cattle Oil Palm TOTAL Fertile soil 9 5 7 12 17 8 No flooding 41 39 40 12 42 38 Clase to 29 28 29 26 33 29 roadlhouse 21 28 24 50 8 25 Flatter areas Fanners reported needing 20 days/ha for slashing (prior to and after felling) and 27 dayslha for felling when clearing forest parcels; and 16 dayslha for slashing and 6 dayslha for felling fallowed parcels (Table 6). Nearly all fanners grew rice in fields cleared from forest in the first year and cassava, maize, pasture, or other crops in the second year (if not fallowed). Fanners sowed rice, rnaize, cassava, and banana in fields cleared from fallows (Table 7). Although fluctuating by year, fanners overall maintained approximately equal areas sown to rice, maize, and cassava. For 1995, the sampled fanners sowed 64 ha of rice, 84 ha of maize, and 57 ha of cassava. For 1996, the same parcels sown to rice were fallowed (31% ), planted to cassava (18%), maize (7%), banana (7%), pasture (4%), ora variety of other uses (32%) not including TR-4 Targeting Research Table 6. Reported labor (days/ha) for clearing forest and fallow by main agricultura! system, Pucallpa. Peru Slash-and-Burn Forest Riverine Toral Cattle Palm TOTAL For Fal For Fa! For Fal For Fal For Fal For Fa! Slash 20 17 22 14 21 15 18 22 12 19 20 16 Fell 30 5 28 6 29 6 22 7 18 5 27 6 Total 50 22 50 20 50 21 40 29 30 24 47 22 Table 7. Percent respondents reponing main crop sown in fields cleared from forest (n= IOO) and from fallow (n= 132), by main agricultura! systems, Puca11pa, Peru Slash-and-B urn Forest Riverine Total Cattle Palm TOTAL For Fa! For Fal For Fal For Fa! For Fal For Fal Rice 88 52 73 49 80 50 50 33 84 50 78 49 Maize 6 30 16 28 11 29 o 47 8 25 lO 30 Cassava 3 lO o 2 1 5 25 o o 17 3 5 Banana 3 3 11 9 8 7 25 7 8 o 9 6 Other o 5 o 12 o 9 o 13 o 8 o lO rice. For parcels sown to maize in 1995, the following year included 49% in fallow, 21% in maize, 9% in banana, and 5% in cassava (and l% in rice). For the 1995 cassava area, 46% was fallowed, 29% was in cassava and 0% was in rice in 1996 (Table 8). Table 8. Changes in land use for lands cropped to rice, maize, and cassava from 1995 to 1996, Pucallpa (% area planted in 1995) Rice Planted 1996 (64 ha) Rice o Maize 7 Cassava 18 Banana 7 Pineapple 2 Pasture 4 Fallow 3 1 Other 32 Total lOO TR-5 Planted in 1995 Maize (84 ha) 1 21 5 9 o 3 49 11 lOO Cassava (57 ha) o o 29 5 8 4 45 100 Targeting Research Farmers reponed actual crop yields for 1995-96: the 1.4 t/h of rice. 1.7 t/h of maize, and 0.2 tlha of beans were lower than their reponed "normal" yields but higher than previous lowest yields of each respective crop (Table 9). Table 9. Respondents ' reponed crop yields (tlha), Pucallpa. 1995-96 ReEorted range Low Normal High Actual 1995-96 Rice 0.9 1.9 2.6 1.4 Maize l.l 2.2 2.9 1.7 Beans 0.1 0.8 0.9 0.2 Fields cleared from forest were cropped for a mean two years: 30% of the respondents cultivated for one year, 44% cultivated for two years. 16% for three years, and lO% for more than three years. Lands cultivated after fallow were cultivated a mean 1.3 years. Respondents discontinued cropping plots cleared from forest due to declining production (major reason cited by 75% of respondents) and weeds ( 46% ). Combining these interview-based resul ts with more informal field observation and discussions with settlers, each of the sub-groups of respondents can be described. Slash-and-burn farmers of the forest. Forest fanners had a high proportion of their lands cleared (73% in 1996), and the highest proponion (39%) of their fanns in fallow or secondary re-growth. Rice was the major crop, and one which suffered from yield-reducing diseases. These fanners had 10% of their lands in perennial crops such as citrus, achiote, cacao, and various fruits. A substantial number grew coca, although demand declined since the end of the domination of the area by the terrorist Sendero Luminoso. To sorne extent it appeared that charcoal production has replaced coca production asan income generating alternative. Riverine slash-and-burn farmers . These fanners had fanns which were 54% cleared and 26% in fallow. Banana was established as a cash crop on the relatively richer soils (compared to upper-forest areas) after initial slash-and-burn rice production. Sikatoka was widespread as a problem in banana, and one that farrners reponed may be exacerbated by the defoliants and herbicides sprayed from helicopters in the Peruvian arrny's efforts to eradicate coca fields. Flooding was a problem and solution (ie the deposits of richer river-borne silts). River farrners also earned incomes from fish and the softwood Guazuma crinita which grew in fallows. Small cattle ranchers. These farmers had the largest parcels (67 ha), the least forest (20%), the lowest proponions of land in annual and perennial crops and in fallow (although they had the largest annual crop fields in absolute terrns), and the highest proportion of land in pasture (54%). For the 20% of Pucallpa respondents having cattle, herd size was a mean 23 head. Pastures were reportedly 40% (of pasture area) Brachiaria spp and 28% Brachiaria spp plus Pueraria phaseoloides, although areas of native pasture (of Axonopus compressus, Paspalum conjugatum, and Homolepis aturensis) were clearly under-reported. Thiny-nine percent of the ranchers reponed using fire for pasture regeneration at a mean interval of every two years; and 68% reponed rotating animals to different pastures at a mean interval of one month. Pressure on pasture resources was low: 73% of these respondents maintained less than one head per ha, 24% had 1-2 head, and only 3% had more than 2 head of cattle per ha of pasture. lt was widely reponed that the period of terrorism by the Sendero ied to substantial declines in cattle numbers and reduced maintenance of fences and pastures. TR-6 Targeting Research Oil palm fanners . These slash-and-burn farrners ha ve accessible upper-area parcels, ha ve taken advantage of local development projects promoting oil palm (Eleais guineensis), and, therefore, have the largest proportion of their farrns in perennial crops ( 17%) and the lowest in pasrure (4%). Success of oil palm will depend on the development of processing infrastrucrure and demand sufficient to maimain prices at profitable Ievels. Conclusions: towards an appropriate research agenda Overall, Pucallpa farrners relied on rice as a majar crop for both sale and consumption. Research to help sol ve upland rice disease problems and the problems of soil nutrient depletion and increases of weeds would benefit many fanners in the area. Pucallpa farrners had a high proportion of their lands in fallow or secondary re-growth. Working with farmers on ímproved fallows using trees and legumes would appear to be reasonable. Upper-area farmers who had earned incomes from coca production (from sales anci/or from wages for weeding and harvesting) were seek.ing new altematives. Charcoal production-given use of selected suitable forest species such as Dipterix odorata-cannot be c!Xpected to be sustainable. Efforts to develop and promete new crops (such as camu camu Myciaria dubia) and agroindustries (eg palm oil) would appear to be reasonable; and research is needed to carefully determine ex ante demand for new altematives. Farmers have had experience with the promotion of supposedly income-generating crops such as citrus and achiote--which unforrunately were market failures. Riverine slash-and-burn farmers were most concerned about diseases affecting their banana plantations (and upland rice). Research to address the problem would be appropriate and needed. Research in Pucallpa has long targeted the cattle ranchers in the introduction and testing of forage and feeding systems altematives including legumes such as Arachis pintoi, Centrosema spp, Desmodium ovalifolium, Cratylia argentea, Stylosanthes guianensis, as well as forage grasses. These settlers, however, may have little interest in more productive forage systems as long as current pasture resources are more than sufficient given the area's reduced herd size. On the other hand, targeted work to increase systems productivity and sustainablity with the few ranchers maintaining more animals per area may be appropriate, as appears to be the case with the current work of the Instituto Veterinario de Investigaciones Tropicales y de Altura (IVIT A) and the CG project, Improved Feeding Systems for Dual Purpose cattle (TROPILECHE). References Brady NC (1996) Alternatives to slash-and-bum: a global imperative. Agriculture, Ecosystems and Environment 58:3-11 Fujisaka S, Escobar G, and Veneklaas E ( 1997) Plant community diversity relative to human land uses in an Amazon forest colony. Forthcoming in Biodiversity and Conservation Loker, WM ( 1993) Medio ambiente y agricultura en la Amazonia Peruana: un experimento metodologico. In: Loker Wlvl and Vosti S (eds) Desarrollo Rural en la Amazonia Peruana. Centro Internacional de Agricultura Tropical (CIA. n and the Intemationa1 Food Policy Research Institute (IFPRI), Cali, Colombia Riesco A and Arroyo M ( 1997) Perfil socio-economice de la region de Ucayali. Unpublished report, Consorcio para el Desarrollo Sostenible de Ucayali/Proyecto de Desarrollo de Agroempresas Rurales del CIA. T TR-7 Additional characterization in Pucallpa in relation to livestock ( 1.1.2. F. Holmann) Targeting Research The objective of this activity is to characterize the existing animal production systems in the benchmark sites at Pucallpa with emphasis on the characterization of existing resources, technologies, constrainsts, and opportunities. In mid-1996, 151 farmers were interviewed in a selected study area in Pucallpa, Perú, to understand current land use dynamics (see section above). About 20% of respondents had cattle with a mean herd size of 23 head. These farmers had the largest parcels (67 ha), the least forest (20%), the lowest proportions of land in annual or perennial crops, and the highest proportion of land in pasture (54%). About 68% of pasture area was established with Brachiaria spp in different stages of degradation. Pressure on pasture resources was low since 73% of the respondents maintained less than one head per ha, 24% had l-2 head, and only 3% had more than 2 head of cattle per ha of pasture. It was widely reported that the period of terrorism by the Sendero Luminoso (Shining Path) led to substantial declines in cattle numbers and reduced maintenance of fences and pastures. Research in Pucallpa has long targeted the smallholder cattle producers through the introduction and testing of new forage species including legumes and pasture grasses. These farmers, however, may ha ve little interest in more producti ve forage systems as long as current pasture resources are more than sufficient given the reduced herd size. Thus, work at this stage should attempt to increase productivity and sustainability in farms with large numbers of cattle per unit area. An additional survey on livestock in Pucallpa has been carried out and is currently being analyzed. TR-8 Integration and analysis of Forest Margins si te data ( 1.2.1; S. Fujisaka) Main achievements • Land use dynamics at three sites in Brazil and Peru were characterized. • Land uses by site after initial slash-and-bum cultivation were described. • Six hypotheses regarding causal factors in land use are offered. Targeting Research Pasture or pennanent crops after slash-&-burn cultivation? Land use choice in three amazon colonies Abstract. Settlers in the Amazon practice slash-and-burn agriculture in forest lands to produce annual crops. After cropping, lands are converted to pasture, planted to perennial crops, or fallowed in anticipation of future annual crop production. Land use was examined in three settlements-Pedro Peixoto in Acre and Theobroma in Rondonia, Brazil, and Pucallpa. Peru. Land use after slash-and-burn cultivation in (orest lands differed among the colonies. Colonists in Pedro Peixoto converted lands to pasture for cattle production; Theobroma settlers adopted a strategy encompassing both dual purpose cattle and perennial crop production~ while the more heterogeneous settlers in Pucallpa included small cattle ranchers and riverine and forest slash-and-burn farmers for whom perennial crops were important. Land uses are described; hypotheses regarding differences are offered; and implications for the adoption of agroforestry are discussed. Introduction Farmer-settlers in the western Amazon have traditionally practiced slash-and-burn agriculture to produce annual crops such as rice, maize, cassava, and beans. In so doing, colonists convert primary tropical forest lands to other uses which include pasture for beef or dual purpose (milk and meat) production, perennial crops, and fallows for future annual cropping. We examined land use in the colonies of Pedro Peixoto, Acre, and Theobroma, Rondonia, in Brazil and in Pucallpa, Peru, where settlement has taken place largely since the late 1960s when road access was improved to each area. Continued slash-and-burn agriculture has contributed to deforestation, carbon emissions into the atmosphere, and losses of biodiversity (Brady, 1996; Fujisaka et al, 1997). Areas converted to pasture--of either native or introduced species--in the Amazon have suffered degradation: many pastures in Acre and Rondonia are infested with Vemoniaferruginea and Imperara brasiliensis (ASB, 1994a); while a third of the pasture area in the cattle zone along the Pucallpa-Lima highway in the Department of Ucayali was considered to be degraded (Riesco et al, 1982). Co1onists' establishment of perennial crops, on the other hand, has been viewed as positive in tenns of diversification of househo1d incomes, andas a relatively sustainab1e land use option with fewer environmental costs compared to pasture or continued slash-and- burn cropping (Sanchez, 1995; Harwood, 1996). Overall, settlers in Pedro Peixoto con verted lands to pasture for beef cattle production, showed little interest in perennial crops, and exploited forest products such as Brazil out, timber, and game. Colonists in Theobroma employed a mixed strategy encompassing dual purpose cattle, perennial crops, and sorne use of forest products. Pucallpa was more heterogeneous, with sorne settlers forming small cattle ranches and others practicing slash-and-burn agriculture with perennial crop establishment and susbstantial proportions of land left in fallows. · TR-9 Targering Research This paper describes the different land uses after initial forest clearing, buming, and cropping in each of the colonies. Hypotheses are offered and discussed regarding the observed diffrences. The sites Pedro Peixoto in Acre and Theobrorna in Rondonia. Brazil, are govemrnent sponsored colonies in the sernideciduous forest zone. Pedro Peixoto was officially established in 1972, covers 370,000 ha divided into 3700 parcels distributed to 3200 farnilies. Lots are located frorn 50 to lOO km frorn the state capital of Rio Branco. Theobrorna was officially estabished in 1979 (both colonies were earlier settled by "spontaneous" arrivals), covers 300,000 ha distributed to 3000 farnilies, and is located about 350 km southeast of the state capital of Porto Velho. The highway BR364 passes through both colonies. Clirnate is wann and hurnid tropical with a rain y season frorn Novernber to June and a dry season frorn July to October in Acre and June to August in Rondonia. Annual rainfall is 2000 mm and mean ternperatures are 22-26°C. Soils in Pedro Peixoto are largely Oxisols and in Theobrorna are Oxisols, Ultisols, and Alfisols (Fujisaka et al, 1996). The Pucallpa site (in the Departrnent of Ucayali which borders the state of Acre to the east) features hurnid tropical forest and is located on an east-to-west gradient leading to the foothills of the Andes along which rainfall ranges frorn 1800 to 3000 mm (mean 2300 mm, with rainfall increasing to the west). Wet months are February-May and Septernber-November; while dry months are June-August and Decernber- January. Mean annual ternperature is 25°C. Soi1s include more favorable alluvia1, riverine systerns where pH is approxirnately 7.7 and availab1e Pis 15 ppm; and higher, well drained forested areas of acidic (pH 4.4), low P (2 pprn) soils. Flatter areas near the city of Pucallpa (but out of our area of interest) are poorly drained (aguajales) and dorninated by Mauritia spp palrns (Riesco and Arroyo, 1997; Loker, 1993). Methods A tearn of researchers representing the Empresa Brasileira de Pesquisa Agropecuaria (EMBRAPA), the Centro Internacional de Agricultura Tropical (CIA T), the Intemational Centre for Research in Agroforestry (ICRAF), and the Intemational Food Policy Research Institute (IFPRI) interviewed 81 settlers in Pedro Peixoto and 74 in Theobroma in late 1994. A similar tearn representing Peru's Instituto Nacional de Investigacion Agraria (INIA), CIAT, and ICRAF interviewed 118 settlers in Pucallpa, Peru, in rnid-1996. Interviews were conducted as a part of characterization activities for the ICRAF-coordinated global project "Altematives to Slash-and-Bum". Interviews dealt with pattems of land use and resource management. Responses were coded and data tabulated and presented in simple descriptive frequencies. Results Settlement. Twenty-seven percent of Pedro Peixoto respondents were bom in Acre; only 1% of Theobroma respondents were from Rondonia; and 27% of the Pucallpa respondents were frorn Ucayali. Respondents irnrnigrating to the study areas arrived a mean 16 years ago in the state of Acre and 9 years ago in the colon y of Pedro Peixoto; 13 years ago in Rondonia and 1 O years ago in Theobrorna, and 17 years ago in both Ucayali and the Pucallpa study area. The studied areas in Brazil have been settled since the 1960s when the highway BR364 was opened and subsequently irnproved; while settlement in Pucallpa started in the 1940s after the construction of the highway linking the city of Pucallpa and Lima on the coast and was facilitated by road irnprovements in the late 1960s. The low proportion of "native" Theobroma settlers seerns to indicare the arrival of a second generation of settlers from other parts of Brazil. TR- IO Targeting Research Overallland use. Pedro Peixoto settlers had parcels of a mean 88 ha. of which 34% was cleared (and 66% was still forested). Of the cleared lands andas a proponion of the total farm. 25% of the area was in pasture, 2% in faJlow, 7% in annual crops. and less than 1% in perenniaJ crops. Theobroma fanners had 76 ha parcels. of which 50% was cleared by 1995. The cleared area included 29% of the parcel in pasture, 4% in fallow, 9% in annuaJ crops. and 8% in perenniaJ crops. Pucallpa settlers had cleared 68% of their mean 35 ha parcels by 1996, with cleared areas including 25% of the land holding in pasture, 29% in fallow, 6% in annual crops, and 8% in perenniaJ crops (Table 1; these land uses were confmned by analysis of satellite images, Fujisaka et al, 1996). Table l. Land use (% area), Pedro Peixoto & Theobroma, Brazil, 1994-95, and Pucallpa, Peru. 1995-96 Pedro Peixoto Theobroma Pucallpa 1964 1995 1994 1995 1995 1996 Forest 69 66 54 50 35 32 Cleared 31 34 46 50 65 68 Pasture 20 25 26 29 25 25 Fallow 6 2 8 4 29 29 AnnuaJ crops 4 7 7 9 5 6 Perennial crops 1 < 1 5 8 6 8 Mean farrn size (ha) 88 76 35 Funher anaJysis of the Pucallpa data shows that its more heterogeneous 1andscape and population features: a) riverine and b) forest slash-and-burn agriculturaJists with about a half to three-fourths of their lands cleared, up to third of their lands in faJlow, lO% of lands in perenniaJ crops, and only 14% of their lands in pasture; e) cattle ranchers with larger farms which were 80% cleared, 54% in pasture, and 21% in faJlow; and a special group of slash-and-burn fanners who established oil paJm (Eleais guineensis) to the extent that they had 17% of their lands in perennial crops (Table 2). Tab1e 2. Land use(% area) by main agricultura! system, Pucallpa, Peru, 1996 S lash-and-Bum Forest Riverine Sub-Total Cattle Oil Palm Total (n=W) (n=7 1) (n=115) (n=23) (n=13) (n=15) Forest 27 46 38 20 51 33 C1eared 73 54 62 80 49 67 Pasture 16 12 14 54 4 25 Fallow 39 26 31 21 24 28 AnnuaJs 8 6 7 2 4 6 Perennials 10 7 10 3 17 8 Mean farm size (ha) 33 26 29 67 32 35 TR- 11 Targering Research Slash-and-burn agriculture. Farmers had cleared a mean L.5 to about 3.0 ha for annual cropping for the season in which interviews were conducted. Farmers cleared forest lands every two to three years, cropped the parcels from two to 2.5 years, and then fallowed lands for 2.5-3.5 years. Labor required for forest clearing was substantially lower in Brazil than in Pucallpa given the greater use of chainsaws in Brazil. Labor for clearing fallowed fields for re-cultivation usually did not require chainsaws and was substantially lower than that required for forest clearing (Table 3). Table 3. Slash-and-burn agriculture, Pedro Peixoto, Rondonia & Theobroma, Acre, Brazil. and Pucallpa, Peru. 1996 Use of forest lands Frequency of forest clearing (yrs) Mean area cleared per year* (ha) Mean years cultivation Mean years fallowed Land clearing Forest: labor (days/ha) Fallow; labor (dayslha) Use of chainsaw (% respondents) * actual area cleared in year of interview Pedro Peixoto 2.1 1.8 2.1 2.4 23 16 90 Theobroma 2.8 2.8 2.5 2.-t 14 14 89 Pucallpa 1.5 1.9 2.1 3.2 47 22 15 Upland rice was the main crop sown in the first year after clearing. In the second year, Pedro Peixoto farmers sowed maize, cassava, and pastures; Theobroma farmers sowed maize and pastures; and Pucallpa settlers sowed cassava, maize, or fallowed their lands. After fallowing, colonists largely planted upland rice and maize in the first year of re-cultivation (Table 4). Table 4. Farmers' (% respondents) dominant crop/land use choice for forest (first & second years after clearing) & fallow parcels (flfst year after re-clearing) Pedro Peixoto Theobroma PucallEa Forest Fallow Forest Fallow Forest Fallow l" 2'¡¡¡ l" 1" 2¡;¡¡ l" l" 2¡;¡¡ l" Rice 92 o 62 70 o 72 78 49 Maize o 42 23 o 44 13 lO 12 30 Cassava o 26 3 o o o 3 36 5 Pasture 17 o 13 25 2 o 5 o Fallow 8 8 29 Other 7 7 12 17 23 13 9 17 16 Total lOO lOO lOO 100 100 lOO lOO lOO lOO Farrners reported harvesting from 1.3-1.5 úha of rice from their forest fields (in the harvest preceeding the interviews). These amounts were s lightly lower than farrners' expected "normal" yields, but higher than their reported lowest yields of a mean 0.7-0.9 t/ha and lower than reported highs of 2.0-2.6 úha (Table 5). TR-1 2 Targering Research Table 5. Farmer-reported rice yields (ha) Pedro Peixoto Theobroma Pucallpa Low* 0.7 0.9 0.9 Normal 1.5 1.6 1.9 High 2.2 2.0 2.6 Actual** 1.3 1.5 1.4 * Farmers esúmated "low", "normal", and "high yields ** Reponed 1993 yields for Pedro Peixoto & Theobroma; 1996 for Pucallpa Land use after annual cropping. Sixty-four percent of Pedro Peixoto farmers, 44% of Theobroma farmers, but only 6% of Pucallpa farmers reported that they con verted lands to pasture as their major land use after cropping. Similarly, 91% of the Pedro Peixoto respondents, 81% of Theobroma settlers, and only 20% of Pucallpa farmers raised cartle. For those having cattle, settlers in Pedro Peixoto and Pucallpa had a mean 23 head, while Theobroma farmers had a mean 30 head. For farmers with cattle, pasture represented 22% of fann area in the Brazilian colonies and 36% of the Pucallpa ranches. Overall, pastures accounted for 25-29% of the total farm areas in the three colonies (Table 6). Table 6. Land use after annual cropping forest lands Land use Pasture & cartle Pasture = main land use after S & B Respondents having Hvestock (%) Mean head of cattle Mean pasture area (ha)* Pasture area (% of all farms) Perennial crops Pereruúals = main land use after S & B Have perennials (% respondents) Land in perennial crops (%) Fallow Fallow = main land use after S & B Land in fallow (% area) * For those respondents having cattle Pedro Peixoto 64 91 23 22 25 o 31 < 1 36 2 Theobroma Pucallpa 44 6 81 20 30 23 22 36 29 25 20 o 69 87 8 8 36 95 4 29 Pastures in Pedro Peixoto were, as reported by area, 57% Brachiaria brizantha, 24% B decumbens, and 10% Brachiaria spp plus Pueraria phaseoloides. Theobroma pastures were largely B brizantha (51 %), and B decumbens (13%). Pastures in Pucallpa were similarly of B brizantha (40%) and Brachiaria spp and P phaseoloides (28%). Respondents tended to under-report areas of native pasture, which in Pucallpa included degraded areas dominated by Axonopus compressus, Homolepis aturensis, and Paspalum conjugatum. TR-1 3 Targeting Research Twenty percent of Theobroma farmers and non e of the Pedro Peixoto and Pucallpa farmers reported thar their main use of lands after slash-and-bum agriculture was for perennial crops. Eighty-seven percent of Pucallpa and 69% of Theobroma fanners had perennial crops, accounting for 8% of their lands; while only 31% of Pedro Peixoto settlers had perennial crops on less than one percent of their \and (Table 6). Majar perennial crops in Pucallpa were banana. citrus, miscellaneous fruit, and oil palm. Theobroma farmers produced coffee, cacao, and rnisce!laneous fruir (Table 7). Table 7. Perennial crops (% respondents having named crop) Perennial Crops Pedro Peixoto Theobroma Pucallpa Coffee 12 46 o Cacao 2 34 3 Banana 15 1 37 Oil Palm o o 11 Ctrus 9 3 17 Other fruits 14 19 15 Softwoods o o 9 Other 1 o 5 Ninety-five percent of Pucallpa farmers reported fallows as their major land use after initial slash-and- bum cultivation in forest parcels; and fallows represented 29% of their lands. Although 36% of Theobroma and Pedro Peixoto settlers reported fallows as their major land use after cultivation, these covered only 2-4% of the farm areas in the two Brazilian colonies (Table 6). Forest products. Almost all Pedro Peixoto fanners gathered Brazil nuts (Bertholletia excelsa); and substantial numbers harvested timber (in small quantities), rubber, fish, and game. Theobroma settlers exploited timber, Brazil nuts, palm heart and fruit (Bactris gassipaes), and fis h and game. Only a quarter of Pucallpa farmers--and mainly those settled along the rivers--reported harvesting timber, mainly the softwood Guazuma crinita, locally bolaina blanca (Table 8). Field interactions revealed that a substantial number of Pucallpa fanners in areas farther from the roads produced coca for the illicit drug trade . Decline ín demand due to control of terrorism and a re tu m of state control has meant that sorne of these producers have turned to charcoal production, exploiting se\ected forest tree species such as Dipterix odorara. Table 8. Respondent's (%) use of forest products Forest product Pedro Peixoto Theobroma Pucallpa Brazil nut 90 36 o Wood 33 50 25 Hunting 18 29 ? Palmito 5 17 2 Fish 6 13 ? Rubber 11 o o Medicinal plants 5 6 o Other 11 16 3 Cash income sources. Fanners sold surpluses of rice, maize, beans (in Pedro Peixoto), and cassava (in Pucallpa). A fourth to a third of Theobroma farmers eamed cash from sales of coffee and cacao; while TR-14 Targeting Research similar proportions received irncomes frorn banana and citrus in Pucallpa. Pedro Peixoto settlers sold cattle (26% of respondents), milk (7%), and poultry and pigs (9%). Theobroma settlers had dual purpose operations, earning incomes frorn meat (22%) and milk (30%). Nineteen percent of Pucallpa settlers sold cattle and 27% earned cash from occasional sales of pigs andJor chickens. Almost half of Pedro Peixoto settlers sold Brazil nuts ; and 17% of Pucallpa settlers earned cash from sales of tirnber. At least half of the respondents in the three colonies reported incornes from off-farrn and non-farrn labor and frorn pensions or other off-farrn cash rernissions (Table 9). Table 9. Respondent's reported majar income sources (% respondents) lncome source Annual crop sales Rice Maize Beans Cassava Perennial crop sales Coffee Cacao Banana Citrus Livestock!livestock products Cattle Mil k Poultry/pigs Forest products Rubber Brazil nuts Wood Labor & pensions Pedro Peixoto Theobroma 50 57 47 16 41 9 4 o 3 36 o 25 o o o o 26 22 7 30 9 o 7 o 44 6 o 9 63 53 Pucallpa 39 37 2 19 o 2 29 25 19 o 27 o o 17 50 Land evaluation. Almost all settlers at each site reported that the values of their lands had risen over their time period of occupation. Perceived values (calculated in terms of head of cattle) increased from 74% per year in Pedro Peixoto to 157% per year in Theobroma. Reasons given for the increases included more pasture, fencing, and watering ponds for cattle in Pedro Peixoto; more pasture, fencing, perennial crops, and more area cleared in Theobroma; and perennial crops, pastures, fencing, and more area cleared in Pucallpa (Table 10). Díscussion and hypotheses Descriptive results showed that although slash-and-bum forest clearing and annual cropping in the three study areas were similar, subsequent land use differed. Pedro Peixoto settlers followed annual cropping with conversion to pasture for beef cattle production, kept little land in fallow, placed little irnportance on perennial crops, and exploited available forest products. TR-15 Targeting Research Table lO. Reported change in land values Pedro Peixoto Theobroma Pucallpa Respondents (%) reporting > value 93 97 95 Mean reported annual increase (%) 74 157 97 Reasons for increased value More pasture 60 50 30 Fencing 56 36 12 W atering pond 30 13 2 Corral 12 16 o More area opened 12 26 15 Perennial crops 12 35 46 House 26 16 28 Roads 25 27 39 Theobroma farrners also placed little land in fallow, but hada more mixed economy, balanced among dual purpose cattle production and perennial crops. Pucallpa farmers-who included forest and riverine slash-and-burn farrners, cattle ranchers, and oil palm producers--had a substantially higher propoction of their lands in fallow relative to the Brazilian colonists, had a much lower proportion of families with cattle; and-except for the cattle ranchers-relied upon perennial crops to an extent similar to Theobroma. These findings give rise to severa! hypotheses for fucther research (hypotheses which our current characterization data were notable to address with sufficient cectainty). Four related hypotheses are of a general nature: l. The more "pioneer" conditions of larger land holdings, greater proportions of forest, and greater isolation from urban markets favor the combination of cattle ranching and forest products exploitation. Pedro Peixoto is located nearly at the end of the highway BR364 far from populated center-south region of Brazil. Settlers with their larger parcels, extensive pastures, cattle and abundant forest lands had little interest in and had had little success with perennial crops due to a lack of markets. There would appear to be few alternatives to the extensive exploitation of land and forest resources of interest to these settlers. 2. Conversely, diminishing pioneer conditions--eg less area in primary forest, greater tiesto markets, decreasing farm size--favor more mixed economies which include perennial crops and dual purpose cattle production. Theobroma and Pucallpa settlers ha ve more limited land and forest resources, are closer to urban markets (the center-south region of Brazil and Lima, respective! y), and appear to be making a transition from more exploitative land use to more diversified economies. Cheese factories in Theobroma and nascent agroindustries in Pucallpa are associated with such diversification in these areas. 3. Land speculation has been a farrn-level factor leading to more land clearing and conversion of forest lands to other uses. TR-16 Targeting Research In each colony settlers perceived that the values of their land assets had substantially increased because of their efforts in establishing more pasture, clearing land, building fences. and planting perennial crops . Related to the first two hypotheses. Pedro Peixoto settlers saw greater retums from pasture and fence development, while Theobrorna and Pucallpa farmers also placed value on their perennial crops. 4. Perennial crop establishment promoted by development projects risk failure where market projections are overly optimistic. In spite of an emerging wealth of publications dealing with potencial new market opportunities for exotic Amazonian fruit and forest products (Toledo,l994; Clay and Clement, 1993), citrus and achiote (Bixa oreilana) in Pucallpa and achiote Pedro Peixoto and Theobrorna were comrnercial failures (although farmers súll harvest and sell citrus at low prices in the city of Pucallpa). More recen ti y, oil palm and carnu camu (Myrciaria dubia) have been prometed in Pucallpa. Theobrorna farmers continue planúng coffee and cacao, and are now faced with the promotion of acerola (Malpighia punicifo/ia) and cupuacu (Theobroma grandiflorum). Whether or not camu camu, oil palm, acerola. and the forest produces palmito and uña de gato ( Uncaria tomentosa) will represent marketing opportunities or disasters for small farmers remains ro be seen. Two additional hypotheses relate to Pucallpa S. Cattle numbers in Pucallpa and the number of settlers maintaining cattle were negatively affected by the period of terrorism associated with the Sendero Luminoso (Shining Path) guerrillas. Respondents in Pucallpa indicated that they had decreased their cattle holdings and had failed to maintain fences during che period of terrorism which ended during the present (Fujimori) govemment administration. The relatively greater proportion of lands fallowed and the relatively fewer respondents with cattle may reflect such a trend (Figure 1). 6. "Altematives" to coca production were not economically attractive when coca demand was high; and that high demand was facilitated by the presence of the guerrillas and absence of govemment control. Informal interviews with small coca leaf farmers in Pucallpa indicated that: a) coca supplied relatively high cash incomes from both sales of leaves by producers and wages for weeding and harvesting; b) demand was high during the Sendero period when the guerrilla protected coca comrnerce; but e) demand dropped when the central govemment regained control of the area and has actively worked to eradicate coca. Conclusions Slash-and-bum agriculture in forest lands has been followed by pasture establishment and cattle, fallows and secondary forest re-growth, and perennial cropping, as well as by pasture degradation, soil degradation, and loss ofbiodiversity (ASB, 1994b; Borit, 1997). The promotion ofmore diverse and potentially sustainable land uses such as agroforescry or agrosilvopastoral systems should consider and, to the extent possible, build on farmers' existing experiences with annual and perennial crops, pastures and livestock, and forest products. On the one hand, there appears to be little current potential for the promotion of more intensive, perrnanent, and diversified (and sustainable) land use in areas such as Pedro Peixoto where abundant land and forest resources combined with isolation from urban markets tend to support the current exploitative pattem of resources use. TR-l7 Targeting Research On the other, decreasing fann sizes (dueto inter-generationalland fragmentation?) combined with access to new market opporrunities appear to be associated with intensification and diversification. The establishment of perennials and the settlers' associated perceptions of the high value in such establishment in Theobroma (where coffee, cacao, and dual purpose cattle have been successful) and Pucallpa (where banana, oil palm, and exploitation of bolaina ha ve been successful) should be seen as locally developed precursors to more sustainable agroforestry schemes. The reported high proportion of improved pastures at the three sites would indicate that the settlers already value and ha ve a tradition of managing such a technical innovation, and that improved agrisilvopastoral systems may be a "logical" next step. Of course, caution is needed in the promotion of new cultivars and products. Many fanners in the region with healthy, productive orchards of citrus or achiote have been more than disappointed by the lack of promised markets. Finally, historically idiosyncratic local development paths must be considered: improved forrage systems are presently of little use to Pucallpa ranchers who ha ve an overabundance of pasture resources as they now rebuild herds reduced during the Sendero period. The development of new Amazonian products was ata mínimum during the same period when coca production was the most attractive altemative for the areas' small fanners. Once herd numbers are rebuilt, however, the Pucallpa area may see an increase in pasture and a decrease in fallow area to proportions of land use more similar to that encountered in Theobroma. If the proposed highway connection between Acre and Peru-and to Lima--is cornpleted, land use in Pedro Peixoto may become more exploitative (eg more deforestation for soybean production) or may become more diversified (and hopefully sustainable) with increased access to new markets . References ASB ( 1994a) Report of research si te selection in Acre and Rondonia states of the Amazon region, Brazil. International Centre for Research on Agroforestry (ICRAF) and the project Altematives to Slash- and-Bum (ASB) ASB (1994b) Alternatives to slash-and-burn: a global strategy. International Centre for Research on Agroforestry (ICRAF) and the project Alternatives to Slash-and-Bum (ASB) Borit HP (1997) Situacion y perspectivas de la seguridad alimentaria en la Amazonia en un marco de produccion agropecuaria y de cooperacion intra-regional, estudio de Peru. Working document, Proyecto GCPfRLA/128/NET, Tratado de Cooperacion Amazonica-Food and Agriculture Organization of the Uníted Nations Brady NC (1996) Altematives to slash-and-burn: a global imperative. Agriculture, Ecosystems and Environmenr 58:3-11 Clay JW and Clement CR (1993) Selected species and strategies to enhance income generation from Amazonian forests. FO:Misc/93/6 Working Paper: Food and Agriculture Organization of the United Nations Fujisaka S, Bell W , Thomas N, Hurtado L, and Crawford E (1996) Slash-and-bum agriculture, conversion to pasture, and deforestation in two Brazilian Amazon colonies. Agriculture, Ecosystems and Environment 59: 115-130 Fujisaka S, Escobar G, and Veneklaas E (1997) Plant community diversity relative to human land uses in an Amazon forest colony. Forthcoming in Biodiversity and Conservation Harwood RR ( 1996) Development pathways towards sustainable systems following slash-and-burn . Agriculture, Ecosystems and Environment 58:75-86 TR-18 Targeting Research Loker, WM ( 1993) Medio ambiente y agricultura en la Amazonia Peruana: un experimento metodologico. In: Loker WM and Vosti S (eds) Desarrollo RuraL en la Amazonia Peruana. Centro Internacional de Agricultura Tropical (CIA 'D and the Intemational Food Policy Research Institute (IFPRI), Cali, Colombia Riesco A, Meini M. de la Torre M, Huaman H, Reyes C, and Garcia M (1982) Analisis exploratorio de los sistemas de fundo de pequenos productores an la Amazonia, region de Pucallpa. NITA-CIID. Lima, Peru Riesco A and Arroyo M (1997) Perfil socio-econornico de la region de Ucayali. Unpublished report, Consorcio para el Desarrollo Sostenible de Ucayali/Proyecto de Desarrollo de Agroempresas Rurales del CIA T . Sanchez, PA (1995) Science ín agroforestry. Agroforesrry Systems 30:5-55 Toledo, JM, ed (1994) Biodiversidad y DesarroLlo Sostenible de la Amazonia en una Economía de Mercado. Stansa, SA, Lima. Peru TR- 19 Targeting Research Latin America - Hillsides Characterization of si te for livestock research in Costa Rica - Tropileche ( 1.1.2, S. Fujisaka, F. Holmann, G. Escobar (CIA T); L. Badilla, Nils Solorzano, Leticia Badilla, Luis Umaña, Vidal Acuña and Marco Lobo (MAG). The objective of this activity was to characterize the existing animal production systerns in the Tropileche benchmark si tes in the Central Pacific region of Costa Rica. with emphasis on the characterization of existing resources, technologies, constrainsts, and opportunities. Main achievements • 44 livestock producers were interviewed • There is demand for new and improved forage alternatives • Research should be conducted with farmer participation Dual purpose cattle in the pacific coast region of costa rica: land use systems & demand for forage alternatives Abstract. Farrner-ranchers of the Pacific Coast Region of Costa Rica were interviewed. Holdings were small (4-20 ha), medium (21-60 ha), large (61-100 ha), and "hacienda" (above 100 ha). About half ofthose interviewed had dual (milk and meat) purpose operations. Other production systerns were mixed crop and livestock, cattle fattening, calf production and sale, and dairy. Respondents reported few problerns with their pastures and livestock. A majar farmer concern was providing sufficient feed in the pronounced dry season. Strategies included renting pastures and providing anirnals with cane and chlcken manure supplements. That nearly two-thirds had established sorne improved pasture (largely Cynodon nlefluensis or Digitaria decumbens) anda similar proportion had more recently sown Brachiaria spp. would argue for continued work on improved pasture systerns. Problerns associated with Brachiaria included high seed costs, low germination, and poor and costly establishment. Experience with forage legumes and trees was negligible. Further adaptive and farmer-participatory research is needed. A small team of researchers from Costa Rica's Ministerio de Agricultura y Ganadería (MAG) and the Centro Internacional de Agricultura Tropical (CIA T) conducted a sondeo to better understand resource use and production systerns in the Pacific Coast region-where livestock systerns domínate land use (Table 1)-and to Table l. Land use, Pacific Coast Region of Costa Rica, 1996* Land use Area (ha) % oftotal Livestock 188,533 85 Oil palm 11 ,130 5 Rice 6,500 3 Sugar cane 5,435 2 Mango 4,474 2 Others** 5 ,040 3 * Source: MAG, unpublished ** Coffee, beans, papaya. melon, watermelon, maize, tizquizque, vanilla, achiote, tomate, chili TR-20 Targeting Research analyze findings in tenns of present and furure demand (and constraints) for pasrurelforage innovations. The work represented an effort by and for TROPU.ECHE and the (CG) system-wide livestock initiative. Results are intended to assist in research planning in the region. Methods The sondeo was conducted 14-22 October 1996 by three to five researchers from MAG and three from CIAT. The group kept a set of guide questions in mind in arder to converse with farmer-ranchers in the regions surrounding Orotina. San Mateo, Esparza. and Miramar. A formal or written questionnaire was not used. The team spent fi ve days visiting fanns and fields, and interviewing 44 opportunistically sampled respondents. Interviewers worked singly or in pairs. Interviews were conducted in the momings and group discussions to review fmdings were held each aftemoon. A fmal day was spent compi1ing data, discussing implications and conclusions, and brainstorming about furure project activities. Find.ings Land use. Size of holdings ranged from four to 450 ha. We classified farms as small (4-20 ha), medium (21-60 ha),large (61-100 ha), and "hacienda" (100+ ha). Forty-one percent ofholdings were small, 34% were mediurn, 16% were large, and 9% represented haciendas (Table 2). Larger proportions of small (72%) and medium (53%) compared to the larger land holders (43-50%) produced annual crops such as beans and maize. Mean areas of annual crops for the small and medium holders having annual crops were 1.4-1.9 ha. At least half of those interviewed raised perennial (largely mango, citrus, cashew) or tree (ie Tectona grandis, Gmelina arborea) crops with the small and medium holders having means of 2.9 and 3.5 ha respectively. The proportion of respondents and areas held in forest or forest fallow increased from 22% of small holders with a mean 1.8 hato 75% of the hacenderos having a mean 64 ha (Tab1e 2). Mean areas of pasture were 8 (srnall), 35 (medium), 54 (large), and 171 (hacienda) ha. Most pasrure was native jaragua (largely Hyparrhenía rufa): 78% of small holders and all of the medium to hacienda holders had native pasrure, with mean areas ranging from 7 to 159 ha. A mínimum of two-thirds of respondents in each farm-size category had improved pasrures of largely Cynodon nlemjluensis or Digitaria decumbens (and of smaller areas of more recently estab1ished Brachiaria spp). As a proportion of pasrure holdings, smal1 holders had 45% of total pasture lands in improved species, medium had 15%, large had 23% and haciendas had 7% (Tab1e 2). D decumbens was produced by sorne for baling and sale. Other improved grasses in the area were estrella mejorada (Digitaria swazilandensis), jaragua mejorada (Andropogon gayanus), king grass (Pennisetum purpureum), and pasto Honduras (lxophorus unisetus). Livestock. Small holders hada mean 13 head of cattle, medium had 50, large had 66, and haciendas had 131. These included means of 11, 34, 37, and 70 adult animals respectively, such that adults per ha of pasture decreased as farm size increased-from 1.6 on small holdings to 0.5 on the haciendas. Mean numbers of cows giving milk ranged from 5 (small holders) to 14 (haciendas), with production large1y in the 6-7 bottleslhead/day range. Operations having 10 or more milk cows accounted for 12% of small, 47% of medium, 71% of large, and 50% of hacienda holders. Forty-four percent of small ho1ders, 53% of mediurn, 71% of large, and 50% of hacienda could be classified as dual-purpose (milk and meat) producers. The others included cattle fattening operations, calf producers, and (two cases) dairies. Many of the smal1-holders (50%) but few of the larger operations (0-14%) relied on non-livestock and off-farro activities such as fruit (eg mango) and crop (eg rice, maize. beans, coffee) production or carpentry or hired labor (Table 2). TR-21 Targeting Research Table 2. Land use, interviewed farmer-ranchers, Pacific Coast Region, Costa Rica, 1996 (n=44) Small Medium Large Hacienda 4-20 ha 21-60 ha 61- 100 ha lOO+ ha (n=l8) (n= 15) (n=7) (n=4) Mean farm size 11 41 76 235 % w/annual crops 72 53 43 50 Mean area (ha)* 1.4 1.3 1.5 13.0 % w/perennial crops** 56 53 43 75 Mean area 2.9 3.5 1.4 13.3 % w/forest/fallow 22 53 71 75 Mean area 1.8 6.3 20.0 63.7 % w/sugar cane 39 40 43 50 Mean area 0.3 0.3 0.9 0.4 Total pasture area 8 35 54 171 % w/improved pasture 72 67 86 100 Mean area 3.6 5.4 12.2 11.4 % tot pasture improved 45 15 23 7 % w/native pasture 78 lOO 100 100 Mean area 7 32 46 159 Mean total head cattle 13 50 66 131 % w/milk cows 85 80 100 100 Mean milk cows 5 11 12 14 Mean bottles/day/cows 6.1 7.7 7.0 4.4 % tot w/10+ milk cows 12 47 71 50 Mean head adults 11 34 37 70 Mean adultslha pasture 1.6 1.2 0.8 0.5 % dual purpose 44 53 71 50 % w/non livestock income 50 7 14 o * Mean area for those respondents having the given item ** Includes afforestation w/commercial timber species Pasture management. All farmers reponed manual weeding of their pastures (by machete); and 79% applied herbicides (2, 4-D, Tordon, Combo) as needed. Thirty percent reponed applying inorganic fenilizers, largely to their Brachiaria spp plots at establishment. A third (32%) of all respondents distributed evenly across classes rented pasture lands for pan of the year-either in the rain y season to allow pastures to recover for the dry season or during the dry season when owned pasture resources were depleted (Table 3). Worst weeds appeared to be Sida sp (locally Escobilla), what was tentatively identified as lschaemum indica (locally Ratana), Mimosa pudica (Dormilon), Paspalum virgatum (Zacate Burro), Rottboellia sp (Zacate de Fuego), Cyperus sp (Piepaloma), and the unidentified Maten and Espino. Other weeds mentioned (but not TR-22 Table 3. Pasrure management: use of herbicides, fenilizers and rented land * Herbicides Fenilizers Rented land % all respondents 79 30 32 * Only one reponed burning; the group thought that more used fire Targeting Research identified) were Casa Gente, Chiguiza. Escobilla Morada, Quiebra Plato, Aorecilla. Churrisata, Oreja de Chancho, Paira, Horguctillo, Muneco, Una de gato, Chapemo, Chan, Bledo, Canilla de Venado, Guizarro, Quebracha. Santa Maria, Achotillo, Vainilla, Ajillo, and Papa Miel. Calapogonium mucunoides (locally Taranta) was considered both a weed and pasture component by many respondents who recognized the benefits of the legume while controlling its domination over other pasture species. Animal feed supplements. Considering all respondents, almost all provided animals with salt and mineral supplements ; 72% provided molasses (largely in the dry season; and 44% provided chicken manure. Other purchased supplements included soybean husks, concentrates, and bone mea!. Thirty-five percent provided sugar cane and 9% gave crop residues produced on-farm (Table 4). Table 4. Use of feed supplements Purchased Salt Minerals Molasses Chicken manure Soybean husks Concentrares Bone meal Palote On-farm produced Sugar cane Crop residues % of respondents 93 84 72 44 16 12 5 5 35 9 Reported problems. Fanners reponed relatively few problems associated with pastures, livestock, or life in general: 60% had no livesrock problems; 70% had no pasture problems; and 70% had no general problems. The few livestock related problems included vampire bats (23%) and lack ofwater (9%). Pasrure problems included weeds (19%, see above) and pests affecting C nlemfluensis. General problems were high input costs (19%) and low output prices (12%, Table 5). Livestock problems-which farmers reponed having solutions-included ticks and fleas, carbon sintornatico, septicima, and mastitis. Exposure to pasture and forrage alternatives or innovations. As mentioned, 62% of the respondents had experience sowing improved pastures other than Brachiaria spp; and the same proportion had ex.perience (albeit more recent and on small areas) with Brachiaria spp. Few had experience with forage trees or hala, TR-23 Targeting Research Table 5. Reported problerrs Problem % respondents Cattle Non e 60 V ampire bats 23 Lack of water 9 Snakes 5 Diseases 5 Pasture None 70 Weeds 19 Pests (in C nlemjluensis) 9 Winds/drying 5 General None 70 High input costs 19 Low output prices 12 Lack of roads 5 Lack of lands 5 Tax.es 5 legumes (Table 6). Farmers commonly recognized that cattle ate the fruit of the tree Guazima ulmifolia. Projects in the area have worked to introduce Gliricidia sepium, Cratylia argentea, Leucaena luecocephala, Table 6. Use of forage technical altematives Altematives % respondents Improved pasture * 62 Brachiaria spp pasture 62 Legume focages 5 Forage trees 5 * Improved pastures other than Brachiaria; legume forages and forage trees were recently established on two farrns each by projects Erithryna spp, and /vlontS sp as multi-purpose species. The use of live fences was universal in the area. A few farmers had recently established Arachis pintoi as a part of other projects and mainly to control weeds and soil erosion under coffee. Based on the interviews and on even more informal follow-up discussions, the following factors were identified as constraints to the adoption of improved pastures: a) the high price (and unavailability) of seed, b) low germination rates of the seed, TR-24 Targeting Research e) difficulties in establishment. d) the need to prepare land/cost of mechanization, e) availability of land to rentas an altemative, f) the slow process of seed releases by the govemment. and g) sorne lack of ex.posure to innovations. On the other hand, other factors were identified as supporting the adoption of pasture and forage innovations: a) lack of feed resources in the dry season, b) possibly increasing prices for supplements such as chicken manure, e) ex.isting substantial adoption of improved pastures, d) practice of establishing improved pasture following intensively rnanaged high-valued annual crops such as watermelon, e) the intensive rnanagement systems of the small-holders, and f) the assumed higher yields associated with the altematives. Conclusions For this draft, the group concluded that: l. There is, especial\ y with the small-holders, experience with and dernand for improved pastures. 2. There would be high dernand for innovations which would increase feed resources in the dry season. 3. There was negligible experience with forage legumes. 4. Almost all of those interviewed were familiar with the constraints to the adoption of improved pastures. 5. We need to know the results and experiences of the various other projects (eg ODA) which have worked in the area on reforestation, improved pastures, multi-purpose trees, soil erosion control, and green manures. 6. Research on the limiting factors mentioned above has already been conducted. 7. Further adaptive research is desirable on fodder and animal yields comparing existing and improved systems. 8. Participatory adaptive research on rnanagement of new altematives and technology transfer (eg farmer-to-farmer) is needed. 9. Among other constraints, researchers now ha ve overly large areas of responsibility (both in terms of geographical areas covered and in fields of expected expertise); and that local MAG technicians lack needed capacity. The main constraint facing livestock operations was the lack of dry season feed resources. Farmers rented lands in the wet season so that pastures would be healthy for the dry or rented in the dry season once owned pastures were depleted. Farmers also purchased supplements such as chicken manure, concentrares, and molasses for their animals, but were worried about high and increasing costs of such inputs. Research needs to address the issue of dry season feed scarcity; and resulting innovations would be of high local dernand. TR-25 Ex-ante analysis of improved forage alternatives (1.3; F. Holmann) Main achievements Targeting Research • Demonstrated 35% of hillsides in dual-purpose fanns could be protected for reforestation if forage production was intensified on lower slopes and still maintain same net income. • New and improved forage alternatives are more profitable than current feeding systems but require matching genetic potential of cows. The objective of this activity was to evaluate altemative feeding systems in dual purpose animal production systems in che benchmark sites by means of: (a) bio-economic optimization models to simulare different management and macroeconomic scenarios, and (b) identification of altematives with potencial for improving existing systems . During the frrst half of 1997 visits were made to benchmark sites in Perú, Costa Rica, and Nicaragua to collect che infonnation needed to do the ex-ame analysis of collaborating farms in arder to determine their current situation in terms of management, use of resources, animal productivity and profitability . This inforrnation will be used to analyse altemative scenarios in terms of animal production, capital investment, and profitability from improved altemative feeding systems being implemented in the benchmark sites. It is expected that che ex-ante analysis will be completed by January 31, 1998. The main conlcusions from a preliminary analysis for Costa Rica are the following: A linear prograrnming model developed to analyze agricultura! production systems at the farm leve! was used. Under average productivities and input-output prices, the most profitable option is to grow sugarcane. However, this activity requires a topography which allows mechanization and soils with good fertility . The second and third bese options are growing cashew and producing mil k under specialized dairy systems. The dual-purpose activity was a viable option only when the price of milk was between US$0.19/kg to $0.13/kg or when che price relationship between beef and milk ranged between 4.5 and 8. Beef production was a viable option only in situations where the price relationship between beef and milk was greater than 8.5: 1, which has never existed in Costa Rica. However, it is an attractive activity in places where there is no infraestructure (ie., roads, electricity) and where the opportunity cost of labor is greater outside the farm since this activity requires little labor which in many cases, can be perforrned by children or adults excluded from the job market. With respect to the forage altematives evaluated, the marginal profitability of improved grasses and legumes is a function of labor wage rate, output price, and productivity. Thus, in dual purpose systems, the legume Cratylia argentea is 47% more profitable than the most comrnon forage altemative (ie., jaragua) when the milk price is US$0.30/kg. However, when the price of rnilk is US$0.20/kg, the marginal profit of Cratylia es practically ni! due to the fact that it utilizes 54% more labor than the Brachiaria+Arachis mix and 233% more labor than Jaragua. Likewise, in dual purpose systems with a productivity of 1,000 kg milk/1 and labor wage rate valued at US$10/day, the Arachis+Brachiaria mixture and Cratylia generates 45% higher net income than the altemative with jaragua alone. However, as labor wage rate increases >$13/day, the altemative with jaragua alone starts to be the best option under these low productivities. If milk productivity increases from l ,000 kg milk/1 to 1,500 kg/1, the improved grasses and Iegumes are more profitable than jara gua under all labor wage rates evaluated in this study. TR-26 Targeting Research In addition, the trade-offs of intensification were estimated for the three animal production systems (ie., specialized dairy, dual purpose, and specialized beef, Tables 1 thru 3). Given current animal productivities, íntensification liberares signifficant areas now under livestock use (ie., from 9% for specialized beef up to 35% for dual purpose syseems). However. intensification increases the production coses in all cases (from 3% for specialized beef to 8% for dual purpose systems). Thus, to obtain the same net income per farm without increasing production coses per unit of product, the new and improved forage alternatives evaluated in this study need not only to increase stocking rate, but also to increase productivity per cow by at least 10%. Table l . Trade-offs of the intensification process in specialized dairy systems to obtain the same annual farm net income with different forage alternatives at two milk productivities. Jaragua B+A B +A Change1 @ 2,000 kg/1 & Cratylia & Cratylia (%) @ ~.ooo kg/1 @ 2.200 kg/1 Area (ha) * Pastures 30 21.7 16.1 -28 (-46) * Conservation o 8.3 13.9 +28 (+46) Net Income * Farrn!yr 3613 3687 3,614 o (0) * ha/yr 120 170 224 +42 (+87) Productivity (kglha) * Milk 971 1,648 1,842 +70 (+90) * Beef 105 178 178 +69 (+69) * Stocking Rate (AU/ha) 0.96 1.63 1.63 +70 (+70) Cost o( Production * Mi1k ($/kg) 0.187 0.203 0.186 +8 (0) * Beef ($/kg) 0.59 0.63 0.58 +7 (0) Use of Resources * Labor (# days/yr) 261 358 266 +37 (+2) * Cows (#) 24.5 30 22.3 +22 (-9) * Capital ($) 14,974 21.662 15,228 +45 (+2) 1 Values in parenthesis show the porcentual change for the productivity level of 2,200 kg/1 with respect to Jaragua. TR-27 Targeting Research Table 2. Trade-offs of the intensification process in dual purpose systerns to obtain the same annual farm net income with different forage alternatives at two milk productivities. Jaragua B+A B+A Change1 @ 1,200 kg/1 & Crarylia & Craryiia (%) @ 1,200 kgll @ 1,300 kgll Area (ha) * Pasrures 30 19.4 16.5 -35 ( -45) * Conservation o 10.6 13.5 +35 (+45) Net lncome * Farm/yr 2.693 2687 2694 o (0) * ha/yr 90 138 163 +53 (+81 ) Productivity (kglha) * Milk 487 812 899 +67 (+85) * Beef 140 233 233 +66 (+66) * Stocking Rate (AU/ha) 0.91 1.52 1.52 +67 (+67) Cost of Production * Milk ($/kg) 0.197 0.203 0.193 +3 (-2) * Beef ($/kg) 0.83 0.86 0 .82 +4 (-1) Use of Resources * tabor (# days/yr) 206 260 221 +26 (+7) * Cows (#) 23.2 25 21.2 +8 ( -9) * Capital ($) 13,446 17574 14487 +31 (+8) 1 Values in parenthesis show the porcentual change for the productivity level of 1,300 kg milk/1 with respect to Jara gua. TR-28 Targeting Research Table 3. Trade-offs of the intensification process in specialized beef production systems to obtain the same annual fann net income with different forage alternatives at two beef productivities. Jaragua B+A B+A Change1 @ 200 & Crarylia@ & Crarylia@ (%) kg/calf 200 kg/calf 230 kg/calf Area (ha) • Pastures 30 27.3 15.9 -9 (-47) • Conservation o 2.7 14.1 +9 (+47) Net Income • Farrnlyr 1292 1,292 1,294 o (0) • halyr 43 47 8 1 +9 (+88) Productivity (kg/ha) *Milk NA NA NA NA • Beef !57 263 280 +67 (+78) • Stocking rate (AU/ha) 0.87 1.45 1.41 +67 (+62) Cost of Production • Mi1k ($/kg) NA NA NA NA • Beef ($/kg) 1.06 l.l5 1.07 +8 (0) Use of Resources *Labor(# days/yr) 134 258 149 +93 (+1 1) • Cows (#) 22 33.5 19 +52 (- 14) • Capital ($) 14660 26,691 15,187 +82 (+4) 1 Values in parenthesis show the porcentual change for the productivity level of 230 kg per weaned calf TR-29 Sites for production systems research in Central America (1.1.4; S. Fujisak.aand P.C. Kerridge) Main achievements • Hillsides sites and farmers in Nicaragua and Honduras were visited. • Contacts were made with project researchers, government officials, others. • A suggested research approach and funding possibilities are discussed. Targeting Research Visit. We visited the hillsides "reference" sites in Nicaragua (San Dionisia, sub-cuenca del rio Calico, Matagalpa) and Honduras (Y arito, Y oro) to explore the need for and possibilities of conducting fanner participatory research to improve agricultural and natural resource use systems. CIATs work in the two hillside sites in Nicaragua and Honduras have featured collaborative efforts strucrured largely around characterization of the watershed sites using GIS analysis, the organization of CIALs (with the Programa Campesino a Campesino in Nicaragua and IDRC-Guelph's Investigacion Panicipativa en Centro America in Honduras), and data gathering to develop indicators of well-being. Funding has come from SDC. IDRC, BID, and Holland. CIA T researchers working in Central Ame rica agreed that additional specific (agronornic. varietal, soils, resource management) research was needed in Central America; and that PES efforts rnight be an appropriate addition to the collaborative teams working in the region. Ronnie Vernooy in Nicaragua and Rector Barecto in Honduras each expressed che same view, ie that additional focused, farmer participatory agricultura! problem solving within the respective sites would be appropriate and desirable. We also discussed future collaborative research possibilities with MAG·Nicaragua. They were emphatic that research on specific, problem-solving agricultura! innovations was needed. The MAG thought that such research should be conducted in a (so far not identified) site in the dryer region around Esteli as well as San Dionisia. Although in agreement that farmer-participatory research was needed, they were less supportive of cornmunity organization efforts per se. Eduardo Marin Castillo, Asesor de Planificacion del Uso de la Tierra. was identified as CIAT's contact person within the MAG. Trip findings. Nicaragua. The local population of San Dionisia appears to be somewhat equally divided becween small dual-purpose cattle ranchers and maize-bean fanners. The ranchers make extensive use of live fences, but ha ve lictle in the way of improved animal feeding resources. Research similar to that conducted by Tropileche would be appropriate for the small ranchers. Farmers plant maize followed by beans on large, sloping fields. For the observed bean crop, fanners planted along the contour and and used herbicides and low levels of tillage at crop establishment and for weed control. Use of moderate amounts of inorganic fertilizer in the maize crop probably benefitted the beans. Farmers' bean fields looked vigorous, with good plant density and few signs of pest or disease problerns (other than an unidentified yellowing or requema in lower, possibly waterlogged areas). Various projects in the area--apparently reacting to the cropping on steep slopes--have worked on the inrroduction of contour hedgerows as a means of soil erosion control. Soilloss rates were not available. Although sorne shallow soils and eroded fields were in evidence, many cropped fields appeared to have relatively deep soils; most fields displayed few signs of serious soil erosion; and in most cases the standing bean crop appeared healthy. We would recommend against our involvement in soil erosion control research. Given an apparent opportunity and need to introduce and test improved bean and maize germplasm, Dr Vernooy is coordinating fanner participatory varietal testing through the CIALs. The three trials visíted were well designed and established. Farmers were enthusiastic; although they mentioned that interest had been insuffcient to involve more community members in conducting trails. Furure work could involve many more fanners as interest grows, however. TR-30 Targeting Research Greater panicpation was demonstrated by a women's CIAL working with vegetables . Nurseries had been established, but these had been placed in open fields far enough from water sources to make the necessary hand irrigation a task requiring substantial coordination among CIAL members. Rabbits, a local pest of beans, would probably become a problemas vegetables emerged. A suggestion was made to establish vegetable nurseries clase to farmers' houses- albeit protection from chickens would then be required. Overall, wider farmer panicipatory testing of improved forages (legumes including trees and grasses) and of annual crops (beans, maize) would appear appropriate to the area. Honduras . Similar toSan Dionisia, Y arito fanners grew maize (May-Sep) followed by beans (Oct-Dec). The area spans from 500 to 1400 masl and includes steep slopes, small areas of river-bottom lands, and wider, flat valley floors. Pine forests on shallow soils are interspersed with agricultura! areas at higher elevations. Most farmers had few animals-at most a pair of oxen or a mule or horse for draft and transport. There are, however, substantial numbers of cattle ranchers and cattle in areas adjacent to the research area. As in Nicaragua, although there was ample evidence of projects working on soil erosion control, we would make a preliminary guess that such work is largely unnecessary (or not serious enough as a problem that farmers would more widely invest in such innovations) given low tillage levels on sloping fields, low total rainfal l, substantial vegetative cover other than at crop establishment. CIALs in the area were testing new bean germplasm and integration of legume cover crops (mucuna) in the maize-bean system. TI~ Program Regional de Maíz was testing maize cultivars on farmers' fields. CIAT has been testing the suitability of different mucuna and canavalia germplasm to different altitudes (given differences in flowering and flowering times by altítude). In coordination with these efforts, we would propase: a) further rapid diagnosis focusing on land use histories of cultivated parcels, systems productivity, crop management practices, crop problems, farmers' problem solving practices, and related farmers' knowledge; and b) farmer participatory testing of a limited number of new maize, bean, cover crop legume, and forage cultivars. Research would be conducted at three to five sites along a transect including lower, middle, and higher altitude areas. Larger numbers of farmers interested in testing new cultivars and systems would be sought. Simple trials of sound experimental design would be necessary. Work could begin in both Nicaragua and Ho1_1duras at the start of the wet season in May-June, 1998. Farmers in Nicaragua and Honduras make extensive use of live fences- largely but not limited to Gliricidia sepium. We are looking for ways to introduce more use of live fences in Pucallpa. Peru. Investigation of the extension programs-if any--needed to get such widespread adoption would be useful. Conclusions There appears to be a need for additional work by the systems project in Central America. Collaboration and coordination will be essential to the success of CIA Ts work in Central America. The ideas and suggestions presented here can be díscussed by CIAT researchers (PE2, PE4. PE5, SNl, SN3, IP I, IP5 , PROFRIJOL, TROPILECHE) working in Central America during the week of the ICR. Norway has indicated that they wil\ discuss details of a propasa! MAG-Nicaragua submitted to Norway on CIAT's behalf. DICTA (Honduras) is interested in developing ajoint project with CIAT in the area of production system research. TR-31 Targeting Research Improved predictability and prevention of erosion ( 1.1.6; Karl Mueller Sarnann, Jesús A. Castillo, Kai Sonder. Ana María Patiño: Collaborators: Edgard Amézquita, Raúl Madriñán, Cesareo Gallego, Luis E. Mina, José Luis Adarve, Disnardo Peña, Universidad de Hohenheim, Gennany; National University of Colombia, Palrnira.) University of Hohenheim- BMZ Project- Soil Degradation and Crop Productivity Research for Conservation Technology Development in Andean Hillsides Farming) Main achievements • Over lOO rainfall events were recorded over the last year for analysis of intensity -energy relationships in the tropical Andes. • Soil parameters like turbimetry, the pattern of aggregate size distribution and hot water extractable carbohydrates (HWEC) could be identified as sensitive descriptors of changes in structural stability properties of tropical inceptisols. • Parameters like turbimetry, content of HWECs and aggregate size distribution were also closely related with differences in soil management. Report. Research on basic processes and the characterization of the physical parameters in volved in soil degradation and erosion lead to a better understanding of soilloss and form the basis for sound decisions and decision supporting models in the areas of applied research and in conservation politics. They thus provide the foundation for targeting the problems in the field. Since 1987 basic research related to soil and water management improvement is carried out on two sites in the Cauca department of southem Colombia. The Universal Soil Loss Equation (USLE), a widely used amount of data on climate, soils, soil dynamics, soil losses and cropping systems performance from over 40 runoff plots were systematically collected. In the last year basic research concentrated on three mayor areas: Characterization of rainfall and rainfall intensity 1 energy relationships on a tropical Andean site ( rainfall factor R). Research on soil inherent pararneters and on soil water balance related with observed management induced changes in soil's susceptibility to erosion (erodibility factor K). The qualitative and quantitative relationships between erosion induced soil degradation and crop productivity and the econornic and ecological appraisal of the costs of soil erosion. a) Drop-size energy load-rainfall intensity relationship Aggressiveness of rainfall, the agent inducing erosion processes at the soil surface varíes a lot with intensity. Higher rainfall intensities like those occurring in tropical rainstorms are related with bigger drop size, higher velocities of raindrops and hence higher kinetic energy. The relationships existing between rainfall intensity, drop-size distribution and the k.inetic energy hitting the soil surface are well documented for temperate climates and the mathematical descriptors developed in North America are used in models pertinent to water erosion. Nevertheless its accurateness has never been tested or verified under conditions of the tropical Andes. TR-32 Targering Research Complementary to the registering rain gauges used since 1987 to determine the amoum and intensity of rainfall, in 1993/94 an electronic raindrop sensor- computer system ( Distrometer) was used to describe rainr'all intensity 1 energy relationships at the Santander de Quilichao site. Preliminary results from 35 measurements done at that time suggested no mayor differences with regard to drop size distribution and imensity, between the tropical Andean site and temperare climates. However observations of only 35 rainfall events is too small a sarnple to draw conclusions and therefore additional mea.surements had to be carried out to examine this hypothesis and base R-factor calculations on a salid and sufficiently representative ser of data. New measurements of rainfall events were started in September of 1996 continued throughout 1997 with a high record density especially in the rainy season that lasted from March until June and during the second rainy period which started in October. During the dry season from June to September 1997 only one rainfall event was recorded and it didn't produce soil Ioss. In total lOO rainfall events were recorded from September '96 until October '97. Due to the climate during this time period most rainfalls were rather moderare ones with maximum intensities never exceeding 80 mm per hour. lt is hoped that the phenomenon "El Niño" will not interfere severely with the rainfall characteristics and - amounts in the remaining time available for this research activity. The raw data collected so far was imported into data sheets (Excel) to facilitare further evaluation, especially to calculare rainfall intensity, the erosivity factor R of the Universal Soil Loss Equation (USLE) and other rainfall characteristics like erosivity índices and to permit easier importation into the mega database on project results and of available figures accumulated in the project over the last ten years. b) Search for data on the R-factor and mapping of erosivity During the recent soil conservation workshop held in CIA T in October 1997, severa! contacts were made with visitors from different Andean countries on the acquisition of rainfall intensity data. The best possibilities are offered by IDEA.ivl, the Colombian institute that manages hydrological and climatic information bases and related rnatters and the coffee-growers federation FEDECAFE. Other possibilities but probably with Iess amount of data exist in Ecuador and Venezuela. Though data about rainfall intensity is rather scarce, as most rneteorological stations are not equipped with a pluviograph it is expected that it will be possible to compile a ser of stations for a better assessment of rainfall erosivity in the Andean region. After the present rainy season a data gathering trip is planned in early 1998 with priority given to data available in Colombia and if resources will allow it also in Venezuela and Ecuador. e) Validation of indicators for soil health with respect to soil erodibility Research on erosion plots over the Iast years clearly showed that erodibility, a soil' s susceptibility to be dispersed and carried away by the action of rainfall is a very dynamic soil characteristic. It tends to increase during prolonged soil exposure and in response to intensive tillage operations. Soil Ioss measurements carried out in previous years in different cropping and managernent sequences also could show, that mayor changes in soil erodibility can occur over short periods of time. These changes are not sufficiently detected and differentiated by the prediction model which takes into consideration soil organic matter (SOM), aggregation degree classification, soil texture and infiltrarían properties. TR-33 Targeting Research A Ph.D. research program was therefore started in 1996 to investigare the impact that tillage, cropping practices and rotation sequences have on soil aggregation, microbial activity and other, biologically influenced pararneters of strucrural stabiliry or soil health with respect to erodibiliry. The research effort airns at the identification of parameters that can be used as indicators for the monitoring and diagnosis of management impact on short- and long term changes in strucrural strength. The objective is to provide tools for the formulation of precise recommendations for land management in tropical hillsides, where soil erosion is the single most important factor of soil degradation. Reproducible indicators, with sorne of them being simple and accurate enough to be used by local scientists and extension services to monitor the efficiency of regenerative techniques, can help to make better use of the high potential to reduce degradation through cropping and management practices. At the same time they can issue an early warning signa! if indicators show that current land use is conductive to an enhancement of erodibility rather than stability. During the last year the influence of severallong term management patterns. applied to the runoff plots on two research sites were evaluated as to its influence on the changeable erodibility of tropical inceptisols. The following parameters were tested for its applicability as quantitative descriptors for erodibility changes: l. Determination of stable aggregates 2. Aggregate size distribution of stable soil aggregates 3. Quantification of the content of hot water extractable carbohydrates (HWCH) 4. Physical rupture and measurement of soil in suspension by turbimetry S. Microbial activity 6. Abundance and diversity of soil fungi 7. Presence and composition of soil meso fauna The numeric values of the abo ve mentioned descriptors which are a function of the contrasting rnanagement history of the plots will be related with other, well understood soil attributes measured in the laboratory. Moreover and- very rarely found- figures will also be related with values of real soil1oss and surface runoff under the influence of naturally occurring rainfall in the field and should lead to proveo indicators for soils resistance to erosion. The following six out of eight trearments were selected for evaluation in this experiments: l . bare fallow after pasture (grass removed in March 1996) - check - 2. crop rotation + chicken manure + tillage 3. crop rotation including cycles of sown grassland + fertilizer + minimum tillage 4. crop roration including weed faJlow + fertilizer + tillage 5. crop rotation always with a legume associated + fertilizer + tillage 6. crop rotation including cycles of sown grassland + ferti lizer + tillage In a first phase methods and parameters were tested and adjusted for further use as descriptors and preliminary results have been obtained that still require verification in the field during the 1997/ 1998 croppmg season. The stability and size distribution of aggregates at the soil surface, two parameters that govern soi1 erosion to a large extent in most soils, were highly influenced by contrasting management patterns. In figure 1 the distribution curves of stable aggregates of a " virgin" soil after a) more than 10 years of improved pasture TR-34 ., "' ! e: ., ~ ., a.. Targeting Research a ) b ) 100 90 80 70 60 50 <40 30 20 10 1 1 1 1 '1 / ~ 100 90 80 ., "' 70 ~ 60 e: ., 50 ~ ., <40 a.. 30 20 1 o li( V ¿ 1/ ...... r- o o 1 O .O 1 0 . 1 1 o o .o 1 o . 1 1 o Equ ivalen t d ia m e ter ( m m ) E q u ivalen! d iameter ( mm ) Figure 1 . Distribution of stable aggregate sizes at the soil surface on the Mondomo site in an a) "virgin" soil after more than 1 O years of impro\ ~d pastores and b) after 1 O years of an intensi ve crop rotation including legume components, tillage and mineral fertilization. and the distribution curve after b) lO years of cultivation and tillage, show marked differences in shape. In the soil after grassland about 50% of the aggregates at the surface have a díameter of more than 5.5 mm, whereas after 10 years of cultivation less than 15 % have a diameter of more than 5.5 mm and more than 50% have a diameter of less than 1.5 mm. The method of comparing the shape of the curve of the aggregate size distribution at the soil surface seems to ha ve potential for being used asan indicator to describe the direction of changes, away from or towards more sustainability. Intensive monitoring of characteristics of aggregation at the soil surface in function of previous management is therefore undertaken in uniformly tilled and standardized plots of all treatments and will be compared against the "virgin" soil check. Another parameter studied on the basis of different management patterns is the content of hot water extractable carbohydrates (HWEC). About lO% of the organic matter of a soil may occur as carbohydrates and the majority of them as polysaccharides. From the point of view of the stability of soil aggregates they constitute an important component , because they actas binding agent for soil particles. In previous studies reported last year and in the present studies only the fraction of HWECs is evaluated. Preliminary results from studies at the Mondomo si te are shown in figure 2. Highest levels of hot water extractable carbohydrates could be observed in the "virgin" soil after recent removal of the improved pasture followed by the soil from a crop rotation with minimum tillage, a conventional rotation with tillage and lowest levels after severa! years of abare fallow treatment. TR-35 rtting Research -Cl -- 1400 Cl =- e 1200 .2 l1l .... 1000 -e - 400 .e o J:l 200 ..... l1l ü o o 2 3 4 5 Treatments Figure 2. Concentration of hot water extractable carbohydrates (HWEC) equivalent to J.1 g glucose per g of soil in aggregates >0.25 mm in an oxic Hurnitropept of the Mondomo site. 1 = "virgin" soil after removal of grassland, 2 = crop rotation with minimum tillage 3 =convencional rotation with tillage 4 = bare fallow for more than 5 years. Similar results have been obtained with the method of monitoring structural stability by turbimetry (results not shown) but the method is still in the process of refinement for evaluating the soil type under investigation. Starting in October 1997 the response values of the selected variables will be related to values of real soil loss from tilled bare fallow plots from "virgin" plots after grassland (best bet reference value) and from plots with variable soil-crop management history. Correlations with other standard parameters of structural stability in the laboratory will also be established. The overall objective of this procedures is to prove their aptitude as indicators of changeable erodibility of a given soil. It is assumed, that as a result of this verification process a set of parameters can be selected for the diagnosis of a soil' s erodibility status. As already mentioned changes in the erodibility of a soil are often attributed to biological agents and their influence on the stabilization of macro aggregates or even the reconstitution of soil structure. However, in spite of their high potencial as indicators for soil health and their influence on structural soil dynarnics, the extent and the narure of biological reconstruction agents is stil! poorly understood. First results from research started in this area a few months ago suggest, that Penicillium, Trichoderma and Rhizopus exert a very positive effect on soil structure in the inceptisols evaluated but both the role of fungi and the role of soil mesofauna (e.g. rnites) is not yet clear and needs further evaluation. A M. se. thesis research project has therefore been started in collaboratíon with the National University of Colombia in Palrnira (Ana Maria Patiño). Samples will be taken from the same treatments of the erosion plots as those used in the Ph. D. thesis of Jesus A. Castillo and will allow to establish relationships TR-36 Targeting Research between management, soil physical properties and the presence and function of the biological agents under evaluation. d) Relation between indicators and soil-crop management As it is shown with the examples presented in figure l and figure 2 most of the pararneters used in this srudy not only showed high correlation values with soils susceptibility to erosion but could at the same be related to variation in rnanagement practices. The parameters therefore fulfill an important requisite for its use as indicators, because gains in strucrural sustainability can be related to soil-crop management practices and hence to recommendations for better land husbandry. References Haynes, R.J. and Francis G.S. 1994. Influence of crop species on microbial biomass and soil aggregation. Transactions 15th world congress of Soil Science. Vol. 4B. Acapulco, Mexico, July 1994 Molope, M.D., Page, E.R. and Grieve, I.C. 1985. A comparison of soil aggregate stability tests using soils with contrasting cultivation histories. Communications in Soil Science an Plant Analysis 16: 315- 322 Pajasok, T and Kay, B.D. 1990. Assessment of a combination of wet sieving and turbimetry to characterize the strucrural stability of moist aggregates. Canadian J. of Soil Science, 70: 33-42 Wischmeier, W.H. and Smith, D.D.l978. Predicting rainfall erosion losses -a guide to conservation planning. US. Dept. of Agriculture. Agriculrural Handbook No. 537. Washington D.C. TR-37 Asia- Forages for smallholders Targeting Research Forages for Smollholders FPR locations The forages for Smallholders Project (FSP) is a Southeast Asían regional project that began in 1995, funded by AusAID and managed by CIA T (Centro Internacional de Agricultura Tropical) and CSIRO Tropical Agriculture (the Commonwealth Scientific and Industrial Research Organisation). Its focus is to develop forage technologies in partnership with smallholder farmers in upland areas. where they are needed for livestock feeding or resourc ;! management (including erosion control, soil fertility improvement, weed control and reducing labour requirements). The FSP is a network of smallholders farrners, development workers and researchers. It is coordinated by national organisations in Indonesia, Laos, Malaysia, Philippines, Thailand, Vietnam and southem China. In 1997, the FSP worked at seventeen on-fann sites in Indonesia, Laos, Philippines and Vietnam (Figure 1). Additional sites will become operational in Laos and Vietnam in 1998. TR-38 Targeting Research Minimum data for site characterization in on-farm studies in S.E. Asia A visir was rnade to Asia by Sam Fujisaka to review the farmer participatory activities within the Forages for Smallholders and the Improved Sustaínability of Cassava-based Systems Projects. Particular attention was given to initial site characterization with respect to assessing future impact. We present both this report and framework that has been adopted by the FSP. CIA T' s farmer participatory research in asia: towards improving trials and assessing impacts ( 1.1.3; S. Fujisaka) Main achievements • Fanner participatory research sites in Asia were visited. • Recommendations are provided for improving FPR in soil conservation and use of forages and legumes for animal feeding systems and improved fallows. • Recommendations are provided for improving impact analysis. Abstract. The International Center for Tropical Agriculture collaborates with national agricultura! research systems in Southeast Asia on farmer participatory research to improve forage and cassava systems. Research includes work on a range of systems in which forages can play different problem-solving roles and on soil erosion control in cassava systems. Several sites were visited and farmers interviewed. Suggestions are provided for the possible improvement of the farmer particpatory research and the measurement of project and research impacts. Suggestions concerned improved targeting by working with individuals and community sub-groups in the diagnostic process, improved trial lay-out to facilitate data recovery and demonstration effects, on-farm monitoring of both both technical performance and farrner feedback, facilitating farmer discovery and innovation in participatory research, and impact and adoption assessment. Introduction Field sites where the Internacional Center for Tropical Agriculture (CIAT) collaborates with the national agricultura! research systems (NARS) of Thaíland, Laos, Indonesia, and the Philippines on farmer participatory research (FPR) were visited in February 1996 to consider the issues of site characterization, farmer-participatory technology evaluation, and impact assessment. Research to improve agricultural systems through use of improved forages is conducted at three sites in Xieng Khouang Province, Laos, at two sites in East Kalímantan, Indonesia, and on Leyte in the Philippines. These sites are among 30 located in Indonesia, Laos, the Philippines, Vietnam, Malaysia. Thaíland, and China where the "Forages for Smallholders Project" (FSP) is managed by CIAT and the Commonwealth Scientific and Industrial Research Organization (CSIRO), with funding from the Australian Agency for Intemational Development (AusAID). Additional research is being conducted at one site in Thailand and two in East Java, Indonesia, to improve cassava-based systems through soil erosion control, improved cassava germplasm, and soil nutrient management. These sites are among 13 located in Thailand, Indonesia, Vietnam, China, and the Philippines; and linked via a research network funded by the Nippon Foundation of Japan. TR-39 Targeting Research Researchers at most of the sites used participatory rural appraisal (PRA) techniques to characterize sites, diagnose problems. and prioritize participatory (and other) research. In terms of FPR implementation, sites ranged from initial sta.rt-up efforts to fairly well-established farmer trials. FPR trials included measurement of soil run-off under different erosion control alternatives, screening of "best bet" forages, and farmer-designed forage trials by "spontaneous" collaborators. Site visits For the FSP forage research sites, Peter Horne (CSIRO/CIAT), Sam Fujisaka, and Phonepaseuth Phengsavanh (FSP) visited severa! sites (Ban Ta, Ban Sang, and Ban Phousi) in Xieng Khouang Province in Laos; Horne, Fujisaka. Nathan Russell (CIAT) Maimunah Tuhulele (Directorate General of Livestock Services, Jakarta),and Ibrahim (East Kalimantan Livestock Service) visited Sepaku and Makroman in East Kalimantan; and Werner Stur (CIA T), Francisco Gabunada (CIAT), and Dolores Alcober (Farm and Resource Management Institure, FARMJ, at the Visayas Stare College of Agriculture, ViSCA) visited the Matalom site(s) in Leyte, Philippines. For the Nippon Foundatation supported network of cassava research sites, Reinhardt Howeler (CIAT) and Fujisaka visited Soeng Saang in Thailand; and Howeler, Fujisaka, Russell. and staff from the Research Institute for Grain Legumes and Tuber Crops and Brawijaya University, borh in Malang, East Java, visired the Blitar and Dampit sites in East Java. CIAT, FSP (in the case of the forage FPR sites), and NARS researchers visited the sites, farmers' fields, and research trials--both FPR and researcher-managed. Farmers were informally interviewed at each site. NARS staff (and staff of a Lao-German project in Laos) and local farmer leaders provided additional information at each site. Impressions gained and suggestions offered reflect short field visits: inaccuracies and misperceptions are hopefully not extensive. Thesites Forage research sites. Forages research in Laos, East Kalimantan, and the Philippines consists of small farmer testing of improved grasses and legumes--including leguminous trees--for livestock feed, improved soil nutrient cycling, fallow improvement, and conversion of areas dominated by Imperara cylindrica. FPR on forages is conducted in collaboration with the Lao-German (GTZ) Nam Ngum Watershed Management and Conservation Project (NA W ACOP). NA W ACOP works at several sites-including Bao Ta, Ban Sang, and Bao Phousi-in Xieng Khouang Province. Farmers at Ban Ta managed smalllowland rice paddies and larger fenced fields for shifting cultivation of upland rice, opium poppy (Papaver somniferum), and other upland crops. Cattle were pastured in common access grazing lands located sorne 5 km from the village. Farmers were starting to test forages, with objectives of improved and more accessible animal feed , fallow improvemem and improved soil nutrient recycling and weed control. Trees for use as live fences will be tested in the future. Farmers were testing forages for livestock feed at Bao Sang on the Plain of Jars. Farmers grazed animals on local grasses (Themeda sp, others). Extremely low available phosphorous (3-5 ppm) was a factor leading to severe animal health problems and to difficulties in producing forages without fertilizer P applications (Gibson 1996). One interviewed farmer lost a cow in the past year. He described symptoms of P deficiency, but had treated the animal with antibiotics which cost more than the triplesuperphosphate (TSP) needed to have cured the animal. TR-40 TargeringResearch Farrners in Ban Phousi had sufficient rainfed lowland rice (and opiurn poppy) that shifting cultivation on their sloping lands was reportedly declining. Forages for carabao (water buffalo) strengthening prior to land preparation, for pregnant and lactating females, and for cattle fattening may be among needed and desired innovations. Carabao in the area appeared healthy and fairly well nourished. FPR on forages is being conducted in East Kalirnantan as a collaborative effort by FSP, CIAT, the Directorate General of Livestock Services, Jak.arta, and the East Kalimantan Livestock Service headquartered in Samarinda Sepaku farmers are transmigrants from Java. They have established small, relatively intensive tree- and crop gardens surrounding their homelots along the unpaved access roads. Most farmers had lowland rice paddies formed along natural watercourses and lower areas near the settlements. The surrounding uplands were dorninated by lmperata grasslands. Wild pigs-which are neither hunted nor consumed by the Muslim population-made food cropping in the grasslands virtually impossible. Cattle were a main system component, with animals now grazed (both tethered and herded) in the grasslands and supplemented by cut-and-carry feeding of the more palatable native grasses. Farmers cornplained that such grasses were now found only at greater distances, making cut-and-carry more time consuming. Cattle appeared relatively well nourished. Makroman farmers produced rnaize, cassava, and lowland rice. Although clearly not distributed evenly across farm families, Makroman's approximately 1100 families had sorne 1000 head of cattle, 1100 goats, 900 ha of lowland rice, and a total village area of 20,000 ha. The landscape included large lower areas were salinity was a problem and sorne upland areas dominated by Imperara. FSP works with two farmer groups: Maju farmers had both goats and cattle, while Sidodadi farmers raised only goats. The few cattle seen appeared to be well-managed and healthy. Use of oversown Centrosema pubescens and Stylosanthes guianensis as a green manure and live rnulch in rnaize and cassava was being tested by at least one farrner. Matalorn, on the island of Leyte in the Philippines, consists of a lower area of acidic soils (from pH 4.5) and an upper area of calcareous soils (up to pH 8.4). Farrners in the lower Barangay San Salvador grew paddy rice, maize, sweet patato, and sorne upland rice. Most had a few cattle, either owned or managed under a local sharing systern (alima) . It appeared that cattle were sold to finance~hildren's educations or to malee needed borne improvements, and that remittances from employed sons and daughters living outside of the community formed a substantial part of local incomes. Farmers complained of decreasing open- or common access grazing areas (where animals were tethered) dueto increasing numbers of animals. Overgrazing, pasture dominance by Axinopus compressus, and under-nourished animals were common throughout the area. Barring reductions in aggregate herd size as a response to declining forage resources ( or to other economic or policy factors), farmers clearly needed additional and better forage resources. Farrners did not supplement animaldiets for periods of peak draft demand or for pregnant or lactating females. Fanners in the upper area (Monte Alegre) of calcareous soils relied largely on cash incomes from production and preliminary processing of abaca (Musa sp) fiber. Sorne families mantained goats, which were botb tethered and penned and fed (using cut-and-carry) leaves of the local Leucaena leucocephala and Gliricidia sepium. Farrners complained about a relatively recent and growing problem of infestation by Chromolaena odorata. Farmers had an insignificant number of cattle. Not observed, the comrnunity reportedly hada substantial number (40) of carabao, apparently in the lower areas. Cassava research sites. Sites visited in Thailand and East Java feature FPR on soil erosion control and both FPR and researcher managed trials/demonstration plots on cassava varieties and soil nutient management. TR-41 Targeting Research Fanners from Ban Noen Sombuun (Soeng Saang District, Korat Province, Thailand) produced cassava for the commercial starch rnarket on gently undulating slopes. The relatively large fields with long slopes were subject to soil erosion, although not as severe as in other more steeply cultivated areas in Thailand (and throughout the region). Fanners reported increasing cassava yields due to adoption of new varieties (Rayong 1 followed by Rayong 5 and Kasetsart 50). Fanners did not seem to be very concemed about soil erosion. Fanners in Blitar, East Java, grew maize, soybean, upland rice, a variety of tree crops, and fast-growing trees for timber. The calcareous soils ranged from shallow with exposed sub-soil rocks to deeper soils in more level areas. Fanners had terraced much of their land, with construction reportedly following granting by the govemment of prívate land tenure only a few years ago. Fanners' terraces, live fences, and hedgerows appeared to control much of the potential in-field soil erosion in this area of sloping but intensive agriculture. Gully erosion by run-off between fields appeared to be a problem, however. Farmers in Dampit, East Java, produced fewer crops and more cassava; and had terraced their lands less than Blitar fanners. Besides terracing, farrners' contour ridging within cassava fields would appear to reduce soil erosion. Fanners were most interested in the cassava varietal testing, secondly in the fertilizer trials, and lastly in the soil erosion control trials. Farmer-participatory research Si te selection. Sites ideally should be selected as those in which given problems to be addressed exist, where problems are serious enough that farmers have sought so1urions to the problems, where potential technical innovations function in the manner intended, and where the problem-solving superiority of the innovation(s) can be clearly demonstrated. Forage research sites were selected to represent agroforestry, upland, plantation, grassland, and lowland farming systems (FSP 1995); and somewhat different forages and uses are being tested as appropriate to each site type. For example, Ban Ta represents fenced shifting cultivation plus extensive cattle grazing where improved fallows for improved soil nutrient cycling and eventual grazing are being tested. Nearby Ban Phousi represents a lowland system having both cattle and carabao feeding needs. Sepaku represents a grassland and cattle system in which increasingly scarce quality grazing may induce fanners to sow forages for cut-and-carry supplementation; while nearby Makroman represents a site with few large animals, but in which soil nutrient cycling and ground cover were among primary concems. Both FPR and researcher-managed trials in these different systems are to eventually be tailored to local systerns needs. Such si te differentiation and research targeting should contribute strongly to the success of the FPR efforts and will facilitate ex post impact assessment. Although the cassava research sites were not sirnilarly classified, the three sítes visíted varied from large field cassava monocropping in Thailand, to a system in which cassava is one of severa! important crops produced on small farrns (Blitar, Indonesia). A possible concern is that the sites where cassava is a majar cash crop did not feature the steep, more erosion-prone slopes found in, for example, northem Thailand and parts of Sumatra. The cassava farrners largely reported increasing yíelds and a lack of concem over their moderate rates of soil loss. Effective FPR on soil erosion control will likely be more difficult in areas where: a) fanners do not perceive productivity losses due to soil erosion and b) soil losses in the demonstration contour treatments may be only slightly less than fanners' current practices. As an arguable idea based on comparing sites around the tropics, perhaps a mínimum of about 80 tlha/year of soilloss is necessary (but not necessarily sufficient) for fanners to be sufficiently concemed about soil erosion such that they would be willing to invest in new soil management practices. TR-42 Targeting Research Site Characterization. Site characterization at most of the sites relied upon PRA tools and interactions with farmer groups. For the NA W ACOP sites in Laos, project personnel conducted participatory "situation analyses" which examined natural resources, local social-instirutional and economic structure, and site-specific factors. The project uses aerial photos for participatory land use planning. Results are available through the NA W ACOP project. PRA was conducted in the East Kalimantan sites as part of a training course in FPR conducted by FSP and the Directorate General of Livestock Services. Written results of these exercises-if they exist-were unavailable, although FSP-DGLS staff were very able to recall PRA results. PRA or "participatory diagnosis" (PD) was similarly conducted in Matalom by FSP and F ARMI-ViS CA as part of a similar training course. Written results are available (Gabunada et al 1991). PRAs were conducted by the cassava researchers at the FPR sites in Thailand and East Java. Written results were either not produced or were not made available. The cassava research, however, implicitly-and probably correctly--concluded that soil erosion was a problem at each site, that farmers needed and wanted new cassava varieties, and that N, P, and K trials were needed to assist farmers in their fertilizer-use decisions. Overall, site characterization (and, as will be discussed, impact analysis) could benefit from interviews with individual farmers. Benefits for characterization would include a better understanding of dífferent strategies encountered within sites--e .g. cattle fattening, calf production, single animal "bank accounts" or draft power in the case of livestock systems and forage needs . Diagnosis. Initial characterization needs to be diagnostic as well as descriptive. Farmers and researchers need to jointly identify and prioritize problems, and to understand interacting problem causes and effects. Farmers' knowledge, perceptions, and problem solving approaches need to be understood. The FSP researchers have done well in combining characterization, often a descriptive process, with adequate problem diagnosis, identification of current farmer problem solving practices, and identification and prioritization of needed further FPR (eg Gabunada et a/1991; Gibson 1995; NAWACOP, personal communication). Although the cassava researchers are clearly correct in their concems regarding soil erosion control, FPR may turn out to be more successful in germplasm screening and selection and soil nutrient management than in soil eros ion control due to the lack of farmer concem over the rates of soil being lost at the sites visited. Fanner participant selection. Farmer participants might best be selected among those who: a) are faced with and are aware of a given problem that the proposed FPR and associated innovations are meant to address, and b) have attempted to solve the problem in the past. "Problem" in this case should be factors leading to reduced yields and/or income. Individuals wanting to join the participatory research after seeing initial field trials should be allowed and encouraged to participate, especially in the more farmer- designed "spontaneous" experiments. It is less desirable to work with farmers whose participation is based on their perceived or real benefits of project participation per se. Charismatic, enthusiastic farmers had Jeading roles at severa! sites. In sorne cases, these early-adoptors appeared to be effective in increasing farmer participation. On the other hand, project staff may become overly comfortable working with these leaders at the expense of working towards more extensive participation and participation of the initially less enthusiastic. TR-43 Targeting Research FPR trial design: forages. At least in the Philippine site, farmers' trials were divided into those in which farmers volunteered to test at least four, usually "best bet" species, and more "spontaneous" trials in which farmers were free to test whatever they requested. Such a division of trials is sensible and useful. Sorne suggescions may be appropriate considering all forage sites visited: a) Sorne of the more "formal" trials were fenced, others not. Unwanted grazing was a "problem" in sorne of the initial trials in Laos. These trials might all be fenced. b) Where appropriate, live fence species trials can be incorporated with the forage FPR trials. e) Line-sown forages might best be planted along the contour where slopes are more than 18 percent. d) Clear and relatively permanent marking of treatments and signs briefly explaining each trial would be usefui for data collection and in promoting demonstration effects of each tri al. e) Fanner-accessible locations would increase the demonstration potential of FPR trials. FPR trial design: soil erosion control. Initial poor layouts supported the idea of increasing supervision on a fewer number of erosion run-off trials in arder to increase effectivity of these trials both as trials and as demonstrations. At the same time, it would be desirable to increase the number of farmers (interested in) testing best-bet or desired contour vegetative strip alternatives. Sorne ideas for the run-off trials, either farmer or researcher managed: a) All vegetative strips must be established along the contour. Simple instruments such as the A-frame or carpenter's leve! on a string can be used to establish contour lines. b) The sides of the treatrnent plot must be perpendicular to the contour. e) Contours should extend in-line across the various treatments so that natural terracing can take over the whole area. d) Run-off trenches need to be placed along a single contour line. e) Trees in hedgerows need to be established at closer distances for effective hedgerow forrnation. t) To increase realism and demonstration effects, treatment lengths should be sufficient to incorporare at least three contour rows, each placed at distances corresponding toa one meter elevation drop. g) Again, soil erosion trials may need something in the order of at least a 20-25% slope, 80 tlha/year soil loss, and shallow soils to be effective as farmer demonstrations. h) Farmer-treatment controls must be accurately replicated in trials. Using a treatment "worse" than actual farrner practice ( eg not using con tour ridges where farmers use su eh a measure) does not allow farmers to accurately compare their practices to tested innovations. A caution, however: as run-off trials are better managed through more researcher supervision, there may be a tendancy for the same researchers to provide recommendations based on trial results. Such recornmendations can work against farmer experimentation, learning, technology adaptation, and eventual adoption of individually or locally appropriate innovations. Trial monitoring and data collection. Among data to be collected as FPR trials are conducted are: a) performance indicators for the forages or erosion control treatrments, b) farmer evaluations and perceptions regarding treatments, and, to the extent possible, e) records or estimates of benefits derived from, and of direct and/or opportunity input costs (ie in terms land, labor, and capital) required by the innovations. Again, G~bunada et al ( 1997) pro vide good examples of "researcher observations" and TR-44 Targeting Research "farmers' comrnents"-both "negative" and "positive". Although it may be too early at most ofthe sites to measure costs and benefits. the issue should be considered (and is discussed in the section on impact assessment). The collection of fanners' appraisals and insights should be done in the fieid with individual fanners. Participating farmers should eacb be encouraged to discuss what they feel are possitive and negative factors associated with particular innovations. Farmers' individual observations can be aggregated and presented in simple descriptive. tabular form. An important activity of the cassava FPR is the farmer field day in which experimental plots are harvested, yields (cassava and intercrop) measured. and results compiled and discussed by participating farmers. Farmers and researchers work with soil erosion, soil nutrient management, and varietal treatments. Farmers' initial evalautions have been used to select treatements for subsequent seasons of FPR. Forros of FPR. Useful approaches to FPR include: a) relatively formal, replicated trials conducted by a few farmers with moderate trial design and organizational inputs provided by researcher-facilitators (eg Ashby 1993), b) more informal trials, often without replication. but with sorne researcher-inputs into "best-bet" treatments and triallay-out (eg the FPR erosion trials and the more formal forage trials discussed in this paper), ande) informal farmer-developed testing of components and adaptations of components and combinations of components. The formal trials produce replicable results and allow for statistical testing of hypotheses, but leave little room for farmer innovation and broad participation. Supervision and organization costs can be high. The simpler trials-as used at the sites visited-may provide findings suitable largely for the formulation of further hypotheses or may suggest trends, but often serve as demonstrations and learning tools for participating farmers. The very informal testing of new combinations by interested farmers can result in greater numbers of participating farmers, locally appropriate problem solutions, and possible adoption of such innovations (Fujísaka 1993). Projects usmi..Ily must encourage and enable such farmer experimentation. through initial provision of seed or, for example, of farmer-to-farmer training. The cassava and the forages FPR appear to have settled on the relatively more informal trials of best bets followed by more extensive, less "supervised" trials by interested farmers. This overall strategy appears to be quite appropriate to the (Asían) sites visited. Towards impact assessment Ex ante impact assessment. Bread ex ante impacts are commonly assessed in the initial development of research projects in the definition of problems and statement of expected outputs. Considerable work has been done to define forage needs and to target forages research in Southeast Asia. Assessments of then future needs made 1 O years ago were accurate: the region's self-sufficiency in meat production was 94% in the period 1973-77, but was expected to drop to 62% by year 2000~ and demand for forages-given increases in ruminant production-was expected to double by year 2000 (Remenyi and McWilliam 1986, cited in Stur l99x). Implicitly, future adoption of improved forages and increased self-sufficiency in meat production (although not necessarily causally linked) would be quantifiable and was to comprise the majar desired impact. TR-45 Targeting Research In terms of soil erosion. the quanriry of sediments eroding into rivers and then flowing into the world's oce:ms is greatest in southeast Asia (Milliman and Mearle 1983). How much of the sediment load can be attributed to natural processes vs human activities, and how much of the erosion caused by human activities is dueto agricultura! practices are not known. Soilloss rates measured on upland agricultural fields in the region vary widely, as do soillosses in cassava growing areas. The FPR on reducing soil erosion in cassava systems would have benefited from initial quantification of soillosses in the different cassava growing areas and general targeting (eg areas where soillosses exceed given levels) in arder to malee sorne general ex ante estimates of expected impacts. Ex post impact assessment. Three levels of ex post impact analysis can be considered: the region (Southeast Asia), the targeted agroecosystem, and the project si te. Impact assessment is most practica! at the site level, however. For the forages research at the regional. Southeast Asia, leve!, broad impacts in terms of increases in herd sizes, liveweight gain rates. adoption of introduced forages and meat self-sufficiency are arnong longer-term objectives. The FSP would not necessarily be concerned with ex post measurement of such impacts as these are expected after broader technology transfer and corresponding implementation of needed policy and institutional innovations. Such impacts, however, are long-term goals and are measurable, usual! y through use of national agricultura! statistics. Sirnilarly, for research on soil erosion at the regional leve!, monitoring of sediment loads in the world's major river systems may provide one way to assess impacts of changing land management trends. Again, this would not be a project activity (albeit a key concern). In targeting different fanning systems the FSP forages research defmed its more specific expected extrapolation domains. The systems were agroforestry, upland, plantations, grasslands, and lowland systems. Countries were Indonesia, Laos, Malaysia, Philíppines, southern China, Thailand, and Vietnam. Given such targeting, it would have been ideal to have data on: a) area of each system by country, b) current herd size by system and country, e) currem adoption of introduced forages by system and country, d) typicalliveweight gains by system and country, and e) numbers of livestock producers per system and country. Impact analysis could then, at a mínimum, assess changes in herd sizes, forage adoption, and liveweight gains. Unfortunately, data to clearly delineate the sub-systems are Iacking, as are ways to disaggregate data on (among others) livestock numbers by subsystem. Data on current levels of adoption of introduced forages is similar! y lacking. For FPR on soil erosion in cassava systems, different cassava producing systems (eg cassava in shifting cultivation, commercial starch producers, cassava in mixed permanent cropping) in Asia were not initially defined and then characterized in terms of respective ranges of soil losses. Impact analysis by the project at the targeted agroecosystem leve! is, therefore, not possible. Impact analysis at the local or site leve!, however, can be done using data is collected at the initiation, conduct, and termination of project activities. For the forages research, useful community leve! data to be gathered initially would include: a) "problem(s)" to be addressed by the research in the community and agroecosystem (ie what type of system does the communuty represent?), and proposed measures of success; b) problems and opportunities defined in terms of different farmer strategies encountered in the community (eg, lowland rice producers with a draft buffalo vs cash crop producers with severa! head of beef cattle vs upland farmers needing improved nutrient cycling); TR-46 Targeting Research e) sinúlarly, problems and opportunities defined in terms of different livestock strategies employed within the comrnunity (eg calf production, dual purpose, steer fattening, draft animal maintenance) and the proportions of persons engaged in each strategy; d) number of families, number of families having livestock, total number of different types of livestock, and ranges of individual livestock holdings; and e) livestock feed resources and different livestock feeding systems employed in the comrnunity (ie open access grazing, grazing on own fields plus cut and carry, types of pasture and fodder resources used, use of supplements, use of fodder trees); and rough proportions of persons engaged in each activity. The above would allow researchers to further define problems and objectives, target individuals and groups within the community, and specify expected achievements. Data gathered from individual fanners representing project-targeted sub-groups could then be obtained: a) The above-mentioned numbers of livestock and feeding systems employed per individual farmer participant; b) Simple analyses of costs and returns of individuals' current livestock systems; e) Use, if any, of innovations (ie introduced forages). Simple costs and retums can be obtained by gathering data from a sample of fanners representative of the larger target group (and of any smaller sub-set of participating fanner researchers). Data would include material, cash, and labor costs and yields and gross returns. Data should reflect annual costs and returns, and should reflect both costs and retums per annimal unit and per farm. Given difficulties of obtaining liveweight gains, increase in price of animal at time of sale vs at the time of purchase can be used to tabulate gross returns. Calculations can then include returns to variable costs and to factors (labor, farm resources, material costs, cash). Ex post impact can then be assessed at the end of the project by exanúning: a) changes in numbers and types of livestock (by both participating and non-participating fanners); b) adoption of innovations: changes in livestock numbers, adoption of new livestock feeding systems, and adoption of introduced forages-expressed in terms of numbers of individuals and areas; e) simple analyses of costs and returns of adoptors' systems; and estimation of differences in income (returns above variable costs) from livestock enterprises, comparing adoptors and non-adoptors or pre-and post-adoption. Adoption (and non-adoption) of innovations should be assessed over the course of the project via experimental plot monitoring and fanner interviews. Plot monitoring was discussed above and reflects the need to measure the performance of innovations and to accurately compare that performance with the performance of existing farmers' systems and practices. Field visits should also serve to assess adoption of innovations by project non-participants. Farmer interviews are necessary to obtain farmer feedback on the innovations tested. Farmers who participate in the project and who adopt ínnovations, non-participants who adopt, participants who do not adopt innovations beyond the experimental plot, and non-participants who do not adopt tested innovations should be interviewed. Simple interviews should include questions regarding: a) farmers' open-ended assessments of the innovation(s); b) adoptors' reasons for adoption; TR-47 Targeting Research e) non-adoptors' reasons for non-adoption (including non-participants' non-farrúliarity with the innovation); d) adoptors' costs (material, labor, cash) and retums; and e) adoptor's modifications or adaptations of the basic innovations tested and reasons underlying such adaptations. While the projects visited rnay choose not to consider all off-farro and non-farm activities in their simple econorrúc analyses, sorne idea of the opportunity costs of cash, labor, and land needs to be obtained in evaluating adoption of innovations. Farmers wil1, of course, invest their resources where gains are highest and/or risks lowest. Finally, sorne idea is needed of larger national and regional trends in assessing local impacts. For example, to what extent have farmers adopted soil erosion control innovations due to increasing population and a closing of the land frontier vs dueto project efforts? Environmental impacts. Differences in soil erosion comparing treatments and using the plastic lined soil collector trenches developed by Howler provide a direct idea of environmental benefits possible if adoption of the respective soil erosion control measures were to take place. Again. measurement of soil erosion in different cassava production systems would be needed to estímate potential environmental benefits. Further thinking about levels of erosion on different types and depths of soils "needed" to induce farmers to want to adopt erosion control measures is needed. For forage innovations, plot leve! studies can indicate changes in soil nutrient status and organic matter levels; and, at a higher scale, researchrs can measure such benefits as decreases in degraded natural pasture areas, reduction in annual grassland buming, and increases in tree cover and use of trees for forages and live fencing. Assessing potential environmental impacts of introduced forages, however, would need to consider whole systems; just as more productive agricultura! systems in the forest margins can contribute to increases in deforestation rates, more productive livestock systems could contribute to either systems degradation or improvement. Conclusions Collaborative farmer participatory research now underway in Southeast Asia seeks to enhance local adaptation and adoption of innovations in animal feeding, soil nutrient management, weed control, fallow improvement, Imperata grassland "reclarnation" for more productive use, intensive livestock systems, soil erosion control, and varietal (cassava and forage) improvement. Farmer participation is important to the extent that severa! innovations are "knowledge intensive" rather than the simpler introduction of new varieties or specific management practices. Systems to control soil erosion, new livestock feeding systems combining legumes (including trees) and grasses and cut-and-carry with grazing, the very idea of sowing and managing forages (in systems where forages have never been sown or directly managed), systems to improve fallows and provide fodder resources, fodder banks, and use of rainfed lowland rice fields in the dry season are all examples in which farmers must gain new forms of knowledge and make systems-level adaptations and adjustments. Participation has been manifested in participatory rural appraisals, in the development and management of on-farm trials, and in the assessment of results. This paper discussed the need to better understand intra-community diversity. Site characterization and diagnosis can be expanded from largely PRA group activities to work with individuals and targeted sub-groups within comrnunities. Farmer participants perhaps need to be those: a) aware of, b) suffering from given problems being addressed by the research, ande) who have tried to solve the given problem in TR-48 Targeting Research the past. Care might be taken such that work with charismatic farmer leaders results in broader participation. Severa! recommendations were provided regarding design, execution, and monitoring of the FPR trials. Continua! monitoring of the performance of the innovations and of farmers' ( open-ended) insights remains the ideal. In the case of the soil erosion control research, care will be needed to maintain a balance between well-supervised demonstration trials and an associated tendency to then provide recommendations to farmers and the heart of farmer participatory research-that of farmer experimentation, discovery, adaptation, and, eventually, adoption. Finally, the measurement of impacts, both estimating ex ante and measuring ex post must be a key part of FPR. Defmition and delineation of targeted systerns (both regional and agroecosystem levels), measures of their current performance (eg in terrns of current soil losses or livestock productivity) and levels of desired performance changes brought about by project efforts are essential for the development of ex ante es ti mates. Site-level success can be estimated more directly using simple rates of adoption of innovations and the costs and benefits associated with such adoption. Such specific e.x post analysis shows what is possible and allows for further ex ante extrapolation, again given addressing of Iarger institutional and policy needs. References Ashby, J. 1993. Insitutionalizing farmers' participation in agricultura! research. Annual report of the Farmer Participation Special Project. Cali: CIA T Fujisaka, S. 1993. A case of farmer adaptation and adoption of contour hedgerows for soil conservation. Experimental Agriculture 29:97-105 Gabunada, FA, EA Balbarino, and AP Obusa. 1997. Farrner participatory research on forages in Matalom, Leyte. Unpublished. Gibson, Trevor. 1997. The Plain of Jars: an example of phosphorus deficiency for forages and Iivestock. SEAFRAD News 4:2 Milliman, JD and RH Meade. 1983. World-wide delivery of river sediment to the oceans. Joumal of Geology 91:1-21 Remenyi, JV and JR McWilliam. 1986. Ruminent production trends in Southeast Asia and South Pacific, and the need for forages. In: GJ Blair, DA Ivory, and TR Evens, eds. Forages in Southeast Asían and South Paci.fic Agriculture. Canberra: Australian Centre for International Agricultura! Research Stur, Werner. 199x. Research and development for forage production and supply in South-east Asia. TR-49 Targeting Research Minimum data set for site characterization in on-farm studies in Southeast Asia (1.1.5; W. Stür, CIAT, P. Horne, CSIRO) Main achievements • Development of a framework for characterization of FSP si tes • Agreement reached on a minimum inforrnation set for site characterization at comrnunity and surrounding area levels • Collectionlassembly of minimum inforrnation has been initiated for all sites Report. Si tes were selected to cover a range of environments and famúng systems, commonly found in upland agriculture in Southeast Asia. Within systems, sites selection was based on the following criteria: 1) Are farmers facing a serious problem that can be addressed by introducing forage technologies and does it affect many farmers? 2) Are there many farmers who perceive it as a serious problem and are they looking for ways to overcome it? 3) Are farmers interested and willing to work with the project? 4) Are our national partners able to support work at these sites? 5) Are local development workers (and support from their office) available to join the project? 6) The existence of effective farmer groups is an advantage in many but not all situations. Si te characterization is carried out for different purposes ( or phases) of the project (Figure 2). These are l. To identify potential sites for farmer participatory forage technology development, 2. To select sites and obtain commitment of farmers and development workers to work together, 3. To plan and work with farmers, 4. To assess of the impact of forage technologies. The methodology of characterization for these different purposes varies. It includes secondary information, participatory diagnosis, participatory planning, interviews with individual farmers, participatory evaluation, surveys and direct measurements. The framework for characterization including target comrnunities, methodology and outputs is described in Figure l. Following are summaries of site descriptions for identification and selection of potential FPR sites. Results of problem diagnosis during participatory diagnosis are discussed in Output 1.2.2, details of activities at si tes in Output 2.2, and assessment of impact of forage technologies in Output 4. Site descriptions for identifying and selecting FPR sites By September 1997, se venteen on-farm si tes were operational. Severa! of the si tes in Indonesia and the Philippines started work in 1996, while most of the sites in Vietnam and Laos were established only recently. Site descriptions for identifying and selecting FPR sites are available foral! sites. However, comparisons across sites are difficult since descriptions have not been based on a comrnon information set. Recently, the FSP has established a mínimum information set for this purpose and this inforrnation is in the process of being assembled foral! sites by national partners (Table 1). The required inforrnation includes sources such as l. data and maps (often available in district or provincial offices) TR-50 -l :;el V. Figure l. Framework for characterization of FSP si tes Farmers participating actively in the FSP Targeting Research Mpthoda: s~ary lnf~ pn.~!i&ta. rnty>s, lntervtews:; OQ#IeJVations) OLIJput: • brlef.$:1escf1p~ ~f cllmat(; sQIIs, ~ ~·. landscape ai1d lañ ".;z . . · with9ut lívestock) .< ''" ~ • 1. Brief description of the area focussing on key issues affecting development 2. Description of topography, soils, local clima te if different from general area 3. Land use systems • relative land area for each use(%) • topographic location of each land-use • what are the main 1and use systems and their benefits/constraints? • what inputs are used in agriculture? • how is non-cropped land used? • what is the land-ownership system? 4. Livestock farming systems • types, number, distribution • why are livestock kept? (e.g. production systems) • what proportion of farmers keep livestock? • ownership (e.g. is shared ownership of livestock common?) • are inputs used in raising livestock? (e.g. supplementary feeding, veterinary chemicals?) • how are livestock managed? (e.g. feeding systems- are animals grazed throughout the year, are they fed cut feed, who in the family is involved and how) • how are livestock marketed? • whatlwhen are the main constrains and opportunities? • how have farmers been dealing with these constraints until now? • how do they want to dea1 with them in future? 5. Trends in the farming system • what changes are happening within the farming system? • what changes are happening within the livestock raising system? 6. What are the main sources of income of farm families? 1. What other rural development programs have been and are currently working in this area? PD PD d,i,PD PD d, i,PD PD PD PD PD · Po PD · PD PD PD i, PD i,PD d = data and maps i = key information/observations PD = participatory diagnosis TR-52 Targeting Research 2. key information from discussions with govemment officers, village heads and key farmers, and personal observations; 3. participatory diagnosis. A summary description of sites showing the range of agro-ecosysterns and environments included in the FSP is provided in Table 2. Full site descriptions will be available at the tíme of the 3rd Regional Meeting of the FSP in March 1998. TR-53 ~ Ul "'" Table 2. S ---------- f si te d 'f - - ---- -----"'r•------ LOCATION 3 Cll -.. < 8 §, SI TES g u " "O "O ü a a ~ 'r:l 'r:l j < ::E Indonesia Sepaku 11, Easr Kalimanran 1•s <100 D-M Makroman, Easl Kalimanlan o .5•s <100 M Goronralo. North Sulawcsi o.s•N <100 M Marenu, Sourh Tapanuli 1°N <200 D Pulau Gambar, North Sumalra 4"N <200 E Saree, Acch 5° N 500 M I.aos Luang Phabang 20"N 300-1000 D Xieng Kl1ouang I9 °N 1,300 D l'hilippines Cagayan de Oro, Mindanao S "N <100 D-M Corabalo, Mindanao 9•N <300 D Davao, Mindanao 9"N <100 E Guba, Cebu IO"N 500 E Malirbog. Bukidnon S "N <300 D Mala1om, Leyre IO"N <300 M Vietnam Xuan Loe 16°N 100 M M'Drak, Daklak I2°N 500 M Vielnam-Swedish Projecr 21 °N 300 E (northcm provinces) -- --- 1 Markel acccss: E= easy (close 10 markers), M= medium, D = difficuh (vcry remole) 2 Soil fertility: L = low (infertile), M= moderate fertility, H = high fertili!y AGUO-ECOSYSTEMS "O "O i a c. " " 'O u u á > > ·~ ·~ <;; E !! !! >( o .S l1l v' v' v' v' v' v' v' v' v' v' v' v' v' v' v' v' v' v' Targeting Rel·earch CUMA TE SOlLS J I s} 'O ª\) ;e E a 1 " ~ ~-s '};. ,e. .D :g ~ ·~ ?;-'O ..2 ;::: :a a o o .S ·~ ] 'r:l -¡¡¡ ·~ .r: .9 " ~e -)!} ·~ a ~ e o ;g ::::2 ~g o Vi "' ¡¡:: Cll Vl 2,200 <3 L 4 .5-50 2,000 <3 M 4.8-5.3 v' 1.300 3-5 M 6-7 <2,000 3-5 L 4.5-5.5 v' 2,000 o M 6-7 1,600 2-4 L 4 .5-5.5 v' 1400 S L-H 5-7 v' 1400 S L-H 4.5-6.5 2,000 2-4 L-M 5.5-6.5 v' >2000 <3 M 6.5-7.5 1/ 2,500 <3 M 6 >2.000 <3 M-H 6-7 <2,000 2-4 L-M 5.9-6.0 2,000 3-5 L-M 4.5-8.5 v' v' 3,200 2-4 L-M 5-6 1400 4 M 4.5-5.5 1700 S M S-6 ---- --- ·-- --- Summaries of rapid PD appraisals ( 1.2.2~ W. Stur, CIAT and P. Horne, CSIRO) Main achievements Targeting Research • participatory diagnoses (PD) ha ve been conducted at all 17 on-farm si tes of the FSP during 1996 and 1997 • reports on PDs are available in local languages~ English summaries are available for most sites and the remaining English language summaries are being prepared by national partners Participatory diagnosis Participatory diagnoses have been found particularly effective in l . Identifying if there are problems which can be addressed with the introduction of forage technologies; 2. Detenniníng if farmers perceive these problems as sufficiently serious to be willing to allocate input (labour and capital) towards trying to salve them; 3. Learning how farrners cope with these problems and what approaches they have tried to overcome them; 4. Understanding problem-cause relationships faced by farmers 5. Sensitising our local technical counterparts to how well fanners understand their famúng systems and how able they are in analyzing problems and potential solutions; 6. Establishing a good working relatíonship with farmers. A summary of majar problems identified by farmers during PDs is presented in Table l. An important part of participatory diagnosis is the analysis of causes of the problems. This helps to understand causes and linkages of the problems, making decisions on how to approach the problem much easier. An example of a problem-cause relationship, as analysed by farmers during a participatory diagnosis on the problems related to raising livestock in Matalom village, Leyte, Philippines, is presented in Figure 1 (Gabunada, Stür and Home, 1997). The farmer group first identified the set of common problems they faced in raising livestock. This random list was then sorted in arder of importance and the interactions between them identified. Next the fanners discussed each problem to identify what actions they had already taken to alleviate that problem. Encouraged by the many steps they had already taken, they finally reassessed each problem to see if there were any other steps they could take to improve the situation. The result is a clear summary of the current status of livestock rearing in that village which the farmers had developed, understood and owned. The whole process took two hours. In this case the farmer group identified "better forage species" as an important step they could take to improve livestock husbandry. The researchers used this as a basis to suggest a range of forages which might fit into their livestock and cropping system. The farmers selected what they wanted to try from the options provided and plans for the tria! were made. TR-55 Targeting Research Table l. Major problems identified by farmers during PDs 1 1 1 l i 1 1 1 1 J ~ 1 .. 1 -;:. :¡ 1 1 ; '"' 1 1 "' ; 1 ~ l 1 "' 8 ~ ;:¡ ,., o -~ :S 1 1 ::¡ .. 1 ! 1 !! ll ~ SITES ~ J ! o ... ~ 1 :lo 1 :!$ :z = ~ ~ : ,., ª ·= ~ "' .. 1 "' o: 1 o: ~ ~ ~~ ] ~ e =- e -a 1 3 i! . o: ~ e - ~ ~ -S ~ 3 1 ,. 1 e ~ ~ ·¡¡ ~ ~ ~ "3 ll 1 ,., ] .& ... .., ~ " ~ ¿¡ ~ j .JI Indonesia 1 1 1 i 1 1 1 ! 1 1 Sepaku II, East Kalimantan 1 , ...... 1 ! v ...... 1 1 Makroman, East Kalimantan 1 1 , .,.... : v i v 1 1 Gorontalo, North Sulawesi i v ! ¡ 1 1 ¡ 1 1 Marenu, South Tapanuli ...... ...... i 1 1 i Pulau Gambar, North Sumatra , ...... í 1 1 1 1 i 1 Saree, Aceh 1 I V 1 1 1 1 ¡ ¡ ' l 1 1 La os 1 Luang Phabang ...... ...... I V 1v ...... Xieng Khouang l v . ...... !V 1 Philippines 1 1 1 Cagayan de Oro, Mindanao l v 1 ¡v 1 Cotabato, Mindanao ...... , ...... 1 1 1 ...... l Davao, Mindanao ...... l 1 1 1 1 1 Guba, Cebu V' 1 1 1 ...... Malitbog, Bukidnon V' , ...... 1 MataJom, Leyte V' ¡ v 1 1 V' Vietnam 1 1 1 1 1 1 Xuan Loe V' 1 M' Drak, Daklak 1 V' l v Vietnam-Swedish Project ...... 1 1 1 ...... TR-56 Limited area for 2rnZing Reduce the number of animals Parasite/disease susceptibility in buffaloes Seek advice from veterinarian Seek knowledge on how to identify diseases t Overwork m buffaloes Hire and Exchange Labour Lack of feeds in the season Low leve! of animal nutrition Pro vide feed supplements ! Thin animals with low wei2ht gains / T argecing Research Drought Insufficient knowledge of new methods Low productivity: few offspring thin offspring c:J Problems D Farmers' response Possible Future Solution Figure l . Feed resources problem diagnosis by smallholder farmers in Matalom, Leyte, Philippines (Wemer Stür and Francisco Gabunada, FSP-CIAT; Peter Home and Phonepaseuth Phengsavanh, FSP- CSIRO, Maimunah Tuhulele, DGLS, Indonesia; Ed Magboo, PCARRD, Philippines; Viengsavanh Phimphachanhvongsod, DLF, Laos, Le Hoa Binh, NIAH, Vietnam) TR-57 New approaches for targeting technology development (1.1.4; S. Fujisaka and R. Howeler) Main achievements Targeting Research • Developed and conducted a course for trainers in FPR in Thailand and Viecnam. • Incorporated lessons from the first course to improve the second course. • Trained researchers and extensionists - Thailand 25, Vietnam 27, Laos 2, China l. Report "Training ofTrainers" courses were conducted in FPR in Paakchong, Thailand (03- 12 September) and in Thai Nguyen, Vietnam (17-26 September), as part ofPE-5's involvement in CIATs Nipon Foundation- supported work in Asia. The course was conducted with funds from the NIPPON funded project 'Improved Suscainability of Cassava-based Cropping Systems' but there was clase collaboration with the 'Forages for Smallholders Project' which provided trainers and materials and sent trainees to the course. The course was organized and coordinated by Reinhardt Howeler. Resource persons were Sam Fujisaka, Hans Dieter Bechstedt (IBSRAM-Bangkok), Francisco ("Papang") Gabunada (FSP/CIAT, Philippines), and Suchint Simaraks (Khon Kaen University) in Thailand, and Peter Horne (FSP-CSIRO/CIAT, Vientiane), Guy Henry (CIRAD), and Nguyen Van Dinh (CARE, Hanoi). Participants in Thailand represented the Field Crops Research Institute and the Field Crops Promotion Division of the Extension Service of the Department of Agriculture, the Thai Tapioca Development Institute, the Livestock Department and the Land Development Department. Vietnamese participants represented the Institute of Agricultura! Science in South Vietnam. universities (of Agriculture and Forestry at Thy Duc, Hue University of Agriculture and Forestry, Thai Nguyen University), and severa! Provinicial Departments of Agriculture. We found that a 10 day course in FPR is possible, useful, but not necessarily easy to design and conduct. Initial Overview. The frrst day of each course was a "show-and-tell" overview targeting "decision makers" and bosses of trainees. The goal was to sensitize bosses to the degree that they would support future FPR activities. We were successful with short, concept-plus-case-example presentations. More attention is needed in the difficult task of involving science administrators. Course Presentations. Initial presentations regarding principies and case studies were largely acceptable. Concise presentations which emphasize or highlight key ideas, concepts, and tools were most needed. Participants needed inputs as to differences and similarities among RRA and PRA or PD, on-farm research and FPR. Handouts in Thai and Vietnamese were appreciated and needed. Trainees received a copy of a FPR manual adapted from CIA T IPRA materials by the FSP and translated into the local languages. Tools and Exercises. Classroom exercises in use of participatory diagnostic and evalutaion tools such as scoring and weighting, mapping, calendars, preference ranking, matrix ranking, and problem diagnosis and diagramming were largely successful, especially with exercises which required trainees to participate based on their own Ji ves and experience. An exercise in working up a partial or enterprise budget from "manufactured" data appeared to be useful. Although necessary, I am not certain that exercises on TR-58 Targeting Research listening, neutrality and cornmunication skills were effective: people are sensitive or insensitive to others/farmers to the degree that such a short course is unlikely to make desired impacts. Field Exercises. Participants practiced the use of participatory diagnosis and evaluation tools at selected project sites. Farmers appeared to ha ve fewer difficulties in working on multidimensional calendars and matrix rankings than did trainees. Trainees need to carefully consider the differences between researchers' and fanners' evaluation criteria. Cassava breeders, for example, were surprised that the Viemamese farmers did not use duration as an evaluation criteria and were not interested in shorter duration cultivars (because of highly diversified farms on which cassava was harvested at different ages between multiple times of peak labor demand). In discussing problems, farmers-especially in Viemam-appeared programmed to say they "needed new technologies". Getting trainees to further elicit farmers' specific problems and solutions was a challenge. Once farmers did turn their attention to real problems, however, they had few difficulties in diagramrning interacting causes and effects. Travel time to sites was excessive: future training courses need to increase field practice time and decrease travel time (which was high dueto travel to actual project sites). Of course, participatory evaluations could only be conducted at project sites where farmers had tested new alternatives. Trainees were not very adept at making whole-systems field observations. For example and perhaps because of their cassava orientation, Vietnamese trainees failed to note the extensive and severe tungro virus in rice, the farmers' main crop or an insect defoliator of tea, another majar crop. Trainees did not appear to have many cornmunication problems with farmers, although a few spoke more than they listened. Trainees were quite good in involving male and female, young and old, and poorer and richer. Making Sense of Data. In feedback (to farmers) sessions, trainees tended to present essentially the same data obtained via participatory diagnosis and evaluation. Trainees did not easily synthesize and informally analyze data. Sorne effort was required in demonstrating how data provided by farmers could be presented in simple graphs and figures and how such data could lead to identification of possible new relationships, hypotheses, questions and conclusions. Trainees were not experienced in making concise, to the point presentations. Participatory Research. We discussed different types of participatory research from relatively informal technology adaptation by many farmers to more controlled varietal or fertilizer trials conducted by fewer farmers (eg as in the CIALs). Trainees needed to be reminded that FPR refers to participatory research and not only participatory diagnosis. We discussed experimental design, especially possible problems and complications associated with the erosion control trials. The need for the "farmers' practice" treatment to be accurate in arder for comparisons to be relevant to farmers was emphasized. Trainees visited farmer participatory research varietal (cassava), fertilizer, and soil erosion control trials. Farmers evaluated tria! results. Varietal evaluations were insightful in terms of farmers' criteria and weighing. The soil erosion control trails were less conclusive (see following). The Research. Lands are gently sloping and per ha soillosses are relatively low in Thailand; and farmers carefully and intensively--in time and space-manage a range of crops and resources in lowland and adjacent upland fields (with practices which include bench terracing and contour ridging) in Vietnam. As such, SF considers that that farmers would not further invest in soil erosion control in these two areas; and TR-59 Targeting Research that farmer participatory research on soil erosion control may be counterproductive in terms of the desired instiruúonalization of the approach by national programs. Proposals. As a last part of the course, participants developed proposals for further work in either or both farmer participatory research and training in FPR methods. The Course. We developed a reasonably useful course through experiences gained during the two courses. From trainees' evaluations we need to spend more time in the field for exercises with farmers and need to provide more written materials. Use of "icebreaking" songs and exercises were useful in creating a team spirit and daily cartoons in demonstrating principies. TR-60 Output 2. Integrated technology and management options Collaborative research and development (2.1, PE-5 team) Technology development The project 'Sustainable systems for smallholders' is aimed at the integration of strategic research outputs from gennplasm improvement and natural resource management in CIA T to improve the productivity and sustainability of farming systems at the farm level. Thus it is essential that we work with the collaboration of other projects in CIA T. Likewise it is our aim that the adaptive research that is undertaken is done so in collaboration and indeed largely through national partners. The degree that integration has been achieved in CIA T varíes. Among the germplasm projects, there is close integration with forages and increasing integration with cassava, in particular, in South East Asia. The forage and cassava scientists in South East Asia are coordinating sorne activities, in particular, methodology development in FPR and impact assessment. It is likely that we will develop close linkages with the rice research project in the proposed adaptive systems project in Pucallpa, Peru. A visit was made to Africa to become familiar with the research in systems where beans are an important crop. The conclusion was that closer linkages should be developed with the systems researchers in Africa but that the research is best managed from there. The following is an extract from the trip report. "The team in Africa expressed a desire for closer involvement with others in CIAT, particularly, in the area of natural resource management. Charles Wortmann has planned activities with PE-2 and PE-5 . It would be desirable if other CIAT scientists could make themselves more familiar with the work in Africa and consider how they might complement activities of the bean team and the other IARC Centers with whom the team collaborates. PE-5, Sustainable Systems for Smallholders, would benefit from closer interaction with the Systems activities within the Bean Research Team for Africa. Firstly, Charles Wortman and others (Carey Farley and Sonii David) working in systems work in Africa could make a useful contribution to those working with small farmers in LAC and Asia. This collaboration will commence with the Systems workshop planned for early December where experíences in FPR and systerns research will be shared. Scientists in different regional areas could arrange ro visit during other travel , e.g. on return to CIA T or associated with horne lea ve. There could be more exchange of reports and draft papers rather than just the tlood of administrative circulars. There is opportunity for CIAT to increase the extent of adaptive systerns work in Africa. Other IARC's are less involved in adaptive systems research than CIAT at the presenttime, though they are involved in IPMIICM studies focusing on their particular commodity mandate. However, systems research should continue to be associated with the bean research team and grow from within this base rather than to establish a separate systems project which might bring resistance from other IARC partners. Leadership should be encouraged from within this team. Adaptive systems research would best be associated with the Ecoregional Program, the African Highlands Initiative (AHI) where it is in higher altitude areas (> 1500 masl). In the mid-altitude areas ( 1000-1500 masl) there might be sorne association with the proposed EU funded project. Areas of mutual interest are: Use of legume as green manure in farming systems Development of FPR methodologies for technology development Developrnent of methodologies tor dissemination of results achieved through FPR Assessrnent of impact of FPR technology development. Policy issues " Close liaison is maintained with the NRM management programs through twice-monthly meetings. Closer links will evolve as we develop adaptive systems research in the reference sites for the Ecoregional Program for Tropical Latín America, in Pucallpa, Peru (Forest Margins), Nicaragua and Honduras TD-1 Technology development (Hillsides) and the Llanos, Colombia (Savannas). Considerable effort was expended during the year towards building collaborative research teams in the Ecoregional Reference or Benchmark sites- Pucallpa, Central American Hillsides and Savannas, with initial emphasis on the Forest Margins site in Pucallpa. A participatory planning workshop was held within CIA T to develop a consensus on overall priori ti es for research in Pucallpa, CIA T initiated a meeting of DG's of ICRAF, CIA T and CIFOR in Pucallpa (to be held 4~6 November 1997) and has taken responsibility for facilitating a participatory planning workshop with all stakeholders in Pucallpa to be called by CODESU and funded by IDRC. in early 1998. The key to continued development of a CIA T research team in Pucallpa will be the ability to raise or transfer funds for research in that site. The specifically funded sub~projects: Forages for Smallholders, Improved Sustainability of Cassava-based Cropping Systems and Tropileche were designed with the concept of clase collaboration. Each of these sub-projects has its specific research network of close and occasional collaborators. The FSP and Tropileche produce six-monthly newsletters. Tropileche is a project associated with the Systemwide Livestock Program coordinated by ll..RI. 1t is seen as a platform on which the IARC' s may interact with nationallivestock programs in the area of feed improvement for livestock in the Latín American tropics. It initially commenced operating with partners in Peru and Costa Rica. However, partnerships are actively sought with other countries and livestock groups. This was ini tiated through a workshop on methodology for research related to dual-purpose cattle production held in July 1996. Nicaragua became a partner of the Tropileche consortia in 1996 and on-farm fo rage technology development was commenced during 1997. Our partner in Nicaragua is the Dairy Development Project (MAG). Nicaragua is fully funding its research with funds from the World Food Program. The project has requested CIA T to expand the Tropileche activi ties in 1998 from one si te (Muy-Muy -Matiguas watershed, an intermediate zone between the humid and the dry lowland tropics) to three sites (dry, intermediate, and humid tropics). Another partner which joined Tropileche in mid-1997 was Honduras through DICTA. Like Nicaragua, Honduras is fully funding its research in Tropileche with funds obtained from the Govemment of Japan. Tree trips were made to Panama during 1997 to establish contacts based on the interest of IDIAP to join Tropileche, but the fi nancia! resources to finance the research costs in Panama have yet to be found. Likewise, a trip was made to Brazil in September 1997 to develop with Embrapa a project proposal to start research activities with CPAC, the University of Uberlandia, and the Cooperative of Milk Producers of Prata, the watershed selected by Brazil as the Tropileche site. If funding becomes available, Tropileche will start research activities with our colleagues in Brazil during 1998. During the year we ha ve developed proposals to establish adaptive research teams in Central America and Pucallpa, Peru. It appears that sorne funding will become available to allow us to expand research acti vities in these places. TD-2 Technology development New crop and livestock technologies (2.2) Legume-based forage components for dual-purpose cattle in tropical Latin America - Tropileche and Nestle projects (2.2.1; C. E. Lascano, P. Argel, F. Holmann) Main achievements • Documented that with the introduction of Arachis pintoi in association with Brachiaria spp in degraded pastures in forest margins there is an increase in milk yield, less weeds in the pasture and more earthworms in the topsoil where compaction is a problem. • Found that performance of pre-weaned calves grazing Stylosanthes guianensis pastures was similar to that in the traditional system (calf with the dam during part of the day) but that there was one more liter of saleable milk. On-farm evaluation of grass-legume mixtures in the NESTLE Project in Caquetá, Colombia (2.2. \.l ; C. Lascano, G. A. Ruiz, F. Holmann and L. Rivas) The NESTLE Project entered into its third year of activities in dual-purpose cattle farms in the piedmont region of Caquetá, Colombia. A total of 102 ha each of Brachiaria alone and in association with Arachis have been established in 16 farms since the Project began in 1995. Pasture establishment: During 1997, two new accession of Arachis pintoi (CIAT 18744 and 18748) were established in the less fertile soils found in the topography named "mesones" were the commercial cultivar (CIA T 17434) did not perform well. A total of 27 ha each of Brachiaria spp alone and in association with Arachis were established in 5 new farms. The area planted with a mixture of Brachiaria species (B.decumbens, B. humidicola and B. brizantha cv Marandu) in association with the legume ranged from 3 to 10 ha. Measurements in the pastures: The botanical composi tion of pastures in participating farms has been continuously monitored using BOT ANAL and results for the grass/legume pastures are shown in Table l. Table l. Botanical composition of legume-based pastures sown in farms participating in the Nestlé Project (Caquetá, Colombia). Farm Months after Botanical composition (%) Brachiaria Arachis Native e:rass Weeds 1 29 46 31 15 8 2 29 49 40 S 6 3 27 47 35 13 5 4 21 37 27 23 13 5 18 39 16 25 20 6 11 29 15 53 3 7 10 49 20 23 8 TD-3 Technology development The legume content has ranged from 15% in newly established pastures to 40% in two or more year- old pastures. The high proportion of Arachis in the biomass on offer is partly due to good seed germination and to early grazing of the pastures to reduce competition from the grass, as has been recommended to farmers. It was interesting to observe that in pastures with Arachis there were less native grasses and broad leaf weeds that in the grass pastures (Table 2). This was particular! y evident in the older pastures (Farms 1, 2 and 3) and c learly indicates that the introduction of Arachis in association with Brachiaria is contributing to more edible biomass and to less need to control weeds either manually or with herbicides with obvious econonúc and environmental benefits. Table 2. Weed content in pastures sown wi th grass alone and with grass in associa tio n with Arachis pintoi in farms in the Nestle Project (Caquetá, Colombia). Farm 1 2 3 4 5 Months after planti ng 29 29 27 21 18 Brachiaria Native Broad leaf grass weeds (%) 28 6 24 10 19 6 12 14 15 9 Brachiaria + Arachis Na ti ve Broad leaf grass weeds (%) 15 8 5 6 13 5 23 13 25 20 Measurements of the soil fauna: ln one of the farms collaborating in the NESTLE Project we surveyed contrasting pasture for earthworms and results are shown in Table 3. The earthworm count in the Arachis based pasture was 4 to 5 times higher than in the introduced Brachiaria and native grass pastures. It was also interesting to note that the highest proportion of earthworms was in the topsoil (0-5 cm), where most of the compaction in the Caquetá soils occurs. Thus the introduction of Arachis in degraded pastures in forest margins is contributing not only to improve milk production, but also to also to improve soil biological acti vity and probably to reduce soil compaction, which causes major interna! drainage problem in these high rainfall areas. Table 3. Earthworm count in grass and grass-legume pastures in farms in the Nestlé Project (Caquetá, Colombia). Pastures Native grass Brachiaria* Brachiaria/Arachis* *Two-year old pastures. No. of samples 6 6 6 0-5 11 21 69 Depth (cm) 5- 10 10-20 Total (No. earthworrn!m2) 3 3 17 5 26 11 80 Milk yield measuremen ts: An important aim in the NESTLE Project is to monitor núlk yield in grass and grass/legume pastures established in the different farms. However, it has been difficult to get farm owners and workers to routine ly measure milk from cows grazing contrasting pastures. Fortunately in one of the farms it has been possible to make detailed measurements of núlk yield in individual cows and results are summarized in Table 4 . Increments in núlk yie ld dueto Arachis are dependent on the cow genotype and TD-4 Technology development Table 4. Mil k yield of cows grazing Brachiaria spp. with and without Arachis pintoi in one farm of the Nestlé Project, (Caquetá, Colombia). Item No. of Pastures observations Brachiaria spp. Brachiaria spp. + Arachis (Milk yield, kg/cow/d) Cows Holstein crosses 403 6.4 b 6.9 a Other crosses 320 4.5 b 4.7 a Stage of lactation 1/3 250 6.1 b 6.6 a 2/3 220 5.9 b 6.2 a 3/3 253 4.9 5.1 a,b Means in the same row with different letters are different (P<0.05). lactation stage. Cows with Holstein blood grazing Arachis-based pastores produced 0.5 liters more/da y on average than when grazing the grass past:ure. In contrast, cows with a different genotype produced only 0.2 liters more of milk /day when grazing the legume-based pasrure. These results are consistent with previous results from controlled grazing experiments in Quilichao that had shown a s ignificant interaction between cow genotype and responses in milk production due to improved feeding. It is evident that the economic benefits to farmers of the Arachis technology will depend toa great extent on the type of cow used in the farm. In most dual-purpose cattle systems in tropical America cows ha ve very low genetic potential to produce núlk and thus it should be stressed to pasture improvement programs need to be accompanied with genetic improvement of the milking herd. Diffusion of Arachis: It was fe lt from the beginning, that the success of the NESTLE Project would depend on ensuring that a majar area of improved pastores were planted in 1 Oto 15 farms of the region. The strategy also considered that the 10-1 S key farmers ínitially selected to participare in the project would act as promoters of the Arachis technology to surrounding farmers. This in tum would ensure that a mínimum of 100 farmers would be exposed and beco me adopters of the new pasture technology at the end of the 4 year-project. The above strategy did not work given the prevalence of absentee owners in the regían, lack of agriculrure machinery and high cost of the Arachis seed available in the market. Consequently an altem ative diffusion approach of the Arachis technology had to be put in place for the extension phase of the project. The strategy promoted by NESTLE consisted of: 1) Creating a technology Transfer Fund managed by NESTLE 2) Conducting a survey among all milk producers that sell milk to NESTLE to define interest in recuperating degraded pastores using Arachis 3) Contracting the multiplication and purchase of cornme rcial seed of A rachis to fulfill demand among interested producers 4) Contracting tractors for time1y 1and preparation 5) Allowing farmers to pay for the total cost of pasture establishment with milk sold to NESTLE, with no interest on the money. During 1997, 50 farmers were involved in the diffusion program launched by NESTLE and most plantings (3- 1 O halfarm) of Arachis in mixture with Brachiaria were successful. The program will TD-5 Technology development continue in 1998 with emphasis on the introduction of new genotypes of Arachis pintoi (CIA T 18744 and 18748) to the less fertile soils in "mesones" which cover over 60% of the piedmont region in Caquetá. Seed multiplication o f the new Arachis ecotypes to support the Nestle Project was contracted by CIA T with SEFO-SAM in Bolivia. During 1997 two Newsletters dealing with the activities of the on-farm activities of the NESTLE Project and recommendations on pasture establishment and management were distributed among technical assistants and producers in the region. In addition, interested fanners were invited to visit Arachis pastures in different fanns participating in the project. Adoption study: During May of 1997 a survey instrument was developed to analyze the demand for Arachis pintoi in Caquetá. Colombia. This study is a follow-up of an earlier ex-ante analysis executed during 1996. The survey formal was tested and person nel were trained in its proper use. A total of 226 producers were surveyed ( 174 randomly selected Nestle's milk producers plus 52 early adopters) between the months of June and July. The information was coded and analyzed during the months of August through November, and it is expected that a first draft of the study wi ll be ready by December of 1997. On-farm evaluation of grasses and legumes in Peru, Costa Rica, Nicaragua and Honduras - Tropileche Project (2.2. 1.2; C. Lascano, P Argel, F. Holmann , K. Reacegui and NARS partners) The on-farm evaluation of improved grasses and legumes in Peru, Costa Rica and Nicaragua form pan of the research activities in TROPILECHE. The on-farm activities include: (a) improved grass and grass- legume pastures for milking cows; (b) dry season forage-based energy and protein supplements for milking cows, and (e) improved feeding/management praccices for pre-weaned calves. In the following sections we summarize the main accivities carried out in the two benchmark sites of TROPILECHE (Pucallpa, Peru-forest margins and Central Pacific region, Costa Rica- subhumid hillsides) and in Nicaragua and Honduras, which are new partners in TROPILECHE. Peru (Partners: M. de la Torre, Cesar Reyes-IVIT A, and J . Vela-IIAP) Establishment of forage alternatives: During 1996 and 1997 acti vi ties were concentrated in the establishment of improved grasses (Brachiaria spp.) associated with Arachis pimoi on 5 farms. Pastures were established after secondary forest, after nati ve grass, and after degraded grass. Dueto erratic weather conditions and poor seed germination, Arachis had to be replanted. It is expected that by the end of 1997 the associated grass-legume paddocks will be fully established to begin recording animal performance ( i.e. milk production) as well as forage attributes (biomass production. feed quality) during 1998. TD-6 Technology development Costa Rica (Partners: C. Hidalgo and M. Lobo -MAG, F. Romero and J . Gonzalez -ECAG, M. Ibrahim and M. H. Franco -CATIE) Establishment of forage alternatives : Fodder banks of Craty/ia argentea and sugar cane were established during 1996 for dry season feeding in 3 of the 7 fanns that are participating in TROPILECHE. Two other fanns were selected to establish Stylosanthes guianensis (CIA T 184) pastures for pre-weaned calves, and three fanns to establish B. brizantha cv La Libertad associated with Arachis pintoi for the milking herd. Fodder banks of Cratylia argentea and sugarcane are well estab1ished in three fanns and evaluations will start in the dry season next year. Pastures of Arachis pintoi in association with B. bryzantha cv La Libe rtad are well established in one fann and at ECAG. In the remaining two fanns the introduction of Arachis in pastures with B. dictyoneura cv Llanero was to be carried out during the rain y season of 1997, but lack of rainfall due to "Él Niño" ha ve delayed the plantings. Thus, it is expected that by the end of 1997, fanns collaborating with TROPILECHE in Costa Rica will have sorne of the new forage a ltematives fu lly established and ready to generate results during 1998 and subsequent years. Management and utilization of Cratylia fodder banks: Banks of the shrub C. argentea are well established at ECAG and in three fanns located in Orotina, Miramar and Esparza. A uniformity cut was carried out in al! the forage banks at the end of September 1997 in order to ha ve vigorous regrowth for the dry period that commences in November 1997 and ends in may 1998. In two fanns the available foliage fo llowing the uniformity cut was made into silage in mixture with king grass and mature material of the grass Jaragua (H. rufa). This is an additional use of C. argentea that was not contemplated initially at the beginning of the TROPILECHE project, and that will he lp to reduce the severe lack of forage experienced in the area during the 6 months dry period. In another very hilly farrn close to Esparza and not presently with the TROPILECHE project, C. argentea has been under evaluation for more than four years (MAG/ODA Reforestation Project). This shrub perfonned better (more DM and more tolerance to drought) than the shrubs L. leucocepha/a CIA T 17263, Morera (Morus sp.) and Guásimo (Guazuma um/eifolia). The fanner has expanded the area of Cratylia and utilizes it to supplement either sugar cane or the Jaragua grass in both the dry and the wet season of the year. Milk production of the cows ( 12 cows presently) ranges from around 3.5 to 5.0 1/cow/day, the farmer states that the soil cover of the paddocks has improved s ince the cows are maintained half of the day in the cattle yard feeding on Cratylia. It is interesting to comment that now that the fanner has deve loped a more efficient system to feed the cows is actively reforesting the more steeped areas (not appropriate for grazing) of the fann. The Centro Agronómico Tropical de Investigación y Enseñanza (CATIE), a member of the Tropileche consortium, has taken advantage of this farm to conduct postgraduate studies on C. argentea cattle intake. Table 5 shows that both wilting and spraying Cratylia and Jaragua grass with molasses increased the DM intake by milking cows. Similar effect was recorded for total nitrogen and metabo lic and digestible energy intake. TD-7 Technology development Table 5. The effect of wilting and the addition of molasses o n DM intake (kg/1 00 Kg L W) of Cratylia argenrea and the grass H. rufa (Information supplied by M. H. Franco, CATIE 1997). Variable Craf)·lia Without molasses With molasses Mean Grass Without molasses With molasses Mean TOTAL Without molasses Wilh molasses Mean Intake (kg DM/100 kg LW) Fresh 0.28 1 0.395 0.338 b 2.075 2.404 2.239 b 2.356 2.799 2.578 b Wilted 0.372 0.432 0.402 a 2.346 2.426 2.386 a 2.719 2.858 2.788 a *Means fo llo wed by the same letter are statistically different, Tukey's test (P<0.05 ) Nicaragua (Partners: T. Fariñas, R. Soza) Mean 0.327 b* 0.4 13 a 2.2 11 b 2.4 15 a 2.538 b 2.828 a Establishment of forage alternatives: The Proyecto de Desarrollo Lechero funded by the World Food Program of the United Nations, is the T ropileche partner for the evaluation of on-farm research components in the Jocalities of Matagalpa and Muy Muy in Nicaragua. Four farms have e ither planted Brachiaria spp./ Arachis pintoi, banks of C. argentea for dry season supple mentation or Stylosanthes guianensis paddocks for pre-weaned cal ves. Replanting during the growing season of 1997 has been impaired because of bad weather conditions dueto El Niño. On-farm pasture evaluation: In one TROPILECHE farm in Esquipulas the farme r wanted to compare mjlk production in his traditional pasture (Hyparrhenia rufa ) against B. briz.antha cv. La Libertad during the rniddle of the dry season. Twenty-two milkin g cows grazed during 27 days a 5-ha paddock of H. rufa, then the same cows grazed during the next 27 days in a 5-ha paddock of B. brizanrha alone. Milk production per cow in B. brizantha increased by 10% (0.72 kg/cow/d) relative to H. rufa. In both pastures, milking cows were offered the same amoum of supplemem, consisting of 0 .5 kg molasses/cow/d, 4.5 kg/cow/d of pou ltry manure, and 1 kg/cow/d of Dolichos Lab Lab. It is expected that the difference in milk production will be larger once Arachis pintoi is associated with B. briz.antha compared to the traditional grass pastures utilized in this region. Honduras ( Partners-G. Galo and C. Burgos, DICTA) Research components of the Tropi leche project are under implementation through the Proyecto de Pastos y Forrajes (PRO PASTO) of the Dirección de Ciencia y Tecnología (DICTA) of Honduras . Following a letter of agreement between CIA T and DICTA a consultancy was made to this country and the foll owing research components recommended: TD-8 Technology deve/opment 1) Evaluation of milk production of dual purpose cows during the dry season fed with sugar cane supplemented with C. argentea. 2) Evaluation of n ilk production of dual purpose cows in grass/legume pastures of B. decumbens, B. dictyoneura/A. pintoi, and 3) Evaluation of liveweight gains of pre-weaned calves on Stylosanthes guianensis pastures. These research components are being implemented in 12 farms disseminated along the humid and subhumid tropics of Honduras in the localities of Catacamas, Juticalpa, Guaymaca, Danlí and San Francisco de Becerra. Planting of the components was initiated in July L 997. On-fann evaluation of legumes for pre weaned calves in Peru and Nicaragua (2.2.1 .3; C. Lascano, F. Holmann and NARS partners) A common practice in dual-purpose cattle systems is for pre-weaned calves to spend the day grazing with the dams , and then separated from the cows during the aftemoon and placed in a corral. Under this traditional system there is high calf mortality and low weaning weights. Peru: In five farms collaborating with TROPILECHE we are evaluating the effect of Stylosanthes guianensis (CIAT 184) on weight gains of pre-weaned calves. Initial results from one farm indicate that with 2.5 to 3.5 month old calves, the highest weight gains were obtained with Style grazing + concentrates +residual milk (Table 6). However, it was observed that weight gains under the traditional system (calves with dam all day long) were similar to those obtained with an improved system (calves drinking residual milk after milking + Style grazing all day long without dam), but that saleable milk increased by 1 kg/cow/day (22% ). Thus, producers could benefit from this technology because it allows an improvement in income dueto more milk without affecting weight gains of calves. The use of Style pastures in these systems could also contribute to the reduction of slash and bum of secondary forest for the establishment of annual crops, since the areas established with Style could form part of a rotation with crops. Table 6. Performance of pre-weaned calves and mil k production in different feeding/ managemenl systems (Farm in Pucallpa, Perú). Variable Treatment 1 (Stylosanthes only) Weight gain of calves (g/d) Saleable rrúlk (1/cow/d) Milk consumed by calves (1/d) *Standard deviations are in parenthesis. 543 (67)* 4.73 (0.57) 0.73 (0.09) Treatment 2 (Stylosanthes +concentrate) 763 (3 1) 4.7 1 (0.95) 0.89 (0. 34) Treatment 3 (control, dam w/calf at foot) 527 (69) 3.86 (0.74) 0.68 (0.17) Nicaragua: In a TROPll...ECHE farm in Muy-Muy, Stylo was used to feed 24 pre-weaned calves during the dry season. Calves were separated by phenotype in four breed groups: Brahman (8), Brahman!Holstein (6), Brahman/Brown Swiss (3), and 7 Ho lsteín/Reyna calves (Reyna is a local syntethic breed). Calves grazed the Stylo pasture for 30 days, and then the farmer rested the paddock to allow the Style to regrow in a rder to collect seed to expand the area during the rain y season in 1997. Results shown in Table 7 indicate that calves gained weight atan average of 300 g/d during the 30-day grazing period, ranging from 227 g/d for Brown Swiss/Brahman crossbred to 323 g/d for Holstein!Brahman calves. In the traditional management system calves usually loase about 20% of their body weight by the end o f the dry season TD-9 Technology developmenc These results indicate that Style is an excellent altemative either to produce more milk in the bucket such as the case in Pucallpa, or in the case .in Nicaragua, for improving performance of pre-weaned calves in dual-purpose cattle systems. Table 7. Weight gains o f pre-weaned calves of different phenotypic groups grazing Srylosanthes guianensis (CIAT 184) during a dry period by (Farm in Muy-Muy, Nicaragua). Phenotype Initial weight Final weight gain Daily gain Total Average (k) (kg) (kg) (g/d) Brahman (8)* 134.6 143.5 8.75 292 (30 .2) (32.7) Brown Swiss/ Brahman (3) 104.3 11 2.0 7.70 257 ( 12.2) ( 14.8) Holstein/ Brahman (6) 132.8 142.5 9.70 323 (47.6) (47.0) Ho lstein/ Reyna (7) 124.7 133.9 9.20 307 (32. 1) (33.7) Overall Mean 127.5 136.5 9.00 300 **Figures in parenthesis are standard deviation of the mean TD-10 Technology development Forage-based systems for dry season supplementation of dual-purpose cattle in tropical Latin America- Tropileche Project (2.2.2) Main achievements • Demonstrated that there was a 25% increase in milk yie ld with the addition of Cratylia argentea in forage-based supplement when forage availability in the pasture was limiting. • Demonstrated that increasing proportions of Cratylia argentea in a forage -based supplement resulted in daily milk yield increases of up to 24% with cows that had the genetic potential to produce milk Evaluation of shrub legumes as dry season supplements for milking cows (2.2.2. 1; P. A vil a, C.Lascano and F. Holmann) An objective of the work carried out in the Quilichao Station as part of TROPILECHE research agenda is to determine the effect on rnilk yie ld of supplementing shrub legumes to cow of different genetic background grazing pastures with contras ting quantity and quality of forage on offer. The forage supplements to carry out this work are King grass and sugarcane as energy sources and the shrub legume Cratylia argentea as prote in source. given that these are the feed resources being tested in fanns at benchmark sites in TROPILECHE. During 1997 an experiment was designed to study milk yield responses to supplements based on King grass alone or in combination with Cratylia argentea fed to cows grazing Brachiaria decumbens pastures wi th contrasting forage on offer. Results shown in Table 8 indicare that there was no effect of including Cratylia in combination with King grass when forage availability in the pasture was not Iirniting (2 A U/ha). However, when forage availability became lirniting ( 4 A U/ha), the inclusion of the shrub legume in combination with the grass resulted in a 25% increase in milk yie ld. Table 8. Effect of forage-based supplements on mil k yield of Holstein crossbreed grazing Brachiaria decumbens at two stocking rates (CIA T's experi ment station in Quil ichao, Colo mbia). Treatment Low Stockin~ Rate (2 A U/ha) - Cratylia + Cratylia Mean Hi~h Stocking Rate (4 A U/ha) - Cratylía + Cratylia Intake of Supplement (kg DM/cow/d) 2.4 3.5 3.0b 4.3 4.5 Mean 4.4 a Milk yield (kglcow/d) 8.8 a 7.9 a,b 6.9 b 8.6 a a,b Means for each response variable wi th d ifferent le tters in the same column are di fferent (P<0.05) - Cratylia = 1.5% B W DM king grass +Cratylia = 1.0% BW DM king grass + 0.5% BW DM Cratylia These results suggest that there will be little benefit of including legume in forage supplements when forage availability in th e pastures is not limiting. In contrast, the inclusion of the legume is justified when forage availabil ity is limited, which is the case under dry season conditions or in situations where TD- 11 Technology development opportunities for grazing in intensive systems are lirnited. In another experiment we studied the effect of increasing the proportion of Cratylia in combination with sugarcane. Results presented in Table 9 show a strong interaction of animal genotype in rnilk yield response to forage-based supplements. With low grade Zebu crosses there was no response to the inclusion of different levels of Cratylia in the diet. In contrast. wi th the high grade Holstein crosses there was a linear response to increasing levels of Cratylia in combination with sugarcane. With the combination of 25% sugarcane - 75% Cratylia there was a 24% ( 1.6 kg rni lkld) increase in mil k yie ld relative to sugarcane a lone . Table 9. Effect of feeding increasing levels of Cratylia argentea (Cratylia) in combination with sugarcane (SC) on mil k yield of high grade Holstei n and low grade Zebu cows grazing Brachiaria decumbens (CIA T' s experiment station in Quil ichao, Colombia). Treatment High grade Holstein 1 - lOO% se - 75% SC- 25% Cratylia - 50% SC- 50% Cratylia - 25% SC -75% Cratylia Mean Low grade Zebu1 Intake of Forage Supplement (kg DM/cow/d) 5.0 4.5 4.0 5.1 4.6 a -lOO% se 3.8 - 75% SC- 25% Cratylia 3.1 - 50% SC - 50% Cratylia 3.5 - 25% SC - 75% Cratylia 4.3 Mean 3.7 b Milk Yield (kg cow/d) 6.6 b,c 7.4 a,b 7.8 a 8.2 a 5.9 c,d 5.8 c,d 5.5 d 5.4 d a, b, e, d means for each response variable with different letters in the same column are d ifferent (p<0.05) 1 Stocking Rate = 2.7 AU/ha These results confi rm previous findings and indicate that the genetic potential of cows to produce rni lk could be an important determinant for adoption by farmers of legume-based feeding systems in dual- purpose cattle systems in LAC. TD-1 2 Technology development Improved livestock feed supplies for smallholder farms in Southeast Asia - Forages for Smallholders Project (2.2.3; W. Stür, CIA T ; Peter Home, CSIRO; Francisco Gabunada, FSP-Philippines; Phonepaseuth Phengsavanh, FSP-Laos, Maimunah Tuhulele, DGLS, Indonesia; Ed Magboo, PCARRD, Philippines; Viengsavanh Phimphachanhvongsod, DLF, Laos, Le Hoa Binh, NIAH, Vietnam) Main achievements • Established 17 sites where the FSP is developing forage technologies with fanners in Indonesia, Laos, Philippines and Vietnam Summary of activities The following forage technologies are being evaluated by farrners at FSP s ites in Indonesia, Laos, Philippines and Vietnam: • Intensively managed plots (grasses) • Protein banks (herbaceous and tree legumes) • Tree legumes in fence lines • Contour hedgerows (grasses, herbaceous and tree legumes) • Legumes in crop fallows or grown in association with crops (herbaceous legumes) • Cover crops (herbaceous legumes) • Grasses and legumes for grazing (mono culture or in association) A summary matrix of forage technologies by site is presented in Table l . At most sites farrners chose to evaluate more than one technology. lntensively managed grass plots near houses or animal sheds are being evaluated at almost all sites. Often farmers identified a \ack of labour as a limitation and they see intensively managed plots asan attractive way of reducing the demand on labour of keeping livestock. In many cases farrners intend to use these intensively managed plots only at specific times. Examples are days when they ha veto go to the market, sorne farnily members are sick, or during periods of peak labour demand by other agricultura) activities. The range of technologies tested by farmers increases as they become familiar with forages and they see more opportunities on their farrns. A list of the main forage species evaluated by farmers is presented in Table 2. The most promising species, so far, are mentioned in the text describing activities at each site. TD-13 Technology development Table l . Forage technologies evaluated by farmers SITES FORAGETECHNOLOGY "' ~ o ¡.., o u Indonesia · ' -·-···----·--·-------------------'----------·---·-----·---_..__._ .. __ Gorontalo, North Sulawesi · t/ 1---·-·------------·--------·--- ---·-__;__. _______ ,_, ____ _ ___ Makrom~m, East Kaliman.::..:t.::.:.a:..:..n _____ ____ v _________ ,_·-·---·-.. ·-·-·--·--""'-------- Marenu,_South T~panuli _ t/ t/ Pulau Gambar, ~~rth Su_111atra -·-···--------~""'--- .;--------~-- .. T ·-----~-~ee, ~~~h--·----·--··- t/ t/ t/ ·-----·--·----·---------- Sepaku II, East Kalimantan t/ ! ., ., La os --=·---------··-·--------·-·---- ----+--~------·--··-----·--·--- -----f _ Luang P~_abang_·--------·-·-- ---~t/--- --: __ e( ___ . __ _.._v ._·---··-···--- Xieng Khouang t/ t/ · Phil~ppine~ ... -----·-----···----·--·--!---- ___ ¡.._._ ____ .. ____ , ... __ _, __ ......;.... ___ _ --~agayaf!_ de Oro, Mindanao t/ t/ t/ '---M' lang_'!_nd Car:t)!en, North Cotabato ----'~.,-----·----·--·----·----+------·--· -;;-~¡ --- Davao, Mindanao t/ r----·------ __________ _ ...;..._ __ . ______ , ____ _ . ___ _ - .. g':!ba, ~~~------.. --------·---+----""' , t/ ___ .,_·_._.,_-_-=~===·-- ~~alitbog! . Bukid_f!_ o u l v In Indonesia, the FSP work is coord inated by Mrs. Maimunah Tuhulele of the Directorate Genera l of Livestock Services (DGLS) overall and locall y by staff of provincial or district Livestock Services Offices. Exceptions are the sites in Marenu and Pulau Gambar which are supervised by Tatang Ibrahim of the Assessment Center for Agricultura! Technologies for North Sumatra in conjunction with the local Livestock Services Offices. TD-15 Technology deve/opment Gorontalo, North Sulawesi In Gorontalo, on-farm activities are carried out in two villages, Reksonogoro and Molalahu in subdistrict Timbawa. Local collaborators are Susilan and Idrus ofthe local Livestock Services office. Farmers are particularly interested in grasses and legumes for grazing. Following a regional evaluation of forages in 1996, promising forages were distributed to two farmer groups for evaluation in 1997. In Molalahu, these were planted by a local farmer group, while individual farmers planted forages in Reksonogoro. Emphasis has been placed on involving more farmers in the evaluation and to extend the range of forage technologies tested . Makroman, East Ka/imantan Activities at this site are reported in section 2.2.4 Marenu, Tapanuli Selatan The FSP commenced development of forage technologies with transmigrant farmers whose livelihood is based on sheep production on a 1 ha infertile upland area in late 1996. About 30 farmers are evaluating intensively managed cut plots to supplement existing feed resources. Sorne farmers are also evaluating tree legumes as protein banks. Promising species include Paspalum atratum and P. guenoarum. Pulau Gambar, North Sumatra The project co llaborates with a women's sheep raisers cooperative in a lowland rice area. About 15 farmers are evaluating different provenances of Gliricidia sepium and other tree legumes, grown as protein banks and fence lines. These are grown to supplement naturally occurring vegetation along fields and roadsides, and crop residues. Saree, Aceh The province of Aceh has large areas of hilly and mountainous areas which traditionally ha ve been used for grazing of cattle. Declining access has resulted in farmers forming cattle cooperatives. These farmer groups can then apply to the government for the granting of grazing leases. The FSP is working wirh one such farmer group in Aceh Besar to develop forage technologies to increase animal production from their grazing area. Farmers are evaluating intensively managed cut plots, protein banks and grasses and legumes for grazing. Promising species include Panicum maximum, Brachiaria spp .. Stylosanthes guianensis, Desmodium rensonii and Gliricidia sepium. Sepaku 11, East Ka/imantan Farmers in this area are transmigrants from Java who arrived in this Imperara cylindrica grassland area the early 1970s. Wild pig and lmperata cylindrica make cropping of the vast upland areas almost impossible and many farmers have resorted to cattle production (mainly breeding) for their livelihood. The FSP works with farmers who are organised in a livestock production group. In 1995-96, the group evaluated a range of forage species on a common area. In 1996-97, farmers planted small are as of forages as intensively managed cut plots and grasses and legumes for grazing on their own land. Many of these farmers are now expanding their areas. A few farmers are also evaluating oversowing of Stylosanthes guianensis CIA T 184 into Imperara grassland with the aim of suppressing the growth of Imperara and to improve the quality of the pasture. Promising species include Stylosanthes guianensis CIA T 184, Brachiaria brizantha, B. decumbens, B. humidicola andAndropogon gayanus. La os In Laos, on-farm evaluation and development of forage technologies commenced at two locations in 1997, under the supervision of Viengsavanh Phimphachanhvongsod and Phonepaseuth Phengsavanh of the Lao Department of Livestock and Fisheries. TD-16 Technology development Luang Phabang On farrn evaluations of forage technologies commenced with 58 individual farrners and 7 groups farrners in eleven villages (Houay Hia, Kiew Nya, Nam Awk Hu, Sen Oudom, Kieuw Talun Nyai. Nasai Chaleun, Tha Po, Chong, Nong Phu, Na Ang and Don Xai) in June 1997. Initially, the main interest of the farrners has been in intensively managed plots for feeding animals at critica! times of year. The most promising species so far have been Stylosanthes guianensis CIAT 184, Panicum ma.ximum T-58, Brachiaria decumbens cv Basilisk and Brachiaria brizantha cv Marandu. The focus for next years ' activities will be on the remoter villages where 1ivestock assume a much greater role in sustaining village live1ihoods. Xieng Khouang On farrn evaluations of forage technologies commenced with 17 farrners in four vi llages (Ta, Phousi, Sang, Piang Louang, Khang Pa Nyian and Nyot Kha) in June 1997. The work is being coordinated by Hong Thong Pimmasan in collaboration with the GtZ Nam Ngum Watershed Development project. Farrners were initially interested in intensively managed plots for feeding animals at critica! times of year, forages for fallow improvement and forages for weed control. The most prom.ising species so far have been Stylosanthes guianensis CIA T 184, Panicum maximum T -58 , Brachiaria decumbens cv Basilisk and Brachiaria brizantha cv Marandu. One cross visit of interested farrners from villages neighbouring Ta village was facilitated. The project will expand the number of farmers in those villages where the success of sorne farmers has generated demand from others for seed . Farrners from nearby vi llages with similar problems will also be invited, after PD, to participate. Philippines In the Philippines, the FSP is coordinated by the Philippines Council for Agriculture and Natural Resources Research and Development (PCARRD). The coordinator is Eduedo Magboo of the Livestock Research Division. At FPR sites, the project collaborates with li vestock officers of the local govemment, uni versities and National Agricultura! Organisations such as the Philippine Coconut Authority and the Philippine Carabao Centers. Cagayan de Oro The area is an extensive, hilly upland area with com being the most commonly grown food crop. Many farrners have 1 or more cattle or buffaloes for land preparation and to generate cash income. The local govemment is assisting poor farrners to acquire livestock through a livestock dispersa! program. The project works with a farmer group which initially planted a range of forages in a common area before individual farrners proceeded to plant forages on their own land. Forage technologies be ing tested include intensively managed plots, hedgerows and legumes for improved fallows. Promising species include Paspalum atratum, Pennisetum hybrids and Stylosanthes guianensis. M'lang and Carmen, North Cotabatu Farrners in two villages, one an upland area and the other a rainfed lowland area, are evaluating forages to supplement existing feed resources . Activities started in late 1996 and no feedback is available so far. Davao, Mindanao The project started to work with Small Coconut Fanners' Cooperatives whose members also belong a da iry cooperative near Davao City in 1997. Increasing pressure on existing feed resources and increasing prices of concentrate supplements are forcing farrners to look for altemative feed supplies for their dairy cows. Farrners are currently in the process of planting a range of forage species in communaJ and individual nurseries to propagate promising species. This work is backed up by a regional evaluation of forages under coconuts by the Philippine Coconut Authority in Davao. TD-17 Technology developmenr Cuba, Cebu Farmers in Guba are located in mountainous areas behi nd Cebu City. They intensively crop sometimes terraced areas producing grapes, flowers and vegetables for the Cebu Market. Most farrners also ha ve l-3 cattle, buffaloes or goats that are kept e ither in a stall or are tethered on their land. The manure from animal production is an important part of the farming system. Forages are grown along contour hedges, intercropped with crops oras cover crops among fruit trees and grapes. The FSP collaborates with farrners belonging to an NGO, the Mag-uugmad Foundation, to develop forage technologies for this intensive system. Promising species include Arachis pintoi, Setaria sphacelara var. splendida, dwarf napier and a variety of tree Jegumes. Malitbog, Bukidnon The FSP works with farrner groups in this extensive, hilly upland area to develop forage technologies fo r intensively managed plots, protein banks and contour hedgerows. Locally, activities are organised by the agricultura) office of the municipality. In 1997, farrners started to plant forages on their own land to individually evaluare forages they selected from a regional evaluation site ( 1995-96) in the area. Matalom, Leyre In Matalom, the FSP collaborates with the Visayan State College of Agriculture (ViSCA). The area is a mixture of extensive and intensive upland fanns. Animals play an important role for draught power and as a source of readily-available cas h when needed by the farnily. On-farrn evaluation of forages follows a regional evaluation of forages in Matalom. Forage technologies tested by farrners include protein banks, intensively managed plots and contour hedgerows. Promising species include Stylosanthes guianensis CIA T 184, Penniserum hybrids and Paspalum atratum. Vietnam On farrn evaluation of forages commenced at three broad locations in Vietnam during 1997, under the overall supervision of Le Hoa Binh from the National Institute of Animal Husbandry. Towards the end of 1997, he will be joined by two new project staff; Le Van An (seconded from the Un iversity of Agricul ture and Forestry, Hue) and Bui Xuan An (seconded from the University of Agriculture and Forestry, Ho C hi Minh City). Both have experience in Farrner Participatory Research methodologies. M 'Drak On farrn evaluation of forages commenced with 30 farmers in Chu' Caroa commune, M'Drak, Daklak province in May 1997. The area is dominated by Imperara grassland. The main interest of the farrners was in grasses and legumes for reclamation of Imperara areas and for providing dry-season supplementation for grazing cattle. The most promising species so far have been Stylosanthes guianensis CIA T 184 and Brachiaria decumbens cv Basilisk. In response to considerable demand from farrners for help with reclamation of Imperara areas, the future focus ofthe project will be both on the Chu' Caroa commune andan area of Imperara to the southwest of M'Drak. In addition to supervising this work, Truong Tan K.hanh from Tay Nguyen University in Buon Ma Thuot, has been conducting an evaluation of forage tree species with potential for dry season feeding. Viernam-Swedish project sites In collaboration with Bui The Hung of the Vietnam-Swedish Mountain Region Development Project, on farrn evaluations of forage technologies commenced with 50 farmers in fi ve northem provinces (Lao Cai, Yen Bai, Ha Giang, Tuyen Quang and Phu To). The most promising species so far have been Panicum ma.ximum T -58, Brachiaria decumbens cv Basi lisk, Styiosanthes guianensis CIA T 184 and Brachiaria brizantha cv Marandu. The most common concem of farmeJS was to find green feed for fish and to supplement pigs. The future focus of the project in this area will be on farrners in Ha Giang and Tu yen TD-18 Technology development Quang, with expansion next year into areas where World Vision is working with the Hmong ethnic minority, who raise cattle in remate highland areas. Xuan Loe On farm evaluations of forage and tree legume technologies eommenced with 8 farmers at Xuan Loe, near Hue. The main coneern of the farmers was to find forages for year-round feeding as their traditional grazing resourees have disappeared as a result of expanded eropping and forestry aetivities. The most promising species so far have been Panicum ma.ximum T-58, Stylosanthes guianensis CIAT 184 and severallines of Brachiaría brizantha. With staff of the University of Hue (Le Due Ngoan, Nguyen Thi Hoa Ly and Ho Trung Thong), the FSP will expand the number of farmers evaluating forage technologies at Xuan Loe and commence on-farm work with farmers at Hong Ha commune in nearby A Luoi district. TD- 19 Technology development Legumes and grasses incorporated by farmers for fallow improvement, green manure, soil cover, erosion control and disease control Activities of the Forages for Smallholders Project • Asia (2.2.4; Ir. Ibrahim, Livestock Services of East Kalimantan, Indonesia; Maimunah Tuhulele, DGLS, Indonesia; Wemer Stür, CIA T) Main achievements • Establishment of FPR si tes where forages are used by farmers for improved natural resource management as well as improving their feed supply (multiple benefits) Forage technologies often have more than one benefit and it is difficult and, in many cases impossible, to separare forages grown for NRM purposes from those grown for improved feed suppl y. Fanners growing, for example, contour hedgerows are looking to obtain feed for their animals in the same way as intensively-managed plots as well as limiting erosion. The relative importance of the two benefits varíes from farm to farm, but no farmers at FSP si tes are growing contour hedges purely for erosion control. Exceptions, to sorne degree only. may be legumes grown for soil improve ment or weed control, and tree legumes grown as fence lines. Soil improvement. weed control and the control of animals may be the overriding motivation for farmers growi ng these forage technologies. However, farmers at FSP sítes, use the forage legumes al so for animal feed . It would be difficult to separate these benefits. Many of the si tes, reported in section 2.2.3, ha ve both an animal feeding component and a NRM component. The only s ite reported here is the village ofMakroman, where the entry point for the FSP was legumes grown for fal low improvement and weed control. However, even at thi s s ite the relative benefits of soil improvement and weed control on one s ide and li~stock feed on the other has become increasingly blurred. Makroman, East Ka/imantan, Indonesia This village is a trans migration area where farmers fromJava were settled in 1974. The area consists of a mixture of hi lly upland and rainfed lowland areas. Land holdings vary from 2-3 ha per family. Most families own 1-2 ha o f lowland and 1 ha of upland. In most years farmer can grow 2 rice crops/year. Apart from rice, farmers grow maize, cassava, sweet potatoes, baby com and peanuts. on thei r upland areas. Farmers also maintain intensive home gardens witb fruit trees and vegetables around their houses. Not all upland areas are planted with crops. Often only 02-0.4 ha are planted to crops while the remaining area is dominated by Imperara cylindrica. Son.! farmers continuously crop the same area with inputs of manure whi le others practice a crop-fallow system. The main limitation to planting a ll of their upland area with crops is labour for land preparation and weeding, and declining soil fertility with continuous cropping. Many families raise L -3 cattle and so me al so raise 5- l O goats. Cattle tend to be kept in stalls or tethered near houses during the night and moming and are tethered in the fanner's own fields in the afternoon. Additionally farmers cut naturally occurring grasses (induding Imperara cylindrica) anywhere in the village areas. Farmers spend 2-4 hours/day cut and carry feed for their animals. Rice s traw or stubbles after harvesting belong to the owner and there is no free ~zing of animals. Goats are kept in pens and farmers feed grasses, lea ves of cassava, jackfruit. giliriciia and banana leaves/stem during the dry season. A PD was organized with members of a farmer group caBed Maju (22 families awning 45 cattle and about 100 goats). It was also attended by severa! farmers frommother farmer group called Sidodadi. The latter TD-20 Tecluwlogy development specialize in goat production and many of their members ha ve no access to lowland areas. The most important problem identified by farmers was lmperata cylindrica invading upland areas . Farmers mentioned severa! other problems, none related to animal feeding. The overwhelming problem was the presence of Imperara requiring enormous labour inputs for upland crops. Coping mechanism were to grow only small areas of upland crops, and to concentrare more on animal production. Now farmers are evaluating growing herbaceous legumes in association with maize and cassava. The legumes are cut regularly during the crop growing season to limit competition with the food crop and the cuttings are fed to their animals. Farmer claim that the legumes reduce their weeding requirements substantially and that the crop yield are as high or higher than those areas grown in monoculture. After harvesting of the food crop, the area is left uncropped for sorne time and the forage legumes are a llowed to grow. These legumes successfull y suppress the growth of Imperara cylindrica during the fallow period, making land preparation for subsequent food crops less labour demanding. The preferred herbaceous legume is Cenrrosema pubescens ClA T 15 160, although sorne farmers are also testing Stylosanthes guianensis ClA T 184. Communally. the fanners are now conducting an experiment eva luating the effect of legume fallow on subsequent crop yield. Emphas is for 1997-98 has been placed on supplying more farmers with seed of he rbaceous legumes for more extensive on-farm testing. lnterestingly, there are now many farmers who also evaluate forages in intensively-managed plots for supplementary feeding .. Although feeding was not seen as a problem during the PD, the entry point for stimulati ng interest among farmers was the reduction in labour requirements for co llecting feed for animals. Legumes for use as green manure or fallow improvement - Cauca, Colombia (2.2.4; L.H. Franco, P.C. Ke rridge a nd G. Sturni-U ni v. Firenze) Main achievement • Canavalia brasiliensis and Mucuna pruriens were shown to be potentially useful as g reen cover c rops in Cauca Introduction Farmers in the hillsides of Cauca (Colombia) utilize a rotational system in which a period of cropping is followed by a natural fallow which may or may not be util ized for grazing. The incorporation of legumes into the soil as green manure can bene fit crops in the zone through increasing available soil nitrogen for the fo llowing crop. An earlier study s howed that the amount of labile N (NOJ + NfL) was doubled in comparison to the amount unde r a natural fallow (Proc. XVIll lnt. Grass. Congr. 16:25-26). Maize yield was increased following fallow improvement with Centrosema macrocarpum and Cajanus cajan in comparison with the natural fall ow. Following this initia l demonstration. an experiment was conducted to identify other legumes that would establish rapidly and minimize competitíon from weeds. TD-2 1 Technology development Locality The experiment was conducted in Pita (, Cauca, Colombia (2° 50' N , 76° 25 ' E, a 1350 m.a.s.l.). The site has a moderare to steep slope and the soil had had been highly eroded and dissected during a cropping period of maize. beans and cassava. At the time of land preparation, it had been left in a fallow condition for severa! years. The precipita tion is bi-modaJ, with and average annual rainfall of 1700 mm and dry periods between June and September and January and M are h. The soil is a clay (54-66%) with pH, 4 .1 , OM, 6.0%, P, 1.2 ppm (Bray), and exchangeable bases (meq/ lOOg) - Ca, 0.37, Mg, 0.35 and K, 0. 11 , wi th Al saturation of 80%. Treatments: A. Legumes: l . Cajanus ca jan CrA T 913 2. Calopogonium mucunoides CIA T20709 3. Canavalia brasiliensis CIA T 17009 4. Crotalaria juncea CIA T 21 709 5. Mucuna pruriens CIAT 9349 6. Pueraria phaseoloides CIA T 7 182 7. Caupi - Verde Brasil 8. Stylosanthes guianensis CIA T 1 1844 9. Coctel : C. mucronata 20553, C. schideanum 15727 , C. macrocarpum 57 13 y S. guianensis 11833 1 O. Testigo : Barbecho Natural Ferti lizer: Phosphorus was applied ar 40 kg/ha P as ei ther single superphosphate or chicken manure (with analysis -3.08% N, 3. 17% P, 2.95% K, 3.64% Ca y 0.7 1% Mg) There were three replications in a randomized block design. The legumes were planted at the end of October 1996 in rows . Results Heavy rain immediately after planting caused sorne loss of sorne seed on the steeper s lopes. The following table gives information on the emergence and density of seedlings. Table l . Emergence densi ty of plants at establishment. Species No. plants /m- Carravalia brasilierrsis 17009 3.2 Cowpea- verde brasil 1.5 Calopogonium mucurroides 20709 J 3. 1 Mucuna pruriens 9349 1.7 Stylosanthes guianensis 11 844 6. 7 Cajanus cajan 9 13 3.2 Pueraria phaseoloides 7182 4. 1 Crotalaria j uncea 21709 4.8 Mixture* * Estimation not made in this treatment dueto the number of species TD-22 Emergence (%) 71 58 46 34 34 29 22 21 Technology development During the period of establishment there were significant differences (P::::::O.O l ) in rate of coverage. At four months after planting, C. brasiliensis 17009 covered 67% cover, followed by M. pruriens 20709 with 43% cover, respectively. The other species established very slowly and competed poorly with the weeds (Tables 3 and 4). However, there was low amount of weeds associated with S. guianensis despite its poor coverage of the soil. Table 2. Coverage of the soil during establishment. Species Date 28-0 1-97 07-03-97 20-03-97 03-04-97 Natural fa llow 80 80 80 84 C. brasiliensis 17009 55 67 76 70 M. pruriens 9349 48 43 47 55 S. guianensis 1 1844 11 26 39 30 C. cajan 913 14 24 37 28 C. mucunoides 20709 28 32 38 45 Mixture 18 23 33 49 P. phaseoloides 7182 7 19 26 27 Cowpea-Verde Brasil 33 33 29 O* C. juncea 2 1709 12 2 1 19 ll *Leaves eaten by insects A s ignificant effect of chicken manure over super phosphate was only observed at the fi nal phase of establishment where the re was a greater percentage cover with chicken manure (Table 3). Table 3. Percentage cover of legumes and weeds after estab lishment Treatme nts Cobertura% P source Species Legumes Weeds (kglha P) Grasses Other Total Manure P.¡o Natural fallow 68 24 92 C. brasi/iensis 17009 76 7 2 85 C. mucunoides 20709 58 27 4 89 M. pruriens 9349 54 17 8 79 Mixture 46 16 12 74 P. phaseoloides 7 182 37 28 6 71 S. guianensis 1 1844 36 33 1 70 C. cajan 9 13 33 24 14 7 1 C. juncea 2 1709 13 49 10 72 Super P4o Barbecho Natural 53 23 76 C. brasiliensis 17009 64 11 7 82 M. pruriens 9349 56 13 10 79 Mixture 52 12 9 73 C. mucunoides 20709 32 39 5 76 S. guianensis 1 1844 24 36 7 67 C. cajan 913 23 35 8 66 P. phaseoloides 7182 18 34 8 60 C. juncea 2 1709 10 39 23 72 TD-23 Technology development The biomass production was significantly different between species but not for P source. The treatments that produced the greatest biomass after 5 months were ' natural fallow ' (2.7 t/ha DM), C. brasiliensis 17009 (2.33 t/ha DM), 'mixture'(2.0 t/ha DM), and M. pruriens 9349 ( 1.5tlha DM). The biomass production of other species was much Iess and below the overall average ( 1.3 t/ha DM), as can be seen in Table4. Table 4. Biomass of legume and weeds after establishment Treatments Yield (DM kg/ha) P Source Species Legumes Weeds (kg/ha P) Grasses Other Total Manure P-10 Barbecho Natural 1950 570 2520 C. brasiliensis 17009 2325 35 o 2360 Coctel 1520 235 240 1995 M. pruriens 9349 1460 133 373 1966 C. cajan 913 11 95 130 300 1625 C. mucunoides 20709 795 475 25 1295 C. juncea 21709 767 453 253 1473 S. guianensis 1 1844 425 370 JO 805 P. phaseoloides 7182 355 270 70 695 Super P40 Barbecho Natural 2225 670 2895 Mixture 2480 245 295 3020 C. brasiliensis 17009 2345 120 40 2505 M. pruriens 9349 1533 107 427 2067 C. juncea 2 !709 853 380 627 1860 C. e ajan 9 l 3 880 11 5 255 1250 C. mucunoides 20709 355 515 40 910 S. guianensis 1 1 844 330 300 130 760 P. phaseoloides 7182 165 51 0 260 935 The performance of the species was affected by the rather dry season in 1997. Curren ti y, a crop of maize has been sown following incorporation of the Iegumes in a rder to evaluate nitrogen input. At the time of incorporation, those treatments with the most standing dry matter were C. brasiliensis, S. guianensis and the Iegume mixture, the others being completely defoliated due to the dry severe dry season still being experienced. S. guianensis CIAT 11 844 y 11 833 which established slowly were vigorous at the time of incorporation while cowpea and C. juncea had completely disappeared. TD-24 Technology development Value added options for farmers- Forages for Smallholders Project (2.2.5; (Ed Magboo, PCARRD, Philippines; Liu Guodao, CATAS, China; Francisco Gabunada and W. Stür, CIA T) Main achievements • Commencement of pilot forage seed production by farmers in the Philippines • Support of leaf meal production of Stylosanthes guianensis in China Summary of activities Seed production in !sabela, Philippines Seed production Stylosartthes guianensis CIA T 184 and Centrosema pubescens CIA T 15 160 commenced at Gamu, a Department of Agriculture livestock station in the province of !sabela, Philippines in 1996. In early 1997, more than 200 kg of seed of these species were harvested on station and distributed to farmers in the regían. In 1997, severa! farmers near the station received seed of Stylosanthes guianensis CIA T 184 and planted 1,000 m2 each with the aim to pilot seed production . If successful, the DA station will buy the seed from farmers and sell on their behalf. This system is based on the smallholder seed production scheme in Thailand . A revolving fund to buy seed from participating farmers is available. Leaf mea! production in Hainan Leaf meal production in Hainan, China is a lucrative enterprise for farmers. lt has been described by Guodao and Kerridge ( 1997) as follows: "Plants are allowed to grow until there is a complete cover of the ground. When mature but still green, plants are then cut with a small tracto r mower or by hand and allowed to dry in the field. The dry material is then passed through a hammer mili. In sorne localities a cooperative suppl ies a portable mili and markets the mea!. In other areas the dried plants are carried toa central point for hammer milling. Stands last for 3 years. Farm yields average 15 t/ha and the mea! is sold for USD 140/ton. Because of this high return, growing S. guianensis for mea! production has spread rapidly in Hainan and Guandong. It is estimated that more than 3,000 ha are grown annually in Hainan and Guandong. The largest use of the mea! is in poultry feed to increase the depth of yolk color and its inclusion has been observed lo reduce cannibalism. Beef and draft cattle can be grazed on a grass-stylo association while dairy cows are fed cut stylo along with other cut grass. though use for cattle feed is not widespread. Pigs can be fed a feed ration containing 10-15% of stylo mea! or green stylo cut into 2-4 cm pieces and cooked with o ther food wastes before feeding. Poultry are fed a ration containing 3-5% stylo mea! and for ducks it is included in a swill. For ponded fish, green stylo material is thrown directly into the pond as with grasses or use 1% leaf mea! in the compound feed . An advantage of this intensive cultivation for cover crop, fodder or mea! production is that S. guianensis maintains high productivity for 3-4 years whereas under grazing it does not usually persist more than 2 years. The limiting factor in mea! production is the rate at which air-dried material can be hammer milled with the limited machinery available. Timing of harvests also depend on availability of machinery and as a result sorne stands become over mature and the protein content of the mea! is low. Thus there is a need to (i) improve harvesting and processing, (ii) increase persistence and leaf proportion through management, and (iii) secure long-term production through the identification of new S. guianensis germplasm with broad-based resistance to anthracnose and alternative legume species for leaf mea! production." The FSP is assisting the Chinese Academy of Tropical Agricultura! Crops (CATAS) through collaborative research on searching for more anthracnose resistant accessions of Stylosanthes guianensis, and on developing management strategies to maximise field persistence of Stylosanthes guianensis. TD-25 Technology developmenr New Cassava Options for Asia NIPON Project - Improved sustainability of cassava-based cropping systems (2.2.6; R. H oweler) Main achievements • Criticalleve ls of soil paramete rs for cassava were determined: 3 .2% organic matter (OM), 7 ppm Bray- ll ac tractable P and 0. 14 me K/IOOg. • Promising new management options to reduce eros io n in cassava fi e lds identified: conto ur hedgerows o f Paspalum atratum, Setaria spacelata or Brachiaria brizantha; and intercropping w ith peanut, pumpkin or cucumber. Strategic and applied research. Strategic and applied research on improved cassava technologies was conducted in collaboration with universities and research institutes in Thailand, Indonesia, Vietnam, China and the Philippines . Long-term fertility trials The objective of these trials, conducted in eight locations in China. Vietnam and Indonesia, is to study the long-term effect of cassava on soi l fertility and to determjne the ferti lizer requirements to maintain soil productivity. Most of these trials have completed 7-8 consecutive cassava cropping cycles. During the latest cycle, harvested in 1996 or 1997, there were significant responses to application of N in six, to P in two, and to K in fi ve of the e ight sites. T his clearly indicates the importance of N and K application in most cassava growing soils in Asia, while P application is seldom necessary. The responses toN and K application tended to increase over time, as indicated by the relative yield in the absence of N and K in a long-term NPK tria! in Nanning, China (Figure 1); it is mainly due to the removal of relati vely Iarge amounts of N and K in the success ive root harvests. This, however, was seldom retlected in a marked decrease in exchangeable K or OM in the soil. Using the combíned results of nine trials conducted between 1993 and 1996, critica! levels of various soil parameters for cassava were estimated to be 3.2% OM, 7 ppm Bray II-extractable P and 0.14 me K/ JOOg (Figure 2) . The leve ls for P and K are slightly higher than those reported earlier, i.e. 4.5 ppm P and 0. 13 me K/ lOO g, but fall within the "medium" range that have been recently reported elsewhere (Howeler, l995a; 1995b; and 1996b). Criticallevels for OM had not previous ly been reported. S imjlarly, cri tica! levels for N and K in youngest full y-expanded cassava leaf blades (YFEL), sampled at 3-4 months after planting, we re estimated to be 5.7% N and 1.9% K (Figure 3), which again are slightly highe r than those reported earl ier, i.e. 4 .6% N and 1.7% K (Howeler, 1995b and 1996b). There was no clear relationship between P response and the P concentration of YFEL-blades, as there were few cases of a significant P response observed in the NPK trials included in the analysis. T he above-mentioned c ritica! levels, which correspond to 95% of maximum yield, are a useful guide in the interpretation o f soil and plant tissue analyses. From soil and tissue ana lyses resul ts, reasonable fertilizer recommendations can be formulated . Erosion control experiments The objective of these trials is main ly to indentify effective crop/soil management practices to reduce erosion when cassava is grown on slopes. Past experiments have shown the effecti veness of fertilizer application, contour ri dging, rrunimum or zero ti llage, intercropping with peanut, planting at closer spacing and the planting of contour barriers of veti ver grass, Tephrosia candida or elephant grass in reducing soil loss (Howeler, l996a; 1996c; 1996d; 1998a; and 1998b). TD-26 Technology development • : NoPoJ. 20 ... Po o X. o "' .. > No ol • 10 • ol u o 100 .. 80 ~ ..... .. 60 .... >. ., > 40 .... ... ol 20 ..... .. ..: o _0 . 20 0'1 o o Critical ~-level ~ ...... ., ~0 . 10 loo: ..... .... o <11 o 20 e p. ~ l>o 10 .... .... o Critical P-level <11 ---------------------------------- o o 1 2 3 4 S 6 7 8 9 Cr op cycle Figure l . Effect of annual applications of N, P and K on cassava root yield, re lative yield (yield without the nutrient over the highest yield with the nutrient) and the exchangeable K and available P (Bray 2) content of the soil during e ight years of continuous cropping at the Gunagxi Subtrop. Crops Research Institute, Nanning, Guangxi , China. TD-27 lOO 80 ~ ~ ... .... -;:. 41 ~ ~ • '¡; c.: 20 • o 1.0 1.5 100 • o • o ~ ... '¡; -;:. 41 o -= • ~ ... c.: 20 00 5 10 100 30 ~ ... ~ ... -;:. .. -= • 40 ... c.: 2A) o o o.os 0.10 • Technology development V V V V • • 2.0 2.5 3.0 3.5 4.0 4.5 Orpnlc matter("JG,) V V V V 15 • o= Nanniac o = G .. aopt.o.a 4 = Daozbou (CATAS) V= IIwlc Loe • = 'l'raco • = UmasJaya • = Yocyakata • = Jatüc.eno •= VaSCA 25 30 Anilable P(ppm, Bray m V i O.IS 0.20 0.25 0.30 35 ·- ·- ·- 0.35 Figure 2. The re lation between the re lative yield of cassava, i.e. the yield without the nutrient as a percent of the highest yield with the nutrient, and the organic matter, available P and exchangeable K contents of the soil in ten long-term NPK trials conducted in Asia from 1993 to 1996. TD-28 Gl 3 .5 4 . 0 \ ... 0 e 0 0 ... e e .o. o o 0 . 30 0 . 35 o .40 \ lOO ~ -so 'O ..... " .... >.. 60 " > .... u "' ...... .e o 41 ao: 20 o 0 . 6 0 . 8 1 . 0 \- K N ... p 4 . 5 5.0 0 -se 205 CATAS • 94 e -se 205 CATAS • 95 .o. -se 124 CATAS 1 95 Technology development • -se 201 Nanning • 95 ••Vinh Phu ACN3 '95 • •XM 60 ACN3 1 95 5. 5 6.0 in YFEL-blades • ·- • • • • • • • 0 . 45 0 .50 0 . 55 0 . 60 in YFEL-blades 1 1 . 2 1 . 4 1.6 1.8 2.0 in YFEL - blades Figure 3. The re lation between the relative yield of cassava, i.e. the yield without the nutrient as a percent of the highest yield with the nutrient, and the concentration of N, P and K in the youngest fully-expanded leaf(YFEL)-blades at 3-4 months after planting of various cultivars planted in four long-term fertility trials in China and Vietnam. . TD-29 Technology development Initial data from an experiment conducted in collaboration with the Field Crops Research Inst. of DOA in Thailand showed that changing the date of planting of cassava from the early to the late rain y season reduced erosion and actually increased yíelds. However, more recent data obtained from the same tria! indicates that soillosses vary markedly from year to year and are mainly related to the amount and intensity of rainfall . Erosion tends to be higher when cassava is planted in the early or mid rainy season, but planting in the late rainy or early dry season can still result in considerable erosion if rainfall is high during the following wet season (Figure 4). The amount of soilloss is also dependent on the extent of canopy cover during the rain y period. o--,.... .. o •A ........... · -~ ....... • - o-loo< ..-.. •• r..-_,.,...... • -Atri ....... Figure 4. Effect of date of planting cassava, cv Rayong 90, on the accumulated dry soil loss due to erosion on 4 .2% slope in Rayong Field Crops Research Center, Rayong, Thailand from 1994 to 1997. Rainfall distribution is shown below. TD-30 Technology development In another experiment, conducted for three consecutive years, it was found that intercropping cassava with peanut was always most e ffective in reducing erosion, while total gross income increased compared with monocropping. Intercropping with pumpkin, muskmelon or cucumber was slighty less effective than peanut in reducing erosion, but greatly increased the total gross income of the intercropping system (Table 1). Although the gross income obtained with pumpkin was less than with cucumber or muskmelon, pumpkin production is less risky as the crop is more tolerant of drought. Table l . Effect of intercropping cassava with various crops on soi l cover, cassava and intercrop yields, gross income, a nd soilloss due to e rosion in Pluak Daeng, Rayong Thailand. Data are average values for three years { 1994-1997). Soil Yield (!lha) Gross income ($/ha) Dry soil Cover loss Cropping systems (%) Cassava lntercrop Cassava lntercrop Total (!lha) l. Cussava monoculture 18.2 1 671 671 42.6 2. C+peanut 67 13.06 0.87 462 322 784 18.1 3. C+watermelon 23 15.51 1.40 582 278 860 35.2 4. C+muskmelon 57 16.81 2,72 595 845 1440 ::!5.1 S. C+cucumber 57 16.79 3.01 629 804 1433 23 .8 6. C+pumpkin 72 16.64 3.09 632 548 1180 24.8 7. C+watermelon (at 1 MAP)2 5 16.84 o 628 o 628 35.9 8. C+muskmelon (at IMAP)2 9 23.79 o 905 o 905 33.5 9. C+cucumber (at 1 MAP)2 9 16.73 o 650 o 650 28.2 10 C+pumpkin (at 1 MAP)2 16 18.61 0.11 672 22 694 28.3 1) Soil cover by intercrop at 2 months after planting (visual estimation) 2) lntercrops planted one month after planting cassava: in Treatments 2-5 planted simultaneously On-farm triaJ s on erosion control practices, conducted in four locations of Rayong province during three consecutive years, indicate that the combination of contour ridges, closer spacing (0.8x0 .8m) and fertilizer application was the best package of practices to reduce erosion and increase yields (Table 2). However, the results of this treatment were not significantly different from the "farmers' practice", because neighboring farmers ha ve over the years aJready adopted most components of these recommended practices. The data indicate that contour ridgi ng was the most effective single practice to reduce erosion, while it also increased cassava yields. Table 2. Effect of various crop/soi l manageme nt practices on cassava yield and dry soil loss due to erosion in on- farm tria ls cond ucted in four locations o f Rayong province ofThai la nd. Data are average val ues for three years ( 1994- 1997). Farmer' s prac tice 1 I .Ox 1.0 m spacing ; no ridg ing: no fertilizers I.Ox0.6 m spacing; no ridging; no fertilizers I .Ox0.6 m spacing; no ridging; with fertilizers2 I .Ox0.6 m spacing; contour ridging; w ith fertilizers2 0 .8x0.8 m spacing ; contour ridging; with fe rtilizers2 Row y ie ld {t/ha) 17.69 11 .26 12.58 15.96 20.34 21.09 Soilloss {t/ha) 19.76 29.33 25 .6 1 34. 15 17.7 1 15.78 1) Farmer' s practices gene rally include con tour ridging, planting at I.Ox0.6 m, a nd application of 200 kg/ha 15-15- 15 fertilizers. 2) With fertilizers is 312 kg/ha 15- 15- 15 TD-31 Technology development Another preliminary tria! conducted in Khaw Hin Som, Thailand, in collaboration with Kasetsart University, indícates that con tour hedgerows of Paspaium atratum BRA 96 1 O and Setaria spacelata actually increased cassava yields as compared to the check of cassava wi thout hedgerows. This might have resulted from better moisture retention during the dry season due to the application of grass mulch from the pruned hedgerows. Moreover, these two species showed little competition with neighboring cassava plants, as cassava yields near the hedgerows were nearly as high as those of plants farther from che hedgerows (Figure 5). The same tria! also showed that tall ancilor highly productive grasses like king grass (Saccarum sinense Roxb.), sugarcane (S. officinarum L.), and dwarf or normal elephant grass (Pennisetum purpureum) are highly competitive, reducing cassava yields not only in neighboring rows, but also in the next row away from the grass. This competition is particularly serious during periods of drought stress. The three varieties of vetiver grass included in the tria! were intermediately competitive. Another experiment conducted in Vietnam in collaboration with the Agro-forestry College of Thai Nguyen, showed that intercropping cassava with peanut reduced erosion losses from 72 to 18 t/ha, whíle increasing gross income from 7.8 to 11.4 mil. dong/ha. This practice combined wi th contour hedgerows of vetiver grass further reduced e rosion losses to only 6.6 tlha, while the gross income remained the same (Table 3). Table 3. Effect of various cropping systems and management practices on che yields o f cassava, incercropped peanuts and hedgerow species, as well as on toca! dry soilloss dueto erosion when cassava was planced on 10% slope at Agro-forestry College of BacThai, Thai Nguyen, Bac Thai, Vietnam in 1996. Treatmenls cassava l. No ridging, no intecrop. no hedgerows 15 .70 2. No ridging, peanut intercrop. no hedgerows 15.42 3. No ridging, peanut intercrop. Tephrosia hedgerows 16 .62 4. No ridging. peanut intercrop, vetiver grass hedgerows 16.60 5. Contour ridging, peanut intecrop, Tephrs.+vetiver hedgerows 15 .48 1) peanut dry pods 2) prices: cassava fresh roots: peanut dry pods: d 500/kg d 5000/kg Yield (t/ha) peanut 1 hedgerows 0.733 0.448 0.86 0 .640 2.00 0.810 0.83 Gross Dry soi l income2 loss ('OOOd/ha) (c/ha) 7.850 72.2 11,375 18.3 10.550 11 .3 11.500 6 .6 11.790 9.4 In a tria! conducted in collaboration with the Bogor Research Institute for Food Crops (BORIF), in Lampung, in the southem part of Sumatra island of Indonesia, it was found that cassava grown in monoculture resulted in more erosion than two consecutively grown crops of maize, soybean or peanut, or rice-soybean re lay crops. However, cassava intercropped with maize and upland rice followed by soybean (after the rice harvest) caused less erosion than any of the crops grown in monoculture (Figure 6). Soil losses were considerably reduced with the application of adequate amounts of fertilizers. Farmers in Indonesia usually intercrop cassava with maize or with maize, rice and legumes (in the wetter areas), but generally do not apply sufficient fertilizers to obtain high yields and reduce erosion. Improved management practices for cassava production in Asia The above-mentioned experiments clearly show that various soil/crop management practices in cassava- based cropping systems have a potentiaJ to not only increase cassava yields or gross income, but also to significantly reduce soil losses by eros ion. These practices include: 1) intercropping with peanut (Vietnam), with peanut or pumpkin (Thailand), or with upland rice, maize anda legume crop (Indonesia); 2) adequate fertilirzer or manure application; 3) contour ridging; 4) planting at closer spacing; and 5) contour hedgerows of vetiver grass or Tephrosia candida, or possibly Paspalum atratum, Setaria sphacelata or Brachiaria brizantha. However, farmers usually face certain constraints which limits the adoption of these practices. For example, in China, intercropping with peanut has not been successful TD-32 Techno/ogy deve/opment Paspalum a ~ ~ u ...----. o JGng grass Sugaccanc: Owañ elephant Normal elepbant :l:::::_ grass grass ~ ~ ~ 1 2 3 2 3 2 3 1 2 3 row row row row 3m & ~ ~ 1 ~va 3 't._ barrier ___t Figure 5. The effect of different grass species used as contour barriers on the fresh root yield of cassava, cv. KU 50, grown in three rows between barriers in Khaw Hin Sorn, Chachoengsao, Thailand in 1996/97. since the peanuts are often damaged by rats. In Thailand farmers seldom intercrop cassava because of scarcity of labor and frequent intercrop failures due to the unpredictability of rainfall. In addition, contour ridging and contour hedgerows make land preparation more difficult and thus more costly. Almost anywhere, cassava farmers are poor and do not ha ve the money to apply adequate amounts of chemical fertilizers, while animal manures are not always available (especially in Thailand). Thus, while there are many options to improve cassava management practices, there are also many constraints to the adoption of these practices. Which practices are most effective and most acceptable to farmers in a particular region can best be deterrruned by farmers themselves through farmer participatory research (FPR), followed by dissemination of the most useful practices by farmer-to-farmer extension. TD-33 Without ferotizers .. .2 o .. Months aftet plandng 400 ;a e ·¡¡¡¡ ~ lOO A J J A Technology development .. cassa.va maize-maiz.e ric:e-soybean ca.ssava maize-maiz.e iice-to~ soybean-soybean peanut-peanut C+M+R-S iatera"op Figure 6. Accumulated dry soil loss due to erosion in various crops grown in monoculture or intercropped with ot without fertilizers on 5% slope in Tamanbogo, Lampung, Indonesia in 1995/96. The rainfall distribution during the cropping cycle is shown below. Arrows indicate time of planting of first and second crop. TD-34 Technology development Integrated soil conservation and soil improvement strategies (2.3) Adoption of Improved Management Practices for Cassava Production in Asia NIPPON Project - lmproved Sustainability of Cassava - based cropping systems (2.2.3; R. Howeler) Main achievements • 13 farmers in FPR pilot sites in Thailand and Vietnam have now planted the best erosion control practices, identified previously in FPR trials, on small areas of their cassava production fields . Farmer Participatory Research In 1996/97 about 128 FPR trials were conducted in nine pilot sites in Thailand, Vietnam, China and Indonesia. Of these, 54 trials were on erosion control practices, 37 on varieties, 26 on fertilizers and 11 on inter-cropping systems (Table 4). These trials were conducted by the farmers on their own fields, but with assistance from researchers and extensionísts involved in the project. Most trials had 3-6 treatments without replication. Farmers in each pilot site had generally decided among themselves which practices or varieties they wanted to test, so within one pilot si te the treatments within each type of trial were the same. Thus, each farmer could be considered as one replication. In the FPR erosion control trials the effect of each treatment on soil loss dueto erosion could be measured by weighing the soil sediments that had collected in plastic-covered ditches along the lower side of each plot. At time of harvest farmers and researchers harvested the plots together. The combined results were presented and discussed with the farmers . Based on the data of yie ld, gross and net income as well as soillosses by erosion, farmers selected the most promising treatments for further testing or adoption. Tables 5 to 8 show sorne of the results of FPR erosion control trials conducted in Thailand, Vietnam and China; in Indonesia farmers did not use the same treatments in all trials, making the interpretation of the resuhs difficult. In the two pilot sites in Thailand (Table 5), most farmers preferred either vetiver grass or sugarcane (for chewing) as con tour hedgerows, planting on contour ridges or intercropping with mungbean. In the three sites in Vietnam (Tables 6 and 7), farmers preferred intercropping cassava with peanut, while planting also contour hedgerows of either vetiver grass or Tephrosia candida. In China (Table 8), farmers in one site also preferred cassava intercropping with peanut and vetiver grass barriers; other intercrops like maize, soybean or sesame did not produce any yield, while Sty/osanthes or lndigofera barriers did not establish well. At the Blitar site in Indonesia most farmers preferred planting Gliricidia sepium as contour hedgerows, since the prunings of these leguminous trees can be used either as animal feed or as greenmanure. Elephant grass is also often used as a contour barrier and animal feed, but this grass competes seriously with neighboring cassava plants. Tables 9 and 10 show examples of FPR variety trials conducted in Thailand and Vietnam, respectively. In Soeng Saang district of Thai land yields were low dueto a drought in June/July 1996, but highest yields were obtained with the variety Rayong 90. This variety was also preferred because of its high starch content. However. in a matrix ranking of varieties, conducted with about 15 farmers in Soeng Saang as part of a practice in PE during the FPR tra ining course in Thailand, farmers indicated a clear preferrence for either Rayong 5 or Kasetsart 50 (KU 50) over Rayong 90. Farmers indicated that Rayong 90 was more difficult to weed and harvest because of bent stems; this variety also had lower germination and less drought tolerance than Rayong 5 or KU 50. In the Wang Nam Yen site in Thailand, farmers also preferred TD-35 ---i o w 0\ Teclmolugy developme11 t Table 4 . Number a nd types of Farme r Partic ipatory Researc h (FPR) tria ls with cassava cond uc ted in four countries in Asia in 1996/97. Type of tri al Erosion control Varicties Fertilization lntercropping Total Socng Saang Nakorn Ratchasima 8 3 8 19 Thailand Wang Nam Yen Sra Kaew 7 6 13 Pho Yen Bac "lñai S )) 6 11 33 Vietnam Thanh Hoa Vinh Phu 7 3 4 14 Luong Son 1-/oa Binh 3 3 3 9 China Baisha Tunchang llainan llainan 4 1 4 1 4 12 2 Indonesia Dampit Wates Malang Blitar 10 9 1 S 12 14 Table 6 . Average res ults of five FPR erosio n contro l trials conduc ted by farmers in Tie n Phong a nd D ac So n villages o f Pho Y e n dis tric t,Bac Thai p rovince, Vietnam in 1996. Dry soil Yield (t/ha) Gross Production Net loss 1) !neo me Costs4> lncome (t/ha) cassava2l intercroE3) <--------------(mi l .don~ha)----- ----- --- -7 l . Farmer's practiccS) 8.33 11 .53 6.92 2.25 2. Tephrosia hedgerows, no ridging, peanut intercrop 6.62 11 .02 0.372 8.47 2.30 3. Vetiver grass hedgerows, no ridging, peanut intercrop 6.34 12.82 0.280 9.09 2.30 4. Tephrosia hedgerows, contour ridges, peanut intercrop 4.85 12.30 0.3 18 8.97 2.30 S. V etiver+Tephrosia hedgerows, no contour ridges, no intercrops 4. 17 12.78 7.67 1.94 1) dry soilloss tluring 1996 . 2) fina l yie ld o f fresh roots 3 ) dry pod s 4) includcs cost of m anure, fertilizers a nd peanut seed 5) m o noculture cassava with 15 t!ha of pig manure, 144 kg urea, 107 SSP a nd 95 KC I/ha 4.67 6. 17 6.79 6.67 5.73 Farmcrs' P rcfcrcnce (%) o o 39 58 3 Technology development Table 5. Average results of six FPR erosion control trials conducted by farmers in Wang Sombuun village ofWang Nam Yen destrict, Sra Kaew province ofThailand in 1996/97. 1 Gross income ('0008 /ha) Cassava Mungbean T 1 = up-and-down ridging 22.10 T 2= con tour ridging 20.67 T 3= vetiver grass hedgerows 18.10 T4= grass mulch 21.45 T 5= mungbean intercrop 12.65 5.38 Prices:cassava:B 0.75/kg fresh roots mungbean: 8.00/kg dry grain 1 US$ = 25 baht 2) Production costs(B/ha) : normal practices: planting/maintenance veti ver grass: application grass mulch: intercropping of mungbean: 3) Out of 12 farmers at me · ti ng Total 22.10 20.67 18.10 21.45 18.03 Production <--('0008/ha)--> 12.50 12.50 13. 12 13.12 13.37 12.500 620 620 875 Net lncome 9.60 8.17 4.98 8.33 4.66 Soil loss (t/ha) 47.79 28.27 10.16 29.14 15.53 Farmer Preference3 o 9 9 6 Table 8. Average results or' FPR erosion control trials conducted in four farmer's fields in Kongba village, Baisha county, Hainan, China in 1996. Practices Farmer's practice (cassava monoculture) C+peanut, vetiver grass barriers C+soybean, !ndigofera barriers C+maize, Stylosantlres guianensis barriers C+sesame, Stylosantlres guianensis barriers C monoculture. contour ridges 1 >prices: cassava fresh roots: Y 0.22/kg peanut dry pods: 4.5/kg 1 US $ is about 8.1 yuan 2>based on only two trials 3)based on only one tria! TD-37 Yield (tlha) Cassava lntercrop 13.5 14.0 0.625 16.5 20.72) 6.53) 13.5 Gross Dry soil lncome1) loss (Y/ha) (t/ha) 2970 125 5892 89 3630 97 45542) 622) 14303) 2163) 3322 77 .....¡ o ~ 00 Technology development Table 7. Average results of an FPR erosion control tria! conducted by three farmers on 16% slopc in Dong Rang village of Luong Son district, Hoa Binh province. Vietnam in 1996. Yield (tlha) Treatments 1) Cassava l . Cassava(C)+taro(T), no fertilizers, no hedgerows 9.00 2. C+ T , with fertilizers, vetiver grass hedgerows 13.02 3. C+ T, with fertilizers, Tephrosia candida hedgerows 14.09 4. C+peanut, with fertilizers, vetiver grass hcdgerows 15.66 5. C+peanut, with fertilizers , Tephrosia candida hedgerows 14.29 1) All plots recceived 5 tlha of FYM; fertilizers = 40N+40P2ü 5+80K20 taro or peanut received seperately: 7N+20P2ü 5+20K20 in all treatments 2) Dry biomasss from peanut and lea ves of hedgerows 3) Prices:cassava fresh roots: d 500 /kg taro fresh corms: 1100 /kg peanut dry pods: 5000 lkg FYM: 100 /kg urea (45%N): 3000 /kg SSP (17%P20 5): 1000 /kg KCI (60%K20): 2200 /kg intercrop 2.260 1.800 1.800 0.660 0.693 Gross Fertilizer Net Biomass Dry soil Farmers' income3) costs3) Income Incorp.2) loss preference <----------(mi 1.dong/ha)----------> (tlha) (%) 6.99 0.74 6.25 - 43.13 8.49 1.53 6.96 0.144 19.67 9.02 1.53 7.49 0.864 15.95 11.13 1.53 9.60 1.570 2.39 + 10.6\ \.53 9.08 2.165 3.99 + Technology development Table 9. Average results of three FPR cassava variety tria ls conducted in Soeng Saang district of Nakorn Ratchasima province and of six trials conducted in Wang Nam Yen district of Sra Kaew province of Thailand in 1995/96. Soeng Saang Wang Nam Yen Root Starch Fanner Root Starch Farmer Yield Content preference Yield content preference 1 ) Varieties (tlha) (%) (tlha) (%) Rayong 1 11 .42 27.1 24.03 26.5 o Rayong 60 9.91 26.5 29.70 27. 1 o Rayong 90 18.40 33.o2> +++ 29.44 28.2 9 Rayong 5 1\ .85 25.8 + 32.82 29.4 5 KU 50 14.56 28.7 ++ 32.90 29.5 1 )out of 12 farmers 2)Average ofonly 2 trials Rayong 90 over other varieties, even though in the FPR trials the yield and starch content of Rayong 90 were slightly lower than those of Rayong 5 and KU 50. Table 10 shows that in Pho Yen district ofVietnam farmers strongly preferred the new variety KM 95-3 even though it produced a similar yield as CM4955-7 or SM98 1-3. A participatory evaluation (PE) of these varieties, conducted as part of the field exercises in the FPR training course in Vietnam, revealed that farrners did not like CM4955-7 because of low starch content, poor plant type and poor storability of planting material. Table 10. Average results of eleven FPR variety trials conducted by farmers in Tien Phong and Dac Son villages of Pho Yen district, Bac Thai province, Vietnam in 1996. Varieties/lines l . Xanh Vinh Phu (local) 2. KM 60 3. OMR 25-33-105 4 . SM 937-8 5) CM 4955-7 6) SM 981-3 7) SM 1557-3 (KM 95-3) 1) Price: cassava fresh roots: d 600/kg Cassava Yield (tlha) 20.22 22.49 21.80 20.77 23.76 23 .35 23.8 1 Gross income Net 2 income <-----(mil. Dong/ha----> 12. 13 10.06 13.49 11.42 13.08 11.01 12.46 10.39 14.26 12. 19 14.0 1 11.94 14.29 12.22 Farmers' Preference (%) o 33 33 o 58 42 100 2) Cost of app1ication of 1 O t FYM+ 174 kg urea+250 kg SSP+ 133 kg KCJ/ha is 2.065 mil. dong/ha TD-39 Technology development Tables l l and 12 show the results offertilizer trials in Thailand and Vietnam, respectively. In Thailand highest yields and net income were obtained with the application of 50 kg/ha each of N, P205 and K20. Table 11 . Average results of six FPR ferti lizer trials conducted by farmers in Noon Somboon village of Soeng Saang district, Nakorn Ratchasima province of Thailand in 1996/97. Fertilizer rate N-P20 5-K20 (kg/ha) 0-0-0 50-0-0 50-50-0 50-0-50 25-25-25 50-50-50 Root yield (t/ha) 20.36 24.02 22.86 2 1.90 25.76 28.82 l)cassava price: B 0.80/kg fresh roots at 30% starch 0.78/kg fresh roots at 29% starch 0.76/kg fresh roots at 28% starch 2)c ost fertilizers: urea 8 350/bag of 50 kg TS P 425/bag of 50 kg KCI 270/bag of 50 kg 15- 15-15 330/bag of 50 kg Starch con test (%) 28.3 28 .2 28.0 28.5 29.6 29.2 Gross Fertilizer Net Incomel) cosr2) Income <------------('000 8/ha)-------------> 15.47 o 15.47 18.25 0.78 17.47 17.37 1.72 15.65 16.64 1.23 15.41 20.61 1.10 19.5 1 22.48 2.20 20.28 Table 12. Average results of four FPR fertilizer trials conducted by farmers in Tien Phong and Dac Son villages of Pho Yen district, Bac Thai province, Vietnam in 1996. Cassava Gross Fertilizer Net Farmers' yield income2> costs3) Income Preference Treatments (t/ha) <------------(mil. dong/ha)------------> (%) l. Farmer's practice 1) 8.93 5.36 1.84 3.52 o 2. 10 tlha ofFYM; 40N+40K20 10.56 6.34 1.41 4.93 o 3. 10 t/ha FYM; 80N+80K20 12.40 7.44 1.82 5.62 79 4. 10 t/ha FYM; 80N+40P20 5+80 K20 13.22 7.93 2.06 5.87 21 l)Average farmer application: 13.3t FYM+115.8 kg urea+39 kg SSP+56.5 kg KCL/ha 2)Prices: cassava fres h roots: d 600 lkg 3)costs: FYM: d 100 /kg urea (45%N): 3000 lkg SSP (17%P20 5): 1000 /kg KC1 (60%K20): 2200 /kg TD-40 Technology development However, a similar net income could be obtained with the application of half this rate. Because of lirnited financia! resources, farmers decided that 25 kg/ha of each nutrient was a more practica! rate. In Vietnam, farmers apply large amounts of pig manure to cassava. Results of the FPR tria! (Table 12), however, indicate that a higher net income could be obtained with a lower rate ( 1 O t/ha) of manure combined with 80 kg N and 80 kg K20/ha applied as chernical fertilizers. This was the fertilization practice most farmers preferred. Of the three cropping systems tested in Vietnam, farmers overwhelrningly preferred the system of intercropping cassava with peanut, as compared to intercropping with black bean or planting cassava in monoculture. After conducting these FPR trials in various sites, ten collaborating farmers in Thailand decided to plant contour barriers of either vetiver grass or sugarcane at 1 m vertical intervals in small areas (about 2000 m2) of their cassava production fields; s imilar! y, three farmers in Vietnam planted vetiver or Tephrosia candida barriers in their fie1ds as a hedge against erosion. Moreover, most of our collaborating farmers in Thailand are now planting new varieties and use about 200 kg/ha of 15-15-15 ferti1izers in their cassava fie1ds; those in Vietnam are planting new varieties, intercrop cassava with peanut and apply chernical fertilizers as well as animal manures. In China many collaborating farmers are now planting new higher- yielding varieties, and are experimenting with the application of a special fertilizer rnix containing NPK and Zn as well as sorne chicken manure. In Indones ia most farmers have not yet implemented any improved practices as few of the tested treatments were clearly superior to their present practices. TD-41 Soil erosion and conservation practices in Cauca (2.3.1; K. Müller-Samann, Felicits Floerchinger, Kai Sonder Technology development Contributors: Wemer Doppler, Cesareo Gallego, Luis.E. Mina , Jose Luis Adarve, Disnardo Peña Universidad de Hohenheim, Germany) - U niversity of Hohenheim Sub-project. Soil Degradation and Crop Productivity Research for Conservation Technology Development in Andean Hillsides Main achievements • En vironmental and economic impact of soil-crop management alternatives in cassava was calculated on the bas is of lo ng term measurements of soil losses by combining an empírica! soilloss model (USLE) with a modified soilloss-producti vity model (PI). • With traditional cassava in tilled p lots, the shallow so ils of the pilot area of the northern Cauca department willloose their productivity w ithin 25 (pessimistic scenario ) and 84 Years ( optimistic scenario ). • Applying soi l conservation practices a llows cassava produc tion for over 90 years (pessimis tic case) or with only minar yield depressions between 4-14 % in 100 years (optimistic scenario). • Based on actual cassava prices and a 25 yea rs planning horizon annual investments between 18 1 to 875 $ U S per hectare are economically justif iable to support conservation practices fo r cassava in the northern Cauca hillside e nvironment. Characteristics of degradation, íts effect on productivity and the economic implications of soil erosion Soil erosion is a major environmental and agricultura! problem worldwide. Each year, 75 billion tons are removed from the land by wind and water erosion, with most corrung from agricultura! land. The loss of topsoil degrades agricultura! land and possibly renders it unproductive. Worldwide, about 12 x l06 ha of arable land are destroyed because of rrusmanagement (Pímentel et al. , 1995). These are alarming facts, especially as in the end of the 20'h century we are facing a remarkable transition in the history of agriculture (Rutton, 1992). Prior ro this century almost all of the increase in food production was obtained by bringing new land into production. Yet in the next century, almost all increases in world food production must come from higher yields - from increased output per hectare. To adequately feed people a 0.5 ha of arable land is needed, yet onl y 0.27 ha per capita is available. In 40 years, only 0. 14 ha per capita will be avai lable because of loss of land and rapíd population growth. In many regions, Iirruted land is already a major cause of food shortages and inadequate distribution. Soil erosion rates are highest in Asia, Africa, and South America, averaging 30 to 40 tons ha·' year· ' (Pimentel et al., 1995). Economic Costs of Soil Erosion Under natural conditions, the topsoil that is lost is largely replenished from the subsoils. Topsoil can be defined as a renewable resource with a threshold leve! below which resource use renders it nonrenewable. Many of the land-use practices adopted in the developing countries appear to be consistent with measures that transform topsoil into a nonrenewable resource. The extraction of a natural resource in the current period reduces net benefits to the future generations. This loss is defined as the user cost. The user cost of soil exploitation is likely to be low in situations where the net retums without soil conservation exceed those with soil conservation for substantial lengths of time. TD-42 Technology development The user concept is clearly relevant in heavily eroding environments where the net retums with conservation are higher than those without conservation. A farmers wealth may influence the adoption of soil conservation practices but decisive are expectations about future income. These expectations are conditioned by the farrner' s planning horizon and discount rate. A farmer who expects the net returns with soil conservation to be lower than those without conservation is certainly likely to postpone conservation. The length of time over which the net returns without conservation exceed those with conservation is often too long for the planning period to be a significant influential variable. If the effects of land degradation are perceived unlikely in the near future, the adoption of soil conservation practices would correspondingly remain unlikely. Also, we can not expect subsistence farmers in developing countries to be concerned about soil degradation and have a long planning horizon if they are struggling for survival, not knowing if they can make a living from this year to the next. In a study undertaken by the Soil Conservation Project in 1994 (CIA T, 1995) in a cassava growing zone of the northem Cauca department, Co lombia, farmers were asked about their perception of soil erosion and their soil management practices. The mean gradient of slopes being cultivated was 35 percent. 97 % of the 60 farmers interviewed asserted to ha ve observed soil being eroded by rainfalls. Almost all of them (96 %) considered soilloss was reducing the productivity of their land. Asked about their soil conservation strategies 3 1 % adrnitted not to employ any conservation method and another 30%mentioned the traditional zone tillage (loosening only the planting site) and the tradition of rotation as the only means to alleviate the extent of soi llosses from their fields. The reasons for not using erosion control practices were the following: No materials available No technical ass istance Too expensive No time Don ' t know 4 % 45 % 22 % 15 % 19 % All those reasons can be summarized in two reasons: No or inadequate technical assistance and costs implied in the application of specific soil conservation technology. The argument ' no material available' signifies that the farmer rnight know conservation practices like living barriers or mulching, but does not know where and how to get mulch material or planting material for barriers. Good technical assistance, developing conservation strategies with the farmers and making available those materials or introducing other control practices like mínimum tillage, planting on contour ridges or whatever rnight be suited for the conditions is therefo re needed to assist farrners in the adoption and development process. 'Too expensive ' and 'no time ' are problems that should be sol ved with the aid of the government, either by giving access to favorable credits, with the development of economically attractive conservation altematives ( see also output 1.3 ) or subsidizing the employment of conservation practices. However befare the latter will happen realistic estimates of erosion 's impact on the productivity and its econornic consequences must be available to fac ilitate political decisions in this respect. TD-43 Technology development Costs of conservation practices The costs of conservation practices are hard to evaluate. The prices of sorne practices are variable enough to make it d ifficult to obtain accurate averages. Opportunity costs are a means of considering many of the less obvious costs of conservation. For example, one must forgo the opportunity to grow a crop on the area where a grassed waterway is planned. The opportun ity to make a profit is real , even though a gully might destroy it later. The costs of conservation do not stop when a practice is installed. Opportunity costs remain like a ghost in the background and someone will surely be tempted to plow up the waterways and steep slopes e ven if the e rosion hazard is great. Maintenance is an obvious continuing cost of conservation. Terrace channels and waterways need to be cleaned and occasionally reshaped. Drainage systems must be kept open. Vegetation must be ferti lized, managed or replanted. Material and methods Topsoil loss by erosion in different growing systems The data for the esti mation of soi l erosion under different management options were obtained from a study conducted at the CIA T (Intem ational Center of Tropical Agriculture) research station at Santander de Quilichao, Cauca department, Colombia, on an Inceptisol. Since 1986 an erosion tria! is run by the joint CIA TI University of Hohenheim Project, where the USLE (Universal Soil Loss Equation) is tested and calibrated for its applicability under tropical conditions. Over the years on severa! test plots soil loss by water erosion has been measured with different tillage and cropping techniques (Reining, 1992; 1995) (Table 1 ). These data were used for the calculation of long term soil and productivity losses by soil erosion. Table l. Soilloss by water erosion of different cassava cropping and cultivation systems in Santander de Quilichao in the 1987 to 1994 growing seasons. Treatment Soilloss (t ha.1) 87/88 88/89 90/91 91/92 92/93 93/94 Cassava tradi tional 5.1 17.2 8.2 4.9 4.6 7.5 Cassava on con tour ridges 3 5.1 n.d.* 0.4 0.5 n.d. Cassava between contour grass strips 5.3 15 .3 1.7 1.6 1.1 *soil loss was not determined The following treatments were considered: - Cassava planted on the flat as asole crop at 1 m x 1 m. - Cassava planted on parallel contour ridges ata distance of 1 m as a sale crop at 1 m x 1 m. - Cassava planted as asole crop ata distance of 0.90 m x 1 m with contour strips of Vetiver grass ( Vetiveria zizanioides) of 1 m width ata distance of JO m. Soil loss was meas u red on plots with slope gradients ranging from 7-13 % and a length of 22.1 m (bare fallow) and 16 m ( cropped plots). Data were corrected for the LS-factor (Length-Slope factor of the USLE; Wischmeier and Smith, 1978) to obtain data corresponding to a plot with 9 % slope gradient and 22.13 m length (LS-factor = 1). According to IGAC ( 1976) 77 % of the soils of the Cauca department are classified as Inceptisols. Usually smallholders in the Cauca department cultivate soils on much steeper and longer slopes. As an average TD-44 Technology developmenr value s lopes with a gradient of 25 % and 50 m length were assumed. In order to esti mare soil losses under smallholder conditions in the same period, data from table 1 were taken as a basis to calculare soil loss on slopes with a gradient of 25 % and 50 m length using the slope-effect chart from Wischmeier and Smith ( 1978). A field with 25 % gradient and 50 m length has a LS-factor of 7.5 and consequently data from Table 1 were multiplied by 7.5 to obtain the estimated soil loss on smallholder fie lds (Table 2). Table 2. Estimated soilloss in different cassava culti vation systems o n plolS with 25 % slope gradient and 50 m length. Treatment Soil loss (t ha.1) 87/88 88/89 90/91 91/92 92/93 93/94 Cassava trad itiona l 38.3 129 62 36.8 34.5 56.3 Cassava on contour ridges 22.5 38.3 n.d.* 3 3.8 n.d. Cassava between contour grass stri ps 39.8 114.8 12 .8 12 7.5 8.3 *soilloss was not determined With an average bulk density of 0 .84 g cm·J (Re ining, 1992) 1000 kg ha·1 of eroded topsoil correspond to a loss in topsoil depth o f 0.0 12 cm. Lowest and highest soil losses of the different cultivation systems were assumed to estimare the impact on producti vity in an optimistic and in a pessimistic case scenario. lmpact of topsoilloss on cassava yields The impact of topsoilloss on cassava yields was estimated with the Producti vi ty Index model. The relation between the produc tivi ty index (PI) and the amount of topsoil loss was calculated with the followingequation (seeCIAT, Project No. lO, Annual Report 1996): PI =0.2 l- 5. 18 * lO·' c m. Then the impact of topsoi l loss on cassava yields was deterrnined with the equation Y = -0. 1 + 8.3 PI- 15.5 p ¡2, where Y is the norrnalized cassava yie ld, with Y = l corresponding to 33 125 kg ha·1• Calculation of the economic impact of topsoilloss Annual gross returns per hectare were calculated using the equation : BR =Y * Pe, where BR = gross returns (US $ ha·'); Y = fresh root yield (kg ha.1); Pe = local market price for cassava (starch factories). With the gross returns for each culti vation system and erosion rate the value of the soil was deterrnined. Following Van Kooten et al. ( 1989) the marginal user cost was calculated. The marginal user cost denotes the impact of current exploitation of the soil on future profits via the level of the soil stock. Once a soil is exploited beyond its regeneration rate, the society looses resources that otherwise could be used in the future. The marginal user cost was calculated with the equation: TD-45 n MUC, = L (BR: - BR,) 1 ( 1 +r)1 , t=l where MUC = Marginal user cost at time t, BRc = Gross retum with soil conservation. BR, = Gross retum without soil conservation, r = Discount rate. Technology development In other words the marginal u ser cost represents the present value of the soil that would be lost if no soil conservation was done. Once the soil value was determined, its equivalent leve! annuity was calculated following Williams and Tanaka ( 1996 ), using the equation: A, = MUC, * [r * ( 1 +r)" 1 ( l +r)" -1] where A,= annualized value of soil (US $ ha'1). The annualized value of soil is equivalent to the amount that could be spent each year to conserve soil to year t. An annual discount rate of 12 % was assumed. Results The highest erosion rates were observed in the traditional cassava growing system. Soil losses in the experimental plots varied consistently from one year to another depending on the climatic conditions. Data of Table 1 were multiplied by the LS factor 7.5 to obtain an estimation of the average soil losses in the Cauca department with average field length of 50 m and a slope gradient of 25 % (Table 2). In the pessimistic case, when the highest erosion rates were taken for the calculation of long term soil losses, after 25 years no yield would be obtained on the traditional cassava plots, after 30 years on the plots with contour grass strips, and after 84 years on the plots with contour ridges . In the optimistic case only a s light yield decline would occur on the plots with soil conservation, but on the traditional system still a substantial yield reduction would occur. TD-46 Technology development Figure l shows the yield decline in the different treatments caused by topsoil erosion. 40000.-------------------------------------. -= =&.::-~- -R:_ -o- -a- -o - D--D--ID-- 30000 --6---6- - -ó.--Óo-ó- --a:.- 7 liS ..e Cl " ~ -o ~ m ·::;. 20000 "-; o '\ e ..e '\ en Q) u: \ 10000 \ .._ \ \ \ o o 20 40 60 80 100 Years _._ Cassava traditional (flat without ridges) -e- Cassava on contour ridges pessimistic case _ ..,__ · Cassava between contour grass strips -o- Cassava traditional -a- Cassava on contour ridges optimistic case - 6- · Cassava between contour grass strips Figure l . Erosion effect on cassava yields in different cropping systems for a time period of lOO years. Slope gradient 25 %, field length 50 m. Optimistic case: lowest soil loss assumed; Pessimistíc case: highest soil loss assumed. Cauca Department. Colombia. Tables 3a and 3b show how soil erosion would influence yields, gross retums, marginal user costs, and equivalent annuities in the pessimistic and optimistic case over a time period of 100 years. The local market price for cassava was assumed to be 0. 155 US $ kg·1 (price paid by the starch factories in 1997). Discussion Comparing the efficiency of the two soil conservation techniques it might be concluded that contour grass barriers are not very effective in erosion control. But the high soillosses in !987/88 and 1988/89 were due to poor establishment of the barriers. Once the grass forms a dense barrier erosion is reduced more efficiently. TD-47 Technology developmenr Table 3a. Cassava yields, gross retum, marginal user cost, and equivalent annuity of two different soil conservation methods (contour ridges, contour grass strips) compared to the traditional method, Cauca Department, Colombia. Highest erosion rates were assumed. Years Yields (kg ha.1) Gross retum Marginal user Equivalent cost Year 1 ton Annuity Year 1 (US $ ha.1) ($ ha. 1) ton ($ ha-1) trad. ridges Strips trad. ridges Strips ridges strips ridges strips o 32S79 32S79 32S79 soso soso soso o o o o S 29323 31797 297S9 4S4S 4929 46 13 694 119 192 33 10 24360 30867 25582 3776 4784 3965 2146 386 380 68 IS 17690 29789 20049 2742 4617 3108 38S2 7 14 S66 JOS 20 93 12 28563 13159 1443 4427 2040 S479 1036 733 139 2S o 27 189 4912 o 4214 761 6866 13 16 875 168 27 o 26598 1233 o 4123 191 7341 1343 924 169 30 o 25666 o o 3978 o 7968 989 40 o 22 177 o o 3437 o 9408 1141 50 o 1809S o o 280S o 10136 1220 60 o 1342 1 o o 2080 o 10474 1258 70 o 8154 o o 1264 o 10622 l27S 80 o 2295 o o 355 o 1068S 1282 83 o 422 o o 6S o 10695 1284 84 o o o o o o Ca1culations were done based on the gross retums because production costs until delivery to the starch faetones vary from farro to farm. Also the costs of soil conservation were not considered, because data are not yet complete and sufficiently representative . The results of this study indicate that remuneration of eros ion control depends on the eros ion rate and the p1anning horizon. ln the pessirnistic case after 25 years no yields are obtained from the traditional cassava plots. If the farmer has a planning horizon of 2S years, the present value of the soil saved by contour ridges is 6866 $ US ha-1• That is the amount that could be invested at the present time to prevent soil erosion. With a discount rate of 12 %, the 6866 $USare equivalent toan annua1 invention of 875 $ US. If erosion rates were Jower, less money could be invested. Assuming the optimistic case and a planning horizon of 25 years, the present val u e of soi l ( = the marginal user cost of not controlling eros ion) would be 1419$ US, corresponding toan equivalent annuity of 181 $ US. Colombia is a country with a relatively high discount rate. The higher the di scount rate, the lower is the appraisal of future values. If the discount rate was lower, the cost of erosion would be much higher, in other words the soil would be much more precious. But even so , Colombia is loosing rnillions of dollars annually by erosion. Baquero (1993) TD-48 Technology development Table 3b. Cassava yields, gross retum, marginal user cost, and equi valent annuity of two different soi l conservation methods (contour ridges, contour grass strips) compared to the traditional method, Cauca Department, Colombia. Lowest erosion rates were assumed . Years Yields (kg ha"1) Gross retum Marginal user Equivalent cost Year 1 ton Annuity Year 1 (US $ ha"1) ($ ha- 1) ton($ ha"1) trad. ridges Strips trad. ridges Strips ridges strips ridges strips o 32S79 32S79 32S79 soso soso soso o o o o S 3 1864 32S23 32427 4939 S04 1 S026 198 169 SS 47 10 31024 3246S 32268 4809 5032 S002 541 464 95 82 1S 300S8 32407 32102 46S4 5023 4976 890 766 13 1 11 2 20 28966 32348 3 1931 4490 50 14 4949 1187 1026 159 137 2S 27749 32288 3 1752 430 1 5005 4922 1419 1230 18 1 157 30 26406 32227 3 1567 4093 4995 4893 159 1 1382 198 172 40 23343 32 102 3 11 77 36 18 4976 4832 1800 1568 2 18 190 50 19777 31974 3076 1 306S 4956 4768 1897 16SS 228 199 60 15709 3 1842 30318 2435 4936 4699 1939 1639 233 203 70 111 38 3 1706 29850 1726 49 14 4627 1957 1709 235 205 80 6065 3 1567 29355 940 4893 4550 1964 17 16 236 206 90 488 3 1424 28833 76 4871 4469 1967 17 18 236 206 lOO o 3 1277 28286 o 4848 4348 1968 17 19 236 206 calculated for the department of Nariño in the Colombian Andes an annual loss of 2.3 to 7 mi Ilion US $ US by not contro lling e rosion. Depending on the erosio n rate and the planning horizon investment in erosion control will pay back sooner or later. The longer the period, the less likely farmers will invest the ir money in conservation technology . Results of this study and the study of Baquero (1993) show clearl y that the Colombian economy is suffering high losses by soil erosion. Policy makers and govemments should possess a longer planning horizon than indi vidual farmers and take into account the needs of future generations. Subsidies and credits for eros ion control and investments in sustainable land management research and conservation politics would benefit the whole nation. Based on the results o f this study a detailed analysis of the Cauca department could be done to obtain an estimation o f the erosion costs in that regía n. lnc lus ion of soil loss data in different crop rotations and the costs of establishment a nd maintenance of erosion contro l techniques would be useful and can in part be provided by the end of this ongoing project. TD-49 Technology development Characterization of conservation practices with respect to soil water balance In order to evaluate the possible effects of different cassava cropping systems on the retention of soil water, measurements of the soil moisture content with a TDR (Time Domain Reflectancy) system were to be performed from April 96 until March 98. Dueto problems with the provider (late and not complete delivery) of the Moisture Point MP-917 TDR system, the soil moisture contents measurements started with a one-year del ay. In April 20, 1997 TDR probes were installed in 6 erosion plots on the field station in Santander de Quilichao. The plots and the respective treatments chosen for evaluation and monitoring are given in the following table. Table 4 Long term treatments, no. of probes and rotation elements chosen for soi l moisture studies on runoff plots in Santander de Quilichao Experimental Station. Long term treatment or cropping pattern Cassava based Traditional rotation Cassava rotation with Minimum tillage Grass/legume ley (true meadow) Cassava with Vetiver (Vetiveria zizanioides) ~rass barriers Number Crop from 4.97-7.97 of probes 4 Cowpea 4 Cowpea 4 Brachiaria decumbens and Centrosema macrocarpum 8 Cowpea TD-50 Crop from 10.97-3.98 Bare fallow Bare fallow Bare fallow Bare fallow -15cm 15-30cm 30-60cm 0-90cm -120cm References Technology development In all plots ( 16 x 8 m ) but the grass barrier treatment, 2 probes were installed in the middle of the plot, one probe being positioned 3 m down the slope from the first one. One repetition was installed. To see possible effects of the Vetiver grass barrieron the subsurface water flow, 4 probes were installed each in two plots with Vetiver hedgerows. Two were placed above the barrier and two below. Due to the high clay content of the soil, especially in the deeper horizons, insertion of the probes was delayed almost two weeks. According to the instructions of the manufacturera hole has to be prefonned by inserting a pilot rod once. At the plot site the pilot rod had to be inserted up to 7 times and still the insertion of the probe itself was difficult. Each probe has a total length of 140 cm and is divided into five measuring segments as illustrated in the drawing on the left . Daily measurements were perfonned from Monday to Friday and after heavy rainfalls also on Saturdays. Routine measurements were perfonned by a field worker and take approximately 2 hours for 200 lectures. The probes are pennanently installed in the soil and custom cables of LO m length allow measurements of the soil moisture content without stepping on the eros ion plots in order to avoid compaction by trampling. From October 1997 on a system comparison will start with a neutron probe system to explore the re lative advantages of the two soil moisture monitoring systems in a tropical soil environment with high contents of sesqui-oxides. The installation of the neutron probe aluminum tubes will be used to calibrate the TDR system both according to the manufacturers instructions to compensate for the probe material and for differences between real soil moisture content and the one measured by the TDR system. From the experiences already made with the equipment, it can be said that the handling of the TDR equipment is far easier and faster than a neutron probe system. More important still , there is no danger to the handler of the equipment as no radiation is involved in the measuring process. Baquero H. , LB. 1993. Urna interpretacao dinamica dos efeitos da modemizacao da agricultura sobre a erosao dos solos na regiao monhosa nariñense. Tese de doutorado. Universidade Federal de Vicosa. Vicosa, M.G. 232 pp. CIAT, 1995. Diagnostico sobre condiciones socio-economicas y de manejo de recursos naturales en una zona yuquera del norte del departamento del Cauca. Documento de trabajo No. 1, Proyecto especial de conservacion de suelos, CIA T. Programa de Yuca. F!Orchinger F. and Müller-Samann K.M. 1996 El efecto de la erosión en la productividad del suelo. Memorias del "Seminario Nacional sobre Actualización en Conservación de Suelos en Ladera. Feb. 6- 8, 1996, Santafé de Bogotá. INA T, JICA, CECIL . pp 13 1-136 Hook, Livingston, Sun & Hook:: Remote Diode Shorting Improves Measurements of Soil Water by Time Domain Reflectometry : SSSJ, Vo1.56#5'92 TD-5 1 Technology development IGAC (Instituto Geografico Augustin Codazzi). 1976. Estudio general de suelos de los municipios Santander de Quilichao, Piendamo, Morales, Buenos Aires, Cajibio y Ca1dono (Departamento de l Cauca). Vol. 12(4). Bogota. 472pp. Pimentel, D., C. Harvey, P. Resosudanno, K. Sinclair, D. Kurz, M. McNair, S. Crist, L. Shpritz, L. Fitton , R. Saffouri, and R. Blair. 1995. Environmental and econornic costs of so il erosion and conservation benefits. Science 267: 1117-11 23. Pierce F.J., W.E. Larson, R.H. Dowdy and W.A.P. Graham 1983. Productivity of soils: Assesing long terrn changes due to erosion. Joumal of Soi1 and Water Conservation, Vol.38: 39-44 Reining, L. 1992. Erosion in andean hillside farming. Hohenheim Tropical Agricultura! Series No. J. Verlag Josef Margraf, Weikersheim, Gerrnany . 2 19 pp. Ruppenthal, M. 1995. Soil conservation in andean cropping systems. Hohenheim Tropical Agricultura! Series No. 3. Verlag Josef Margraf, Weikersheim, Gerrnany. !lO pp. Rutton, V .W. 1991. Constraints on sustainable growth in agicultural production: into the 21 51 century. Outlook on Agriculture 20(4): 225-234. Schofield T .G., G.J. Langhorst, G. Trujillo, K. V. Bostick, W.R. Hansen: Comparison of Neutron Probe and Time Domain Reflectometry Techniques of Soil Moisture Analysis: In Proceedings of the Symposium and Workshop on Time Domain Reflectometry in Environmental, lnfrastructure, and Mining Applications, Northwestern Uni versity, Evanston, Illinois, September 7-9. 1994 Topp,G.C., S .J. Zegelin and l. White: Monitoring Soil Water Content using TDR: an Overview of Progress. In Proceedings of the Symposium and Workshop on Time Domain Reflectometry in Environmental, lnfrastructure, and Mining App1ications, Northwestem University, Evanston, Illinois, September 7-9, 1994 Van Kooten, G.C., W.P. Weisensel, andE. de Jong. 1989. Estimating the costs of soi1 erosion in Saskatchewan. Canadian Joumal of Agricultura! Economics 37:63-75 . Williams, 1 .R. , and D. L. Tanaka. 1996. Econornic evaluation of topsoilloss in spring wheat production in the northern Great Plains, USA. Soil & Tillage Research 37: 95- 11 2. Wischmeier, W.H .. and D.D. Smith. 1978. Predicting rainfall erosion losses - a guide to conservation planning. U.S. Department of Agriculture, Agriculture Handbook No. 537. Washington D.C. TD-52 Dissemination Output 3. Enhanced ability to promote adoption of sustainable land use practices Methods developed for increasing adoption of sustainable land use practices (3 .1) Activities contributing to this output in 1997 have been i) Continued development and adaptation of Farmer Participatory Research principies that are applicable in different cultural contexts (see section in Targeting Research- New approaches for targeting technology development). ii) A workshop on Systems Research in CIA T is being organized for l-2 December 1997. This wi ll be one of the topics discussed. Integrated resource models to facilitate extension of results (3.2) Model for resource allocation at the farm level (3.2.1; A. Gijsman) Main achievements • Linkage being established between the CENTUR Y soil organic model and the DSSA T model . • A legume cover crop/green manure model is being developed in conjunction with CIMMYT. DSSA T extended with a legume cover model and an improved soil-organic matter model The DSSAT model is the most widely used model for si mulating crop yields and economic retums. Modules ex ist for most commercial crops but not for cover crops/green manure. As the ultimate aim is to use DSSAT to simulate smallholder cropping systems and link it to physical GIS databases, it is imperative to have a model that included the use of green cover/green manure. A training workshop on the application of the DSSA T model was attended, during which it was discussed extensively with severa! DSSAT authors what kind of modifications to the model would be needed for making it applicable to smallholder systems with legume cover crops or green manure. Three topics carne to the front, around which the workplan was developed i) The DSSA T model does not simulate the growth and development of green-manure/cover-crop legumes; it does, however, simulate severa! food legumes (soybean, bean, peanut). Since these crops do not differ much in the ir growth and development pattem, it was decided co use one of the existing legume sub-models and reformulate it for a green-manure/cover-crop legume. An experiment is being conducted to obtain detailed growth and development data for these cover/green manure Jegumes in order co provide parameters for the model. The followi ng Iegumes are being grown - Mucuna pruriens 9349, Centrosema pubescens 151 60, Canavalia brasiliensis 17009, Cowpea, var. Verde Brasil and Common bean BA T 477, as reference ero p. D- 1 Dissemination This work is done in collaboration with Dr. Jeff White and lr. Dewi Hartkamp (CIMMYT), who are applying DSSAT to similar green-manure/cover-crop systems, and with scientific support from Dr. Ken Boote (Univ. Florida) and Dr. Gerrit Hoogenboom (Univ. Georgia), co-authors of existing legume sub- models (Dr. Jeff White is also coauthor). Elemencs of che experiment will be replicated at other locations (e.g. Mexico, by CIMMYT), so as to obtain a wide range of temperature and photoperiod conditions. ii ) For simulating low-input systems in developing countries, it is absolutely crucial that the model is able to make a rel iable estímate of the nutrient release from soil organic matter (SOM). The present SOM sub- model of DSSA T is too meager for this, since DSSA T was developed for systems where most nutrients are derived from che m.ical fertilizers. Given our earlier experience with the CENTURY SOM model, it was decided to link these two mode ls. This is done in collaboration with Walter Bowen (CIP) and Paul Wilkens (IFDC), authors of severa! DSSAT soil modules, and Bill Partan (Colorado State Univ.), the lending CENTUR Y author. iii) Presently, DSSA T only simulates the nutrient nitro gen, an important limitation for its application to systems that may not be primarily N limited (e.g. the low-P soils of South America). The University of Michigan (Dr. Samira Daroub) just started a collaborative project with IFDC and Dr Dennis Friesen of CIAT's soils' group to also incorporate phosphorus in DSSAT. Although initially the focus will be on inorganic (rock) phosphorus, the above-mentioned linkage between DSSAT and CENTURY facilitates a potencial widening of the scope to also include organic P. because the CENTURY SOM sub-model al ready does include organic P. In earlier work we made suggestions on how to modify the P sub-model of CENTURY to accommodate the strongly P-sorbing soils we are dealing with (Gijsman et al. 1996- Agronomy Joumal 88: 894-903). Besides these three tapies, a fourth issue is the lack of a livestock grazing option in DSSAT, because many of the smallholder systems we are dealing with ha vean important livestock e lement. Contacts ha ve been made with NRI (UK: Dr. Peter Thome) and ILRI (Dr. Philip Thomton), who are working on such models, to see how we can incorporate this in DSSA T. For che coming months, the focus will initially be on data collection from the legume experiment, and on the development of a modified DSSA T model that includes a green-manure/cover-crop legume option and has an improved SOM module . Data on weather, soi l type, farm management, etc. will be collected for the sites where the model will be applied. Model application will follow these deve lopments. Funding is being sought for continued research on model application in system anaJysis through a joint application with Colombian coffee research station CENICAFÉ (Dr. Jaime Arcila) for a collaborative project on modification and appl ication of DSSAT. Contacts have been made with severa! donors. Models on Iivestock adapted to dual-purpose systems (3.2.2; F . Holmann and R.D. Es trada) Main achievement • Farm simulation model expanded and validated. The farm simulation model developed initially by CATIE and RISPAL was expanded this year in CIAT with Ruben Daría Estrada to incorporate addit ional agricultura! (i.e. four crops instead of three), reforestation (two activities instead of one), and livestock activities ( i.e. in addition to dairy, dual-purpose and beef activities, fattening was also incorporated). Tbe model was validated by Ruben Darío in D-2 Dissemination Honduras in July of this year, and later by F. Holmann during the training of three CORPOICA colleagues in October of 1997. Training approaches and materials on technology diffusion for use by farmers and tecbnicians - Forages for Smallholders Sub-project (3.3; W. Stür, CIA T ; P .Home, CSIRO; F. Gabunada (FSP, Philippines); Ed Magboo, PCARRD, Philippines) Main achievements • Produced training modules for in-country training o n "Developing forage technologies with smallholder farmers". • Conducted the first in-country training course on "Developing forage technologies with smallholder farmers" in the Philippines. • Provided practica! training for forage scientists and develo pment workers from Southeast Asia Summary of activities Modulesfor in-country training courses on "Developingforage technologies with smallholder farmers". Following a workshop on "Forage agronomy, seed supply systems and seed production" in Thailand in late 1996, a training manual was developed entitled "Developing forage technologíes wíth smallholder farmers". The workshop in Thailand was attended by representatives from a ll countries participating in the FSP and it was designed to discuss content and training methods of subsequent in-country courses. This manual is now being translated into locallanguages by our national partners and wi ll fonn the basis for a series of in-country courses. First in-country training course on "Developing forage technologies with smallholder fa rmers" in the Philippines. The first training course was held for 15 participants from FSP sites and related agencies in Los Baños from 3-15 August 1997. This course served as a pilot course for the in-country courses and the manual was edited based on feedback from participants and experience from this course. Practical training for forage scientists and development workers from Southeast Asia Practica( training was provided to all collaborating forage research and development workers at FSP sites by Wemer Stür and Peter Home. Additionally, two key partners, Mr. Ibrahim of the Livestock Services of East Kalimantan, Indonesia and Mr. Viengsavanh Phimphachanhvongsod of the Department of Livestock and Fisheries, Laos rece ived hands-on training in another country in the region. Mr Ibrahim, who is responsible for FSP activities in Makroman and Sepaku 11 in East Kalimantan, spent three weeks working with Mr Francisco Gabunada in the Philippines. He received training in forage agronomy and seed production, and participated in FSP site visits in the Philippines where he assisted with PD and PP sessions. Mr Viengsavanh, who is the FSP country coordinator for Laos, participated in a Fanner Participatory Research Training course arranged by CIA T in Thailand. 0 -3 Output 4. Known impact and utility of new technologies and strategies Economic impact and market opportunities evaluated (4.1 ) Latin America (L. Rivas and F. Holmann) lmpact During May of 1997 a survey instrument was developed to analyze the demand for Arachis pintoi in Caquetá, Colombia. This study is a follow-up of an earlier ex-ante analysis executed during 1996. The survey was tested and personnel were trained in its proper use and 226 producers were surveyed ( 174 randomly setected Nestle's milk producers plus 52 early adopters) between the rnonths of June and July. The information was coded and analyzed during the months of August and October, and it is expected a first draft of the study by December of 1997. Asia-Forages for Smallholders Sub-project (W. Stür, CIA T ; P. Home, CSIRO; S. CIAD Main achievements • Initiated discussion with national partners on methodology of impact measurements through the visit of Sam Fujisak.a, CIA T to FSP si tes in Indonesia, Laos and Philippines Summary of activities The following paragraphs describe the current state of discussion on assessing the impact of forage technologies on human and natural resources, animal production and farm income within the Forages for Smallholders Project. These discussions were initiated with the visir and trip report of Sam Fujisaka, CIA T and will be further discussed during the Annual Review of PE-S in December 1997. Assessment of impact of forage technologies can occur only when the technology has been adopted by farmers. In most cases, farmers working with the FSP are still in the process of evaluating and modifying technologies in relative small areas befo re deciding if they will plant larger areas. In 1997, farmers at sorne sites decided to expand their area of forages from small plots to larger areas (>0. 1 ha). These farmers will be the first to experience a significant impact of forage technologies on human and natural resources, animal production and farm income. Sites where this is occurring are Sepaku ll and Marenu in Indonesia. In 1998, other sites such as Cagayan de Oro. Malitbog and Matalom in the Philippines and Makroman, Aceh and Gorontalo in Indonesia are likely to get to the same stage of development. The key indicator of success of the FSP is adoption of forage technologies by farmers . This can be quantified at all FSP sites by estimating the number of farms using forage technologies and the areas of forages grown on each farm. Quantifying the impact of these technologies on people and environment is a more difficult task with many potencial impacts and ways to quantify these impacts. 1-1 Impact The approach of the FSP has been to work in partnership with farmers in developing forage technologies, and this approach has proven successful. There is no reason why farmers should not be fully involved in assessing the impact of the forage technologies they adopted. Clearly, farmers will only adopt forage technologies on a larger scale if they experience a positive impact of the technology on their lives. They are in the best position to define their reasons for adopting and to prioritize these reasons. This information may help to lirnit the number of indicators to be quantified. Assessment of impact will require measurements of input and opponunity costs of the various forage technologies adopted by farmers. Also, indicators of impact of technologies must be compared with production systems that have not adopted these technologies (section 1.1.5, Figure 2). These may be (a) baseline data of the community collected befare adoption occurred, (b) non-adopters in the same area, or (e) case studies of individual farms from testing to adoption . The ability of carrying out impact assessment at FSP sites is, at sorne s ites, constrained by the lack of suitable local development workers who are able to carry out the necessary measurements. A careful selection of si tes and indicators is needed to provide information on impact across the range of forage technologies and agro-ecosystems covered in the project. lmpact assessment will be one of the main topics for discussion of the next Annual RegionaJ Meeting of the FSP in March 1998. Below is a prelirninary list of potential indicators, which may be of use in assessing impact of forage technologies, for discussion: Forage indicators: • Area of forage technologies grown • Productivity of forages • Contribution of forage technologies towards total feed requirements Animal indicators: Animal productivity • Liveweight gain of small ruminants sheep and goats (difficult with large rurninants) • Indirect measurements of productivity of large ruminants via sale price achieved, usefulness as draught animal , body condition, etc. • Reproductive performance (calving interval, litter size, etc.) • Off-spring mortality and growth • Animal health (e.g. egg counts of interna! parasites) Human resources indicators: Labour requirements (by family members) for • Cutting naturally occurring forages along roads, etc. • Herding cattle for grazing or tethering • Weeding crops following legume fallow or with companion legumes • Land preparation following legume fallow vs. natural fallow Natural resources indicators: • \ mount and quality of manure produced • .)oil fertility through yield of subsequent crop • Soil structure and biology (e .g. earthworms) • Weed population • Soil erosion 1-2 lmpact Farm income: • Sale of animals • Manure saJes (in cash or aJtemative use of manure; e.g. how much forage can 50 kg manure grow and how much L WG could this produce?) • Sale of forage Environmental impact on soil, vegetation, water and atmosphere (4.3) Asia - Forages for Smallholders Sub-project (W. Stür, CIA T ; P. Home, CSIRO; S. Fujisaka, CIA T) Main achievements • initiated discussion with national partners on methodology of impact measurements through the visit Sam Fujisaka, CIAT to FSP sites in Indonesia, Laos and Philippines Summary of activities • These ha ve been included under Targeting Research under Mínimum data set for s i te characterization in on-farm studies. Personnel trained (4.4) Latín America (F. Holmann) During October 1997 three research colleagues from CORPOICA (Osear Duarte and José Pulido from headquarters in Tibaitatá and Jorge Silva from Valledupar's regional office) received training on the proper use of a farm simulation model with special emphasis on the livestock component taking as a case study the Valle del Cesar. The training activity was divided in two parts: ( 1) An initial trip to Valledupar in June with the three research colleagues to understand the agricul tu ra! production systems of the Valle del Cesar' s watershed. During this visit we held interviews with processi ng plants, milk producers, agricultura) input stores, and research officers, plus the collection of information (farm gate prices, productivity, constraints, etc); and (2) The training period itself, which lasted 2 weeks in October 1997. During this period the objective was to understand the rationale of the simulation model, Jearn how to input data, and finally, interpret the results in order to apply the model to different scenarios in other watersheds. Asia-Forages for Smallholders project (W. Stur, P. Home) Details of training workshops were reported under Dissemination in the section, 'Training approaches and materiaJs on technology diffusion for use by farmers and technicians'. I-3 Asia-Improved Sustainability of Cassava-based Cropping Systems (R. Howeler) Details of training were provided under Targeting Technology in the section 'New approaches for targeting technology development'. Results of research communicated (4.5) Latin America - Tropileche project (F. Holmann) Tropileche Newsletter lmpact The Tropileche Consortia has produced three newsletters. Publication dates are March and October. The objecti ve of this newsletter is to inform about the activities of the Con~ortia, on-going research trials, research results being produced at the different benchmark s ites, and any other news our partners consider useful to inform. These newsletters can be obtained free of charge through the Tropi leche HomePage on the Internet (see below). Tropileche data base on research results from dual-purpose cattle The Consortia developed in October 1996 a data base with research results generated since 1960 in tropical Latín America on dual-purpose cattle. Themes include nutrition and feeding, forages (grasses and legumes), genetic improvement and reproduction, animal health, economics, and extension, transfer, and adoption of technology. There are more than 1,300 references and about 100 additional ones are updated each month. All references include basic descriptors and about half of them also include an abstract. This data base was developed in micro CD/ISIS and follows the normative of the information system AGRIS-CARIS from FAO. This database is now operational and available through the Tropileche HomePage on the Internet (see below). Tropileche on Internet The Tropileche Consortia has just developed its own HomePage on the Web. which contains the two news letters that have been produced as well as the data base containing research results generated in tropical Latín America. This HomePage can be accesed through the CIA T HomePage (hnp://www.ciat.cgiar.org) either through the "Soil and System" icon or through the "Information and Documentation Unit". In addition, this HomePage has a list of researchers with affinities in research on dual purpose cattle in LAC with addresses to be contacted. Thus, researchers can access Tropileche from anywhere in the world and consult the database, solicit information, and communicate and interact with other colleages. I-4 Impact Workshop Proceedings During July 1996 Tropi1eche held a workshop in CIA T about on-fann research methodologies in duaJ- purpose cattle farms in tropical Latín America. These proceedings were edited throughout 1996 and 1997 and are curren ti y in the print shop. It is expected the procedings will be available in November 1997. Latín America - BMZ Soil Conservation Project, Cauca (K. Muller-Samann) A manual on soil conservation was produced. A database on results from strategic and applied research on soil conservation and management in Cauca has been maintained. It will be expanded to include the data on the K-factor from the research of Jesus Castillo and that on the R-factor by Kai Sonder. Asia - Forages for Smallholders project (W. Stür, CIAT) Main achievements • Held the second Annual Regional Meeting of the Forages fo r Smallholders Project in Hainan, China • Published two issues of the SEAFRAD newsletter Summary of activities The second AnnuaJ Regional Meeting of the Forages for Sma1lholders Project was he1d at the Chinese Academy of Tropical Agricultura! Sciences, Hainan. China from 19 to 24 J anuary 1997. The meeting was attended by 15 of our research partners, 1 O Chinese representatives, Bryan Hacker, Pe ter Kerridge, Wemer Stür and Chunji Liu (CSIRO). Twe1ve of the participants were either partly or totally sponsored by their own organisations. The meeting consisted of two days of presentations and discussions, followed by a field visit to see Stylosanthes seed and leaf mea! production systems. The proceedings of this meeting will be published as a Technical Report of the FSP. Editing is in progress and publication is scheduled for early 1998. SEAFRAD, the Southeast Asían Forages and Feed Resources Research and Development Network is a network of researchers and development workers who are working with forages. The secretariat of SEAFRAD and editorship of the SEAFRAD newsletter rotate between member countries on an annual basis. In 1996, it was based at the Department of Livestock and Fisheries. Lao POR; in 1997 in the Livestock Research Center, MARDI, Malaysia; and in 1998 it wilt be based in the Directorate General of Li vestock Services, Indonesia. In 1997, two issues of the SEAFRAD newsletter were produced and distributed in February and June. A further issue is in preparation by the editor for 1997, Dr. Wong Choi Chee, MARDI. I-5 Asia-Improved Sustainability of Cassava-based Cropping Systems (R. Howeler) Activities lmpact Results from our collaborative experiments and from the FPR project were presented by national program colleagues at the 5th Regional Cassava Workshop, held in Nov 1996 in Hainan island of China. Most of these papers have now been edited and the Proceedings of this Workshop will be printed and distributed in early 1998. A paper on sustainable cassava production practices was presented at the Intemational Cassava Starch Symposium, held in Nov 1996 in Nanning, Guangxi , China. This paper will appear in the Starch Symposium Proceedings in early 1998. A paper on the CIA T Cassava and Forages-for-Smallholders projects was presented atan Ecoregional Planning Workshop for the Red River Basin in Vietnam, organized by IRRI and held in Oct 1997 in Hanoi, Vietnam. I-6 Publications Publications 1997-1998 Journal papers Barahona Rolando, Lascano C.E., Cochran Robert, Morrill Jim, and Titgemeyer Evan. C.,l997. Intake, digestion, and nitrogen utilization by sheep fed tropical legumes with contrasting tannin concentration and astrigency. J of Animal Science 75:1633-1640 Chong, D.T. Tajuddin, L, Samat, Abd . M .S., Stür, W.W. and Shelton, H.M. 1997. Stocking rate effects on sheep and forage productivity under rubber in Malaysia. J. Agríe. Sci. (Camb.) 128, 339-346. Fujisaka, Sam. 1997. Pasture or permanent crops after slash-and-bum cultivation? Land use choice in three Amazon colonies. Submitted to Agroforestry Systems. Fujisaka, Sam, German Escobar, and Erik Veneklaas . 1997. Plant community diversi ty relat ive to human land uses in an Amazon forest colony. Biodiversity and Conservation Vol. 6: Fujisaka, Sam, and German Escobar. 1997. Towards a practica! classficati on of slash-and-bum systems. ODI Rural Development Forestry Network Paper 2 lc Fujisaka, S, C Castilla, G Escobar, V Rodrigues, E Veneklaas, R Thomas, and M Fisher. 1997. Impacts of forest conversion : estimates of carbon emiss ions and plant species loss in a Brazilian Amazon colony. Subrnitted to Agriculture,Environment, and Ecosystems. Fujisaka, S., F. Holmann, G. Escobar, N. Solórzano, L. Badilla, L. Umaña. and M. Lobo. 1997. Sistemas de producción de doble propósito en la región Pacífico Central de Costa Rica: Uso de la tierra y demanda de alternativas forrajeras. Pasturas Tropicales Vol. 19( 1 ): 55-59. Gijsman A.J. and Sanz J .l. 1997 S oí l organic matter dynarnics in a volcanic-ash soi 1 under fallow or cultivation with applied chicken manure. European Joumal of Soil Science ( revised and resubmitted). Gijsman A.J. 1997 Freezing roots for decomposition studies : an altemative to drying. Communications in Soil Science and Plant Analysis ( submitted). Gijsman A.J ., Alarcón H.F. and Thomas R.J. 1997 Root decomposition in tropical grasses and legumes, as affected by soil texture and season. Soil Biology & Biochemistry Biochemistry 29: 1443-1450. Gijsman A.J. , Oberson A., Friesen D.K., Sanz J.I and Thomas R.J . 1997 Nutrient cycling through rnicrobial biomass under rice-pasture rotations replacing native savanna. Soil Biology & Biochemistry 29: 1433-1441 . Guodao, L., Phaikaew, C and Stür, W. W. 1997. Status of Stylosanthes development in other countries. II. Stylosanthes development and utilisation in China and sourh-east Asia. Tropical Grasslands 31, ( in print). Publications Hacker, J.B., Phimphachanhvongsod, V. , Novaha, S., Kordnavong, P., Veldkamp, J. and Simon, B. K. 1997. A guide to the grasses of Xieng Khouang province, Lao POR, and sorne notes on ecology of grazing lands in the province. Genetic Resources Communication (in press) (Australian Tropical Forages Genetic Resources Centre, CSIRO Tropical Agriculture, Brisbane). Hess H.D. y Lascano C.E. 1997. Comportamiento del consumo de forraje por novillos en pasturas de gramínea sola y asociada con una leguminosa. Pasturas Tropicales 19( No.2): 12-20 Horne P .M. and Stür, W.W. 1997. Current and future opportunities for improved forages in south-east Asia. Tropical G rasslands, Special Issue 2, 117- 12 1 (in press). Müller-Samann, K. M. 1997 y Castillo, J . A. 1997. Labranza mínima, una tecnología para las laderas de Colombia? CONSERVANDO, Vol. 1( 1): 4p.- Continuado en Vol. 1(2):3p. Ng, K.F., Stür, W.W. and Shelton, H.M. 1997. New forage species for integration of sheep in rubber plantations. J. Agríe. Sci. (Camb.) 128, 347-356. Putthacharoen, S., R.H. Howeler, S. Jantawat and V. Vichukit. 1998. Nutrient uptake and soil erosion losses in cassava and six other crops in a Psamment in eastern Thailand. Accepted for publication in Field Crops Research Ruppenthal, M., D . E. Leihner, N. Steinmüller and M. A. El-Sharkawy 1997. Losses of organic matter and nutrients by water erosion in cassava-based croppi ng systems. Expl. Agric., Vol. 33: 1-12. Workshop and conference papers Argel, P.J ., Pizarro. E. A. and Kerridge, P.C. 1997. Ara chis pintoi- a mulri-purpose legume for sustainable land use. Proceedings of the 18th Intemational Grass lands Congress, Canada, June 8-19, 1997. (in press) . Blai r, G. and Kerridge, P.C. 1997. Nutrient management in tropical forage systems - What should be, and what is practised. Proceedings of the 18th Intemational Grasslands Congress, Canada, June 8-19, 1997. (in press). Barahona R., Lasca no C. E. Cochrane R.C, Morrill 1 .L and Tiggemeyer E.C. 1997. Condensed tannins in tropical legumes: Concentration, aststrigency and effects on the nutrition of ruminants. In: XVill Int. Grass l. Cong.June 8-19, Winnipeg, Manitoba and Saskatoon, Saskatchewan, Canada. Session 8, 7-8. Binh, L.H. 1997 . FSP activities in Vietnam. In: W.W. Stür (ed) Feed Resources for Smallholder Livestock Production in Southeast Asia. Proceedings of a regional meeting in Vientiane, Lao POR, 16-20 January 1996. CIA T Working Document No. 156, Los Baños, Philippines, 47-48. Chen C.P., Chee, W.C., Abdullah, A. 1997. FSP Activities and results in Malaysia. In: W.W. Stür (ed) Feed Resources for Smallholder Livestock Producrion in Sourheasr Asia. Proceedings of a regional meeting in Vientiane, Lao POR, 16-20 January 1996. CIAT Working Document No. 156, Los Baños, Philippines. 53-54. Cheng, Y. and Home, P.M. 1997. Field experiments with forages and crops: practica! tips for getting it right the first time. (FSP, Laos) 48p. 2 Publications Faylon, P.S . and Lanting, E. 1997. FSP activities and results in the Philippines. In : W.W. Stür (ed) Feed Resources for Smallholder Livestock Production in Southeast Asia. Proceedings of a regional meeting in Vientiane, Lao POR, 16-20 January 1996. CIAT Working Oocument No. 156, Los Baños, Philippines, 5-1 O. Florchinger F. and MüHer-Samann K.M. 1996 El efecto de la erosión en la productividad del suelo. Memorias del "Seminario Nacional sobre Actualización en Conservación de Suelos en Ladera. Feb. 6-8, 1996, Santafé de Bogotá. INAT, JICA, CECIL . pp 131-136 Gabunada Jr., F.G. 1997. Farmer participatory research activities at two si tes in the Philippines - initial results and leamings. In: W.W. Stür (ed) Feed Resources for Smallholder Livestock Production in Southeast Asia. Proceedings of a regional meeting in Vientiane, Lao POR, 16-20 January 1996. CIAT Working Oocument No. 156, Los Baños, Philippines, 31-36. Gabunada, Jr. F.G. and Stür, W.W. 1997. Leucaena foundation collection at IRRI, Los Baños, Philippines. In: Co-ordination Meeting of ACIAR Project 9433 "New Leucaenas for Southeast Asian, Pacific and Australian Agriculture" held at the University of Queensland, Australi a, 1-4 April 1997,2: 1- 16. Gabunada Jr., F.G., Stür, W.W. and Home, P.M. 1997. Oevelopment of forage components through farmer participatory research . Proceedings of the 18th Intemational Grasslands Congress, Canada. June 8-19, 1997. Session 24.33-34. Guodao, L. and Kerridge, P.C. 1997. Selection and utilization of Stylosanthes guianensis for green cover and feed meal production in China. Proceedings of the 18th Intemational Grasslands Congress, Canada, June 8- 19, 1997. Session 19, 49-50. Guodao, L., and Huaxian , H. 1997. FSP activíties and results in China. In: W.W. Stür (ed) Feed Resources for Smallholder Li vestock Production in Southeast Asia. Proceedings of a regional meeting in Vientiane, Lao POR, 16-20 January 1996. CIAT Working Oocument No. 156, Los Baños, Philippines, Si-52. Guodao, L. and P.C. Kerridge. 1997. Selection and utilization of Stylosanthes guíanensis, for green cover and feed mea! production in China. Proceedings of the 18th Intemational Grasslands Congress, Canada, June 8-1 9, 1997. Session 19, 49-50. Hacker, J.B. and Kerridge, P.C. 1997. The Forages for Smallholders Project- Aims, Activities and Achievements. In: W.W. Stür (ed) Feed Resources for Smallholder Livestock Production in Southeast Asia. Proceedings of a regional meeting in Vientiane, Lao POR. 16-20 January 1996. CIAT Working Document No. 156, Los Baños, Philippines, 5-10. Ho1mann. F. 1997. Dua1-purpose cattle production in Pucallpa. CIAT Workshop Document no. 167. Workshop proceedings on integrated conceptual framework for tropical agroecosystem research based on complex systems theories. Cali. Holmann, F. 1997. Reflexiones sobre la competitividad de distintos modelos de producción de leche en América Latina tropical. Paper presented at the VI Pan American Dairy Federation, April 23-25, Buenos Aires, Argentina. 3 Publications Holmann, F. and C. Lascan o. 1997. The contribution of Tropileche to the Livestock Research Agenda of Latín America. Paper presented at the Annual Programme Meeting of ILRI in Addis Ababa. Ethiopia, Se pte mber 20-24. 1997. Home, P.M ., Stür, W .W., Gabunada, F. and Phengsavanh, P. 1997. Prospects for introducing forages in smallholder fanning systems in Southeast Asia. In : W.W . Stür (ed) Feed Resources for Sma llholder Livestock Production in Southeast Asia. Proceedings of a regional meeting in Vientiane, Lao POR, 16-20 January 1996. CIAT Working Document No. 156, Los Baños, Philippines, 11-1 9. Home, P. M. 1997. Securing the live lihoods of farmers in upland areas of Lao POR: the role of livestock and opportunities for forage development. Proceedings of the ACIAR workshop "Upland farming systems in the Lao POR: Problems and opportunities for li vestock", V ientiane. 19-23 May, 1997. (ACIAR, Canberra; in press) Home P. M. 1997. Farmer-developed forage manage ment strategies for stabili sati on of shifting_ cultivation systems. Proceedings of the ICRAF workshop "Indigenous strategies for intensification of shifting cultivation in southeast Asia", Bogor, Indones ia, June 23-27 1997 (ICRAF, Bogor, Indonesia; press) Home, P. M. 1997. Let them eat grass: experiences in us ing partic ipatory approaches to develop forage technologies . Proceedings of the ACIAR meeting " Red soil forages- from research into practice", Jianyang, Fuj ian Province, China. 6 to 9 October 1997 (ACIAR, Canberra: in press) Howeler, R.H. (Ed.). 1998. Cassava Breeding, Agronomy and Farrner Participatory Research in Asia. Proc. 5th Regional Workshop, held Nov 3-8, 1996 in Dangzhou, Hainan, China. (in press) Howeler, R.H. 1996a. Cassava agronomy research in As ia, 1987-1992. In : R. H. Howe ler (Ed.). Cassava Production. Processing and Marketing in Vietnam. Proc. Workshop held Oct 29-3 1, 1992 in Hanoi, Vietnam. pp. 2 55-290. Howeler, R.H. 1996b. Mineral nutrition of cassava. In: E.T. Craswell, C.J . Asher and J.N.O'Sullivan (Eds.) Mineral Nutrient Oisorders of Root Crops in the Pacific. Proc . Workshop he ld April 17-20, 1995 in Nuku'alofa. Kingdom of Tonga. A CIAR Proceedings No.65, Canberra. Australia. pp. 1 10- 116. Howe ler, R. H. 1996c. The use of farmer participatory research methodologies to l:!nhance the adoption of soil conservation practices in cassava-based cropping systems in Asia. In : S. Sombatpanit et al . (Eds.). Soil Conservation Exte nsion : From Concepts to Adoption . Proc. Intem . Workshop held June 4-9, 1995 in Chiangmai, Thailand . Dept. Land Deve lopment, Bangkok, Thailand. pp. 159-168. Howe ler, R.H, Nguyen The Dang and W. Vongkasem. 1997. Farmer participarory selecrion of vetiver grass as the most effective way to control erosion in cassava-based cropping systems in Vietnam and Thailand. In: Proc. Intem. Conference on Veti ver, held Feb 4-8, 1996 in Chiang Rai, Thailand ( in press). Howeler, R.H. 1998a . Cassava agronomy research in Asia- An overview. Paper presented at 5th As ían Cassava Research Workshop, held Nov 3-8, 1996 in Dangzhou, Hainan. China.( in press) Howeler, R.H. 1998b. Cassava production practices in As ia - Can they maintain soil productivity? Paper presented at lntem. Symposium on Cassava, Starch and Starch Derivatives. held Nov 11-16, 1996 in Nanning, Guangxi, China. (in press) 4 Publications Ibrahim, Lanting, E .• Khemsawat, C., Wong, C.C., Guodao, L., Phimphachanhvongsod, V., Binh. L.H. and Home, P.M. 1997. Forage grasses and legumes with broad adaptation for Southeast Asia. Proceedings ofthe 18th Intematíonal Grass lands Congress, Canada, June 8-19, 1997. Session 1, 51- 52. Jones, P.G., Sawkins, M.C., Maass, B.L. and Kerridge, P.C. 1997. Gis and genetic diversity - Case studies in Stylosanthes. Proceedings of the 18th Intemational Grasslands Congress, Canada, June 8- 19, 1997. (inpress). Jones, P.C.; Rebgetz, Ross; Maass, B.L.; Kerridge, P.C. 1996. Genetic diversity in Stylosanthes species: A GIS mapping approach. Paper presented at the VIII Congreso sobre el Cerrado and the First Intemational Symposium on Tropical Savannas, Brasília, DF, Brazil. 24-29 March 1996. 5 p. Kerridge, P.C. and Franco, L.H. 1997. The use of legumes for improvement of fallow land in smallholder farmi ng systems. Proceedings of the 18th Intemational Grass lands Congress, Canada, June 8- 19, 1997. Session 16, 25-26. Lascano, C. and F. Holmann, eds.. 1997. On-farm research methodologies for dual purpose cattle farms. Workshop proceedings. CIAT, Cal i, Colombia. Lascano C. E. Ruiz G.A . Velasquez 1. and Rozo J. 1997. Validation of Arachis pintoi as a forage legume in commercial dual-purpose cattle farms in forest margins o f Colombia. In : XVlli Int. Grass l. Cong. June 8-19, Winnipeg, Manitoba and Saskatoon. Saskatchewan., Canada. Session 24,3 1-32. Leihner, D. E., J . A. Castillo y K. M. Müller-Samann, 1997. Cuantificac ión y desarroll o de la erodabilidad de los inceptisoles Andinos en el suroccidente Colombiano. Paper presented at an Intemational Workshop he ld in CIA T, 13-17 October 1997 (proceedings in preparation). Maass Brigitte, Franco, L.H., Ramirez, G., Lascano, C.E. and Velasquez, J.E. 1997. Codariocalyx gyroides: A new forage option for the humid tropics. In: XVIII Int. Grassl. Cong. June 8-19, Winnipeg, Manitoba and Saskatoon, Saskatchewan, Canada. Session 1: 53-54 Müller-Samann; K.M . 1997. Hacia conceptos integrales en la conservación de suelos y aguas en la zona Andina: - más interacción. más adopción, más impacto. Paper presented at an intemational workshop held in CIA T , 13-1 7 of October. 1997 (proceedings in preparation). Müller-Samann, K. M. 1997. Labranza mínima en ladera. Una opción para un manejo más sostenible de suelos y aguas. Conference he ld at the CORPOICA/PRONATA training course on June 17'h, 1997 "Curso de Manejo y Conservación de Suelos bajo el Enfoque de Sistemas de Producción". Müller-Samann, K. M. 1997. Agronomic options for soil conservation based on results from research and extension. Conference held at the Hohenheim University, post graduate training course "Soil Degradation and Conservation in the Tropics". July 10-30, 1997. Müller-Samann, K. M. , Castillo. J. A. , 1997. Experiencias en conservación de suelos en el Centro Internacional de Agricultura Tropical. CIA T . Conference held at "II Semana Técnica de Ingeniería Agrícola". Universidad Nacional de Colombia, Sede Bogotá, 5-8 de Noviembre. 1997. 5 Publications Mullen, B.F., Shelton, H.M ., Gabunada, F., and Stür, W.W . 1997. Psyllid res istance in the leucaena genus. Proceedings of the 18th Intemational Grasslands Congress, Canada, 1 une 8-19, 1997. Session 6, 23- 24. Nursahramdani, E .. and Ibrahim, Ir., 1997. FSP activities and results in East Kalimantan. In : W.W. Stür (ed) Feed Resources for Smallholder Livestock Production in Southeast Asia. Proceedings of a regional meeting in Vientiane. Lao POR, 16-20 January 1996. CIAT Working Document No. 156, Los Baños, Philippines, 41 -43. Pham Van Bien, Hoang Kim and R.H. Howeler. 1996. Cassava cultural practices in Vietnam. In : R.H. Howeler (Ed.). Cassava Production, Processing and Marketing in Vietnam. Proc. Works hop held Oct 29-31, 1992 in Hanoi, Vietnam. pp. 58-97. Phaikaew, C., Guodao, L., Abdullah, A, Tuhulele. M., Magboo, E., Bouahom, B. and Stür, W.W. 1997. Tropical forage seed production in southeast Asia: Current status and prospects. Proceedings of the 18th International Grasslands Congress, Canada, June 8-1 9, 1997. Session 28, 7-8. Phaikaew, C., Nakamanee, G. , and Klum-em, K. 1997. FSP activities in Thailand. In: W.W. Stür (ed) Feed Resources for Smallholder Livestock Production in Southeast Asia. Proceedings of a regional meeting in Vientiane, Lao POR, 16-20 January 1996. GAT Working Oocument No. 156, Los Baños, Philippines, 49-50. Phengvichith, V. and Hacker, J.B . Diversity and Sustainabi1ity of the Pek Savannas of the Lao PDR. Proceedings of the 18th International Grasslands Congress, Canada, June 8- 19, 1997. Session 1, 9- 10. Phimphachanvongsod, V. 1997. FSP activities in Lao POR. In : W.W. Stür (ed) Feed Resources for Smallho1der Livestock Production in Southeast Asia. Proceedings of a regional meeting in Vientiane, Lao POR, 16-20 January 1996. CIAT Working Document No. 156, Los Baños, Philippines, 45-46. Schmidt Axel, Lascano C.E., Shultze-Kraft R. and Maass B. 1997. An approach to define G x E interactions in a core collection of Desmodium ovalifolium. In: XVill Int. Grass l. Cong. June 8- 19. Winnipeg, Manüoba and Saskatoon, Saskatchewan, Canada (in press). Stür , W.W. (ed.) 1997. Feed Resources for Smallholder Li vestock Production in Southeast Asia. Proceedings of a regional meeting in Vientiane, Lao POR, I 6-20 January 1996. CIA T Working Document No. 156, Los Baños, Philippines. 93 pp. Tuhulele, M. 1997. FSP activities and resu1ts in Indonesia. In: W.W. Stür (ed) Feed Resources for Smallho1der Li vestock Producti on in Southeast As ia. Proceedings of a regional meeting in Vientiane, Lao POR, 16-20 January 1996. CIAT W orking Document No. 156, Los Baños, Philippines, 37-40. Book chapter Howe1er, R. H. 1996d. Diagnosis of nutritional disorders and soil fertility management of cassava. In : G .T. Kurup et al. (Eds.). Tropical Tuber Crops. Prob1ems. Prospects and Future Strategies. Oxford and ffiM Publishing Co., New Delhi, India. pp. 181-1 93. 6 Publications Stür, W.W. 1997. Pasture seed production and marketing. (1997). In: J .V. Lovett and J.M. Scott (eds) Pasture Production and Management. Inkata Press, Melbourne, Australia, pp. 155-170 Other publications: Müller-Samann, K. M. 1997. La erosión de los suelos y sus efectos en la producción agrícola en zonas de ladera. "Alternativas tecnológicas para disminuir el efecto de la erosión en cultivos de ciclo corto". (16 Sept., 1997). FIDAR/U-Hohenheim training course on Environmental Education for Teachers of Rural Colleges. 15-18 Sept. 1997. Santander de Quilichao. Müller-Samann, K. M. 1997. Conservación de suelos en fincas de ladera, una cosa sencilla y muy difícil de hacer. Conference held at the training course of GTZJIICA. Orientación de la investigación agraria hacia el desarrollo alternativo. Müller-Samann, K. M. 1997. Propuesta técnica para la producción de mora (Rubus glaucus) y de pasto Imperial como componentes de conservación de laderas. (Recommendations for the production of Blackberry (Rubus glaucus) and of Imperial grass (Axonopus scoparius) as components for soil conservation in hillsides. Documento de Trabajo No. 3, Proyecto de Conservación de Suelos. CIAT/Universidad de Hohenheim, 1997. pp. 16. Villada, D., Cifuentes, R.O., Ayala, V., Doering, H., Müller-Samann, K. M. 1997. Uso de Brachiaria dictyoneura (CIA T No. 6 133) y de Desmodium ovalifolium (CIA T No. 13089) en la estabilización de cárcavas y terrenos degradados (the use of B. dictyoneura and D. ovalifolium in the stabilization of gullies and on degraded slopes). Documento de Trabajo No. 4. Proyecto de Conservación de Suelos. CIAT/Universidad de Hohenheim, 1997, pp. 19 (also accepted for publication in the journal of the Colombian Soi l Conservation Network, "CONSERVANDO"). 7 AusAID Funding: Main contact:: List of Donors Forages for Smallholders Project (FSP) Mr. Keith England Manager, South East Asia Regional Section AusAID GPO Box 887, Canberra ACT 2601 Australia Oxford Forestry Institute (OFI) List of Donors Funding: Providing supplementary funds to FSP for added e mphasis on tree Jegumes with funds from DFID (Philippins and Central America) Main contact:: Mr. Alan Pottinger Oxford Forestry Institute Dept. of Plant Sciences, University of Oxford South Parks Road, Oxford OX 1 3RB United Kingdom University of Queensland Funding: Supporting Leucaena evaluations as part of ACIAR PN 9433 ("New Leucaenas for Southeast Asían, Pacific and Australian Agriculture Main contact:: Max Shelton A CIAR Funding: Main contact: : Department of Agriculture T he University of Queensland Brisbane, Qld 4072 Australia Supporting publication of the booklet "Field experiments with forages and e . )ps" Mr. Peter Lynch A CIAR GPO Box 1571 Canberra, ACT 260 1 Australia AusAID - Laos office Funding: Supporting publication on the booklet "Field experiments with forages and crops" through a community activities scheme grant Main contact:: Mr. Robert Jauncey AusAID Australian Embassy, Vientiane Lao PDR The Nippon Foundation, Tokyo, Japan Funding: Improving agricultura! sustainability in Asia Main contact: Ms Ayako Sono Chairperson The Nippon Foundation Senpaku Shinko Bldg 1-15-16 Toranomon, Minato-ku, Tokyo 105 Japan Colombia- The Government of Colombia Funding: Production systems components Main contact: Dr. Alvaro Uribe Nestlé of Colombia Funding: Main contact: Director Unidad Estudios Agrarios Departamento de Planeación Nacional Santafé de Bogotá, Colombia Improved legumes and grasses for small milk producers in Caquetá Dr. Miguel Jervis P.O. Box 5959 Santafé de Bogotá, Colombia Systemswide Livestock Program Funding : Legume-based forage systems for dual-purpose cattle Main contact: Dr. Hugo Li-Pun BMZ Funding: Main contact: Intemational Livestock Research Center, ILRI P.O. Box 5689 Addis Ababa, Ethiopia Soil conservation Dr. Jochen de Haas Agriculture and Rural Development Federal Ministry of Econornic Cooperation and Development Friedrich - Ebert Allee 40 D-53113 Bonn Germany Interamerican Development Bank (BID) Funding: Legume-based forage systems for dual-purpose cattle farm Main contact: Dr. Rubén Echeverria Econornist, Agriculture Div. PRA 1300 New York A ven u e NW Washington DC 20577 USA 2 List of Donors StaffList Senior Staff PETER C. KERRIDGE, Ph.D., Forage Agronomist, Project Leader SAMUEL FUJISAKA, Ph.D., Agricultura{ Anthropologist FEDERICO HOLMANN, Ph.D., Animal Scientist, agricultura{ economist PETER HORNE, Ph.D., Forage Agronomist (Stationed in Vientiane, Lao PDR- CSIRO) REINHARD HOWELER, Ph.D., Agronomist (Stationed in Bangkok, Thailand) CARLOS E. LASCANO, Ph.D., Ruminant Nutritionist KARL MUELLER-SAEMANN, Ph .D., Agronomist (University of Hohenheim) WERNER STÜR, Ph.D., Forage Agronomist (Stationed in Los Baños, Philippines) Consultant PEDRO ARGEL, Ph .D., Agronomist (Stationed in San José Costa Rica) Research Fellow ARJAN GUSMAN, Ph.D., Soil Scientist/Modeller Research Associates PHANOR HOYOS, Agronomist LIBARDO RIV AS, Economist Research Assistants PATRICIA A VILA, Zootechnist JESUS CASTILLO, Agronomist GERMAN ESCOBAR, Bio1ogist FRANCISCO GABUNADA Jr. LUIS H. FRANCO, Agronomist LILIANA HURTADO OTONIEL MADRID, Statistician NORBEY MARIN, Statistician (1eft October 1997) EMMA LUISA ORENCIA PHONEPASEUTHPHENGSAVANH GERARDO RAMIREZ, Statistician KENETH REA TEGUI, Agronomist GUSTAVO RUIZ, Zootechnist DINH XUAN TUNG (1eft in Ju1y 1997) Administrative Assistant MARIA DIADEMA BONll..LA Staff List Technicians JOSE ADARVE JOSE GALLEGO MILLER GOMEZ HERMANUSMA Secretaries BLANCA NHORA AGUIRRE YUVIZA BARONA CHINT ANA CHANHDENG MARGARITA PULGARIN Staff List 2