agriculture Article Combined Effects of Drought and Soil Fertility on the Synthesis of Vitamins in Green Leafy Vegetables Taewan Park 1, Sahrah Fischer 1 , Christine Lambert 2 , Thomas Hilger 1,* , Irmgard Jordan 3 and Georg Cadisch 1 1 Institute of Agricultural Sciences in the Tropics (Hans-Ruthenberg-Institute), University of Hohenheim, 70593 Stuttgart, Germany 2 Department of Nutritional Science, University of Hohenheim, 70593 Stuttgart, Germany 3 Alliance Bioversity International and CIAT, Food Environment and Consumer Behaviour Lever, 00621 Nairobi, Kenya * Correspondence: thomas.hilger@uni-hohenheim.de Abstract: Green leafy vegetables, such as Vigna unguiculata, Brassica oleraceae, and Solanum scabrum, are important sources of vitamins A, B1, and C. Although vitamin deficiencies considerably affect human health, not much is known about the effects of changing soil and climate conditions on vegetable vitamin concentrations. The effects of high or low soil fertility and three drought intensities (75%, 50%, and 25% pot capacity) on three plant species were analysed (n = 48 pots) in a greenhouse trial. The fresh yield was reduced in all the vegetables as a result of lower soil fertility during a severe drought. The vitamin concentrations increased with increasing drought stress in some species. Regardless, the total vitamin yields showed a net decrease due to the significant biomass loss. Changes in vitamin concentrations as a result of a degrading environment and increasing climate change events are an important factor to be considered for food composition calculations and nutrient balances, particularly due to the consequences on human health, and should therefore be considered in agricultural trials. Keywords: drought; environment; food composition; food security; human health; nutrition security; plant nutrition; soil fertility; vitamin A; vitamin B1; vitamin C Citation: Park, T.; Fischer, S.; Lambert, C.; Hilger, T.; Jordan, I.; Cadisch, G. Combined Effects of Drought and Soil Fertility on the Synthesis of Vitamins in Green Leafy 1. Introduction Vegetables. Agriculture 2023, 13, 984. Changes in crop growth conditions do not only affect yields, they also influence the https://doi.org/10.3390/ nutrient content of the crop. In areas with a high prevalence of nutritional deficiencies in the agriculture13050984 human population, even small differences in the amounts of vitamins due to less favorable Academic Editor: Marcin Rapacz growing conditions may influence the deficiency rate. Low (pro-)vitamin A (including pro-vitamin A carotenoids, such as β-carotene) intake is one of the most significant threats Received: 14 March 2023 to health in Sub-Saharan Africa (SSA). In this region, up to 20% of children below five Revised: 22 April 2023 years of age suffered from a vitamin A deficiency in 2016 [1]. Deficient intake of vitamin Accepted: 27 April 2023 A impairs visual functions and growth, and increases the risk for respiratory infections, Published: 29 April 2023 which increases the risk of morbidity and mortality [2]. Vitamin B1 (thiamine) is a com- mon deficiency in staple food-dominated diets, exacerbated by practices such as milling grains [3]. A vitamin B1 deficiency can lead to severe neurological and cardiovascular Copyright: © 2023 by the authors. effects and even death [3]. Vitamin C (ascorbic acid) deficiency is relatively common in Licensee MDPI, Basel, Switzerland. middle- to low-income countries. In Kampala, Uganda, for example, 70% of pregnant This article is an open access article women were deficient in vitamin C [4]. Vitamin C deficiencies in the diet, in addition to the distributed under the terms and negative effects of an actual vitamin C deficiency, may contribute to a lower bioavailability conditions of the Creative Commons of essential minerals from diets, such as Fe and Zn [5]. Attribution (CC BY) license (https:// Green leafy vegetables are rich in micronutrients and other phytochemicals, such as creativecommons.org/licenses/by/ vitamins, which are vital for human health. Many studies show the benefits of increasing 4.0/). green leafy vegetable consumption on human health. An increased and regular intake of Agriculture 2023, 13, 984. https://doi.org/10.3390/agriculture13050984 https://www.mdpi.com/journal/agriculture Agriculture 2023, 13, 984 2 of 15 green leafy vegetables is, for example, associated with a lower risk for cardiovascular dis- eases (CVDs) [6]. The authors of a nutrition study in Tanzania confirmed that an increased consumption of green leafy vegetables can contribute to a lower prevalence of anemia and micronutrient deficiencies, particularly in resource-poor communities [7]. Plants contain a wide range of metabolites that have antioxidant potential [8,9] and are pre-cursors of vitamins, essential to human health. Plants produce co-enzymes and antioxidants, such as ascorbic acid, thiamine, and β-carotene, for a specific purpose [10], e.g., to protect themselves from environmental stressors, such as drought or poor soil fertility. Drought causes the production of reactive oxygen species (ROS), which, in turn, can lead to oxidative damage to the photosynthesis apparatus, and, hence, to reduced primary production [11]. Different carotenes and ascorbic acid are the main ROS detoxifying compounds [10,11]. Thiamine is also an important component of plant stress response, in addition to its other functions [12]. For lack of a unifying term in plant nutrition, the vitamin pre-cursors will be referred to as vitamins in this study. Soil fertility provides the basis of plant production by providing plants with sufficient nutrients and water. Varying levels of soil fertility affect food yields, as well as food quality in terms of mineral nutrient concentrations [13]. Furthermore, vitamin concentrations were found to be inversely affected by nitrogen (N) fertilizer application [14,15], thereby suggesting the possibility that soil fertility could also affect vitamin concentrations. Climate change has led to an increase in extreme weather events, such as prolonged and more frequently occurring droughts [16]. While the effects of different climate change variables (increasing temperatures, drought, increasing CO2 and O3) on mineral and macronutrient concentrations in different plant parts have been measured [17–19], the effects on vitamin concentrations have not been covered thus far. Both climate change and soil degradation represent two of the biggest challenges for the sustained production of high quantity and quality foods [20], while concurrently being partially responsible for both occurrences. This paper will focus on the following research question: (i) Does soil fertility, (ii) water stress (drought), and the combination of drought stress and varying soil fertility significantly affect the contents of β-carotene, ascorbic acid, and thiamine in the leaves of three green leafy vegetables (Brassica oleraceae L., Solanum scabrum L., Vigna unguiculata L.)? As the production of vitamins is a stress reaction, especially to stressors such as drought, an increase in vitamin concentrations in all green leafy vegetables with increasing drought is expected. Soil fertility is key for plant nutrient intake, and largely provides the nutrients needed to produce vitamins. Therefore, it is expected that in high soil fertility, the vitamin concentrations will be higher in all the drought severities than in low soil fertility of the same drought severity. The fresh weight is expected to decrease with increasing drought and decreasing soil fertility. 2. Materials and Methods This study was part of the project “Education and Training for Sustainable Agriculture and Nutrition in East Africa (EaTSANE)”. Since the main research area for the EaTSANE project was East Africa, namely the areas of Teso South, Kenya (a region with low soil fertility) and Kapchorwa, Uganda (a region with high soil fertility), the trial was set up to mimic the local conditions. Three local green leafy vegetables were analysed: sukuma wiki (Brassica oleraceae L.), black nightshade (Solanum scabrum L.), and cowpea (Vigna unguiculata L.), using soils with properties similar to the soils found in the East African research areas. The temperature in a greenhouse of the University of Hohenheim, Germany was adjusted to 22 ◦C, which is the mean of the average temperature during the growing season in Kapchorwa and Teso South (Table S1). Daylight was allowed from 6 am to 6 pm, as in the target areas. The greenhouse temperature and humidity were monitored during the entire trial using a TGP-4500 Tiny Tag Plus 2 (Gemini Data Loggers, Ltd., Chichester, UK). Two soils were selected to represent the soils of Kapchorwa, Uganda and Teso South, Kenya [21]. The first soil collected showed similarities to the soils with lower Agriculture 2023, 13, 984 3 of 15 fertility ferralsols in Teso South and was classified as an endostagnic alisol collected in Tauchenweiler, Germany (48◦47′13.3” N and 10◦02′18.9” E). In this paper, the soil with lower fertility will be referred to as “infertile”. The second soil, comparable to a higher fertility nitisol, such as from Kapchorwa, Uganda, was an endoleptic cambisol, and was collected in Höwenegg, Germany (47◦54′55.7” N and 8◦44′25.7” E) (Table S3). In this paper, the soil with higher fertility will be referred to as “fertile”. The soils gathered in Germany show similar properties to the soils of Teso South, Kenya and Kapchorwa, Uganda. The low fertility soils both feature sandy soils, whereas the high fertility soils feature loamy clays. While the pH of both of the low fertility soils is acidic (German alisol: 4.0; Kenyan ferralsol: 4.94), the pH of the high fertility soil was 5.6 in both cases [13]. Both of the low fertility soils showed a very low amount of soil organic matter, whereas the soils of higher fertility showed a higher amount. The three green leafy vegetables were planted into pots on 17 September 2019. The plastic pots (37.5 cm height and 16 cm diameter) were filled with 1.5 kg gravel as drainage, with a depth of 5 cm. The two soil types were dried and sieved (0.9 cm2 sieve type), and 5.5 kg of each soil was placed into the different pots. After sowing, the soil was covered by a thin layer of sand (~10 mm, 80 g) to prevent excessive evaporation and soil cracking. Each pot was sown with four seeds of each green leafy vegetable, and then later thinned out, leaving one seedling per pot. The pot water capacity (PC) was analysed using the gravimetric methods provided by [22]. Three treatments in the watering regime were used: 75% PC as control, 50% PC as mild stress, and 25% as severe drought stress. A total of 144 pots were used, 48 per plant species, with two soil fertilities (fertile and low fertility soil), and three drought intensities (control 1, control 2, mild, and severe) (Table S2). The double control group was used to improve the statistical power [23]. Each treatment had six replicates organised into a randomized complete block design (Figure S1). After plant germination, all the pots were weighed and watered every two days to maintain the assigned drought conditions. Fifty-one days after sowing, flower buds of the control plants became visible. The entire aboveground part of all the pots was harvested and frozen after recording their fresh weight. The belowground biomass (roots) were washed, then dried with paper towels, and the fresh weight recorded. The fresh leaves were stored at −80 ◦C and then lyophilized using a freeze dryer (LyoQuest laboratory freeze dryer, Telstar, Spain) for 24 h, and the dry weight was recorded. The samples were then ground, homogenized, and stored in the dark at −20 ◦C. The plant water use efficiency (WUE) was calculated as the ratio between the above- ground biomass (yield) and water use (evapotranspiration). WUE is generally high when the plants are exposed to drought conditions or are drought tolerant [24], and is, therefore, used to evaluate drought resistance [25]. The WUE of the yield was determined by the division of the fresh yield by water consumption [24]. The concentrations of thiamine, β-carotene, and ascorbic acid were measured in the leaf samples using high performance liquid chromatography (HPLC) at the Institute of Nutritional Sciences, University of Hohenheim. The HPLC data were recorded and anal- ysed using the Shimadzu LabSolutions Software (Version 5.54, Shimadzu Deutschland GmbH, Duisburg, Germany). The chromatographic analysis of the vitamins was conducted using the Shimadzu HPLC system. The Shimadzu HPLC system consisted of a DGU 20A3 Degassing Unit, an LC-20AT Pump, a SIL-20AC HT AutoSampler, and a CBM-20-A Commu- nication Module. The method for thiamine analysis was based on the European Standard (DIN EN 14122:2014) of vitamin B1 determination [26]; however, minor adaptations were made. β-carotene was measured using the method of [27]. The retinol equivalents (REs) of β-carotene were calculated using a factor 1 µg RE = 6 µg β-carotene [28]. Ascorbic acid was measured using the method of [29]. Since C3 plants close their stomata to reduce water loss during water-limiting condi- tions, 13CO2 fixation is decreased and δ13C is discriminated due to reduced CO2 diffusion in and out of the leaves [30]. Therefore, δ13C isotope discrimination has been used to measure Agriculture 2023, 13, 984 4 of 15 water stress in C3 plants [31]. The δ13C contents of green leafy vegetables were analysed by comparing the control (75% PC) to the drought treatments (50% and 25% PC) using randomly selected subsamples (n = 4 per PC of each species) of the green leafy vegetable samples. The selected subsamples were measured with a Euro EA Elemental Analyser (Euro Vector, Pavia, Italy) coupled to a Finnigan Delta IRMS (Thermo Fischer Scientific, Waltham, MA, USA) at the core facility of the University of Hohenheim. The fulfilment of the recommended nutrient intake (RNI) of vitamin A, ascorbic acid, and thiamine by consuming 150 g of green leafy vegetables (fresh weight) as an average serving size was calculated for female adults (19–50 years) [28]. SAS (SAS® University Edition, SAS Institute Inc., Cary, NC, USA) was used for the statistical analysis. The fresh yield, belowground biomass, number of nodules, irrigation water added, water use efficiency (WUE), pot capacity (PC), thiamine, β-carotene, and ascorbic acid were subjected to analysis of variance (ANOVA) for each treatment. ANOVA using PROC GLIMMIX was used to compare the treatments between the plant species and the significance of factors by an F-test at α = 0.05. A two-factorial model was fitted as an equation (SUPPL MAT). The interaction of the treatment and species was significant; therefore, the cell means were compared using a SLICE statement with the SLICEBY options, i.e., SLICEBY = treatments and SLICEBY = species, in the GLIMMIX procedure. The means of yield and WUE differences between the different soil fertility and drought conditions were compared. The vitamin concentrations were expressed on a fresh weight basis. The absolute vitamin amounts (mg per pot) were calculated for thiamine, β-carotene, and ascorbic acid by the multiplication of the fresh leaf yield (g/pot) by the vitamin content (mg/100 g FW) and divided by 100 [32]. The vitamin data were used to analyse whether the changed vitamin contents compensate for the treatments’ changed yield. An analysis of covariance (ANCOVA) using PROC GLM was used to compare the treatment means adjusted for a covariate soil fertility within the plant species. 3. Results 3.1. Plant Yield Analysis All the fresh leaf yields of the three green leafy vegetables were significantly higher (B. oleraceae: p < 0.01, V. unguiculata: p < 0.001, S. scabrum: p < 0.0001) in fertile soil than in low fertility soil with the same watering regime, with the exception of B. oleraceae with a 75% pot capacity (PC), where the difference was not significant (Table 1). Regardless of the soil fertility, the fresh leaf yields of the green leafy vegetables decreased with an increasing level of drought (Table 1 and Figure S1). However, the yield losses caused by drought did not differ significantly between the fertile and low fertility soils (Figure 1). Table 1. Fresh leaf yield (g/pot), belowground biomass (g/pot FW), nodule number per pot, total irrigation water added (ml/pot), and water use efficiency (WUE: fresh yield/water use, g/L) in Vigna unguiculata, Brassica oleraceae, and Solanum scabrum under two soil fertilities with three watering regimes (75% pot capacity (PC), 50% PC, and 25% PC). Fresh Yield Belowground No. of Irrigation Water WUE (g/pot) (g/pot FW) Nodules Added (mL/pot) (g/L) Vigna unguiculata Fertile soil 75% PC 25.2 ± 1.3 c 12.0 ± 1.1 bc 11.2 ± 3.4 a 2882 ± 115 a 8.8 ± 2.0 def 50% PC 18.6 ± 1.9 d 9.0 ± 1.5 cd 6.3 ± 2.1 b 1871 ± 162 bc 10.1 ± 2.9 cdef 25% PC 12.2 ± 1.9 ef 5.2 ± 1.5 de 1.2 ± 0.5 bc 832 ± 162 de 15.3 ± 2.9 bcd Infertile soil 75% PC 13.8 ± 1.3 e 2.0 ± 1.1 ef 0.0 c 2192 ± 165 b 6.3 ± 2.0 ef 50% PC 7.9 ± 1.9 fg 1.0 ± 1.5 ef 0.0 c 1239 ± 163 d 6.2 ± 2.9 ef 25% PC 2.5 ± 1.9 h 0.9 ± 1.5 f 0.0 c 405 ± 162 ef 8.8 ± 2.9 cdef Agriculture 2023, 13, 984 5 of 15 Table 1. Cont. Fresh Yield Belowground No. of Irrigation Water WUE (g/pot) (g/pot FW) Nodules Added (mL/pot) (g/L) Brassica oleraceae Fertile soil 75% PC 41.5 ± 1.7 a 9.4 ± 1.3 c - 2842 ± 142 a 14.9 ± 2.5 bcd 50% PC 32.8 ± 2.1 b 10.2 ± 1.6 c - 2080 ± 179 bc 16.4 ± 3.2 bc 25% PC 12.6 ± 2.1 ef 1.4 ± 1.6 ef - 398 ± 179 ef 43.2 ± 3.2 a Infertile soil 75% PC 40.3 ± 1.4 a 3.9 ± 1.1 ef - 2928 ± 120 a 13.9 ± 2.1 bcde 50% PC 15.0 ± 1.9 de 1.0 ± 1.5 ef - 1204 ± 162 d 11.8 ± 2.9 cdef 25% PC 4.7 ± 1.9 gh 0.6 ± 1.5 f - 342 ± 162 f 14.1 ± 2.9 bcde Solanum scabrum Fertile soil 75% PC 35.0 ± 1.3 b 21.5 ± 1.1 a - 2844 ± 115 a 12.4 ± 2.0 cde 50% PC 25.9 ± 1.9 c 15.4 ± 1.5 b - 1783 ± 162 c 14.8 ± 2.9 bcd 25% PC 17.4 ± 1.9 de 10.4 ± 1.5 c - 917 ± 162 d 20.6 ± 2.9 b Infertile soil 75% PC 16.5 ± 1.3 de 3.0 ± 1.1 ef - 1804 ± 115 c 9.0 ± 2.0 cdef 50% PC 4.8 ± 1.9 gh 0.9 ± 1.5 f - 1013 ± 162 d 4.8 ± 2.9 f 25% PC 1.1 ± 1.9 h 0.7 ± 1.5 f - 209 ± 162 f 14.4 ± 2.9 bcd Analysis of variance (ANOVA) using PROC GLIMMIX was used to compare the treatments between plant species and the significance of factors by an F-test at α = 0.05. LS-means ± SE within the column with the same letter are not significantly different (p < 0.05). B. oleraceae showed the highest yield loss with increasing drought regardless of the soil fertility, as evidenced in the steeper slope increase in the linear regression (Figure 1). Nevertheless, B. oleraceae obtained a higher total fresh yield under no drought (control, 75% PC) (p < 0.0001) and mild drought (50% PC) (p < 0.0001) conditions than the other two green leafy vegetables (Table 2). In contrast, S. scabrum had the highest yield loss with decreasing soil fertility under all watering regimes, shown by the highest interceptor differences between the regression lines of fertile and low fertility soils, while B. oleraceae showed the lowest yield loss by different soil fertility (Figure 1). The belowground biomass was significantly lower in the low fertility soil than in the fertile soil for all the green leafy vegetables (all p < 0.0001) (Table 1). Only severe drought (25% PC) significantly (p < 0.05) decreased the belowground fresh weight of all the species in the fertile soil (Table 1). Table 2. Levels of β-carotene contents and retinol activity equivalents (RAE) of V. unguiculata, B. oleraceae, and S. scabrum under two soil fertilities with three watering regimes (75% pot capacity (PC), 50% PC, and 25% PC). β-Carotene RAE β-Carotene RAE (mg/100 g FW) (µg/100 g FW) (mg/Fresh Yield) (µg/Fresh Yield) Vigna unguiculata Fertile soil 75% PC 7.37 ± 0.26 bc 614 ± 21 bc 1.86 ± 0.09 bc 155 ± 7 bc 50% PC 7.64 ± 0.36 bc 636 ± 30 bc 1.42 ± 0.13 de 118 ± 10 de 25% PC 8.04 ± 0.36 b 670 ± 30 b 0.97 ± 0.13 f 80 ± 10 f Infertile soil 75% PC 6.89 ± 0.26 cd 574 ± 21 cd 0.97 ± 0.09 f 81 ± 7 f 50% PC 6.41 ± 0.36 de 534 ± 30 de 0.57 ± 0.13 g 47 ± 10 g 25% PC 4.92 ± 0.36 gh 410 ± 30 gh 0.12 ± 0.13 h 10 ± 10 h Brassica oleraceae Fertile soil 75% PC 4.02 ± 0.32 hi 335 ± 26 hi 1.66 ± 0.11 cd 138 ± 9 cd 50% PC 4.11 ± 0.40 ghi 343 ± 33 ghi 1.36 ± 0.14 de 114 ± 12 de 25% PC 3.66 ± 0.40 i 305 ± 33 i 0.46 ± 0.14 gh 38 ± 12 gh Agriculture 2023, 13, 984 6 of 15 Table 2. Cont. β-Carotene RAE β-Carotene RAE (mg/100 g FW) (µg/100 g FW) (mg/Fresh Yield) (µg/Fresh Yield) Brassica oleraceae Infertile soil 75% PC 4.03 ± 0.27 hi 336 ± 22 hi 1.62 ± 0.10 cd 135 ± 8 cd 50% PC 3.48 ± 0.36 i 290 ± 30 i 0.55 ± 0.13 g 46 ± 10 g 25% PC 3.61 ± 0.36 i 301 ± 30 i 0.17 ± 0.13 h 14 ± 10 h Solanum scabrum Fertile soil 75% PC 7.59 ± 0.26 bc 632 ± 21 bc 2.66 ± 0.09 a 221 ± 7 a 50% PC 8.09 ± 0.36 b 674 ± 30 b 2.11 ± 0.13 b 176 ± 10 b 25% PC 9.48 ± 0.36 a 790 ± 30 a 1.62 ± 0.13 cd 135 ± 10 cd Infertile soil 75% PC 7.08 ± 0.26 cd 590 ± 21 cd 1.17 ± 0.09 ef 97 ± 7 ef 50% PC 5.99 ± 0.36 ef 499 ± 30 ef 0.31 ± 0.13 gh 26 ± 10 gh 25% PC 5.13 ± 0.45 fg 428 ± 37 fg 0.06 ± 0.16 h 5 ± 13 h Analysis of variance (ANOVA) using PROC GLIMMIX was used to compare the treatments between plant species Agriculture 2023, 13, x FOR PEER REVIEW 6 of 16 and the significance of factors by an F-test at α = 0.05. LS-means ± SE within the column with the same letter are not significantly different (p < 0.05). Figure 1. Fresh yFieilgdu (rge/p1o. tF) roefs they igerleden(g l/eapfoy tv)eogfetthabelegsr eBe. nolleeraacfeyaev, eVg. eutnagbuleicsulBa.tao,l earnadc eSa.e s,cVab. ruunmg,u iculata, and S. scabrum, along with the paolto ncagpwaciitthy t(h%e) pfort cthape afceirtyile( %(L))f aonrdth ienfertiille (SL)) saonild. Ainnfaelrytislies (oSf) csovila.rAiantea l(yAsNis-of covariate (ANCOVA) COVA) was perfworamsepde. rTfhoer mreegdre.sTsihoen rleingerse isns ieoanchli npleasnitn speeaccihesp alraen ptasrpaellceile. sRaegrerepssairoanl lfeolr.mRuelgarse ssion formulas and R2 and R² are givena. Treheg itvriepnle. Tashteertirsikp ilnedaiscatetersis skiginnidfiiccaantcees asti gpn