Sustainable use of land and water under rainfed and deficit irrigation conditions in Ogun-Osun River Basin, Nigeria

<strong>Summary</strong> Human population is increasing faster than ever in the history. There is an urgent need to scale up food production in order to meet up with food demands, especially in Sub-Saharan Africa. In Ogun-Osun River Basin, Nigeria, more than 95% of the crop production is done under rainfed conditions. Fluctuation in rainfall as a result of climate change is a major challenge in the recent times in the basin. Land productivity can be greatly improved by using affordable water conservation practices by peasant farmers who produce crops in the basin. Similarly, water saving measures would have to be adopted by using drip irrigation and application of water at critical stages of growth of crops. Fertility of the soil needs to be maintained by cultivating crops that naturally replenish soil nutrients. Such measures will go a long way in ensuring sustainable use of land and water in Ogun-Osun River Basin. An indeterminate cultivar of Soybeans TGX 1448 2E was cultivated at the Teaching and Research Farms of Obafemi Awolowo University, Ile-Ife, Nigeria during the rainy seasons from May to September, 2011 and June to October, 2012. Similarly, the crop was drip irrigated for two dry seasons from February to May in 2013 and from November, 2013 to February, 2014. The purpose of conducting the experiments in the rainy and dry seasons was to compare the yields and their components and to evaluate the performances of the crop in terms of water use and productivity. The experimental field during the dry season was located at about 1 km from the field used during the rainy season due to the nearness to the source of water. During the experiments in the four seasons, key biometric data of the crop were taken from emergence to physiological maturity. The crop cycle during the rainfed experiment lasted for 117 and 119 days in 2011 and 2012 respectively, while in the dry season it lasted for 112 days in the first season and 105 days in the second season. The lengths of the crop cycles in the four seasons differed a little bit. This is attributed to environmental factors such as weather conditions, nutrient availability in the soil and period of cultivation. During the rainy seasons, six water conservation treatments were used namely Tied ridge, Mulch, Soil bund, Tied ridge plus Soil bund, Tied ridge plus Mulch, Mulch plus Soil bund and Direct sowing without water conservation measure (conventional practice), which was the control treatment. The treatments were placed in a randomised complete block design with four replicates in an area of 31 by 52 m (1,612 m2) and standard agronomic measures were taken. Soil water balance approach was used in determining evapotranspiration during the rainfed and irrigation seasons. Seasonal evapotranspiration was partitioned into the productive transpiration from the plants and non-productive evaporation from the soil. Seasonal average canopy extinction coefficients were 0.46 and 0.51 respectively in the rainy seasons of 2011 and 2012, while in the dry seasons of 2013 and 2013/2014 they were 0.43 and 0.49. The plant height ranged from 51.3 cm for Soil bund to 67.8 cm for the conventional practice in 2011 while in 2012, it ranged from 60.3 cm for Tied ridge plus Soil bund to 80.3 cm for Mulch plus Soil bund. The minimum fraction of Intercepted Photosynthetically Active Radiation was 0.13 during establishment for Tied ridge plus Soil bund while the peak fraction was 0.97 during seed filling for Soil bund during the rainy seasons. Similarly, the minimum and peak leaf area indices were 0.13 m2 m-2 for Tied ridge plus Soil bund during establishment in 2011 and 6.61 m2 m-2 for Soil bund during seed filling in 2012. There were strong and significant correlations between the fraction of Intercepted Photosynthetically Active Radiation and the leaf area indices (LAI) (0.70 ≤ r2 ≤ 0.99) in 2011 and (0.93 ≤ r2 ≥ 0.99) in 2012 by using an exponential model. Seasonal rainfall in 2011 and 2012 was 539 and 761 mm respectively. Seasonal water storages in the soil in 2011 ranged from 407 mm for the conventional practice to 476 mm for Tied ridge plus Mulch, while in 2012 it ranged from 543 mm for Tied ridge to 578 mm for Tied ridge plus Soil bund. Radiation Use efficiency was determined by plotting dry above ground biomass measured at intervals of seven days against the Daily Photosynthetically Active Radiation from Solar radiation and the Instantaneous Photosynthetically Active Radiation measured near solar noon for all the treatments. For the Photosynthetically Active Radiation obtained from solar radiation, Radiation Use Efficiency of the crop ranged from 1.18 g MJ-1 for Tied ridge to 1.98 g MJ-1 of Intercepted Photosynthetically Active Radiation for Tied ridge plus Soil bund in 2011, while in 2012 it ranged from 1.45 g MJ-1 for Tied ridge to 1.92 g MJ-1 for Mulch. There was no significant difference in the average seasonal Radiation Use Efficiency in the two seasons. By using instantaneous measurement of the Photosynthetically Active Radiation, Radiation Use Efficiency ranged from 0.80 g MJ-1 of Intercepted Photosynthetically Active Radiation for Tied ridge to 1.65 g MJ-1 for Tied ridge plus Soil bund in 2011, while in 2012 it ranged from 0.94 g MJ-1 for Tied ridge to 1.24 g MJ-1 for Soil bund. The two approaches gave relatively similar values of Radiation Use Efficiency. Positive -correlation coefficients (0.50 ≤ r2 ≤ 0.89) were found among the treatments between the dry above ground biomass simulated by using a light model and those measured in the field in the two seasons. The seasonal crop water use ranged from 311 mm for Mulch plus Soil bund to 406 mm for Tied ridge plus Soil bund in 2011, while in 2012 it ranged from 533 mm for Mulch plots to 589 mm for Soil bund. Seasonal transpiration ranged from 190 mm for Tied ridge plus Mulch to 204 mm for Soil bund in 2011 while in 2012 it ranged from 164 mm for Tied ridge plus Mulch to 195 mm for Mulch plot. Seasonal evaporation was higher in 2012 ranging from 338 mm for Mulch plots to 408 mm for Soil bund while in 2011 it ranged from 311 mm for Mulch plus Soil bund to 406 mm for Tied ridge plus Soil bund. Water storage in the soil and seasonal crop water use are significantly related. Similarly, the seasonal crop water use, Intercepted Photosynthetically Active Radiation and Radiation Use efficiency were highly related for the crop over the two seasons. Marketable seed yield ranged from 1.68±0.50 t ha-1 for Tied ridge to 2.95±0.30 t ha-1 for Tied ridge plus Soil bund in 2011, while in 2012 the yield ranged from 1.64±0.50 t ha-1 for the conventional practice to 3.25±0.52 t ha-1 for Mulch plus Soil bund. In 2011, seed yield for Tied ridge plus Soil bund was 15.6, 15.9, 25.4, 28.5, 43.1 and 47.1% higher than seed yield for Mulch plus Soil bund, Soil bund, Mulch, Tied ridge plus Mulch, Tied ridge and conventional practice respectively. In 2012, seed yield for Mulch plus Soil bund was 7.4, 21.8, 32.0, 32.3, 43.7 and 49.5% higher than the seed yields for Soil bund, Tied ridge, Mulch, Tied ridge plus Mulch, Tied ridge plus Soil bunds and Direct sowing respectively. Average seasonal seed yield of the crop was significantly related to the Total Intercepted Photosynthetically Active Radiation but not to the Radiation Use Efficiency. Harvest indices ranged from 47.4±4.5% for Tied ridge to 57.6±1.1% for Tied ridge plus Soil bund in 2011 and 53.1±3.0% for Soil bund to 58.1±2.3% for Tied ridge 2012. The highest harvest indices were obtained in Tied ridge plus Soil bund and Tied ridge in 2011 and 2012 respectively. Harvest index was not significantly related to both Intercepted Photosynthetically Active Radiation and Radiation Use Efficiency of the crop. Average seasonal transpiration efficiencies - the ratio of the dry above ground biomass at harvest to the seasonal transpiration - for all the treatments were 7.0 kg ha-1 mm-1 in 2011 and 14.9 kg ha-1 mm-1 in 2012. Transpiration efficiency of the crop was strongly related to Intercepted Photosynthetically Active Radiation but not to Radiation Use Efficiency under field conditions in the rainy seasons. The peak water productivity for seed was 7.99 kg-1 ha-1 mm-1 in 2011 and 5.76 kg-1 ha-1 mm-1 for Mulch plus Soil bund in 2012. Water productivity for seed was strongly and significantly related to Intercepted Photosynthetically Active Radiation. However, it was not significantly related to Radiation Use Efficiency. These findings will provide information to the crop yield modellers during the simulation of yields of Soybeans under water conservation practices. The construction of ridges and Soil bund especially for Tied ridge, Mulch plus Soil bund and Tied ridge plus Soil bund increased the average seasonal cost of production by 28.9% compared with Mulch and conventional practice and by 10.1% compared with Soil bund. In addition, economic water productivity was 3.90 US$ ha-1 mm-1 for Mulch plus Soil bund while for Soil bund and conventional practice, it was 3.30 and 2.27 US$ ha-1 mm-1 respectively. Due to increase in demand for food, there is the need to produce more crop per drop of water under rainfed conditions and to manage water for agriculture at basin scale. The key priority in the study area was to increase the seed yields, water and economic productivity and the financial benefits at the end of a cropping season. The results show that the use of Mulch plus Soil bund had the average maximum transpiration efficiency, seed yield, water and economic productivity, and revenue of 1,630 US$ per ha. By comparing the average seasonal transpiration efficiency, crop water use, yield, water productivity and costs of production for the six conservation practices with those of the conventional practice in the two rainy seasons, Mulch plus Soil bund had the maximum average seed yield, water and economic productivity. Mulch plus Soil bund is hereby recommended for the cultivation of the crop in the study area. Other conservation practices, such as Soil bund, also performed satisfactorily in terms of seed yield and water productivity, although with a slight reduction in revenue. The use of these water conservation practices will not only increase the yields of the crop, but reduce depletion of water in the soil, which could initiate or increase land degradation in the study area to the barest minimum. Hence, sustainability of land and water in Ogun-Osun River Basin can be ensured. These findings demonstrate that land and water productivity of Soybean under rainfed conditions can be significantly improved with water conservation practices under the current fluctuations of rainfall and competition for land resources between agriculture and urban land use in Ogun-Osun River Basin. Field trials were also conducted for two irrigation seasons from February to May, 2013 and November, 2013 to February, 2014. The crop was planted in a Randomized Complete Block Design with three replicates and in-line drip irrigation was applied to supply water to the crops. Five treatments were selected and these are: (i) full irrigation, skipping of irrigation every other week during (ii) flowering; (iii) pod initiation; (iv) seed filling and (v) commencement of maturity. Biometric data, which are number of leaves, plant height, leaf area indices and dry above ground biomass, were taken and recorded every week from sowing until maturity in the two irrigation seasons. Soil moisture contents were taken at the root zone of the plants prior to irrigation in order to determine the net irrigation water requirements at each stage of growth. Harvest indices were determined for each treatment. Number of pods per plant, number of seeds per pod and yields under each treatment were determined after physiological maturity in each season. Regression equations were generated for: (i) yield; (ii) number of pods per plant; (iii) number of seeds per pod; (iv) number of leaves; (v) seasonal transpiration and leaf area indices. Similarly, regression equations were generated for: (i) plant heights; (ii) seasonal transpiration; (iii) number of pods per plant; (iv) number of seeds per pod; (v) dry above ground biomass. Linear regressions were also fitted to the yield, dry above ground biomass and seasonal crop water use. The crop response factor was determined. Water productivity and Irrigation water productivity were computed and compared for each treatment. Linear models were fitted to the water productivity, irrigation water productivity and harvest index. Rainfall contribution to the crop water use was 262 and 50 mm for 2013 and 2013/2014 irrigation seasons respectively. Maximum Leaf Area Index in the 2013 irrigation season was 7.10 m2 m-2 for full irrigation during seed filling, while in the 2013/2014 irrigation season, it was 3.44 m2 m-2 for full irrigation during flowering. The dry above ground biomass after maturity ranged from 359 g m-2 where irrigation was skipped every other week at the commencement of maturity to 578 g m-2 for full irrigation. The seed yields ranged from 1.81 t ha-1 when irrigation was skipped every other week during seed filling to 3.11 t ha-1 for full irrigation. Average seasonal seed yield for full irrigation was 18.8, 21.8, 24.4 and 47.9% higher than yields for treatments where irrigation was skipped every other week during flowering, pod initiation, commencement of maturity and seed filling respectively. Seasonal transpiration ranged from 217 mm when irrigation was skipped every other week during seed filling to 409 mm for full irrigation in the 2013 irrigation season, while in the 2013/2014 irrigation season it ranged from 28 mm for the treatment where irrigation was skipped every other week during seed filling to 223 mm for full irrigation. Seasonal crop water use ranged from 463 mm when irrigation was skipped every other week during flowering to 523 mm for full irrigation in the 2013 irrigation season, while in the 2013/2014 irrigation season it ranged from 364 mm when irrigation was skipped every other week during seed filling to 507 mm for full irrigation. Harvest indices ranged from 56.0% when irrigation was skipped during seed filling to 65.9% when irrigation was skipped during flowering in the 2013 irrigation season, while in the 2013/2014 irrigation season, it ranged from 43.2% when irrigation was skipped during seed filling to 63.9% for full irrigation. Water productivity for seed production ranged from 3.89 kg ha mm-1 when irrigation was skipped during seed filling to 5.95 kg ha-1 mm-1 for full irrigation in the 2013 irrigation season while in the 2013/2014 irrigation season, it ranged from 1.93 kg ha mm-1 when irrigation was skipped during seed filling to 3.00 kg ha-1 mm-1 for full irrigation. Irrigation water productivity ranged from 8.90 kg ha mm-1 when irrigation was skipped during seed filling to 14.0 kg ha-1 mm-1 when irrigation was skipped during flowering in 2013, while in the 2013/2014 irrigation season, it ranged from 2.24 kg ha-1 mm-1 when irrigation was skipped during seed filling to 3.32 kg ha-1 mm-1 for full irrigation. Leaf area indices and yield, number of leaves, number of pods per plant, number of seeds per pod and seasonal transpiration were significantly correlated. Similarly, dry above ground biomass and seasonal transpiration, number of pods per plant, number of seeds per pod were significantly correlated. The crop response factor (Ky), a measure of the relative decrease in seed yield due to relative decrease in evapotranspiration, was 2.24. It indicates that the deficit irrigation imposed on the crop was high and that relative decrease in yields due to deficit irrigation was higher than relative decrease in evapotranspiration. Results show that skipping of irrigation at any growth stage of the crop led to reduction in the leaf area indices, dry above ground biomass and seasonal crop water use. Deficit irrigation had significant effects on both the dry matter and yields. The effect of deficit irrigation was more pronounced on seed yields than on dry matter. Severity of the effects of deficit irrigation depended on the stage of growth and its duration. Deficit irrigation reduced significantly dry matter at flowering and pod initiation. However, deficit irrigation did not affect the plant height. Number of seeds per plant at flowering and commencement of maturity were reduced significantly by deficit irrigation. The number of seeds per pod was significantly reduced when irrigation was skipped at pod initiation only. Seed yields were significantly reduced when irrigation was skipped during seed filling. In the 2013 irrigation season water productivity when irrigation was skipped during flowering was 2.3, 16.1, 23.5, and 36.1% higher than water productivity for full irrigation, when irrigation was skipped during pod initiation, commencement of maturity and seed filling respectively. In the same season, irrigation water productivity when irrigation was skipped during flowering was 15, 20, 29.3 and 36.4% higher than for full irrigation, when irrigation was skipped during pod initiation, commencement of maturity and seed filling respectively. In the 2013/2014 irrigation season, however, water productivity for full irrigation was 8.7, 16.3, 24.7 and 35.7% higher than when irrigation was skipped during pod initiation, commencement of maturity, flowering and seed filling respectively. Similarly, irrigation water productivity was 7.2, 15.4, 24.1 and 32.5% higher than when irrigation was skipped during pod initiation, commencement of maturity, flowering and seed filling respectively. In addition, irrigation water productivity for full irrigation was 24.1 and 32.5% higher than when irrigation was skipped during flowering and seed filling respectively. Stage of growth, its duration, water requirements and seasonal environmental conditions influenced the seasonal water use, water productivity and irrigation water productivity of Soybean. Maximum water productivity and irrigation water productivity were obtained when irrigation was skipped every other week during flowering only in the first season, whereas in the second season full irrigation gave the peak water and irrigation water productivity. This suggests that irrigation water productivity of Soybean can be improved upon by skipping irrigation during flowering and pod initiation. In this study, the costs of production for all the irrigation scenarios were high. This is due to the high cost of water, which constituted between 54 to 59% of the production cost if water is purchased and cost of drip irrigation equipment, which constituted between 75.6 to 76.7% of the total cost of production if water would be given without financial implication. Under the prevailing price and economic conditions after harvest, the use of in-line drip irrigation does not offer economic benefit to peasant farmers, who are the predominant growers of the crop in the study area. Economic benefit may be achieved after long periods of usage with proper maintenance of the irrigation facilities and elimination of the fixed cost from the total cost of production. The water driven crop model AquaCrop was calibrated and validated to predict canopy cover, dry above ground biomass, seed yield, evapotranspiration, soil moisture content and water productivity of the crop. The simulated and measured data compare adequately except for water productivity that was over predicted in the validation data set. The AquaCrop model predicted canopy cover with error statistics of 0.93 ≤ E ≤ 0.98 for both full and deficit irrigation and the degree of agreement d = 0.99 with 4.3 ≤ RMSE ≤ 5.9 (root mean square error) for full irrigation while for deficit irrigation, 0.96 ≤ d ≤ 0.99 with 5.3 ≤ RMSE ≤ 5.8. Dry above ground biomass was predicted with error statistics of 0.08 ≤ RMSE ≤ 0.14 t ha-1 with 0.98 ≤ d ≤ 0.99 for full irrigation, while for deficit irrigation it was 0.06 ≤ RMSE ≤ 1.09 t ha-1 with 0.85 ≤ d ≤ 0.99. One in every five predictions of the above ground biomass was outside 20% deviation from the measured values. The seed yields were predicted with error statistics of RMSE = 0.10 t ha-1 and d = 0.99 and one in five predictions was outside 15% deviation from the measured data. The prediction error statistics for seasonal crop water use for both full and deficit irrigation treatments was 15.4 ≤ RMSE ≤ 58.3 in the two seasons. The AquaCrop model over predicted percolation also in the validation data set. These observations suggest that the percolation components of the model need to be adjusted to ensure better performance. The performance of the AquaCrop model in predicting canopy cover, seed yield and other quantities in this study are commendable and satisfactory. Specific and distinct features, such as the use of canopy cover rather than leaf area index, make the model suitable for developing countries like Nigeria, where researchers may not have access to state-of-the-art equipment for measuring the leaf area index. Similarly, water productivity that is normalized for atmospheric demand and carbon dioxide concentration and its focus on water makes it suitable for diverse locations. Over the years, it has been observed that no model is universal in its ability to take into consideration all differences in cultivar, environment, weather and management conditions. Other cultivars of Soybeans in Nigeria and other agro-climatic environments need to be tested and fine-tuned in the model, in order to ascertain the accuracy of the model. generally, the model predicted the stated parameters with reasonable degree of accuracy and is hereby recommended for use in Ile-Ife and other parts of Ogun-Osun River Basin and Nigeria. Although land, water, and economic productivity of the crop were higher where water was conserved under rainfed conditions, treatment of the soil to conserve water and regular maintenance increased the average seasonal cost of production compared with the conventional practice. High cost of production may reduce the benefits obtained by the crop growers, except when there is improvement in the market price. Therefore, sustainable practice of the water conservation measures must be accompanied with lower cost of production. Under irrigation conditions, the land and water productivity are lower compared with rainfed cultivation. The productivity in the dry season reduces with the severity of the water stress. Average crop water productivity and economic water productivity of all the six water conservation measures in the rainy season were higher than with full irrigation in the dry season. The costs of production of the crop in the dry season were significantly above the cost during the rainfed conditions. Higher water productivity under rainfed conditions in this study is in agreement with the finding that in a significant part of the least developed and emerging countries there is larger opportunity for improving water productivity under rainfed conditions compared to irrigated agriculture. Expansion of arable land may not be feasible in Ile-Ife because of the huge investments involved. Thus, the focus of efforts to expand food production in the area would have to be on raising land productivity on the existing arable lands and improving production efficiencies, outcomes that can only be achieved by using improved cultivars together with improved agronomic practices. Agronomic practices, especially under rainfed conditions, would have to be designed to improve water productivity. Improving water productivity requires vapour shift (transfer) whereby soil physical conditions, soil fertility, crop varieties and agronomy are applied in tandem and managed to shift the evaporation into useful transpiration by plants. During the dry season, the crop would have to be irrigated in order to achieve maximum land and water productivity. Skipping of irrigation during seed filling would have to be avoided in order to prevent significant reduction in yield. Irrigation at the commencement of maturity after the pods have been completely filled with seeds can be skipped. Under water limiting conditions, the amount of water saved by skipping irrigation during flowering, pod initiation, seed filling and maturity can be used for cultivating other crops and thereby increasing the opportunity cost. Incidental rainfall during the dry season would have to be used in order to increase irrigation water productivity of the crop.