Stephan M. Haefele

Improved soil fertility and weed management is profitable for irrigated rice farmers in Sahelian West Africa

Abstract Gains in yield and profitability of irrigated rice ( Oryza sativa L.) in the Senegal river region resulting from improved fertilizer and weed management practices were studied in collaboration with farmers’ organizations, national research institutes and farmers. The profitability of different fertilizer and weed management treatments was compared to farmers’ practice using partial budgeting techniques and the net benefit of all treatments was estimated. Improved fertilizer management increased grain yield by 0.9xa0txa0ha −1 , while the recommended weed management resulted in a yield increase of 1.0xa0txa0ha −1 . The effect of recommended management practices was additive and gave a mean yield increase of 1.8xa0txa0ha −1 compared to farmers’ practices. Principal weed species in rice were Echinochloa colona (L.) Link, Diplachne fusca (L.) P. Beauv. ex Stapf, Sphenoclea zeylanica Gaertner, Bolboschoenus maritimus L., Schoenoplectus senegalensis (Steudel) Raynal, Cyperus iria L. and Cyperus difformis L. Indications of propanil-resistant E. colona were found. Weed infestation was severe and weed biomass comprised between 24 and 49% of the combined biomass of weeds and rice at panicle initiation. Average losses in rice yield caused by weeds were 0.56xa0txa0ha −1 per 10% of relative weed biomass. The value/cost ratios were between 2.1 and 4.6 for the improved treatments, and improved soil fertility and weed management resulted in an increase in net revenues of 40–85% compared to farmers’ practice.

Internal nutrient efficiencies, fertilizer recovery rates and indigenous nutrient supply of irrigated lowland rice in Sahelian West Africa

Abstract Grain yield and plant nutrient accumulation in aboveground plant dry matter were measured at physiological maturity of irrigated rice in on-station and on-farm experiments in four West African countries between 1995 and 1998. The complete database included 261 observations from farmers’ fields, 261 observations from unfertilized sub-plots in the same fields and 120 observations from on-station experiments. This data base was used to calibrate the model Quantitative Evaluation of the Fertility of Tropical Soils (QUEFTS) for irrigated rice in Sahelian West Africa. The model requires the estimation of borderlines of maximum and minimum internal N, P and K efficiency (IE; kilogram grain per kilogram nutrient in the aboveground dry matter), average recovery fractions of applied fertilizer and indigenous nutrient supply. Using the complete database, IE borderlines were 48 and 112xa0kgxa0grainxa0kg −1 N, 211 and 586xa0kgxa0grainxa0kg −1 P, and 32 and 102xa0kgxa0grainxa0kg −1 K, excluding observations with a harvest index −1 . At each site, results for indigenous supply of N, P and K were very variable and no significant relation between indigenous nutrient supply and soil type, cropping intensity or dominant texture group could be detected. The presented data base and resulting parameters can now be used to develop more adequate fertilizer strategies for irrigated rice in Sahelian environments.

Timing of weed management and yield losses due to weeds in irrigated rice in the Sahel

Abstract The effects of differing periods of weed management were studied in experiments in five farmers’ direct-seeded, irrigated rice (Oryza sativa L.) fields during the 1999 wet season (1999WS) and seven fields during the 2000 dry season (2000DS) in the Senegal River delta. Ten weed management treatments were used to identify critical periods of competition and to enable the development of more precise management recommendations. Rice yields in 1999WS ranged from 3.2xa0txa0ha−1 with no weed control to 6.3xa0txa0ha−1 in weed-free plots; a yield loss of 49%. Major weed species encountered in 1999WS were Bolboschoenus maritimus, Oryza longistaminata, Cyperus difformis and Echinochloa colona. In 2000DS, C. difformis dominated the weed flora, and rice yields ranged from 3.7xa0txa0ha−1 with no weed control to 7.9xa0txa0ha−1 in weed-free plots; a yield loss of 47%. To assess the effects of weed biomass on the rice crop over time a linear mixed model was used. This highlighted the importance of E. colona and C. difformis in the wet and dry seasons, respectively, and of O. longistaminata in both seasons. The presence of E. colona early in the crop seemed related to greater levels of rice biomass in 1999WS, while O. longistaminata always had a negative effect on crop growth even at relatively low levels of infestation. Gompertz and logistic equations were fitted to data representing increasing periods of weed-free growth and weed interference, respectively. Critical periods for weed control, obtaining 95% of a weed-free yield, were estimated as between 29 to 32 days after sowing (DAS) in the WS and 4 to 83 DAS in the DS. The lower temperatures in the early DS resulted in slower growth of rice and weeds, delaying the effects of competition. Weed competition either before or after these critical periods had negligible effects on crop yield.

Long-term fertility experiments for irrigated rice in the West African Sahel: effect on soil characteristics

Abstract Long-term fertility experiments (LTFEs) are a tool to investigate the sustainability of cropping systems. The present study analyzed two LTFEs for intensive rice-based irrigated systems in the Senegal River valley at Ndiaye and Fanaye (Sahel savanna). The trials were established in 1991, contain six different fertilizer treatments and rice is grown two times per year. Soil types are a typical Orthithionic Gleysol and an Eutric Vertisol at Ndiaye and Fanaye, respectively. The objectives of the presented study were to analyze the effect of intensive irrigated rice cropping on the soil resource base by studying the changes of soil characteristics over time and by comparing soil N, P and K pools in different fertilizer treatments. In the LTFE at Ndiaye, topsoil pH values increased significantly from 5.5 to about 6.5 and electrical conductivity was high but remained stable. Soil organic carbon (SOC) and total soil nitrogen (TSN) dropped slightly after 16 consecutive seasons but the difference was statistically not significant. At both sites, exchangeable N ranged between 1.6 and 2.8% of TSN and fixed N accounted for 5.5–8.2% of TSN, with slightly higher values in Fanaye. Treatment differences in N dose had no significant effect on these parameters. Results of δ13C analysis showed a decrease due to rice cropping at both sites, and the measurements indicate high turnover rates of soil organic matter. Soil analyses of total soil P and K and of different pools indicated only small changes when these elements were applied at medium quantities. In contrast, treatments with N application only showed considerable soil P and K depletion, and rice cultivation without P and/or K application cannot maintain soil fertility. The soil mining process is relatively quick for P due to the naturally low soil P status, whereas the high soil K reserves buffer even important negative K balances for decades. It is concluded that irrigated rice cultivation in the region can maintain soil fertility if at least medium P doses are applied together with nitrogen.

A framework to improve fertilizer recommendations for irrigated rice in West Africa

A framework to optimize soil fertility management in irrigated rice, based on soil and climate variability, fertilizer and paddy prices and farmers objectives is presented. The framework uses three models in succession. The dynamic ecophysiological ORYZAS model simulates potential rice yields under irrigation, based on weather conditions, cultivar choice and sowing date. This yield potential is used in the static FERRIZ model, together with site specific information on recovery efficiency of applied N, P and K, indigenous soil N, P and K supply, and maximum N, P and K accumulation and dilution in rice dry matter. Resulting outputs are required fertilizer doses to obtain different target yields depending on yield potential and soil nutrient supply. The framework allows sensitivity analysis and agro-economic evaluation of different fertilizer options. In a last step, the dynamic decision tool RIDEV is used to simulate optimal timing of different management actions such as fertilizer application, weeding, and harvest. Resulting integrated crop management recommendations ensure high resource use efficiency adapted to local socio-economic and bio-physical conditions. Provided necessary input data are available, this framework can be applied to field, perimeter or regional scale. The framework was applied to the Office du Niger in Mali, using field data of 58 farmers. In the wet season, fertilizer doses adjusted to three soil fertility classes outperformed current uniform recommendations only slightly, except on soils with very low K supply (10 cases). Only on these soils was application of K profitable. Profit-optimizing fertilizer doses had high costs and low value/cost ratios. Target yields close to potential yield increased farmers risk. Adjusting fertilizer doses to the lower yield potential in the dry season reduced costs and risk without reducing profit, resulting in better value/cost ratios. New simple recommendations for both wet and dry seasons were formulated.

Explaining yield gaps on farmer-identified degraded and non-degraded soils in a Sahelian irrigated rice scheme

The objective of this study was to identify to what extent rice productivity problems are caused by soil quality problems (soil alkalinity) and to what extent by sub-optimal crop management. The study area used is the Foum Gleita irrigation scheme, Mauritania

Long-term fertility experiments for irrigated rice in the West African Sahel: agronomic results

Abstract Long-term fertility experiments (LTFEs) are a tool to investigate the sustainability of cropping systems. Several LTFEs for intensive irrigated rice cropping were established in Asia, but those are rare in Sub-Saharan Africa. Two such trials are presented, both located in the Sahel savanna vegetation zone in Senegal. The trials were established in 1991, contain six different fertilizer treatments and rice is grown two times per year. Soil type at Ndiaye is a typical Orthithionic Gleysol and an Eutric Vertisol at Fanaye. Average grain yields without fertilizer application were 3.4xa0Mgxa0ha−1 per season in Ndiaye and 2.9xa0Mgxa0ha−1 per season in Fanaye. In 20 consecutive seasons best treatments at both sites and in both seasons yielded on average between 6.7 and 7.6xa0Mgxa0ha−1 per crop. Yield components were influenced by cultivar, site, season and fertilizer treatment. The mineral composition of grains was homogenous between cultivars and similar to results from Asia, Australia and the USA. Significant differences among cultivars were found for the mineral composition of straw, especially for Si, Fe, Zn and K. Nitrogen, P and K fertilizer treatments increased the mineral concentrations of N, P, Mg, Ca, Mn and Cu. Highest yields in the LTFE indicate a not significant yield decline of −27xa0kg per season in Ndiaye and a significant increase in Fanaye (+86xa0kg per season). The crop model ORYZAS simulated that potential yields declined by −50xa0kg per season in Ndiaye and increased by +16xa0kg per season in Fanaye (not significant). Simulated yield trends were not always mirrored by trends of average seasonal radiation, which is due to the influence of unfavorable temperature extremes on yield and yield simulations. It is concluded that observed yield trends can largely be explained by climatic influences. The agronomic analysis did not indicate a negative impact of intensive irrigated rice cropping on the soil resource base.

Estimation of vegetation indices for high-throughput phenotyping of wheat using aerial imaging

BackgroundUnmanned aerial vehicles offer the opportunity for precision agriculture to efficiently monitor agricultural land. A vegetation index (VI) derived from an aerially observed multispectral image (MSI) can quantify crop health, moisture and nutrient content. However, due to the high cost of multispectral sensors, alternate, low-cost solutions have lately received great interest. We present a novel method for model-based estimation of a VI using RGB color images. The non-linear spatio-spectral relationship between the RGB image of vegetation and the index computed by its corresponding MSI is learned through deep neural networks. The learned models can be used to estimate VI of a crop segment.ResultsAnalysis of images obtained in wheat breeding trials show that the aerially observed VI was highly correlated with ground-measured VI. In addition, VI estimates based on RGB images were highly correlated with VI deduced from MSIs. Spatial, spectral and temporal information of images contributed to estimation of VI. Both intra-variety and inter-variety differences were preserved by estimated VI. However, VI estimates were reliable until just before significant appearance of senescence.ConclusionThe proposed approach validates that it is reasonable to accurately estimate VI using deep neural networks. The results prove that RGB images contain sufficient information for VI estimation. It demonstrates that low-cost VI measurement is possible with standard RGB cameras.

Overexpression of the class I homeodomain transcription factor TaHDZipI-5 increases drought and frost tolerance in transgenic wheat

Summary Characterization of the function of stress‐related genes helps to understand the mechanisms of plant responses to environmental conditions. The findings of this work defined the role of the wheat TaHDZipI‐5 gene, encoding a stress‐responsive homeodomain–leucine zipper class I (HD‐Zip I) transcription factor, during the development of plant tolerance to frost and drought. Strong induction of TaHDZipI‐5 expression by low temperatures, and the elevated TaHDZipI‐5 levels of expression in flowers and early developing grains in the absence of stress, suggests that TaHDZipI‐5 is involved in the regulation of frost tolerance at flowering. The TaHDZipI‐5 protein behaved as an activator in a yeast transactivation assay, and the TaHDZipI‐5 activation domain was localized to its C‐terminus. The TaHDZipI‐5 protein homo‐ and hetero‐dimerizes with related TaHDZipI‐3, and differences between DNA interactions in both dimers were specified at 3D molecular levels. The constitutive overexpression of TaHDZipI‐5 in bread wheat significantly enhanced frost and drought tolerance of transgenic wheat lines with the appearance of undesired phenotypic features, which included a reduced plant size and biomass, delayed flowering and a grain yield decrease. An attempt to improve the phenotype of transgenic wheat by the application of stress‐inducible promoters with contrasting properties did not lead to the elimination of undesired phenotype, apparently due to strict spatial requirements for TaHDZipI‐5 overexpression.

Participatory evaluation guides the development and selection of farmers’ preferred rice varieties for salt- and flood-affected coastal deltas of South and Southeast Asia

Rice is the staple food and provides livelihood for smallholder farmers in the coastal delta regions of South and Southeast Asia. However, its productivity is often low because of several abiotic stresses including high soil salinity and waterlogging during the wet (monsoon) season and high soil and water salinity during the dry season. Development and dissemination of suitable rice varieties tolerant of these multiple stresses encountered in coastal zones are of prime importance for increasing and stabilizing rice productivity, however adoption of new varieties has been slow in this region. Here we implemented participatory varietal selection (PVS) processes to identify and understand smallholder farmers’ criteria for selection and adoption of new rice varieties in coastal zones. New breeding lines together with released rice varieties were evaluated in on-station and on-farm trials (researcher-managed) during the wet and dry seasons of 2008–2014 in the Indian Sundarbans region. Significant correlations between preferences of male and female farmers in most trials indicated that both groups have similar criteria for selection of rice varieties. However, farmers’ preference criteria were different from researchers’ criteria. Grain yield was important, but not the sole reason for variety selection by farmers. Several other factors also governed preferences and were strikingly different when compared across wet and dry seasons. For the wet season, farmers preferred tall (140–170 cm), long duration (160–170 d), lodging resistant and high yielding rice varieties because these traits are required in lowlands where water stagnates in the field for about four months (July to October). For the dry season, farmers’ preferences were for high yielding, salt tolerant, early maturing (115–130 d) varieties with long slender grains and good quality for better market value. Pest and disease resistance was important in both seasons but did not rank high. When farmers ranked the two most preferred varieties, the ranking order was sometimes variable between locations and years, but when the top four varieties that consistently ranked high were considered, the variability was low. This indicates that at least 3–4 of the best-performing entries should be considered in succeeding multi-location and multi-year trials, thereby increasing the chances that the most stable varieties are selected. These findings will help improve breeding programs by providing information on critical traits. Selected varieties through PVS are also more likely to be adopted by farmers and will ensure higher and more stable productivity in the salt- and flood-affected coastal deltas of South and Southeast Asia.