Abstract
Groundnut production is limited in Sub-Saharan Africa and water deficit or “drought,” is often considered as the main yield-limiting factor. However, no comprehensive study has assessed the extent and intensity of “drought”-related yield decreases, nor has it explored avenues to enhance productivity. Hence, crop simulation modeling with SSM (Simple Simulation Modeling) was used to address these issues. To palliate the lack of reliable weather data as input to the model, the validity of weather data generated by Marksim, a weather generator, was tested. Marksim provided good weather representation across a large gradient of rainfall, representative of the region, and although rainfall generated by Marksim was above observations, run-off from Marksim data was also higher, and consequently simulations using observed or Marksim weather agreed closely across this gradient of weather conditions (root mean square of error = 99 g m-2; R2 = 0.81 for pod yield). More importantly, simulation of yield changes upon agronomic or genetic alterations in the model were equally predicted with Marksim weather. A 1° × 1° grid of weather data was generated. “Drought”-related yield reduction were limited to latitudes above 12–13° North in West Central Africa (WCA) and to the Eastern fringes of Tanzania and Mozambique in East South Africa (ESA). Simulation and experimental trials also showed that doubling the sowing density of Spanish cultivars from 20 to 40 plants m-2 would increase yield dramatically in both WCA and ESA. However, increasing density would require growers to invest in more seeds and likely additional labor. If these trade-offs cannot be alleviated, genetic improvement would then need to re-focus on a plant type that is adapted to the current low sowing density, like a runner rather than a bush plant type, which currently receives most of the genetic attention. Genetic improvement targeting “drought” adaptation should also be restricted to areas where water is indeed an issue, i.e., above 12–13°N latitude in WCA and the Eastern fringes of Tanzania and Mozambique.
Highlights
Groundnut is cultivated in environments where it seems the crop may experience water deficits, but there is no thorough georeferenced assessment of where such water deficits are a problem for peanut production
Marksim generated rainfall were in agreement with the observed rainfall (R2 = 0.82; Figure 1A) and if Marksim generated weather that was wetter than in the observed locations, this discrepancy was fairly consistent across a gradient of rainfall from less than 200 mm to more than 1000 mm during the season
The pod yield predicted from the observed weather was on average 257 g m−2 across the 12 locations and this was in close agreement with an average grain yield of 224 g m−2 simulated from the Marksim weather (RMSE = 99 g m−2)
Summary
Groundnut is cultivated in environments where it seems the crop may experience water deficits, but there is no thorough georeferenced assessment of where such water deficits are a problem for peanut production. Peanut transpiration water needs to grow a fully irrigated crop in a regular rainy season was about 30 L plant−1, which could be extrapolated to 450 mm at the sowing density used in these trials (Halilou et al, 2015). This study proposes to test to what extent and where water is a substantial limitation for growing peanut. We hypothesized that this restriction might only prevail in the driest part of the semi-arid tropics, especially in the North Sahelian zone
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
