|Title||Risk management options in maize cropping systems in semi-arid areas of Southern Africa|
|Author(s)||Masvaya, Esther N.; Nyamangara, Justice; Giller, Ken E.; Descheemaeker, Katrien|
|Source||Field Crops Research 228 (2018). - ISSN 0378-4290 - p. 110 - 121.|
Plant Production Systems
|Publication type||Refereed Article in a scientific journal|
|Keyword(s)||Grain requirement - N mineralisation - N stress - Runoff curve number - Water stress|
Although rainfed cropping in semi-arid areas is risky due to frequent droughts and dry spells, planting early with the first rains is often expected to result in yield benefits. We hypothesised that planting early leads to yield benefits if the planting coincides with a mineral N flush at the start of the season but leads to crop failure if there is a false start to the cropping season. The effects of different management options, including tillage (ploughing and ripping), mulch (two levels 0 and 2 t ha−1) and fertility amendments (five levels: 0; 20 and 40 kg N ha−1; 5 t manure ha−1 and 5 t ha−1 manure + 20 kg N ha−1) on grain yields were simulated using the calibrated and tested APSIM model over a 30-year period (1984–2015). Yields were simulated and compared across seven planting date scenarios (1 November, 15 November, 30 November, 15 December, 31 December, 15 January and planting when cumulative rainfall of 20 mm was received in three consecutive days). Planting with the first rains with manure + 20 kg N ha−1 resulted in the best average yield of 2271 kg ha−1 whilst the poorest average yields of 22 kg ha−1 were observed with planting on 15 January with no fertility amendment (0 kg N ha−1). Planting early (1 Nov to 15 Nov) and with the first rains resulted in exceeding the food self-sufficiency threshold of 1080 kg ha-1 in 40–83 % of the cases if fertility amendments are applied, as well as a low probability of complete crop failure, ranging from 0 to 40%. Grain yield penalties due to a false start followed the trend: ripper + mulch > plough + mulch > ripper (no mulch) averaging 256, 190 and 182 kg ha-1 respectively across all the fertility treatments. The model was able to simulate the occurrence of the mineral N flush with the first rains. Its coincidence with planting resulted in average yield benefits of 712, 452, 382 and 210 kg ha-1 for the following respective planting dates: 1 Nov, 15 Nov, 30 Nov, variable date when >20 mm rainfall was received. Early planting, in combination with reduced tillage, mulch and N containing fertility amendments is critical to reduce risk of crop failure in the smallholder cropping systems of semi-arid areas of southern Africa and achieve the best possible yields.