Ecological characteristics and cultivar influence optimal plant density of East African highland bananas (Musa spp., AAA-EA) in low input cropping systems

Numerous studies have been conducted on the effects of plant density on growth and yield of dessert bananas in the humid tropics, but effects of plant densities in relations with ecological characteristics in low input East African highland banana (Musa spp., AAA-EA genome) cropping systems have not been reported. On-station field experiments were conducted in three contrasting agro-ecological sites of Rwanda (Kibungo low rainfall with medium soil fertility, Rubona high rainfall with low soil fertility and Ruhengeri high rainfall with high soil fertility) to explore germplasm×environment interactions. Five different plant densities (plantsha−1): 1428, 2500, 3333, 4444 and 5000 and two cooking (“Ingaju”, “Injagi”) and one beer (“Intuntu”) cultivars were investigated. The effect of plant density on plant performance (growth and yield) over two cropping cycles in low input systems was determined. The effects of site×cultivar and site×density interactions on yield traits were significant (p<0.05). Annual yield increased with increasing plant density but strongly depended on agro-ecological site (from 6.1 to 9.2tha−1yr−1 at Kibungo, 9.5 to 21.5tha−1yr−1 at Rubona and 7.0 to 25.0tha−1yr−1 at Ruhengeri). Yields of beer cultivars increased with density, but those of cooking cultivars decreased. Maximum yields were attained at 4444plantsha−1 at Kibungo and Rubona whilst yields increased linearly beyond this level at Ruhengeri. Crop cycle duration was prolonged with increasing plant density. Relationships between bunch yield, the total above ground dry matter yields and soil chemical properties suggest that nutrient deficiencies were larger at Kibungo (e.g. K) and Rubona (e.g. K, P, Ca and Mg) compared with Ruhengeri, where yield correlated significantly with leaf area index (LAI). LAI increases up to 4, where 95% of solar radiation was intercepted by the crop canopy, indicating that increasing the LAI above 4 would have little effect on production. Evaporation was much greater at lower rainfall areas (e.g. Kibungo) and accompanied by negative annual water deficit (−135mmyr−1) than at high rainfall areas (e.g. Ruhengeri) with positive water surplus (382mmyr−1). Growing degree days from planting to bunch harvest were higher at Kibungo (3675°C days) but much less at the Ruhengeri cooler site (1729°C days), implying temperature is not restrictive at Ruhengeri. This study showed that the optimal density for bananas depends on water availability, soil fertility and cultivar, which serves as an entry point to maximize yield potential for the East African smallholder farmers rather than using a uniform blanket recommended density. We suggest that agronomic optimal plant density is lower (<4444plantsha−1) in low rainfall (<1000mmyr−1) and less fertile areas but seem to be higher (>5000plantsha−1) in areas with high fertility which receive high rainfall (>1300mmyr−1).