An experiment was conducted in Nobéré, Burkina Faso, West Africa to assess the effect of shade of two indigenous fruit trees, Adansonia digitata (Baobab) and Parkia biglobosa (Néré) on a shade-tolerant crop called taro (Colocasia esculenta) in comparison with millet (Pennisetum glaucum), a shade-intolerant crop, with the aim of optimising the productivity of agroforestry parkland systems. Eight trees of each of néré and baobab were randomly selected on farmers lands for the study. To assess the effect of tree shade on associated crops, the area under each tree was subdivided into three tree influence concentric zones: zone A= from tree trunk to half radius of the crown; zone B=from half radius to the edge of the crown and zone C=from the edge to 3 m away. A control plot was established for each tree in an open area. The performance of crops in terms of growth, yield, dry matter production and nutritional composition was assessed during three cropping seasons between 2006 and 2008 according to concentric zones around trees and the control plot. In addition, biophysical factors including photosynthetically active radiation (PAR), transpiration, photosynthesis, soil physical and chemical properties and root length density were measured under the trees and the control.
Finally, the socio-economic implications of optimizing crop production under trees using taro was evaluated using cost-benefit analysis. Millet performed better under baobab (806.1±121.48 kg ha¯¹) than under néré (320.2±59.91 kg ha¯¹) which was less than the yield in the control plot (595.8±83.43 kg ha¯¹). In contrast, the yield of taro was better under néré (4124.0±469.05 kg ha¯¹) than under baobab (2738.3±595.61 kg ha¯¹) which was almost similar to the yield in the control plot (2336.9±617.04 kg ha¯¹). The higher yield of taro under néré, which had dense and wide crown, confirms its shade tolerance, which was further ascertained by the significantly negative correlation found between PAR availability and the yield and dry matter production of taro. In contrast, the performance of millet increased when PAR availability increased. Results of the nutritional composition showed that tree shade had no effect on the quality of either millet or taro. In terms of yield and nutritional composition, using taro immediately under tree crowns seems to be appropriate strategy for optimising crop production in agroforestry parkland systems. According to the results of the cost-benefit analysis, taking into account the traditional cropping practice in agroforestry parkland systems where the average tree density is 14 trees ha¯¹ and the common crops grown included maize, millet, sorghum, cowpea and peanut, the net present value was estimated at 24317.2 FCFA per hectare (655 .957 FCFA is equivalent to 1 €). This was increased to 84696.1 FCFA per hectare if taro is grown under trees at the average tree density of 14 trees ha¯¹• The net present value increased further with an increase in tree density. Therefore, it was concluded that by using shade-tolerant crops such as taro beneath tree crowns farmers would be able to maintain high density of trees on parklands and improve the systems productivity.
Key words: Adansonia digitata, Colocasia esculenta, Parkland, Parkia biglobosa, Pennisetum glaucum, shade tolerance