N(2)-fixation and N contribution by grain legumes under different soil fertility status and cropping systems in the Guinea savanna of northern Ghana

Continuous cereal-based cropping has led to a rapid decline in soil fertility in the Guinea savanna agro-ecological zone of northern Ghana with corresponding low crop yields. We evaluated the effects of cropping system and soil fertility status on grain yields and N(2)-fixation by grain legumes and net N contribution to soil fertility improvement in contrasting sites in this agro-ecological zone. Maize was intercropped with cowpea, soybean and groundnut within a row, with a maize stand alternated with two equally spaced cowpea or groundnut stands and in the maize-soybean system, four equally spaced soybean stands. These intercrops were compared with sole crops of maize, cowpea, soybean and groundnut in fertile and poorly fertile fields at sites in the southern (SGS) and the northern (NGS) Guinea savanna. The proportion of N derived from N(2)-fixation (%Ndfa) was comparable between intercrops and sole crops. However, the amount of N(2)-fixed was significantly larger in sole crops due to a greater biomass accumulation. Legumes in poorly fertile fields had significantly smaller shoot δ(15)N enrichment (−2.8 to +0.7‰) and a larger %Ndfa (55–94%) than those in fertile fields (−0.8 to +2.2‰; 23–85%). The N(2)-fixed however was larger in fertile fields (16–145 kg N ha(−1)) than in poorly fertile fields (15–123 kg N ha(−1)) due to greater shoot dry matter and N yields. The legumes grown in the NGS obtained more of their N requirements from atmospheric N(2)-fixation (73–88%) than legumes grown in the SGS (41–69%). The partial soil N balance (in kg ha(−1)) was comparable between intercrops (−14 to 21) and sole legumes (−8 to 23) but smaller than that of sole maize receiving N fertiliser (+7 to +34). With other N inputs (aerial deposition) and outputs (leaching and gaseous losses) unaccounted for, there is uncertainty surrounding the actual amount of soil N balances of the cropping systems, indicating that partial N balances are not reliable indicators of the sustainability of cropping systems. Nevertheless, the systems with legumes seem more attractive due to several non-N benefits. Our results suggest that soybean could be targeted in the SGS and cowpea in the NGS for greater productivity while groundnut is suited to both environments. Grain legumes grown in poorly fertile fields contributed more net N to the soil but growing legumes in fertile fields seems more lucrative due to greater grain and stover yields and non-N benefits.