On the complexity of agroforestry systems

Abstract

Tropical soils are prone to rapid degradation if not managed well, and agroforestry systems have the potential to restore degraded soils and support agricultural production together with other ecosystem services. In Brazil, an increasing number of pioneering farmers are establishing agroforestry systems on previously cleared farmland. However, while there are a wide range of agroforestry systems, this diversity has hardly been quantified, and it is not clear how these systems differ in their capacity for nutrient cycling to reverse soil degradation. The objectives of the study were to assess innovative agroforestry systems in terms of taxonomic and functional diversity, spatial structure and management, and to assess how these systems differ in terms of structural complexity and their potential for nutrient cycling. We assessed a LiDAR-derived stand structural complexity index (SSCI), interrow spacing, stem density, tree species richness and diversity, community weighted means (CWM) of foliar nitrogen and wood density, livestock density, pruning and mowing regimes in 30 agroforestry systems in the state of São Paulo, Brazil. We used N, P, K, Ca and Mg stocks in litter as a proxy for nutrient cycling. The agroforestry systems could be broadly categorized into silvopastures, multistrata and successional agroforestry systems. These types spanned a gradient of structural complexity, and this complexity was positively associated with tree species richness and planting density. Litter nutrient stocks were positively associated with pruning and mulching, and negatively associated with CWM of wood density, indicating the importance of pioneer trees. Overall, our results suggest that densely planted, pruned agroforestry systems that contain high species richness, including pioneer trees, contain relatively high amounts of N, P, K, Ca and Mg in their litter. These findings provide insight in the key characteristics of agroforestry systems to support nutrient cycling, and can inform the design of agroforestry systems for the regeneration of degraded agricultural land.