Growth, yield and system performance simulation of a sugarcane-eucalyptus interface in a sub-tropical region of Brazil

The sugarcane (Sacharum officinarum) monocropping has had a great socio-economic and environmental impact in Brazil, and agroforestry (AFs) has been considered as an alternative to its sustainable production. However, there is a lack of field experiments results under such conditions. Therefore, yield measurement across transect in the tree-crop interface in on-farm conditions, as well as the use of simulation models, may allow the evaluation of biophysical interactions between trees and crops and system productivity. In this work, plant growth and yield in a sugarcane-eucalyptus (Eucalyptus grandis) on-farm interface were evaluated. The experimental site had a Chromic Ferralsol soil and it is located in a sub-tropical region of Brazil. Availability of solar radiation for the crop along the transect was estimated and its effect on sugarcane dry matter production was evaluated. Using such relations, two sugarcane-eucalyptus AFs cycles were simulated to estimate system productivity. The field results showed that the trees presented a higher growth in the AFs, while the crop growth and yield were inversely proportional to their distance from the trees. The eucalyptus wood volume increased from 0.15 to 0.29 m3 per tree from monocropping to AFs. The sugarcane dry matter decreased from 35.1 to 8.70 t ha−1 from the furthest to the closest position from tree along the transect. Simulations suggested that light was the main cause of the crop yield reduction, but the importance of competition for water and nutrients increased inversely according to tree distance. However, both simulated AFs provided land equivalent ratio (LER) similar to one (1). Field measurements and simulations indicate that agroforestry systems are a sound alternative for sugarcane cultivation in the studied region. Combination of on-farm studies and simulation models provided conditions to understand tree-crop interactions and to extrapolate field results so that the performance of a new system could be estimated.