Robusta coffee (Coffea canephora Pierre ex Froehner) is an important cash crop in the lowland tropics. Similar to Arabica (C. arabica L.), agroforestry is common in these systems, though comparatively little is known about how nutrient inputs and leguminous shade trees influence Robusta nutritional status and yield. We evaluated how shading with leguminous trees across four input systems, defined largely by quantity and form (compost vs. synthetic) of nitrogen (N), phosphorus (P), and potassium (K), influenced Robusta-based cropping systems in the Ecuadorian Amazon. Soil fertility was largely unaffected by compost and synthetic fertilizer (NPK) inputs or by shade species, and was weakly related to coffee productivity based on principal component analysis (PCA). Though litter appeared to be an agronomically appreciable source of N and P, this was only somewhat reflected in soil extractable N and P, presumably due to nutrient losses under high rainfall conditions (>3,000 mm yr −1 ). Patterns of variation suggest transfer of N and P from litter to soil to leaf. Conversely, litter K was unrelated to soil K, but soil K was closely related to leaf K. Robusta yields appeared to be largely N limited and were more influenced by input type (i.e., synthetic vs. organic nutrient sources) than by shade species or amount of N inputs in the form of compost or synthetic N, P, and K fertilizers. Additionally, yields across all treatments were significantly higher than regional averages, suggesting that slight investments in inputs can substantially increase productivity in maturing Robusta systems, irrespective of shade tree species selection.
INTRODUCTIONCoffee (Coffea spp.) is a globally important cash crop produced in the tropics (Lewin et al., 2004). Worldwide, coffee production is dominated by two species, Arabica (C. arabica L.) (57%) and Robusta (C. canephora Pierre ex Froehner