Tree crop interactions in an Acacia auriculiformis agroforestry system in Gunungkidul, Java, Indonesia

Agroforestry systems comprising combinations of trees and agricultural crops provide smallholder farmers with opportunities to improve their income base and economic resilience through combining production of food and forest products in complementary ways within their existing land base. However, combining trees and crops in agroforestry systems does not always yield net gains in total productivity and the outcome can be dependent on a range of factors including local climatic and edaphic factors, physiological ecology of the species involved, and the management of the system. Mechanisms and patterns of competition for a limited supply of natural resources are often the primary drivers of these outcomes. This study investigated whether competition for water in an environment that is characterised by a 4-month dry season (receiving less than 60 mm in total over the 4 months) is the key factor determining growth and productivity in an Acacia auriculiformismaize/soybean agroforestry system in Gunungkidul Regency, Java, Indonesia. The study focussed on the utilisation of A. auriculiformis as the tree species, which has strong potential to be widely planted in the Gunungkidul area, because it produces high quality solid wood on short rotations of around 7 years. This study consists of two field experiments and a modelling analysis. The experimental work was undertaken at Indonesia's Centre for Forest Biotechnology and Tree Improvement Field Centre in Playen District, Gunungkidul Regency, Yogyakarta Province. An Acacia auriculiformis agroforestry system field trial was established in two experiments; an alley-cropping system and an interface planting layout (adequate fertiliser was applied in both systems). In the alley cropping experiment, a randomised complete block design with five replications, each containing three treatments was used. The treatments were (1) tree monoculture, (2) agricultural crop monoculture and (3) an agroforestry plot combining A. auriculiformis and maize/soybean. In the interface planting experiment, five replicated plots with five rows of trees planted on one side of the plot was utilised. The impacts of trees on crop water status and yield were explored at different distances from the tree interface; -6, -3, 0, 3, 6, 9, and 14.5 m, where '0 m' represents the tree-crop interface. The negative sign refers to the crops cultivated between the tree rows and the positive sign to those cultivated in the open area. In each cropping season, maize was grown early in the wet season (~November-February) followed by soybeans (~February-April) for three cropping seasons up to 27 months after tree establishment. The interactions between trees and crops were explored using the process-based model, Agricultural Production Systems Simulator Next Generation (APSIM X). In the alley cropping study, tree growth up to age 27 months was unaffected by the presence of agricultural crops. Maize biomass or yield in the second cropping season was also unaffected by the presence of trees. However, the maize yield/biomass was significantly lower in the agroforestry system than in the monoculture in the third growing season. Soybean above-ground biomass in the first two cropping seasons and soybean yield in all three cropping seasons were unaffected by the presence of trees in the agroforestry system. Soybean aboveground biomass was significantly lower in the agroforestry system than in the monoculture in the third growing season, but this did not translate into lower yield as the soybeans increased their harvest index to compensate. Differences in soil water deficit, and leaf water potentials (both pre-dawn and mid-day in A. auriculiformis and soybean) were generally not significant between treatments in all three growing seasons. Given the adequate supply of nutrients, these results suggest that tree shading rather than competition for water was primarily responsible for any decline in crop productivity, and that this effect was greater on maize than soybean. In the interface planting agroforestry system, proximity to the trees did not affect either soybean or maize productivity for the first two cropping seasons. However, the agricultural crop productivity was significantly decreased near the tree-crop interface in the 3rd cropping season. Differences in soil water deficit, and leaf water potentials (pre-dawn and mid-day) in soybean were not significantly different between treatments in all three growing seasons. It is thus very likely (but not demonstrated) that light competition was the key factor that resulted in decreased crop productivity in the potential competition zone, in the 3rd planting season, 22 months after tree planting, and this competition zone extended to 9 m from the tree-crop interface. The absence of effects of the trees on crop productivity in the alley-cropping or in the interface planting agroforestry systems were due to an absence of competition for light, water or nutrients up to tree age 15 months. With increasing tree height, the commencement of light competition was likely to be the key reason for the observation of reduced crop productivity 22 months after tree establishment. It was concluded that for the first two cropping seasons the farmers will be able to get both optimum crop productivity and tree growth from this Acacia agroforestry system, but beyond the first 2 cropping seasons, there is a trade-off between the trees and crops that the farmers will need to account for. APSIM Next Generation was used to simulate this Acacia agroforestry system to: (i) evaluate the capability of this model to predict crop yield and soil moisture in the 3rd cropping season, (ii) explore the potential impacts of a full tree rotation (~7-year) on crop productivity, and (iii) simulate the effects of the drier climate scenario on crop productivity. Crop yields were moderately well predicted, but there was not a good correlation between predicted and observed soil water content. The competition zone (i.e. where the presence of trees resulted in a reduction in crop yield) was predicted to extend to 9 m from the tree-crop interface in the 3rd cropping season, extending up to 50 m in the 8th cropping season (tree age of 7 years). Even when rainfall was simulated to be 75% of the median (dry climate scenario), the width of competition zone was not predicted to be impacted. However, lower crop yield was predicted in the drier climate scenario. This suggests that farmers can utilize some general rules around the width of the competition zone, irrespective of climate, to avoid planting within the zone that will result in reduced economic returns. These assumptions need to be tested further to gain more confidence in the model predictions and application to smallholder farmers.