Belowground resource sharing in mixed tree– crop systems: methods to better understand belowground interactions

In: Book chapters
Year published
Authors
Bayala J ,  
Oborn I , Dupraz C
Access
Region

Belowground resource sharing in mixed tree– crop systems: methods to better understand belowground interactions

Abstract: 

Cropping systems based on carefully designed species’ mixtures over time (in terms of crop sequences) and/or space (within a farm or landscape) reveal many potential advantages under various conditions, both in temperate and tropical agriculture1,2,3,4. In general, annual crops are expected to be relatively shallow-rooted while perennial plants, including trees, can have roots extending deep below the crop root zone, giving a foundation to the safety-net hypothesis5. The safety-net hypothesis (intercepting mobile nutrients leaching from crop root zones) complements the nutrient-pump hypothesis (uptake of deep soil resources of relatively immobile nutrients)6,7. However, the actual situation of relative root distributions is more complex8,9,10 and dynamic with seasonal shifts in the soil depth from which water and nutrients are taken up11. In some situations, trees and crops compete for nutrients and water in the same soil layer12,13, even though the impact on crop performance and yield may vary according to rainfall14 and nutrient availability15,16,17. Therefore, the potential benefits of trees in mixed systems depends on complex spatial and temporal interactions involving a large number of factors18,19,20. Strong positive effects (for example, through increased nutrient availability) can be offset by strong negative effects (for example, via shading), making optimization complex and context dependent21.The past decades of agroforestry research have revealed many interacting processes in the sharing of, and competition for, belowground resources, made progress in their quantification and established tools to study mixed tree–crop systems, as this chapter shows. However, the manner in which net effects depend on context still requires empirical verification of simulation models.