Data from: Neighbourhood-dependent root distributions and the consequences on root separation in arid ecosystems

1. Inter-specific root separation is an important example of spatial niche differentiation that drives species coexistence in many ecosystems. Particularly under water-stressed conditions, it is believed to be an inevitable outcome of species interactions. However, evidence for and against this idea has been found. So far, studies aiming at reconciling the debate mainly focus on abiotic determinants. It remains unclear if and to what extent root separation depends on the type and growth form of interacting plants. 2. We conducted a detailed field study in three adjacently located (with pairwise distances < 500 m) arid patchy communities where a common tussock grass species Achnatherum splendens grew in association with either a tree (Elaeagnus angustifolia), a shrub (Nitraria tangutorum) or a perennial forb species (Sophora alopecuroides). In each community, roots and soils were sampled along the soil layers from five depths (0-10, 10-30, 30-60, 60-100 and 100-150 cm) in the patches and in the adjacent bare ground outside the patches. 3. Significant vertical inter-specific root separation occurred in the species-association patches of tree-grass and forb-grass communities, but not in the shrub-grass community. As the neighbour changed going from trees to shrubs and to forbs, rooting profiles of the grass Achnatherum became progressively deeper, with progressively less roots allocated in the upmost 10 cm soil layer and more in the subsequent two layers. After controlling for the differences in soil water and nutrient conditions among the three communities, the effects of neighbour type on grass rooting profiles remained robust. 4. Synthesis. We found that the root distributions of plants in the dryland strongly depend on the type of neighbour plants, which can, at least partially, determine the extent of inter-specific root separation at the community scale. Our work poses new questions about plasticity in root distribution and helps to better understand species interactions and coexistence under stressful conditions.