Abstract1. When placing roots in the soil, plants integrate information about soil nutrients, plant neighbours and beneficial/detrimental soil organisms. While the fine-scale spatial heterogeneity in soil nutrients and plant neighbours have been described previously, virtually nothing is known about the spatial structure in soil biotic quality (measured here as a soil Biota-Induced plant Growth Response, or BIGR), or its correlation with nutrients or neighbours. Such correlations could imply trade-offs in root placement decisions.2. Theory would predict that soil BIGR is (1) negatively related to soil fertility and (2) associated with plant community structure, such that plants influence soil biota (and vice versa) through plant-soil feedbacks. We would also expect that since plants have species-specific impacts on soil organisms, spatially homogeneous plant communities should also homogenize soil BIGR.3. Here, we test these hypotheses in a semi-arid grassland by (1) characterizing the spatial structure of soil BIGR at a scale experienced by an individual plant and (2) correlating it to soil abiotic properties and plant community structure. We do so in two types of plant communities: (1) low-diversity patches dominated by an invasive grass (Bromus inermis Leyss.) and (2) patches covered mostly by native vegetation, with the expectation that dominance by Bromus would homogenize soil BIGR. 4. Soil BIGR was spatially heterogeneous, but not autocorrelated. This was true in both vegetation types (Bromus-invaded vs. native patches). Conversely, soil abiotic properties and plant community structure were frequently spatially autocorrelated at similar scales. Also, contrary to many studies, we found a positive correlation between soil BIGR and soil fertility. Soil BIGR was also associated with plant community structure.
5.Synthesis. The positive correlation between soil BIGR and some soil nutrient levels suggests that plants do not necessarily trade-off between foraging for nutrients vs.biotic interactions: nutritional cues could rather indicate the presence of beneficial soil biota. Moreover, the spatial structure in plant communities, coupled with their correlation with soil BIGR, jointly suggest that plant-soil feedbacks operate at local scales in the field: this has been identified in modelling studies as an important driver of plant coexistence.
K E Y W O R D Sinformation integration, mutualisms, nutritional cues, phenotypic plasticity, plant coexistence, plant-soil feedbacks, root foraging behaviour, soil biota, soil micro-organisms, spatial scale