Spatial variation in plant-soil biofeedbacks: A cautionary tale

Through the generation of positive or negative feedbacks, invasive species can indirectly impact the native plant community.  As these feedbacks develop over time, they have the potential to impact a system not only temporally but spatially as well.  As soil properties and biotic communities are likely to vary over small spatial scales, the net direction and magnitude of feedbacks would also vary dramatically. Most studies have not examined spatial variation within soil biofeedbacks, but have instead focused on the influence of pooled soil communities.  We specifically focus on the role of soil biofeedbacks in explaining the success of the invasive understory grass Microstegium vimineum relative to a commonly associated native forest herb, Eupatorium rugosum.   To evaluate the role of soil biofeedbacks in exotic plant invasions, we conducted a soil inoculum experiment with replicate soil collections from several locations to address spatial variability.  Biomass of Eupatorium rugosum varied dramatically across sites and with inoculum sterilization, but shows no consistent influence of invasion.  Plants grown in sterilized soil performed better overall, irrespective of Microstegium invasion. The influence of invasion and sterilization varied spatially among sites.  Individual sites had remarkable variation in the direction and significance of invasion and sterilization treatments, as well as their interaction.  Some sites showed no invasion effects while others showed facilitative or inhibitive effects of Microstegium.  This study shows that spatial variation can exist over an otherwise homogenous landscape and pooling of soils may obscure the local variation in soil communities that plants encounter in natural systems.  We suggest caution should be taken with regards to interpreting soil-biofeedback studies that involve pooled soil inocula, until the ubiquity and the magnitude of spatial variation is known.