PLoS ONE (Jan 2012)
Conspecific plant-soil feedbacks of temperate tree species in the southern Appalachians, USA.
Abstract
Many tree species have seedling recruitment patterns suggesting that they are affected by non-competitive distance-dependent sources of mortality. We conducted an experiment, with landscape-level replication, to identify cases of negative distance-dependent effects and whether variation in these effects corresponded with tree recruitment patterns in the southern Appalachian Mountains region. Specifically, soil was collected from 14 sites and used as inocula in a 62 day growth chamber experiment determining whether tree seedling growth was less when interacting with soil from conspecific (like) than heterospecific (other) tree species. Tests were performed on six tree species. Three of the tree species had been previously described as having greater recruitment around conspecifics (i.e. facilitator species group) compared to the other half (i.e. inhibitor species group). We were then able to determine whether variation in negative distance-dependent effects corresponded with recruitment patterns in the field. Across the six species, none were negatively affected by soil inocula from conspecific relative to heterospecific sources. Most species (four of six) were unaffected by soil source. Two species (Prunus serotina and Tsuga canadensis) had enhanced growth in pots inoculated with soil from conspecific trees vs. heterospecifics. Species varied in their susceptibility to soil pathogens, but trends across all species revealed that species classified as inhibitors were not more negatively affected by conspecific than heterospecific soil inocula or more susceptible to pathogenic effects than facilitators. Although plant-soil biota interactions may be important for individual species and sites, it may be difficult to scale these interactions over space or levels of ecological organization. Generalizing the importance of plant-soil feedbacks or other factors across regional scales may be especially problematic for hyperdiverse temperate forests where interactions may be spatially variable.