Background/Question/Methods One of the primary mechanisms hypothesized to explain the maintenance of plant species diversity is through negative feedback, where the accumulation of species-specific enemies favors heterospecific juveniles by suppressing the recruitment of conspecific juveniles.  Analyses of patterns of growth and survival of both temperate and tropical trees often reveal demographic signatures of negative density- or distance-dependence that are consistent with predicted outcomes of negative feedback.  Although the power of demographic analyses lie in their ability to examine patterns of large arrays of species within and across forests, these analyses are unable to identify the principle mechanisms underlying such patterns.  In the current study, we conducted fully-reciprocal shadehouse and field experiments to test whether outcomes of interactions between adult tropical trees and their associated soil biota exhibit sufficient specificity to generate negative feedback, and whether variation in the strength of feedback is related to abundances of adult tropical trees.    
Results/Conclusions In both the shadehouse and in the field, we found that interactions between adult tropical trees and their associated soil biota were species-specific and resulted in negative plant-soil feedback.  In a fully-reciprocal shadehouse experiment, soil biota originating from under adult trees significantly reduced the growth of their conspecific seedlings relative to heterospecific seedlings in four of the six species examined.  In a fully-reciprocal field experiment, seedling growth of all five species transplanted near their own adults was significantly reduced relative to heterospecific seedlings.  Such reduction in the field could not be explained by aboveground enemies.  Variation in the strength of negative feedback detected in both pots and in situ was strongly correlated with relative abundance of naturally-occurring adult trees.  Tree species that exhibited strongest negative feedback were less common as adults than those exhibiting weaker negative feedback.  Our results demonstrate that the outcomes of interactions between tropical trees and their soil biota are species specific, and that negative feedback driven by soil organisms may maintain species diversity and determine relative abundance patterns in tropical forests.