Species-herd protection hypothesis predicts that individuals survive better when surrounded by more heterospecific neighbors. Empirical evidence for this hypothesis in natural forests is scarce. We expect that closely related heterospecific neighbors may promote rather than thwart natural enemy spreading based on the findings that natural enemies usually attack closely related hosts. However, once heterospecific neighbors reach adequate phylogenetic distance to their focal individuals species-herd protection effect may emerge, preventing the spread of natural enemies. Thus, considering phylogenetic structure of neighbors is critical in assessing species-herd protection effect. We established 600 seedling plots at two filed stations with marked altitude difference in a subtropical forest of Southern China. We developed a novel model, where a single heterospecific neighbor's effect was linearly related to its phylogenetic distance to its corresponding focal seedling and neighbor effects were additive. We tested (1) whether there was species-herd protection effect at the two stations of different altitudes separately, and (2) If there was species-herd protection effect, at what phylogenetic distance between a focal individual and its neighbor the effect would emerge.
Results/Conclusions
We found that average seedling density, soil moisture and light condition at lower altitude was different from those at higher altitude. At the lower altitude, we found significant negative conspecific neighbor effect (p<0.001) and that heterospecific neighbor effect displayed significant phylogenetic signal (p<0.001), i.e. a more phylogenetically distant neighbor was more likely to increase the survival odds of the corresponding focal seedling. At community average level, a neighbor with phylogenetic distance >146 mys improved its corresponding focal individual’s survival odds. On the opposite, a neighbor with average phylogenetic distance <146 mys would decrease its corresponding focal individual’s survival odds. At the higher altitude, the model failed to detect any kind of neighbor effects. In this research, our results support the species-herd protection hypothesis in the community at low altitude. We suggest that it’s necessary to take phylogeny information into consideration when detecting neighbor effects. The different results of the communities at the two different altitude may result from different environmental conditions and seedling density, both of which can affect the spread of natural enemies.