Effects of trees on their recruits in the southern Appalachians, USA

Abstract

The negative effect of conspecific trees on seedling recruitment in temperate forests has been well documented at the population level for several common species. In 2007, we estimated the survival of 2210 recently germinated seedlings of nine tree species transplanted near conspecific and heterospecific trees, a surrogate for describing distance-dependent mortality, as part of an experiment with landscape-level replication across eight mixed-deciduous forests in the southern Appalachian Mountains of North Carolina. Forest composition was variable but they had a number of woody species in common. Prior to establishing the field experiment, we used a forest inventory database for the region to classify the recruitment patterns of tree species and formulate predictions for species. The field experiment, conducted as a drought was progressing, revealed that four of the nine species planted had variable survival around conspecifics compared with heterospecifics suggesting variation in distance-dependent mortality. Acer saccharum and Tsuga canadensis both had greater mortality near conspecifics than heterospecifics, while Fagus grandifolia and Prunus serotina showed the opposite pattern. Species classified as having greater recruitment around conspecifics, according to the forest inventory data, suffered greater overall levels of mortality in our field experiment. Possibly because of the progressing drought, none of the four species predicted to be most affected by distance-dependent sources of mortality based on the forest inventory data exhibited the predicted patterns of survival near conspecific vs. heterospecific trees in the field experiment. Furthermore, two of the four species ( A. saccharum and T. canadensis) classified as being least affected by conspecific trees actually had greater survival near heterospecifics than conspecifics. Although we identified effects of canopy tree type in four of the nine comparisons, negative effects of conspecific trees were observed for only two ( A. saccharum and T. canadensis) of nine species and mostly contradicted predictions based on patterns from forest inventory data. The inconsistency between patterns from the forest inventory data and from experiments indicates that there may be localized, complex interactions that make generalizations about neighbor effects on tree seedling survival difficult.