Prevalence and strength of density dependent tree recruitment in eastern US forests

Background/Question/Methods

Density dependence could maintain diversity in forests, but studies disagree on its importance. Part of the disagreement results from the fact that different studies evaluate different responses (per-seedling or per-adult survival or growth) of different stages (seeds, seedlings, or adults) to different inputs (density of seedlings, density or distance to adults). Most studies are conducted on a single site and thus are difficult to generalize. Using USDA Forest Service’s Forest Inventory and Analysis (FIA) data, we analyzed over a million seedling-to-sapling recruitment observations of 50 species for both per-tree (adult) and per-seedling recruitment rates, controlling for climate effects in eastern US forests. We focused on per-tree recruitment as it is most likely to promote diversity at the population level, and it is most likely to be identified in observational or experimental data. To understand the prevalence of density dependence, we quantified the proportion of species with significant positive or negative effects. To understand the strength of density dependence, we determined the magnitude of effects among conspecifics and heterospecifics, and how it changes with overall species abundance.

Results/Conclusions

The majority of the 50 species have significant density dependence effects, mostly negative, on both per-tree and per-seedling recruitment. Per-tree recruitment is positively associated with conspecific seedlings, saplings, and heterospecific saplings, negatively associated with heterospecific seedlings, conspecific and heterospecific trees. Per-seedling recruitment is positively associated with conspecific and heterospecific saplings, but negatively associated with conspecific and heterospecific seedlings and trees. Furthermore, for both per-tree and per-seedling recruitment, density dependence effects are stronger for conspecific than heterospecific neighbors. However, the strength of these effects does not vary with species abundance. We conclude that density dependence is pervasive, especially for per-tree recruitment, and its strength among conspecifics and heterospecifics is consistent with the predictions of the Janzen-Connell hypothesis.