Seedling recruitment and survival rates can both create and maintain broad plant community patterns. One such mechanism is through frequency-dependent mortality such as in Janzen-Connell effects. When species-specific enemies accumulate around mature plants, seedlings farther from conspecific adults are more likely to survive. These Janzen-Connell effects have been widely studied in tropical forests and may contribute to high biodiversity. Despite initial assertions that Janzen-Connell effects would only occur in tropical forests, there is growing support for the existence of Janzen-Connell effects in temperate forests, as well.
This study investigates the existence of Janzen-Connell patterns in a mapped Piedmont forest plot in the Georgia State Botanical Garden, USA. All trees and saplings within a 1 ha area were identified, mapped, and measured. A 3m2 seedling monitoring plot was established within each 10m2 plot grid cell. Within each seedling plot, all seedlings of woody species < 1 m tall were mapped, measured, and identified to genus. Seedling growth and survival were observed over a 1 year period.
Results/Conclusions
A total of 1153 seedlings in 18 genera were monitored between 2014 and 2015. Preliminary logistic regression indicates that, overall, seedlings mortality rates were not significantly higher in plots with greater density of conspecific trees (χ2 = 0.30, p = 0.58). For some genera, however, seedling mortality rates were related to other environmental characteristics. Fagus seedlings were more likely to survive in plots with greater heterospecific seedling density (χ2 = 202.43, p<0.001), while Nyssa and Quercus seedlings were more likely to survive in plots with greater conspecific seedling density (Nyssa: χ2 = 4.50, p<0.05; Quercus: χ2 = 4.72, p<0.05). These preliminary results suggest that Janzen-Connell effects are not active in this Piedmont forest. Seedlings do not appear to be significantly limited by frequency-dependent mortality near conspecific trees. Future analyses will incorporate spatially-explicit models to take advantage of the mapped forest dynamics plot.