Publication Information

In the Northern Hemisphere, the surface of south-facing slopes orients toward the sun and thus receives a greater duration and intensity of solar irradiation, resulting in a relatively warmer, drier microclimate and seasonal environmental extremes. This creates potentially detrimental conditions for evergreen plants which must endure the full gamut of conditions. I hypothesize that (1) increased southerly aspect will correlate negatively with evergreen understory plant distributions; (2) derived environmental variables (summer and winter light and heat load) will predict variance in evergreen distributions as well as topographic position (aspect, slope, and elevation) and (3) winter light will best predict evergreen understory plant distributions. In order to test these hypotheses, survey data were collected characterizing 10 evergreen understory herb distributions (presence, abundance, and reproduction) as well as the corresponding topographical information across north- and south-facing slopes in the North Carolina mountains and Georgia piedmont. The best predictive models were selected using AIC, and Bayesian hierarchical generalized linear models were used to estimate the strength of the retained coefficients. As predicted, evergreen understory herbs occurred and reproduced less on south-facing than north-facing slopes, though slope and elevation also had robust predictive power, and both discriminated well between evergreen species. While the landscape variables explained where the plants occurred, winter light and heat load provided the best explanation why they were there. Evergreen plants likely are limited on south-facing slopes by low soil moisture combined with high temperatures in summer and high irradiance combined with lower temperatures in winter. The robust negative response of the understory evergreen herbs to increased winter light also suggested that the winter rather than the summer (or growing season) environment provided the best predictive power for understory evergreen distributions, which has substantive implications for predicting responses to global climate change.