Patterns of demographic response of American ginseng (Panax quinquefolius L.) populations to climatic variation

Background/Question/Methods Classic bioclimatic envelope approaches to projecting species response to climate change assume a range-wide relationship of fitness to climate. However, temperature optima may be population-specific if populations within a species are adapted to local climate. While local adaptation to climatic variables has been experimentally demonstrated in several species, it has rarely been possible to examine demographic data for evidence of local climatic adaptation. In this study, we use a long-term, spatially extensive demographic dataset for American ginseng to ask ‘Does the demographic response of ginseng populations to annual temperature variation suggest local adaptation, or conversely, does it suggest a species-wide, Gaussian relationship to temperature?’ Twelve populations spanning a broad growing season temperature gradient (24 – 31^oC) in eastern North America were censused for 5-11 years to parameterize 9-stage transition matrices. Population growth rate (l) was regressed on temperature across all transition years, and repeated for individual populations. ANCOVA was used to test for heterogeneity of slopes among populations at 2°C ‘data-windows’. Additionally, when parabolic functions were fit to population-level regressions of l on temperature, temperature optima were calculated, and then regressed on 30-year mean maximum temperatures. 

Results/Conclusions There was no statistically significant relationship of l and temperature, when population was not included in the model. Populations responded differentially to temperature in 1 of the 5 data windows. In the remaining data windows, patterns suggested differential response to temperature among populations, however insufficient statistical power likely precluded detecting statistical significance. There was a direct linear relationship between demographic temperature response curve optima and 30-year mean maximum temperatures at each population. Overall, demographic patterns suggest that ginseng populations are locally adapted to climate. This has profound implications for ginseng response to climate change. First, if ginseng populations are locally adapted, populations across ginseng’s entire range will be negatively affected by directional climate change. Secondly, smaller increases in temperatures will push population growth rates farther from local optima for locally adapted populations than if all populations shared a range-wide optimum. This implies that climate change may impact ginseng populations sooner, and ultimately more dramatically, than would be predicted with a bioclimatic envelope approach.