Plant functional traits have the potential to generalize ecological analysis methods and explain patterns of variation across and within species. However, all trait-based approaches make the untested assumption that traits regulate demography (survival, growth and reproduction) to determine population growth, and most assume limited trait variability within a species. We used the non-destructive technique of reflectance spectrocoscopy to measure ten different functional traits, including leaf nitrogen, leaf-mass-area, and defensive compounds (i.e. cardenolides) of over 200 milkweed plants in situ from four populations for which we also have demographic data. Using Principal Component Analysis (PCA), we then quantified within-species variation in these traits. In order to understand how individual variation in these traits interacts with herbivory to influence population dynamics, we present an Integral Projection Model approach that uses plant size, herbivory severity, and functional traits as continuous predictors of ramet population growth rate in Asclepias syriaca.
Results/Conclusions
Through PCA, we identified two significant axes (accounting for 67.1% total variance), with the first axis corresponding to carbon allocation (cardenolides, leaf-mass-area, cellulose, fiber, lignin), and the other to photosynthetic activity (leaf N, photoreflective index, chlorophyll). We found a trade-off between growth and defense in the leaf traits: plants with higher levels of cardenolides had lower levels of structural carbon compounds (P < 1x10-5). Cardenolides also had a significant negative impact on plant demography: higher concentrations of cardenolides decreased the probability of flowering (β = -0.75, P < 0.01) and seed production (β = -0.97, P < 1x10-4) as well as survival to reproduction (β = -0.55, P = 0.01). Leaf mass area increased survival (β = 0.03, P < 0.01) and decreased seed production (β = -0.03, P = 0.01). Photosynthetic traits did not affect plant demography. We then constructed an Integral Projection Model (IPM) that predicts population growth based upon plant height, herbivory intensity, and cardenolides and leaf-mass-area. Our research is a novel and powerful extension of IPMs to include plant functional traits and herbivory intensity as continuous predictors in demographic functions that bridges trait-based approaches in community ecology with demography and population dynamics.