Experimental demography of a woody invasive plant and its native analog; building life histories from short-term datasets using integral projection models

Background/Question/Methods

Invasive species’ distributions are often not at equilibrium in their invasive ranges. Therefore, inferring population dynamics based on current locations of populations may under- or over-estimate population growth rates and, thus, potential spread. We took an experimental approach to investigating the demographic processes that underlay population dynamics across a range of environmental conditions hypothesized to be invasible. Linking the environment to demography is important to determine species range limits and to provide a mechanistic basis for understanding invasions. This link is particularly important for understanding invasive species distributions and for designing appropriate management actions because only mechanistic models allow us to confidently extrapolate population level patterns to a new set of conditions or novel landscape. We investigated establishment dynamics of the woody invasive species Berberis thunbergii and a native ecological analog, Lindera benzoin, using hierarchical Bayesian integral projection models (IPMs) to predict potential establishment risk across an environmental gradient. The IPMs allowed us to combine multiple, relatively short-term data sets of woody individuals to make estimates of population-level patterns.  

Results/Conclusions

Establishment risk varied widely by environment. Throughout the range of both species, the demographic variation at the population level differentiated potential dispersal limitation from environmental limitations in their distributions. For the native L. benzoin, the estimated establishment risk corresponded to current known distributions for the species; positive population growth was found under conditions consistent with the species’ current range, while negative population growth was found in environments outside their current distributions, thus, indicating an environmental limitation to the distribution. The invasive, on the other hand, showed positive establishment risk outside current distributions suggesting that B. thunbergii is primarily dispersal limited. The differences between natives and invasives in the vital rates to which lambdas are sensitive, illustrate that the invasions are based in demography. The links between environment and demographic rates can help us predict establishment risk and viability as species continue to spread into new environments.