Comparative Demography of Anthriscus caucalis across plant communities in its novel range: Inferences for population spread

Background/Question/Methods

Recent efforts to produce a conceptual framework for the study of biological invasions have identified population demography as central to any explanation of fundamental ecological and evolutionary processes that contribute to invasion.  Comparative demography studies offer an efficient method of identifying life history traits that lead to rapid population increase or dominance at local scales, and to identify mechanisms or processes that lead to range expansion when coupled with investigation of dispersal processes.  Few studies have utilized this approached in an invader’s introduced range. 

            The goal of this study was to compare the demography of Anthriscus caucalis, an exotic winter annual forb in canyon grassland ecosystems of the Inland Northwest, across a range of plant communities.  We conducted a four-year demographic study within two bunchgrass community types and two shrub community types at two sites. Intra-annual censuses focused on aboveground measures of abundance and associated vital rates, fecundity and juvenile survival.  Specifically, we tested for differences in means, demographic variance, and environmental variance.  Population matrix models were constructed with vital rates for each community type, and coupled with independent seedbank parameter estimates, to simulate stochastic transient (5 yr) population growth rates (log λ_t) under different dispersal rate scenarios.

Results/Conclusions

Mean juvenile and adult densities differed (P < 0.05) among communities within sites.  Juvenile survival rates did not differ among communities, but demographic variance was significantly lower in shrub communities. Mean fecundity was greater in high shrub (Celtis reticulata; CERE) communities compared to others, whereas demographic and environmental variance was significantly lower in CERE communities compared to others. Under assumptions of no seed dispersal, growth rates (log λ_t) for two of four bunchgrass sites declined (log λ_t < 0; bootstrapped 95% confidence intervals) and each shrub community increased.  CERE communities maintained positive growth rates under assumptions of 60% seed dispersal, while other communities continued to decline.

            This study informs management efforts for: (1) control programs for areas of local dominance, and (2) monitoring programs to prevent range expansion. Stable populations are more likely to occur in high shrub (CERE) communities, which are also likely sources of population spread.  A. caucalis seed morphology indicates that animal dispersal is a likely method of spread. Consequently, the relative amount and physical layout of plant communities may elucidate dispersal patterns.  Our results are being integrated with topographic and remotely sensed data for use in species distribution models (SDMs) at landscape scales within canyon grasslands.