An experimental demographic approach to exotic plant success : Cynoglossum officinale in its native and introduced ranges.

Invasive species often face novel abiotic and biotic environments with different selective regimes where they are introduced.

How these changed conditions influence individual life-history traits, and what particular factors spur increases in population abundance in the introduced versus native range, are not well understood.

I conducted parallel experiments in both the native and introduced ranges of a widespread plant invader in North America, houndstongue (Cynoglossum officinale).

I combined these experimental results with demographic monitoring in each range, and population modeling, to explore how introduction has affected houndstongue demography and life-history evolution, and to determine the role of specialist herbivores and altered responses to disturbance in affecting plant population growth in both ranges.

From reciprocal common gardens in each range, I found substantial population-level plasticity in size and fecundity between native and introduced populations of houndstongue, but no significant genetically based differences in morphology.

Differentiation of native populations in the magnitude of plasticity were much stronger than that of introduced populations, suggesting an important role for founder effects.

From demographic data collected in each range, I found that both survival and growth were higher in the introduced range, where size at flowering was larger and iteroparity more common.

Since iteroparity conferred higher fitness in both ranges, my results imply severe constraints on the evolution of this life-history strategy in the native range, potentially because specialist herbivores select for plants that flower only once.

Finally, results from manipulative experiments at multiple sites in each range involving suppression of insect herbivore pressure and creation of small scale disturbances revealed several important results.

First, specialist herbivores reduced plant size and fecundity in Germany, but generalist herbivores had no effect on plant performance in Montana.

Second, in both ranges, seedling recruitment responded positively to disturbances, but seedling survival was more positively affected in Montana.

Integrating these results into integral projection models of population growth suggest that while escape from enemies may contribute slightly to the increased abundance of houndstongue in North America, it is the differences in response to small disturbances that leads to higher abundance in the novel range compared to at home.