Matrix Models as a Tool for Understanding Invasive Plant and Native Plant Interactions

Demographic matrix models are an increasingly standard way to evaluate the effects of different impacts and management approaches on species of concern. Although invasive species are now considered among the greatest threats to biodiversity, matrix methods have been little used to explore and integrate the potentially complicated effects of invasions on native species. I developed stage-structured models to assess the impacts of invasive grasses on population growth and persistence of a federally listed (U.S.A.)endemic plant, the Antioch Dunes evening primrose (Oenothera deltoides subsp. howellii [Munz]W. Klein). I used these models to evaluate two frequently made assumptions: (1)when rare plant populations decline in invaded habitats, invasive species are the cause and (2)invasive plants suppress rare plants primarily through direct resource competition. I compared two control and two removal matrices based on previous experimental work that showed variable effects of invasive grasses on different life-history stages of O. deltoides. Matrix analysis showed that these effects translated into substantial changes in population growth rates and persistence, with control matrices predicting a mean stochastic population growth rate (ë)of 0.86 and removal atrices predicting growth rates from 0.92 to 0.93. Yet even the most optimistic invasive removal scenarios predicted rapid decline and a probability of extinction near one in the next 100 years. Competitive suppression of seedlings had much smaller effects on growth rates than did lowered germination, which probably resulted from thatch accumulation and reduced soil disturbance. These results indicate that although invasive grasses have important effects on the population growth of this rare plant, invasion impacts are not solely responsible for observed declines and are likely to be interacting with other factors such as habitat degradation. Further, changes in the disturbance regime may be as important a mechanism creating these impacts as direct resource competition. My results highlight the value of demographic modeling approaches in creating an integrated assessment of the threats posed by invasive species and the need for more mechanistic studies of invasive plant interactions with native plants.