The role of landscapes in invasions: Landscape characteristics and invader traits interact to determine invasion risk in models of a potentially invasive bioenergy crop (Miscanthus × giganteus)

Background/Question/Methods

Large-scale production of biofuels offers a path to developing carbon-neutral energy sources building on current agronomic knowledge. There is significant interest in using perennial crops such as Miscanthus × giganteus for this purpose as they produce abundant biomass, have low nutrient requirements, and are drought and cold tolerant. However, the same properties that make species excellent biofuel candidates also make them likely to become invasive. Invasion risk is a function of both species traits and the landscape into which individuals could invade. Incorporating landscape attributes into measures of invasibility is critical for the responsible management of potentially invasive biofuels, and may also be valuable to the management of invasive species in general. Using a spatially explicit simulation model of plant population dynamics we predicted invasion risks on landscapes derived from remotely sensed data for M. × giganteus, as well as theoretical species with different demographic traits. We analyzed the ability of different metrics of landscape composition and spatial structure to predict the risks estimated by our simulations to evaluate how landscape characteristics influence invasion. Using a statistical model, we then predicted invasion risks across the entire United States as part of a cost-benefit analysis for siting options for biofuel cultivation.

Results/Conclusions

While demographic parameters and the biotic resistance of extant communities are key to the invasive potential of specific species, the risks of specific landscapes are largely influenced by three major factors: the abundance of high quality habitat to colonize, dispersal corridors that can speed the spread of invaders and the particular spatial structure of habitat patches. Habitat composition is generally the strongest driver of invasion risk but the relative importance of the different factors varies as a function of invader traits and the timescale of interest. Over long timescales as systems move toward equilibrium, habitat drives patterns, but on shorter timescales, other factors such as the presence of dispersal corridors and patch shape and orientation that support spread play a larger role in determining the speed and extent of invasion. Similarly, because the “natural” mode of spread for invaders is through the production and dispersal of seeds, as either the dispersal capability or fecundity of invader populations decreases, the relative importance of dispersal corridors and spatial pattern increases. Large-scale risk analyses show that optimal biofuel cultivation sites may exist predominantly in the Appalachian regions of the southeastern US where invasion risks are limited and crop yields are high.