Invasive plants continue to spread in riparian ecosystems, causing both ecological and economic damage. This research investigated the impacts of common reed, purple loosestrife, riparian shrubland, and riparian woodlands on the quality and quantity of sandhill crane roosting habitat in the central Platte River, Nebraska, using a discrete choice model. A more detailed investigation of the impacts of common reed on sandhill crane roosting habitat was performed by forecasting a spread or contraction of this invasive plant. The discrete choice model indicates that riparian woodlands had the largest negative impact on sandhill crane roosting habitat. The forecasting results predict that a contraction of common reed could increase sandhill crane habitat availability by 50%, whereas an expansion could reduce the availability by as much as 250%. This suggests that if the distribution of common reed continues to expand in the central Platte River the availability of sandhill crane roosting habitat would likely be greatly reduced.

Nomenclature: Common reed, Phragmites australis (Cav.) Trin. ex Steud.; purple loosestrife, Lythrum salicaria L.

Interpretive Summary: This research demonstrates the suitability of discrete choice modeling for quantifying the impacts of invasive plants on sandhill crane roosting habitat. Moreover, this research successfully applied a discrete choice model to forecast the implications that a spread (unsuccessful management) or contraction (successful management) of the invasive plant common reed (Phragmites australis) could have on the distribution and abundance of sandhill crane roosting habitat. The methods presented in this research could easily be adopted to provide managers with useful information that aids in the management of invasive plants for the benefit of native species' habitat.

For example, managers could use the modeling and forecasting methods developed in this project to quantify the degree to which invasive species impact native species' habitat, prioritize the location of invasive plant management, and forecast which invasive species are most likely to impact native species. In addition, the discrete choice model forecasting methods described in this research allow the user to quantify the change in abundance and distribution of native species habitat as a result of invasive species spread or contraction. As the availability of digital land cover data grows and knowledge of invasive species' impacts expands, discrete choice modeling and forecasting simulations based on discrete choice models could be instrumental in aiding a field practitioner's management of invasive plants.