Graduate student Utah State University, United States
Invasive species are a major conservation threat. One approach for their management is early identification and targeting of invasion “hotspots” (areas of high abundance and/or richness of invasive species) to limit further spread. Identifying hotspots and their drivers can also be useful for answering ecological questions: e.g. Why are some areas more invaded than others? Do alien species facilitate one another? Etc. However, most studies on invasion hotspots are based on species distribution models, where the relationship between each species and its environment is modelled, and the models are then used to predict areas of high habitat suitability for multiple species. While this is useful for predicting future hotspots, it has limited utility for identifying current hotspots. It may also be inaccurate when applied to emerging invasive populations, which are unlikely be in equilibrium. Therefore, we developed a new methodology to identify current hotspots and their drivers, using tests of spatial clustering, correlograms, and ordinations. We applied this methodology to emerging plant invaders in Utah’s rangelands, using data from the state’s Range Trend program. This program involves systematic long-term monitoring of rangeland vegetation, and the use of this data thus reduces biases inherent in citizen science and herbarium records.
The spatial distribution of the emerging invasive species was significantly clustered, with areas of highest invader abundance appearing close to the largest urban centers in the state. An unconstrained ordination was used to identify species with the highest contribution to the variation in invader composition across sites. These were Aegilops cylindrica (jointed goat-grass), Carduus nutans (musk thistle), Convolvulus spp. (field bindweed), and Cynoglossum officinale (houndstoungue). All of these major players had high abundances, wide-spread distributions, or recent increases in population size, indicating that they could have high ecological impacts. Finally, we investigated the drivers of hotspots using a constrained ordination, with climatic factors and proxies of anthropogenic disturbance as predictors. The strongest predictors were minimum precipitation and road density. This is line with studies that indicate a positive association between anthropogenic disturbance and invasion, and a negative association between harsh environments and invasion.
Thus, our methodology appears to be successful in detecting hotspots, and the sites and species identified here can be targets of future management actions. We were also able to address ecological questions on factors associated with site invasibility. Therefore, we believe that our novel approach has applications for both basic and applied questions in invasion ecology.