Potential habitat distribution for the freshwater diatom Didymosphenia geminata in the continental US

© The Ecological Society of America www.frontiersinecology.org E change in North America has reinforced the importance of habitat modeling, to determine the habitat preferences and potential geographic distributions of invasive species in terrestrial (Stohlgren et al. 2006) and aquatic (Williamson et al. 2008) systems. The diatom Didymosphenia geminata is a single-celled alga (Bacillariophyceae; Figure 1) that is becoming increasingly prevalent in North America (Spaulding and Elwell 2007) and is invasive in New Zealand (Kilroy et al. 2008). This diatom has been reported in the western US for over 100 years, but more extensive, nuisance growths have recently become common; nuisance growths are also appearing with greater frequency in the eastern US. In New Zealand, this species was initially discovered on the South Island in 2004 and is now present in over 21 rivers (Duncan 2007), where it forms large growths at several sites. D geminata has the potential to generate serious ecological and economic impacts in both these countries. Unpublished studies in the US and published studies in New Zealand (eg Kilroy et al. 2006) indicate large increases in algal biomass at sites impacted by D geminata, and shifts in algal species composition. There are also differences in the major invertebrate groups between nonimpacted and impacted sites (Kilroy et al. 2006). Predictive modeling provides an opportunity to examine the role of specific environmental variables that may be associated with the distribution of a species at various spatial scales and can help to determine appropriate management actions. Our objectives were to: (1) predict the potential suitable habitats for D geminata; (2) compare four species distribution modeling techniques for predicting suitable habitats for D geminata; and (3) determine D geminata’s response to bioclimatic, topographic, geologic, and hydrologic variables in the continental US.