Modeling invasive species risk from established populations: insights for management and conservation

Abstract

Ecologists commonly use ecological niche models (ENMs) to undertake invasive species risk assessments;however, knowledge shortfalls introduce bias in these models and increase uncertainty while addressingquestions in biogeography. Therefore, our objective was to investigate how the lack of information relatedto population viability impairs invasive species risk assessments. We built ENMs for the invasive sliderturtle (Trachemys scripta) and compared the native and invaded portions of its niches. Both analyses weregenerated based on two approaches: a conservative one, which excluded occurrence records where theestablishment of invasive populations was not confirmed; and a non-conservative one, which encom-passed all occurrence records. Under the conservative approach, the niche similarity test revealed thatthe similitude between native and invasive populations was not different than the expected by chance.Conversely, under the non-conservative approach, the test revealed that native and invasive popula-tions use a similar ecological niche, despite that the occupied portion of the species’ niche during theinvasion was quite larger than the native one. In fact, non-conservative models projected wider areaswith high risk of invasion that were not detected by the conservative approach. While models’ outcomeswere markedly different, both provide valuable information in terms of evolution and conservation. Wefound that information about population viability is really valuable and should be incorporated in riskassessment, invasive records without this information should not be discarded under any point of view.Finally, we discussed the best way to consider this kind of information to assess the management ofinvasive species.