Niche shifts and environmental non-equilibrium undermine the usefulness of ecological niche models for invasion risk assessments

Abstract

niche shifts and environmental non-equilibrium in invading alien species undermine niche-based predictions of alien species' potential distributions and, consequently, their usefulness for invasion risk assessments. Here, we compared the realized climatic niches of four alien amphibian species (Hylarana erythraea, Rhinella marina, Hoplobatrachus rugulosus, and Kaloula pulchra) in their native and Philippine-invaded ranges to investigate niche changes that have unfolded during their invasion and, with this, assessed the extent of niche conservatism and environmental equilibrium. We investigated how niche changes affected reciprocal transferability of ecological niche models (ENMs) calibrated using data from the species' native and Philippine-invaded ranges, and both ranges combined. We found varying levels of niche change across the species' realized climatic niches in the Philippines: climatic niche shift for H. rugulosus; niche conservatism for R. marina and K. pulchra; environmental nonequilibrium in the philippine-invaded range for all species; and environmental non-equilibrium in the native range or adaptive changes post-introduction for all species except H. erythraea. niche changes undermined the reciprocal transferability of ENMs calibrated using native and Philippine-invaded range data. Our paper highlights the difficulty of predicting potential distributions given niche shifts and environmental non-equilibrium; we suggest calibrating ENMs with data from species' combined native and invaded ranges, and to regularly reassess niche changes and recalibrate ENMs as species' invasions progress. The large-scale redistribution of alien species-i.e., species whose presence in a region is attributed to human activities that enabled them to overcome fundamental biogeographical barriers (sensu Richardson et al. 1)-is a defining feature of the Anthropocene 2,3. Alien species' invasions can alter the ecology of recipient environments 4,5 and have socioeconomic impacts on recipient jurisdictions 6. Recognizing these impacts, world nations have committed to develop and implement science-based biosecurity policies and strategies in response to ongoing and future alien species invasions 7,8 , for biodiversity conservation 9 and sustainability 10. Invasion risk assessments assess the ecological and socioeconomic impacts of alien species' invasions, producing the needed information to prioritize alien species and areas for biosecurity intervention 7,11. Invasion risk assessment activities include predicting the potential distribution (i.e., climatically suitable areas) of invading and/or potentially invasive alien species 11. These predictions can be made through ecological niche modelling (ENM; aka species distribution modelling) 12-15-a correlative statistical tool that quantifies species-environment

Climatic niche shifts in 815 introduced plant species affect their predicted distributions

“…Unfilling is thought to result from a lag between colonization and range expansion (Petitpierre et al., 2012), so it is unsurprising that niche contractions in climatic space corresponded to predictions of lesser geographic spread, particularly in introduced species. Our study provides further support (Camenen et al., 2016; Li et al., 2014; C. Liu et al., 2020; Strubbe et al., 2013) for the hypothesis that niche dynamics in introduced species are dominated by unfilling (Petitpierre et al., 2012; Václavík & Meentemeyer, 2012), and that this is borne out in geographic models (but see Early & Sax, 2014; Pili et al., 2020).…”

Smaller climatic niche shifts in invasive than non-invasive alien ant species

“…Many studies have used SDMs to predict potentially suitable areas for invasive species 45 . However, if niche shifts occur during invasion, be they realized or fundamental niche shifts, these SDM predictions will not be reliable 46 . But for invasion biologists, it is encouraging that the most invasive species showed the smallest niche shifts, indicating that SDM predictions or climate-matching risk assessments for species which pose the greatest environmental risks may be the most reliable.…”

A Landscape Epidemiological Approach for Predicting Chronic Wasting Disease: A Case Study in Virginia, US

“…In the context of CWD in Virginia, our jackknife analysis identified that every CWD case influences the amount of risk predicted. Hypervolumes never reaching a complete overlap (i.e., a Jaccard index value of 1) could suggest that variation seen in environmental space with new CWD cases could stem from disease non-equilibrium, meaning CWD current distribution may not be exhausting its potential occupancy of environmental conditions (74). Under this finer-population scale, this would not be surprising given what is known about nearby CWD cases outside of Virginia withheld from this analysis, the range of landscape conditions that CWD has been identified worldwide (e.g., Scandinavia) (75), and the environmental hardiness of prions in general.…”