Assessing species invasions as a cause of extinction

Abstract

In a recent Opinion in TREE [1], Gurevitch and Padillaquestion the generalization that biological invasions area leading cause of species extinctions. The authors notethat declines of native species frequently overlap in spaceand time with invasions by alien species, and theseco-occurrences are sometimes used to infer a causalrelationship – a potentially erroneous conclusion giventhat a common factor, such as physical habitat alteration,might promote both extinction and invasion. The authorsfurther point out that extinctions are often attributable tomultiple causes without any indication of the relativeimportance of invasions, which might be ‘merely corre-lated with other problems’. But I believe they chose a poorexample to highlight this point.Citing the case of native unionid mussels in NorthAmerica, Gurevitch and Padilla assert that the role ofzebra mussels Dreissena polymorpha in their decline isunclear because unionid declines began long before zebramussels were introduced. It is indisputable that mostNorth American unionid species have been declining forover a century owing to various anthropogenic stressors.However, empirical modeling [2] suggests that densezebra mussel colonization has accelerated the localextinctionofunionidspeciesbyafactorof10.Forexample,after having survived decades of environmental degra-dation in the St Lawrence River, several unionid popu-lations declined precipitously and were wiped out within afew years after invasion by zebra mussels [3]. Unionidpopulation mortality across a range of sites in the GreatLakes–St Lawrence River basin is strongly correlated tothe intensity of zebra mussel fouling on unionid shells [4].Such fouling has been shown to interfere with normalunionidmetabolism[5],andunionidsurvivorshipincreasesafter experimental removal of attached zebra mussels [6].Furthermore, mark–recapture studies reveal that unionidpopulations can maintain high annual survival rates atsites wherezebra mussels arescarce orabsent,but declinerapidly with increasingzebra mussel densities across sitesin the same habitat [7]. Thus, although no global extinc-tion has yet been attributed to zebra mussel invasion,there are several lines of evidence demonstrating therole of zebra mussels as a major cause of unionidpopulation extinctions.The example above raises another issue. Although Iagree with Gurevitch and Padilla that a critical synthesisof data is needed to assess the relative importance ofinvasions as a cause of extinction, I would add that such asynthesis should focus on extinction at the populationlevel. Examining population extinctions would enable usto quantify the effect of invasions on the rate at which aspecies is proceeding towards global extinction [2] andwould also result in a more complete understanding oftheir impact on biodiversity loss. A species might undergoa significant range contraction, losing many distinct popu-lations in the process, without becoming a global extinc-tion statistic. Given that a species typically has hundredsof genetically distinct populations [8], any analysis thattallies only global extinctions will overlook a substantialportion of biodiversity loss caused, in whole or in part, byspecies invasions.