Although body size has been found to be positively correlated with increased vulnerability in several insect groups, including hoverflies (Sullivan et al. 2000), carabid beetles (Kotze and O’Hara 2003) and butterflies

(Shahabuddin and Ponte 2005), our results are consistent with other studies on butterflies and moths that reported no relationship between body size and threatened status or risk of population extinction (Thomas and Morris 1995; Nieminen 1996; Koh et al. 2004; Kotiaho et al. 2005; Mattila et al. 2006). Variability, extrinsic factors, and the prediction of vulnerable endemic taxa The goal of this Blasticidin S analysis was to identify the life history traits of endemic species that correlate with the greatest risk of population declines or PI3K inhibitor extinction. Our results indicate that among endemic Hawaiian arthropods, low population density and carnivory are risk factors, especially when co-occurring. Many additional species were negatively impacted by invading ants, however, indicating that the explanatory factors examined had relatively weak predictive power for a substantial subset of

arthropods. Among non-rare species, for example, the best model only explained about 21% of the variation in average population response. For rare species, predictive power was better, but the best model AG-881 datasheet still correctly classified only 42% of vulnerable species. Examination of trends among taxonomic orders was not overwhelmingly helpful. Endemic beetles and spiders showed the most consistency in their negative responses to ants (Tables 3, 4), as has been noted previously (Perkins

1913; Cole et al. 1992; Gillespie and Reimer 1993; Liebherr and Krushelnycky 2007). Spiders are all carnivores, but the beetles included three trophic classes, suggesting that endemic beetles share other traits that make them inherently vulnerable to invasive ants. Non-rare endemic moths were also consistently strongly impacted by ants (as in Cole et al. 1992), but this was not true of rare moths. For most of the remaining orders, a range of responses was observed and strong trends were not evident. It is Sclareol possible that the consideration of additional intrinsic factors could improve predictive ability, although many traits are not relevant, known, or easily measured across the wide range of orders considered here. For example, several studies have suggested that taxa possessing thick exoskeletons may be more resilient to invasive ants (Human and Gordon 1997; Hoffmann and Parr 2008). Similarly, Cole et al. (1992) made the point that two heavily sclerotized species, an introduced isopod and an endemic millipede, were found in higher abundance within ant-invaded areas at two of the same Hawaiian study sites used here. However, degree of sclerotization is difficult to quantify, and we did not find a consistent effect for this trait.