Propagule pressure does not consistently predict the outcomes of exotic bird introductions

Abstract

Some have argued that the role of propagule pressure in explaining the outcomes of bird introductions is well-supported by the historical record. Here, we show that the data from a large published database (including 832 records with propagule information) do not support the conclusion that propagule pressure is the primary determinant of introduction success in birds. A few compendia of historical reports have been widely used to evaluate introduction success, typically by combining data from numerous species and introduction locations. Very few taxa, other than birds, have usable spatially explicit records of introductions over time. This availability of data inflates the perceived importance of bird analyses for addressing factors related to invasion success. The available data allow limited testing of taxonomic and site-level factors of introduction outcomes. We did find significant differences in effort and success probabilities among avian orders and across highly aggregated spatial regions. As a test of a standard and logical expectation of the propagule pressure hypothesis, we concentrated on introductions with the smallest propagules, because it is for these the hypothesis is most likely to be correct. We analyzed the effect of numbers released in small propagules (two through 10) for 227 releases. Weighted linear regression indicated no significant effect of propagule size for this range of release size. In fact, the mean success rate of 28% for propagules of 2–10 isn’t significantly different than that of 34% for propagules of 11–100. Following the example of previous analyses, we expanded the statistical test of propagule pressure to include the full range of release numbers. No significant support for the propagule pressure hypothesis was found using logistic regression with either logit or complementary log-log link functions.

The relationship between propagule pressure and establishment success in alien bird populations: a re-analysis of Moulton & Cropper (2019)

“…Exceptional claims require exceptional evidence. Against a backdrop where the large majority of statistically robust studies find a significant positive relationship between founding population size and establishment success for introduced alien populations (see the meta-analysis by Cassey et al, 2018), Moulton & Cropper's (2019) claim that this does not hold for a large sample of bird introductions appears exceptional, particularly given their results are in stark contrast to previous analyses of the same data (Sol et al, 2012;Duncan et al, 2014). This discrepancy motivated our re-analysis of the data.…”

“…The appropriate comparison is thus one where outcomes for small founding population sizes are compared to larger ones, not to restrict analysis to very small populations, which we know are likely all to have a low establishment probability. Moulton & Cropper's (2019) analysis based on all of the data they take from Sol et al (2012) is flawed for different reasons. Moulton and Cropper argue that these data suffer from a bias due to the aggregation of releases.…”

“…The dataset published by Sol et al (2012) includes information on a wide variety of traits that might affect establishment success, but we focus on the subset of 832 separate introductions of bird populations worldwide analysed by Moulton & Cropper (2019), and presented by them as an appendix to their paper. This subset includes information on taxonomic order, family and species, group (four categories, though it is not stated to what these groups refer), propagule pressure (number of individuals introduced), and introduction outcome (0 = failure, 1 = success).…”

“…To this dataset we added information on region of introduction (Australasia, Ethiopian, Nearctic, Neotropical, Oriental, or Palaearctic) and location of introduction (a total of 46 different country or island territory locations, nested within region) from the original source (Sol et al, 2012). Moulton & Cropper (2019) updated the taxonomy used by Sol et al (2012), and we use their revised groups.…”

“…Here, we revisit the Sol et al (2012) data, and the analyses presented by Moulton & Cropper (2019), to explore whether they fail to show the propagule pressure effect, as Moulton & Cropper claim.…”