Propagule pressure is an important predictor of spatial and temporal patterns of invasions by non-native species. However, most studies of the effects of propagule pressure have used only current, not historical, data. We compared the effects of past and present propagule pressure on present distributions of woody species in central Texas. To estimate past propagule pressure, we used surrogate variables such as the age of nearby development to represent how long propagule pressure had been present. To estimate the effect of present propagule pressure we used surrogate variables such as distance to the nearest developed area. We quantified the relationships between these surrogate measures of past and present propagule pressure, a selected set of environmental variables, and the distributions and abundances of native and non-native woody plant species. We sampled randomly-located plots in three plant communities, and used the results to construct statistical models predicting native and non-native species richness and the occurrence of five common non-native species from five development-related variables and six ecological variables. We compared models using the corrected Akaike information criterion (AICc).
Results/Conclusions
Age of residential development was the best predictor, other than community type, of non-native species richness. Because each of the woody non-native species that was common in our plots is a landscaping plant, it is not surprising that areas near older developments had more non-native species than areas near newer developments. However, it was unexpected that both age of nearest development and average city age, two measures of the age of source populations, were much better predictors of non-native species richness than distance to source populations. Age of source population was somewhat correlated with distance to source population, but we were able to statistically separate their effects. We conclude that in some published studies the reason that distance to a source population has been found to be a successful predictor of invasion may be because it is correlated with length of time of exposure to a source population. Future studies of invasions would therefore benefit from taking into account past as well as present propagule pressure. Our method of estimating the duration of propagule pressure, based on publically-available house construction dates, is readily transferable to any location in which such records are available.