Population persistence of non-native invasive plants

Background/Question/Methods

Natural history records have been shown to be useful in both mapping current distributions of non-native invasive species (NNIS) and predicting their future spread. However, due to factors such as the Allee effect, pathogens, herbivory, land use change, and succession, plant populations do not always persist. Given sufficient time, might NNIS populations respond to such factors with localized extinctions and if so, how long would this take? Furthermore, what might this mean for the usefulness of older natural history records? To explore these questions we are revisiting 400 known invasive species locations in southern Illinois to see if the populations still persist today. Locations were selected from a database maintained at Southern Illinois University that contains over 10,000 records for non-native plants. The majority of these records came from herbarium specimens.  Eighteen species were selected for our study based primarily on the temporal distribution of records and their abundance in the database. GPS coordinates were estimated for each site with Google Earth based on the location information given in the record. Sites were then visited and surveyed for a maximum of one hour each.  The relationship between presence-absence data from these surveys and record age was analyzed with logistic regression.

Results/Conclusions

Target NNIS were absent from thirteen of the 176 sites visited to date. Eight of the absences were in areas that had been recently mown and two were at sites where the target species is under active management. Preliminary results indicated no significant relationship between NNIS occurrences and record age (χ²=0.5606, df=1, P=0.4540), suggesting that for the eighty years of historic records we used, the populations described in older records are just as likely to be relocated as those found in recent records. Many of the species we used, such as Rosa multiflora and Lonicera japonica, are among the most widespread and problematic NNIS in southern Illinois. These populations do not appear to be dying out once established, except on sites where some human action is taken. This finding emphasizes the need to seek out and control NNIS before they can establish over a large range. Our results validate use of older NNIS records in distribution mapping efforts. If these data have a “shelf life” it is greater than the eighty years we examined.