Lengths and correlates of lag phases in upper-Midwest plant invasions

There are often lag phases in plant invasions, seemingly dormant periods between arrival in a new range and rapid population growth. Lags impede prioritization of invasive-species control efforts: when eradication is most feasible, it is often unclear whether a species is benign or a potentially harmful "sleeper weed." I used herbarium records to estimate lag phases for invasive or potentially invasive plant species in three regions of the upper Midwest. I tested whether factors related to species' invasion epidemiology, traits, or the habitats they invade were correlated with lag lengths. From an initial pool of 151 species, there were sufficient records to test for lags in 76 for northern Wisconsin, 90 for southern Wisconsin, and 91 for the southern Lake Michigan region. Lags were identified in 77% (197) of these 257 datasets and ranged from 3-140 years with a mean of 47.3 ± 34.6 (SD). Lags differed by native range, introduction pathway, growth form and habit, dispersal mode, flowering phenology, pollination mode for a subset of species, and breadth, light availability, and water availability of invaded habitats. However, estimated lags were highly variable and tested factors did not have strong explanatory power. Exotic species comprised an increasing proportion of total herbarium records. Of the species with known introduction pathways, 85% were intentionally introduced, mainly as ornamentals. The long durations, high variability, and low predictability of lags, along with human culpability for an increasingly non-native flora, support a cautious approach to species introductions.