Preventing the establishment of a non-native species is critical for ensuring the species does not become invasive, yet most non-native species will have little impact on their environment. Despite this, little is known about what influences whether a species will remain relatively benign, or whether it will cause economic or ecological harm. Understanding a plant’s microhabitat provides insight into the necessary conditions for establishment and the current distribution limitations of a population. We investigated microhabitat preference of the non-native natal grass (Melinis repens (Willd.) Zizka) in Florida scrub using microhabitat sampling to measure vegetation composition. We examined the extent to which microhabitats were associated with natal grass presence and biomass in invaded disturbed scrub and roadside plots using backwards stepwise logistic regression and general linear models to identify significant microhabitat variables. We further compared these plots with those in undisturbed, uninvaded scrub to characterize vegetation across habitat types, and used our model to predict the probability of natal grass invasion in undisturbed scrub. Natal grass preferred microhabitats with high litter volume and distance to shrubs and intermediate cactus, graminoid, and vine cover. Roadside natal grass achieved higher biomass and was less microhabitat limited than disturbed scrub natal grass. We determined that undisturbed scrub plots represent distinct microhabitats that natal grass is unlikely to invade. Microhabitat sampling provides land-managers a non-intrusive technique to assess potential habitat suitability based non-native plant preferences before a costly invasion occurs.
Read full abstract