County-scale mapping of the invasion stage and invasiveness of major nonnative invasive plants in the upper Midwest forestlands, USA

The determination of invasion stages and the degree to which an invasive plant (non-native invasive plant, NNIP) has become established and spread in an ecosystem ("invasiveness") is essential for developing methods of mitigation and control. We mapped the invasion stages and quantified the invasiveness of four NNIPs of great concern, multi ora rose (Rosa multi ora Thunb), nonnative bush hon- eysuckles (including four species from the Lonicera spp family), common buckthorn (Rhamnus cathartica L.) and garlic mustard (Alliaria petiolata [M. Bieb.] Cavara & Grande) in the Upper Midwest forestlands. Specifically, we used the product of the estimated presence probability and mean cover rate of an NNIP from a group of Forest Inventory and Analysis (FIA) plots in a county to represent its severity or area occupied. We then calculated the empirical cumulative density function (ECDF) of the occupied area and used classification and regression tree (CART) to classify the ECDF into a number of disjoint segments to spatially represent invasion stages of an NNIP. The invasiveness of an NNIP in three major forest type groups was then investigated via regression analysis of the change in the estimated mean cover rate with the estimated presence probability across the mapped invasion stages (a proxy for invasion time). This study demonstrates the feasibility of using data from a single time period for determining invasion stages and invasiveness of NNIPs for the deployment of controlling or eradicating measures.