A model that will assist on-ground practitioners in their quest to protect biodiversity values in natural ecosystems

Abstract

There has been ongoing discussion around the necessity for quantitative models in ecology. The use of quantitative modeling is well established in some areas of endeavour, such as Before-After-Control-Impact (BACI) studies, but not in others, in particular the field of invasion ecology. In weed risk analysis, semi-quantitative models (scoring systems, with or without weighting procedures) help policy makers to assess the risk (hazard) posed by individual weed species. Such systems are available to assess weed risk management feasibility at larger geographic scales. However, nothing is available to assist on-ground practitioners in prioritising weed control at the individual site level. Interestingly, the fundamental problem of model choice was solved in the early 2000s by sociological researchers (Dana and Dawes 2004), who demonstrated that qualitative models actually outperformed quantitative models, as long as all of the important factors in the system had been identified. An earlier attempt to establish this finding in the weed invasion literature (Panetta and Cacho 2014) has not been successful. In this paper, I use the results from an ongoing project (“Future-proofing Australia’s National Post-Border Weed Risk Management System”) to develop a model that combines both qualitative and semi-quantitative approaches. This model should be fit-for-purpose by practitioners at the site level, as well as by policy makers charged with allocating scarce resources at larger geographic scales