Developing methods to assess the extent of naturalized slash pine populations and their habitats characteristics in South East Queensland, Australia

Slash pine is an exotic commercial species that is widely planted across South East Queensland’s coastal region. Several studies have observed naturalized slash pine populations around plantations, although there have been minimal efforts to control reported ecological impacts. The aim of this project was to perform a pilot study within Bribie Island to develop methods for mapping slash pine wildings and defining their habitats characteristics; the final goal being to suggest management prioritization approaches. Cost and time effective large scale mapping requires a remote sensing based approach: this study applied the mixture tuned match filtering algorithm to a SPOT 5 image. SPOT data was retained since it was the most appropriate imagery available. The result of the MTMF classification was yet not satisfactory because of SPOT low spectral resolution. In order to determine habitats characteristics, a map of pine occurrence was however produced by stereoscopy. A logistic regression was then performed to predict pine occurrence as a function of distance to plantation, wind direction and ecosystem type. The three explanatory variables were obviously correlated to pine occurrence: slash pines were found within a 500 m buffer zone downwind to plantations and in a limited number of ecosystems. Distance was however the only significant variable for the model since pine presence was not adequately represented in the sample. In order to take the three variables into account, a qualitative approach based on a decision tree was performed to build a map of probability of pine occurrence. Risk maps were then derived combining the probabilities with the Biodiversity Status of the areas concerned. These maps provided the basis for a management prioritization strategy by defining the sensitive edges of the plantations where natural resource managers and the plantation company should concentrate their actions. Suggested management measures include the implementation of buffer zones (e.g. of slash pine clones or tall tree with dense foliage) to prevent seed dispersal, combined with remediation measures (i.e. slashing and burning). Concentrating management actions to high risk zones will maximise efficiency whilst limiting costs.