ProfessorMichael Charleston

Project Total: $510,832; The project aims to address the issue of automatically monitoring multiple threatened Tasmanian forest species in real time. Current machine learning approaches process acoustic data collected offline requiring a large amount of resources and incurring long delays in taking effective decisions. We will design novel low-powered wireless acoustic monitoring devices for deployment in native forests to automatically detect and report on the presence and activity of threatened species, which will allow near-continuous monitoring. This will enable collection of multi-year conservation and forest management data at substantially reduced cost and assist in minimising environmental impact of commercial forestry. Funded by: Forest & Wood Products Australia Limited - Grant - Research, Development and Extension (RD&E) ($510,834); Hancock Victorian Plantations Pty Ltd ($44,281); Forico Pty Ltd ($44,281); Sustainable Timber Tasmania ($44,281); Department of Natural Resources and Environment Tasmania ($10,000); University of Tasmania ($122,879).

Project Total: $380,024; There is a growing interest in the ways in which biological and socioeconomic heritage can shape vulnerabilities to disease. Once viewed as primarily a product of recent conditions such as lifestyle choices, it is now evident that health outcomes can also be shaped by intergenerational mechanisms. Analysis of these in current populations is impractical given the considerable time it would take for a prospective study to unfold. The analysis of historical populations, however, presents an opportunity to circumvent this obstacle. Using data for male and female convicts and their descendants, this project seeks to determine the extent to which disadvantage experienced by one generation impacted on the life expectancy of those that followed. Funded by: Australian Research Council - Grant-Discovery Projects; Monash University; University of New England; University of Guelph; University of Tasmania.

Project Total: $328,117; Our goal is the development of feasible in situ control of sarcoptic mange in wombat populations. Globally important, the Sarcoptes scabiei mite infects >100 mammal species and is among the 50 most common human diseases, causing health, welfare and population impacts. This infection is treatable, and we will test a new treatment (fluralaner), develop new models to guide management, and conduct replicated field trials. This will enable science-based guidelines, advancing disease control, local eradication, and regulatory approval for wombats. Our research framework is adaptable to other mange-impacted species, and advance methods and theory for control of treatable disease in wildlife. Funded by: Australian Research Council - Grant-Linkage Projects; Hydro Electric Corporation; Department of Natural Resources and Environment Tasmania; University of Sydney; University of the Sunshine Coast; Highland Conservation Pty Ltd; Bonorong Wildlife Sanctuary Pty Ltd; Water NSW; MSD Animal Health; University of Tasmania.

Project Total: $397,000; This project investigates the dynamics of the spread of infectious disease in wildlife, derived from incomplete information about contact networks. Infectious diseases in wildlife are difficult to track and control, because it is not feasible to monitor each individual in a population and know all its contacts - the contact network for a population. The project will create ways to best utilise incomplete observational data of contact networks to develop robust predictions of disease spread and population fate, and to reliably predict the outcomes of management interventions. These robust prediction methods will provide better insights for conservation of Australian wildlife. Funded by: Australian Research Council - Grant-Discovery Projects ($397,000); University of Tasmania.

Project Total: $4,000; Conservation efforts for large ecosystems require monitoring in order to determine how well they are working. In particular, terrestrial ecosystems such as production forests, which are heavily regulated for conservation, require detailed monitoring in order to fulfil federal requirements of the sustainable production of wood - DAFF 2008 policy Australias Sustainable Forest Management Framework of Criteria and Indicators?: Criterion 1. Conservation of biological diversity; and Criterion 3. Maintenance of ecosystem health and vitality speak specifically to the need for cheap and reliable monitoring methods. Since bird populations are a well recognised key indicator of ecosystem health we are leveraging long-term environmental audio recordings of forests under conservation management in order to assess biodiversity and ecosystem health. Such acoustic data need to be annotated with which bird species are present, before ecological research can be done, but engaging experts to perform annotation at these scales is prohibitively expensive and time-consuming. This is a pattern recognition task that has proven difficult to automate, but humans are naturally well equipped to perform. Since there are many enthusiastic amateur birdwatchers, crowd-sourcing may be a solution. The application of crowd-sourcing to the task of bird recognition will involve recruiting citizen scientists to use our custom made annotation web application Birdsong (https://tasbirdsong.duckdns.org/). In image annotation for example, combining multiple overlapping non-expert annotations (e.g., by majority vote) has been shown to be comparable to expert annotations. We will investigate to what degree the idea of Wisdom of the Crowd applies to annotation of ecological field recordings. To the best of our knowledge there are no active efforts to crowd-source bird call recognition, or any research published on bird call annotation/annotator quality or how learning affects bird call recognition in citizen scientists. What we learn from this pilot project will guide us in how to recruit, train and retain citizen scientists. It will produce valuable annotations and meta-data for research in the fields of ecology, human cognition, and machine learning and statistics. Further, it will inform future research efforts towards community involvement in conservation research and the efficient production of annotated data for ecological monitoring. Funded by: University of Tasmania; University of Tasmania.

Project Total: $52,000; A PhD scholarship: a study that evaluates the performance of machine learning algorithms in analysing digitally-recorded data to classify bird calls Funded by: Forest Practices Authority ($4,000); Forestry Tasmania ($15,000); VicForests ($15,000); Southern Regional Natural Resource Management Association Inc (NRM South) ($9,000); Tasmanian Land Conservancy ($9,000); University of Tasmania.

Project Total: $650,933; Funded by: Australian Research Council - Grant-Linkage Projects Round 1 ($410,933); Forestry Tasmania ($120,000); VicForests ($120,000); University of Tasmania.

Project Total: $469,800; Funded by: Australian Research Council - Grant-Discovery Projects; University of Tasmania.