Fluoride Action Network



  • Exposure models can predict toxic effects of fluoride consumption in wildlife.
  • Exposure models that vary in intensity of field data collection warrant comparison.
  • Simple spatial metrics can predict fluoride accumulation in a free-ranging mammal.
  • Complex exposure models may not perform better than simple spatial metrics.

Vegetation contaminated by industrial fluoride emissions can cause disease in herbivorous mammals. Spatially explicit exposure models offer a quantitative approach for evaluating and managing the potentially toxic effects of chronic fluoride consumption on wildlife. We monitored eastern grey kangaroos (Macropus giganteus) inhabiting a high-fluoride environment in the buffer zone of an aluminium smelter in southeastern Australia between 2010 and 2013. We measured fluoride levels at 19 pasture sites and determined the foraging range of 37 individual kangaroos. A series of generalised linear models were developed to estimate bone fluoride accumulation as a function of pasture exposure. Model outputs were compared to identify the most appropriate predictive tool for kangaroo bone fluoride accumulation relative to exposure. Accounting for age there was a negative association between bone fluoride concentration and distance of the central emission point from both the mean centre of foraging range and the point of death. The mean foraging range centre was the best predictor, with point of death just as suitable (and simpler), whereas more complex parameters such as monthly and cumulative fluoride exposure were poor predictors of bone fluoride concentration. The more complex dietary fluoride exposure estimates did not improve predictive capability compared with the simple, spatial models. We conclude that in actively managed wildlife populations, simple, locally validated models can provide estimates of bone fluoride accumulation sufficient to support decision-making.