Remotely-sensed foliage cover and ground-measured stand attributes are complimentary when estimating tree hollow abundances across relictual woodlands in agricultural landscapes

Hollows, also known as tree cavities, are critical to the survival of many animal species but are too poorly mapped across landscapes to allow for their adequate consideration in regional planning. Managing cost is important, so we tested whether freely available satellite-derived foliage projective cover and field-measured stand attributes could be used separately or combined to predict tree hollow abundance in relictual Australian temperate woodlands. Satellite-derived foliage projective cover revealed variation in woody vegetation densities both within mapped woodland remnants and cleared areas of the agricultural matrix. Plot-based field assessment of the actual number of hollows in each one-hectare site (n = 110 sites) revealed a relationship with foliage cover. Improvement of the model was achieved if site-based estimates of the proportion of the canopy due to Eucalyptus species and the number of mature trees per hectare were included. Remotely sensed foliage cover can improve on traditional vegetation mapping for predicting hollow-bearing tree and hollow abundances at landscape scales when managing hollow-dependent fauna habitat across relictual woodlands in temperate Australian agricultural landscapes. At finer scales, the addition of other predictors is necessary to raise the accuracy of the predicted hollow densities.