Birdseye in the sky: the relationship between fractional cover, rainfall and woodland birds in a partially modified tropical savanna

Remotely sensed data are commonly used for mapping landscape changes, but are being used increasingly as a surrogate for biodiversity and habitat condition. In this study, we examined bird data collected from long-term monitoring in 60 sites sampled seven times between 2004 and 2016. The sites represent three levels of habitat modification, namely, intact, thinned and cleared. We investigated the relationship between fractional cover measures (green and non-green) and preceding 12-month rainfall, using hierarchical generalised linear mixed models, to see whether these metrics had a relationship to woodland bird species abundance. In total, 121 species were recorded. We were able to model the relationship between the abundance of 57 species and our environmental variables. There were a mixture of responses recorded, including species associated with higher green and non-green cover, but not rainfall changes (e.g. Brown Treecreeper, Climacteris picumnus; Striped Honeyeater, Plectorhyncha lanceolata), species associated with lower fractional cover and higher rainfall (e.g. Galah, Eolophus roseicapillus; Zebra Finch, Taeniopygia guttata) and species with more variable relationships (e.g. Crested Bellbird, Oreoica gutturalis; Weebill, Smicrornis brevirostris; Grey-crowned Babbler, Pomatostomus temporalis; and Jacky Winter, Microeca fascinans). We found that there was a strong relationship between many species of woodland birds considered to reflect an intact and good condition community and different combinations of three remotely sensed variables. Remote sensed data have a role to play, along with field surveys, in assessing bird community condition, for programs such as nature repair markets.