Assessing the spatial and temporal organization of Red Kangaroo, Western Grey Kangaroo and Eastern Grey Kangaroo populations in eastern Australia using multivariate autoregressive state‐space models

Summary To identify patterns in the temporal dynamics and spatial organization of kangaroo subpopulations that are commercially harvested, we fitted multivariate autoregressive state-space (MARSS) models to time series of kangaroo abundance from the main harvest zones of eastern Australia (1990–2019 for New South Wales and 1984–2019 for Queensland). We found that the responses of most populations to process variation, that is variation due to environmental or demographic stochasticity were correlated and that populations responded synchronously to environmental changes. Furthermore, we examined the influence of the covariates lagged rainfall, Southern Oscillation Index (SOI) and harvest rate on kangaroo abundance. We found that lagged rainfall had a positive influence, SOI and harvest rate had negative influences but the influence of covariates was not consistent across species or between subpopulations. In terms of population spatial structure, the analysis identified two subpopulations of Red Kangaroo (Osphranter rufus) in New South Wales and four subpopulations of grey kangaroos (a combination of Eastern Grey Kangaroo (Macropus giganteus) and Western Grey Kangaroo (Macropus fuliginosus)), which appeared to be associated with a rainfall gradient from east-west. In Queensland, separate subpopulations were identified in each of the three main harvest management zones for both Red Kangaroo and Eastern Grey Kangaroo. The implications of these results for the management of commercially harvested kangaroos are discussed.