ABSTRACT
Globally, hundreds of mammal species face the threat of extinction in the coming decades, and in many cases, their ecology remains poorly understood. Fundamental ecological knowledge is crucial for effective conservation management of these species, but it is particularly lacking for small, cryptic mammals. The Julia Creek dunnart (Sminthopsis douglasi), a threatened, cryptic carnivorous marsupial that occurs in scattered populations in the central west of Queensland, Australia, was once so poorly studied that it was believed extinct. Sporadic research since its rediscovery in the early 1990s has revealed that S. douglasi is distributed across land at risk from many threats. Fundamental knowledge of S. douglasi population density is urgently required to inform conservation management at key sites, yet the species has historically proven hard to detect. Indeed, the status of the largest known population of S. douglasi, in Bladensburg National Park, is unknown. Here, we conducted a population study on S. douglasi at two sites within Bladensburg National Park via live mark-recapture surveys during 2022 and 2023. From likelihood-based spatially explicit capture-recapture (SECR) modelling we provide the first estimates of density and population size for S. douglasi. Live trapping resulted in captures of 49 individual S. douglasi (with 83 captures total, including recaptures). We estimated S. douglasi to occur at a density of 0.38 individuals ha-1 (0.25-0.58) at one site and 0.16 individuals ha-1 (0.09-0.27) at another site, with an estimated mean population size in suitable habitat at Bladensburg National Park of 1211 individuals (776-1646). Our S. douglasi density estimates were similar to that reported for other threatened small mammals in Australia. We also found evidence of extreme S. douglasi population fluctuations over time at Bladensburg National Park, which is of concern for its future conservation. Our study has provided the first estimate of density for S. douglasi, a threatened dasyurid species from the Mitchell Grass Downs of central western Queensland, Australia. Our research provides crucial population data to assist the management of this poorly studied species. We demonstrate a method that can be applied to species with low detection probability to ultimately help address the mammal extinction crisis faced by Australia and the rest of the world.
