Search

The Carpentarian rock-rat (Zyzomys palatalis) is a critically endangered endemic rodent known from only four sandstone gorges in the southeast Gulf of Carpentaria, Northern Territory, Australia. These gorges harbour thickets of monsoon rainforest and broadleaf woodland, surrounded by a Eucalypt savanna matrix. The long-term persistence of Z. palatalis is threatened by altered fire regimes, grazing by feral animals and stock, weed intrusion, and the stochastic hazards associated with small, fragmented populations. To assess the relative importance of these threats and develop practical management options, a population and habitat simulation model was developed, based on the best existing data. Population viability was predicted to be highly sensitive to the frequency of hot, late dry-season fires. Progressive habitat degradation (due predominantly to intense late dry-season fires) is likely to substantially reduce population size and lead to the probable extinction of the species within the next 100 years. The most effective management strategy to counteract this threat would be regular, controlled, fuel reduction burns in the vegetation around the gorge entrances during the early dry season. Establishing a new population (through translocation of captive-bred individuals) would not appreciably reduce extinction risk, but could provide valuable additional data on the impact of threats, if conducted as an adaptive management experiment. [ABSTRACT FROM AUTHOR]

Aerial survey methods for census of wildlife populations are widely used to manage species. There has been considerable research to develop methods to cope with imperfect detection during aerial surveys, however misidentification remains an unresolved but significant issue. Using data from a comprehensive dugong aerial survey of the coastline of the Northern Territory, Australia, we assessed the influence of observer, sighting and environmental variables on probabilities of both observers in double-count, dual-observer teams detecting dugongs and dolphins ('duplicate sighting probability'), the confidence with which observers could distinguish between dugongs and dolphins, and the reliability of post-detection observations by comparing data between observers. Dolphin duplicate sighting probability varied dependent on group size, Beaufort sea state and the lateral position inside the strip-width transect in relation to the aircraft. Dugong duplicate sighting probability varied across dual-observer teams. Counts of individuals in dugong and dolphin groups were unreliable past ≈5 individuals. Observer confidence in taxon identification varied by individual observer and throughout the duration of the survey (dolphin sightings) and potentially by the lateral position inside the transect strip-width in relation to the aircraft (dugong sightings). The probability of taxon identification disagreement between observers for duplicate dugong and dolphin sightings varied with group size, and markedly if one observer was uncertain (i.e. not 100% confident) of their assigned taxon. We believe this is predominately due to the presence of Australian snubfin dolphins (Orcaella heinsohni) in the survey area, which look like dugongs when seen from a distance, momentarily. We further show that application of different data filtering scenarios for uncertain taxonomic identifications and taxon identification disagreements between dual observers produces highly variable estimates of observer-specific dugong detection probability (ranging from 0.26 to 0.80 in a single observer) and dugong abundance (x̅ ± S.D., 40% ± 25%: difference between highest and lowest estimate, relative to the lowest estimate, across different data filtering scenarios). These results are particularly relevant not only to researchers conducting dugong aerial surveys, but also for other marine mammal aerial surveys in passing mode that target sympatric, morphologically similar species. We recommend development of operating procedures and analytical methods to quantify and adjust for misidentification and uncertain taxon identifications. • Accurate species identification (ID) and counts are essential for aerial surveys. • Species detection and accurate ID varies between observers and with other variables. • Detection & Abundance estimates vary depending on how ambiguous data are filtered. • Surveys of sympatric, morphologically similar species may be similarly affected. [ABSTRACT FROM AUTHOR]