Assessment of survey protocol for estimates of abundance for elusive nocturnal primates
2020
Abstract Context Estimates of a species abundance and habitat preferences provide vital information on their status and the appropriate conservation management. For nocturnal arboreal primates, obtaining reliable estimates of these parameters is particularly challenging because of their cryptic behaviour, often resulting in a small number of detections. Although techniques are available for assessing the abundance of species with a low probability of detection, most require strict assumptions that are difficult to meet. Aims Here, we aimed to explore the possibility of improving nocturnal-primate abundance estimates when detection probability is low and to determine the minimal effort required to calculate reasonable estimates of their overall abundance and the effect of habitat type on abundance estimates. Methods We used count data obtained from spotlighting along line transects for estimating density of Bengal slow loris (Nycticebus bengalensis) in north-eastern Thailand with N-mixture hierarchical modelling, to run simulations of varying survey parameters and asses the effort needed to produce robust estimates based on the relative bias from each simulation. Key results N-mixture analysis showed that the data obtained from our study were still biased (9%), with a lambda of 1.79 lorises, detection probability of 0.11, 50 survey sites and a maximum of 12 sample occasions. The simulation results found that increasing the number of sample occasions to 14 per transect would produce an acceptable bias (<5%). Conclusions We recommend that future studies on nocturnal arboreal species should use preliminary surveys to gauge the specific lambda and probability of detection so as to establish the effort needed to produce reasonable estimates of abundance. Implications Our study showed that count data obtained from spotlighting can be used to produce robust abundance estimates of nocturnal arboreal species. Unlike simple encounter rate, this method incorporates detection probability and habitat preferences, yet does not require additional trained field technicians.
Keywords:
- Correction
- Source
- Cite
- Save
- Machine Reading By IdeaReader
48
References
0
Citations
NaN
KQI