I have the following covariates in my study.
distance to Water
distance to Road
distance to Trail
distance to Human – habitation
Elevation
Habitat
presence of Human Litter
Canopy Cover
Annual Rainfall
Annual Temp
distance to Agriculture
Hi Hannah, 2:4 and 7 i would consider bias variables and follow renner et al (with warton i guess around 2013), i.e. calibrate normally and then predict to stack where bias variables are constants across study area to remove bias.
Cheers, R
You could add evidence of domestic dogs, hunters, or other predators. Are you using camera traps? Use their presence as a covariate.
Also time of year may affect detection rates.
Also, anything that affects abundance will affect detection rate P, so using a Royle-Nichols model (2003) which models a per individual detection rate R may reduce the number of covariates that affect detection.
Your list is good and you can also add presence of feral dogs, hunters or other anthropogenic activities detected as suggested William
William V Bleisch and Oluseun Adesola Akinsorotan i would add presence of domestic dogs however it was only a month long study with 18 camera traps and 6 sites, so i am unsure how reliable the results would be, in addition to not being able to do time of year. Predator wise there are no apex predators for the opossum in grenada. The only other mammals are mongoose, feral cats, stray dogs, mona monkey and armadillo. I lowered my final list down removing human litter and human density as the results were producing unusual estimates for occupancy and detectability. I stuck with water source, trail, road, canopy cover, buildings and elevation. They produced the most reliable results. Trail, road and buildings relating to human disturbance, and canopycover, elevation and water source for environmental factors. Land cover were not within the final AICc models with a delta of less than 2 following literature. AICc was chosen due to the small sample size.
William V Bleisch when you mean use their presence as a covariate, would that be for detection? IE. plotting a graph showing higher the detection, the more opossums were found?
Sorry for the confusion. My meaning was that you could use presence of dogs and/or hunters as a covariate for opossum presence, and perhaps also for opossum detection.
William V Bleisch no worries. I added the variable in, however it was not included in any of the top models for detection. Still had a relationship though
- Badges
- Science topic