IUCN / SSC Cat Specialist Group - Digital Cat Library |
|
||||||||||
|
||||||||||
|
|
| Newey, S.; Potts, J.; Irvine, R.J. | |
| Simulation study to inform the design of wildcat camera trap monitoring protocols | |
| 2015 Full Book | |
The Scottish Wildcat Conservation Action Plan aims to halt the decline in Scottish wildcat numbers and to implement actions focussed on improving the conservation status of the Wildcat in Scotland. Key to assessing the outcome of conservation actions is the need monitor wildcat populations in prioritised areas. Camera trapping to collect images and identify individual wildcats for capture-recapture analysis to obtain density estimates has been proposed as the main method for monitor wildcat populations and to assess conservation measures. Here we assess, through a simulation based study, the probable effectiveness of different camera trap survey designs for monitoring wildcat population densities in designated priority areas. Main findings Based on current knowledge of wildcat ecology we assessed how: 1) survey design; (a) number of camera traps (81, 100, 125), (b) trap spacing (750, 1,000, 1,250), and (c) survey duration (60 and 80 days), and 2) wildcat ecology; (a) population density (D=1km-2), (b) capture probability (g0=0.01, 0.02), and (c) home range (sigma = 300, 500, 700 m) affect the statistical power to detect a 25% change in wildcat density over 5 years with 80% power. Based on the knowledge of wildcat ecology available at the time work was carried out simulations showed that; a) the number of individuals caught and recaptured tended to increase with size of trap grid, camera spacing and sigma (representing home range size), b) the number of individuals caught and recaptured also increased with survey duration and capture probability, and c) for any given grid size, sigma, capture probability and survey duration, the number of spatial recaptures (individuals caught at different locations) decreased with increasing trap spacing. Simulations here were limited to surveys where the assumed wildcat cat density was 1km-2 which is much higher than most reported densities for wildcat populations in Scotland, but simulations using lower densities failed to run due to too few encounters. The performance of density estimates assessed by the relative bias, relative standard error, and coverage were generally poor for surveys based on small trap grids when sigma was small and when capture probability was low with correspondingly low poor statistical power to detect the required level of change. The performance of density estimates increased with number of traps, capture probability, survey duration and increasing sigma with correspondingly improved statistical power. The power to detect decreases in density was greater than to detect an increase, and in both cases statistical power was highly dependent on the combination of survey and ecological parameters. The power of simulated survey designs increased markedly with increasing sigma. Where sigma was 300 m none of the survey designs had sufficient power (>80%) to detect the required 25% change in population density. However, when sigma was 700 m, surveys with larger camera trap grids and/or longer survey duration and when detection probability was 0.02 had sufficient power to detect a decrease and, to a lesser extent, an increase in population density. For any given combination of sigma, detection probability larger camera trap grids and more camera traps increased power, as did longer survey duration. Similarly, for any given combination of camera trap grid, sigma, and survey duration, higher power was associated with the higher value of detection probability. Sufficient statistical power to detect the required level of increase in wildcat density was limited to the trap grids with trap grid with 132 camera traps, where trap spacing was 1,250 m, and where sigma = 700 m. However, the utility of camera trapping to monitor wildcat density is dependent on the assumptions surveyors are prepared to make about wildcat ecology particularly home-range, but also probability of capture and probable density. New information emerging on wildcat ecology in Scotland may offer new data to inform camera trap survey studies. |
|
 |
PDF files are only accessible to Friends of the Cat Group. Joining Friends of the Cat Group gives you unlimited access and downloads in the Cat SG Library for one year, and allows you to receive our newsletter Cat News (2 regular issues per year plus special issues). More information how to join here |
(c) IUCN/SSC Cat Specialist Group ( IUCN - The World Conservation Union) |