The evaluation of indices of animal abundance using spatial simulation of animal trappingDave Ramsey A D , Murray Efford C , Steve Ball B and Graham Nugent B
A Landcare Research, Private Bag 11052, Palmerston North, New Zealand.
B Landcare Research, PO Box 69, Lincoln, New Zealand.
C Landcare Research, Private Bag 1930, Dunedin, New Zealand.
D Corresponding author. Email: firstname.lastname@example.org
Wildlife Research 32(3) 229-237 https://doi.org/10.1071/WR03119
Submitted: 22 December 2003 Accepted: 21 February 2005 Published: 22 June 2005
We apply a new algorithm for spatially simulating animal trapping that utilises a detection function and allows for competition between animals and traps. Estimates of the parameters of the detection function from field studies allowed us to simulate realistically the expected range of detection probabilities of brushtail possums caught in traps. Using this model we evaluated a common index of population density of brushtail possums based on the percentage of leg-hold traps catching possums. Using field estimates of the parameters of the detection function, we simulated the relationship between the trap-catch index and population density. The relationship was linear up to densities of 10 possums ha–1. We also investigated the accuracy (bias and precision) of the trap-catch index for possums to estimate relative changes in population density (relative abundance) under conditions of varying detection probability, and compared these results with those obtained using a removal estimate of the population in the vicinity of trap lines. The ratio of trap-catch indices was a more precise estimator of relative abundance than the ratio of removal estimates but was positively biased (i.e. overestimated relative abundance). In contrast, the ratio of removal estimates was relatively unbiased but imprecise. Despite the positive bias, the trap-catch index had a higher power to determine the correct ranking between population densities than the removal estimate. Although varying detection probability can bias estimates of relative abundance using indices, we show that the potential for bias to lead to an incorrect result is small for indices of brushtail possum density based on trapping.
We thank Greg Arnold for his assistance with the algorithm for the stochastic generation of g(0) and σ used in the half-normal detection function.
Anderson, D. R. (2001). The need to get the basics right in wildlife field studies. Wildlife Society Bulletin 29, 1294–1297.
Anderson, D. R. (2003). Response to Engeman: index values rarely constitute reliable information. Wildlife Society Bulletin 31, 288–291.
Ball, S. , Ramsey, D. , Nugent, G. , Warburton, B. , and Efford, M. (2005). A method for estimating wildlife detection probabilities in relation to home-range use: insights from a field study on the common brushtail possum (Trichosurus vulpecula). Wildlife Research 32, 217–227.
Davis, S. A. , Akison, L. K. , Farroway, L. N. , Singleton, G. R. , and Leslie, K. E. (2003). Abundance estimators and truth: accounting for individual heterogeneity in wild house mice. Journal of Wildlife Management 67, 634–645.
Don, B. A. C. , and Rennolls, K. (1983). A home range model incorporating biological attraction points. Journal of Animal Ecology 52, 69–81.
Efford, M. (2004). Density estimation in live trapping studies. Oikos 106, 598–610.
| CrossRef |
Efford, M. , Dawson, D. K. , and Robbins, C. S. (2004). DENSITY: software for analysing capture–recapture data from passive detector arrays. Animal Biodiversity and Conservation 27, 217–228.
Engeman, R. M. (2003). More on the need to get the basics right: population indices. Wildlife Society Bulletin 31, 286–287.
Engeman, R. M. , Allen, L. , and Zerbe, G. O. (1998). Variance estimates for the Allen activity index. Wildlife Research 25, 643–648.
| CrossRef |
McKelvey, K. S. , and Pearson, D. E. (2001). Population estimation with sparse data: the role of estimators versus indices revisited. Canadian Journal of Zoology 79, 1754–1765.
| CrossRef |
Otis, D. L. , Burnham, K. P. , White, G. C. , and Anderson, D. R. (1978). Statistical inference from capture data on closed animal populations. Wildlife Monographs 62, 1–135.
Ruscoe, W. A. , Goldsmith, R. , and Choquenot, D. (2001). A comparison of population estimates and abundance indices for house mice inhabiting beech forests in New Zealand. Wildlife Research 28, 173–178.
| CrossRef |
Skalski, J. R. , Robson, D. S. , and Simmons, M. A. (1983). Comparative census procedures using single mark–recapture methods. Ecology 64, 752–760.
Skalski, J. R. , Simmons, M. A. , and Robson, D. S. (1984). The use of removal sampling in comparative censuses. Ecology 65, 1006–1015.
Steiner, A. J. (1983). Comtrap: a trapping simulation interactive computer program. Journal of Wildlife Management 47, 561–567.
Warburton, B. , Barker, R. , and Coleman, M. (2004). Evaluation of two relative-abundance indices to monitor brushtail possums in New Zealand. Wildlife Research 31, 397–401.
| CrossRef |
Zarnoch, S. J. (1979). Simulation of effects of learned trap response on three estimators of population size. Journal of Wildlife Management 43, 474–483.
Zippin, C. (1956). An evaluation of the removal method of estimating animal populations. Biometrics 12, 163–169.
Zippin, C. (1958). The removal method of population estimation. Journal of Wildlife Management 22, 82–90.
Appendix 1. Stochastic algorithm to simulate the parameters of the half-normal detection function g(0) and σ based on empirical estimates from seven study populations (n = 192) given in Fig. 1
Initial analysis of the distribution of σ and log(σ) indicted non-normality. However, there was evidence that a good approximation could be gained by fitting a mixture of two normal distributions to log(σ) using maximum likelihood, giving the estimates shown in Table A1 (below)
which indicated ~78% of the mass around . The conditional distribution of g(0), given σ, was estimated by fitting a generalised linear model and assuming g(0) was inversely related to σ and hence, follows a gamma distribution giving a relationship for 1/g(0) = 0.26σ – 1.08. Unfortunately, taking this relationship and the estimate of the dispersion parameter (0.35) did not seem to adequately reproduce the observed inverse relationship. However, changing the rate parameter to 25 gave satisfactory results. Hence, g(0) had a gamma distribution with shape = 25/(0.26σ – 1.08) and scale = 1/25. Simulated results are presented in Fig. A1.
Fig. A1. Plot of simulated values of g(0) and σ (x-axis) versus observed (field) estimates of g(0) and σ from data presented in Fig. 1 (y-axis). Solid line represents a 1 : 1 relationship.