Abstract
Noninvasive DNA sampling has become increasingly popular in wildlife research and conservation because it allows scientists to gather valuable genetic information without disturbing or harming the animals. However, the correct identification of the species or individuals in the sample is virtually impossible when using this kind of sampling. Consequently, it becomes essential to consider the errors hidding true identities in order to improve the quality of the data. Errors, if left unaddressed, can have a considerable impact on the accuracy of statistical inferences drawn from the data. This paper endeavours to review some research about misidentification problems and how Bayesian models and Markov Chain Monte Carlo (MCMC) methods can be applied. In addition, a hypothetical scenario is presented to illustrate how genetic material can serve as unique identifier of individuals, and to highlight the potential difficulties that may arise if a proposal distribution for the MCMC simulations is inappropriately chosen.
Publisher
Universidad Nacional de Colombia
Reference22 articles.
1. Barker, R. J., Schofield, M. R., Wright, J. A., Frantz, A. C. & Stevens, C. (2014), 'Closed-population capture-recapture modeling of samples drawn one at a time', Biometrics 70(4), 775-782.
2. Berg, N. & Lien, D. (2009), 'Sexual orientation and self-reported lying', Review of Economics of the Household 7(1), 83-104.
3. Biffi, D. & Williams, D. A. (2017), 'Use of non-invasive techniques to determine population size of the marine otter in two regions of Perú', Mammalian Biology 84, 12-19.
4. Bran, P. (2018), Properties of Gibbs samplers for inference in genetic markrecapture models, PhD thesis, University of Otago, Dunedin, New Zealand.
5. Chib, S. & Greenberg, E. (1995), 'Understanding the Metropolis-Hastings algorithm', The American Statistician 49(4), 327-335.