Quantitative Biology > Quantitative Methods
[Submitted on 3 Apr 2023]
Title:Population size estimation with capture-recapture in presence of individual misidentification and low recapture
View PDFAbstract:While non-invasive sampling is more and more commonly used in capture-recapture (CR) experiments, it carries a higher risk of misidentifications than direct observations. As a consequence, one must screen the data to retain only the reliable data before applying a classical CR model. This procedure is unacceptable when too few data would remain. Models able to deal with misidentifications have been proposed but are barely used. Three objectives are pursued in this paper. First, we present the Latent Multinomial Model of Link et al. (2010) where estimates of the model are obtained from a Monte Carlo Markov Chain (MCMC). Second we show the impact of the use of an informative prior over the estimations when the capture rate is low. Finally we extend the model to the multistate paradigm as an example of its flexibility.
We showed that, without prior information, with capture rate at 0.2 or lower, parameters of the model are difficult to estimate i.e. either the MCMC does not converge or the estimates are biased. In that case, we show that adding an informative prior on the identification probability solves the identifiability problem of the model and allow for convergence. It also allows for good quality estimates of population size, although when the capture rate is 0.1 it underestimates it of about 10%. A similar approach on the multistate extension show good quality estimates of the population size and transition probabilities with a capture rate of 0.3 or more.
References & Citations
Bibliographic and Citation Tools
Bibliographic Explorer (What is the Explorer?)
Litmaps (What is Litmaps?)
scite Smart Citations (What are Smart Citations?)
Code, Data and Media Associated with this Article
CatalyzeX Code Finder for Papers (What is CatalyzeX?)
DagsHub (What is DagsHub?)
Gotit.pub (What is GotitPub?)
Papers with Code (What is Papers with Code?)
ScienceCast (What is ScienceCast?)
Demos
Recommenders and Search Tools
Influence Flower (What are Influence Flowers?)
Connected Papers (What is Connected Papers?)
CORE Recommender (What is CORE?)
arXivLabs: experimental projects with community collaborators
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.