A Bayesian Capture–Recapture Population Model With Simultaneous Estimation of Heterogeneity

Abstract

We develop a Bayesian capture-recapture model that provides estimates of abundance as well as time-varying and heterogeneous survival and capture probability distributions. The model uses a state-space approach by incorporating an underlying population model and an observation model, and here is applied to photo-identification data to estimate trends in the abundance and survival of a population of bottlenose dolphins (Tursiops truncatus) in northeast Scotland. Novel features of the model include simultaneous estimation of time-varying survival and capture probability distributions, estimation of heterogeneity effects for survival and capture, use of separate data to inflate the number of identified animals to the total abundance, and integration of separate observations of the same animals from right and left side photographs. A Bayesian approach using Markov chain Monte Carlo methods allows for uncertainty in measurement and parameters, and simulations confirm the model's validity.

Report of the Working Group on Population Parameters and Demography of Tursiops truncatus in the Southwest Atlantic Ocean

“…Lodi et al (2014) discuss that such a low survival probability could represent both a gradual abandonment of the study area or a high mortality outside this area. In contrast, the apparent adult survival estimates for Laguna and Patos Lagoon Estuary bottlenose dolphin populations are close to the values reported in the literature where overall adult survival probabilities ranged from 0.92 to 0.97 (Wells and Scott, 1990;Corkrey et al, 2008;Currey et al, 2008;Silva et al, 2009). The disparity found between adult survival estimates for Laguna (0.917) and Patos Lagoon Estuary (females = 0.97; males = 0.88) populations is likely reflecting differences in study length and other methodological differences.…”

Integrated modeling of bilateral photo‐identification data in mark–recapture analyses

“…However, this is not as efficient as conducting an integrated analysis with both types of encounters, given that sample sizes may be considerably reduced, resulting in lower precision (e.g., Madon et al 2011). Corkrey et al (2008) proposed a novel approach for modeling left-and right-sided encounters of dolphins in the context of state-space models, and this approach has recently been extended by Hiby et al (2013). However, Corkrey et al (2008) make a number of assumptions that will often be violated, including (1) the probability of being encountered on both sides during the same sampling occasion is the product of two independent events (encountered on left side, encountered on right side) and (2) records of individuals only seen on the left side are independent of those only seen on the right side.…”

“…Corkrey et al (2008) proposed a novel approach for modeling left-and right-sided encounters of dolphins in the context of state-space models, and this approach has recently been extended by Hiby et al (2013). However, Corkrey et al (2008) make a number of assumptions that will often be violated, including (1) the probability of being encountered on both sides during the same sampling occasion is the product of two independent events (encountered on left side, encountered on right side) and (2) records of individuals only seen on the left side are independent of those only seen on the right side. While the former assumption may be reasonable under certain sampling scenarios, the latter seems unlikely to hold in practical applications.…”

“…Second, we adapt the approach of Link et al (2010) to the analysis of bilateral photographic encounters, including methods needed to set up an efficient Markov chain Monte Carlo (MCMC) sampler for posterior simulation. We then conduct a small simulation study to examine possible biases and anticipated efficiency gains when our integrated approach is compared with the usual approach of analyzing encounters from one side only and with the approach of Corkrey et al (2008). Finally, we apply our integrated framework to estimate population size from bilateral photo-identification records of bobcats in southern California.…”

Bayesian Estimation of the Size of a Street-Dwelling Homeless Population

“…Castledine, 1981;Smith, 1991;George & Robert, 1992;Fienberg, Johnson, & Junker, 1999;Basu & Ebrahimi, 2001;King & Brooks, 2001;Tardella, 2002;Manrique-Vallier & Fienberg, 2008;Corkrey et al, 2008). This article is organized as follows: First, the authors describe the methodology and modelling assumptions.…”