Hidden Markov Model for Dependent Mark Loss and Survival Estimation

Abstract

Mark-recapture estimators assume no loss of marks to provide unbiased estimates of population parameters. We describe a hidden Markov model (HMM) framework that integrates a mark loss model with a Cormack-Jolly-Seber model for survival estimation. Mark loss can be estimated with single-marked animals as long as a sub-sample of animals has a permanent mark. Double-marking provides an estimate of mark loss assuming independence but dependence can be modeled with a permanently marked sub-sample. We use a log-linear approach to include covariates for mark loss and dependence which is more flexible than existing published methods for integrated models. The HMM approach is demonstrated with a dataset of black bears (Ursus americanus) with two ear tags and a subset of which were permanently marked with tattoos. The data were analyzed with and without the tattoo. Dropping the tattoos resulted in estimates of survival that were reduced by 0.005-0.035 due to tag loss dependence that could not be modeled. We also analyzed the data with and without the tattoo using a single tag. By not using.Supplementary materials accompanying this paper appear on-line.

Phenotypic selection and covariation in the life‐history traits of elephant seals: heavier offspring gain a double selective advantage

“…We used multievent capture-recapture models to estimate three demographic parameters: apparent survival probability (φ) (hereafter survival), recruitment probability (ψ), and recapture probability (ρ). In addition, we incorporated tag loss (τ 21 and τ 10 ) within the multievent framework, avoiding pitfalls associated with post hoc correction of survival estimates (Laake et al 2014). This parameter was needed given that seals may lose both the tags they were marked with at weaning, at which time they become unidentifiable and appear 'dead' within the capturerecapture context (Oosthuizen et al 2010).…”

Flipper tag loss in Steller sea lions

“…In conclusion, results of our study and those of Johnson et al (2016) suggest that, high flipper tag loss coupled with tag dependence will prevent monitoring of agespecific vital rates of sea lions unless a sufficient subsample of animals are also marked with an independent third mark (Laake et al 2014). Depending on what is feasible for a species and study, the third mark may be a brand, tattoo, natural markings, or genetic ID (Laake et al 2014).…”

“…Depending on what is feasible for a species and study, the third mark may be a brand, tattoo, natural markings, or genetic ID (Laake et al 2014). The use of passive integrated transponder (PIT) tags as third marks often requires animals in hand due to the short range of these instruments, and the longevity and efficacy of PIT tags in marine mammals has not yet been well studied (reviewed by Walker et al 2012).…”

“…When we fitted Model 5 (Table 1) with the tag dependence term removed, double-tag loss estimates for pups tagged with classic medium and classic sheep tags were underestimated by 28.4% and 8.6%, respectively, and for nonpups tagged with those tags, by 34.1% and 68.5%. Other studies have demonstrated notable bias in / estimates when failing to account for dependence in tag loss (Diefenbach and Alt 1998, Schwarz et al 2012, Laake et al 2014. To examine its effects on / estimates for SSL, we performed a post hoc analysis of "tag-only" capture histories.…”

A model for first‐estimates of species‐specific, age‐specific mortality from centralized band‐recovery databases

“…3. Mark loss-There are well-developed models for mark loss within Jolly-Seber (JS) models, but these rely on a portion of the marked population being double-marked (Table 1; Laake et al 2014). For many of the species we wish to model, there is no doublemarked segment of the population.…”