Multivariate State Hidden Markov Models for Mark-Recapture Data

Abstract

State-based Cormack-Jolly-Seber (CJS) models have become an often used method for assessing states or conditions of free ranging animals through time. Although originally envisioned to account for di↵erences in survival and observation processes when animals are moving though various geographical strata, it has evolved to model vital rates in di↵erent life-history or diseased states. We further extend this useful class of models to the case of multivariate state data. Researchers can record values of several di↵erent states of interest, e.g., geographic location and reproductive state. Traditionally, these would be aggregated into one state with a single probability of state uncertainty. However, by modeling states as a multivariate vector, one can account for partial knowledge of the vector as well as dependence between the state variables in a parsimonious way. A hidden Markov model formulation allows straightforward maximum likelihood inference. The proposed HMM models are demonstrated with a case study using data from a California sea lion vital rates study.

Sex‐biased survival contributes to population decline in a long‐lived seabird, the Magellanic Penguin

“…, Johnson et al. ). We sexed individuals using one of four methods, all of which we assumed were certain: (1) genetic testing, (2) cloaca size around egg laying (Boersma and Davies ), (3) breeding behavior (copulation, arrival, nest defense, etc.…”

“…We used Hidden Markov mark-recapture models developed with the package marked (Laake 2013, Johnson et al 2016, 33 yr of banding data for known-aged penguins, and perturbation of population matrix models to examine sex-specific demography in this species. Because mortality is highest, and our knowledge of this species is lowest, during the nonbreeding season (Stokes et al 2014), our analysis focused on this period.…”

“…We developed a Hidden Markov mark-recapture model using the mvmscjs function in the R package marked (R Version 3.5.1, marked Version 1.2.2; Johnson et al 2016) for the Punta Tombo population of Magellanic Penguins based on the 44,374 recapture histories. The statistical foundation of marked is detailed in Laake (2013), Laake et al (2013), and Johnson et al (2016). Our model included two states: sex and breeding state.…”

“…, Johnson et al. ), 33 yr of banding data for known‐aged penguins, and perturbation of population matrix models to examine sex‐specific demography in this species. Because mortality is highest, and our knowledge of this species is lowest, during the non‐breeding season (Stokes et al.…”

“…Developing reliable sexing techniques based on genetics (Sabo et al 1994, Faux et al 2014, morphometrics (Boersma 1977, Hanners and Patton 1985, Boersma and Davies 1987, Cappello and Boersma 2018, and/or behavioral traits has long been a focus of study in the seabird community, but many individuals remain unsexed throughout even long-term seabird studies (Ancona et al 2017). Fortunately, there are now statistical approaches to estimating sex-specific demography when the sex of some individuals is unknown (Nichols et al 2004, Choquet et al 2009, Genovart et al 2012, Johnson et al 2016).…”