A hidden Markov movement model for rapidly identifying behavioral states from animal tracks

Abstract

Electronic telemetry is frequently used to document animal movement through time. Methods that can identify underlying behaviors driving specific movement patterns can help us understand how and why animals use available space, thereby aiding conservation and management efforts. For aquatic animal tracking data with significant measurement error, a Bayesian state‐space model called the first‐Difference Correlated Random Walk with Switching (DCRWS) has often been used for this purpose. However, for aquatic animals, highly accurate tracking data are now becoming more common. We developed a new hidden Markov model (HMM) for identifying behavioral states from animal tracks with negligible error, called the hidden Markov movement model (HMMM). We implemented as the basis for the HMMM the process equation of the DCRWS, but we used the method of maximum likelihood and the R package for rapid model fitting. The HMMM was compared to a modified version of the DCRWS for highly accurate tracks, the DCRWSNOME, and to a common HMM for animal tracks fitted with the R package . We show that the HMMM is both accurate and suitable for multiple species by fitting it to real tracks from a grey seal, lake trout, and blue shark, as well as to simulated data. The HMMM is a fast and reliable tool for making meaningful inference from animal movement data that is ideally suited for ecologists who want to use the popular DCRWS implementation and have highly accurate tracking data. It additionally provides a groundwork for development of more complex modeling of animal movement with . To facilitate its uptake, we make it available through the R package .

The Conditionally Autoregressive Hidden Markov Model (CarHMM): Inferring Behavioural States from Animal Tracking Data Exhibiting Conditional Autocorrelation

“…These states are typically assumed to follow an area-restricted search pattern, whereby foraging patches are characterized by shorter step lengths occurring in diffuse directions, and are interspersed with periods of directed travel consisting of longer step lengths directed straight ahead (see e.g. Whoriskey et al (2017)). While these states can be directly inferred from the state-dependent distributions of the HMM, the interpretation of these state estimates resulting from the DCRWS is less straightforward.…”

“…Usually (again see e.g. Whoriskey et al (2017)), high γ values are interpreted as highly persistent movement (indicative of transiting) and low γ values constitute highly random movement (representing foraging). As a result, transiting and foraging are not necessarily delineated by longer and shorter step lengths.…”

“…The work of Morales et al (2004) along with that of Jonsen et al (2005), popularized the state-switching model framework into the de facto way of analysing animal movement data in discrete time (McClintock et al, 2012;Whoriskey et al, 2017;Patterson et al, 2017). While not all animal movement models which incorporate state-switching into the movement process have distinct behavioural states, the ones that do generally fall under the hidden Markov model (HMM) framework.…”

momentuHMM: R package for generalized hidden Markov models of animal movement

“…While statisticians have been applying HMMs to telemetry data for decades, R (R Core Team, ) packages such as bsam (Jonsen et al., ), moveHMM (Michelot, Langrock, & Patterson, ), and swim (Whoriskey et al., ) have recently helped make these models more accessible to the practitioners that are actually conducting telemetry studies. While these contributions represent important steps forward, the models that can currently be implemented remain limited in many key respects.…”

Animal movement models for migratory individuals and groups

“…These approaches include the use of spatial and temporal covariates and clustering methods to understand the portions of animal trajectories that indicate distinctly different patterns (e.g. Whoriskey et al., ). For example, potential function specifications in stochastic differential equations (SDEs; Brillinger, ) have facilitated the explicit inclusion of covariates in continuous‐time models.…”