arXiv subjects

Mark V Bravington

Publications and source records attributed to Mark V Bravington.

3 recordsLinked to original sources

arXiv2020

Variance propagation for density surface models

Spatially-explicit estimates of population density, together with appropriate estimates of uncertainty, are required in many management contexts. Density Surface Models (DSMs) are a two-stage approach for estimating spatially-varying density from distance-sampling data. First, detection probabilities -- perhaps depending on covariates -- are estimated based on details of individual encounters; next, local densities are estimated using a GAM, by fitting local encounter rates to location and/or spatially-varying covariates while allowing for the estimated detectabilities. One criticism of DSMs has been that uncertainty from the two stages is not usually propagated correctly into the final variance estimates. We show how to reformulate a DSM so that the uncertainty in detection probability from the distance sampling stage (regardless of its complexity) is captured as an extra random effect in the GAM stage. In effect, we refit an approximation to the detection function model at the same time as fitting the spatial model. This allows straightforward computation of the overall variance via exactly the same software already needed to fit the GAM. A further extension allows for spatial variation in group size, which can be an important covariate for detectability as well as directly affecting abundance. We illustrate these models using point transect survey data of Island Scrub-Jays on Santa Cruz Island, CA and harbour porpoise from the SCANS-II line transect survey of European waters.

stat.ME↗

arXiv2015

Efficient MCMC implementation of multi-state mark-recapture models

Inherent differences in behaviour of individual animal movement can introduce bias into estimates of population parameters derived from mark-recapture data. Additionally, quantifying individual heterogeneity is of considerable interest in it's own right as numerous studies have shown how heterogeneity can drive population dynamics. In this paper we incorporate multiple measures of individual heterogeneity into a multi-state mark-recapture model, using a Beta-Binomial Gibbs sampler using MCMC estimation. We also present a novel Independent Metropolis-Hastings sampler which allows for efficient updating of the hyper-parameters which cannot be updated using Gibbs sampling. We tested the model using simulation studies and applied the model to mark-resight data of North Atlantic humpback whales observed in the Stellwagen Bank National Marine Sanctuary where heterogeneity is present in both sighting probability and site preference. Simulation studies show asymptotic convergence of the posterior distribution for each of the hyper-parameters to true parameter values. In application to humpback whales individual heterogeneity is evident in sighting probability and propensity to use the marine sanctuary.

stat.AP↗

arXiv2015

Modelling latent individual heterogeneity in mark-recapture data with Dirichlet process priors

The natural subgroups often seen in mark-recapture studies and the complexity of real mark-recapture data means that parametric and discrete style models can be insufficient. Non-parametric models avoid these often restrictive assumptions. We consider the non-parametric Dirichlet process for modelling latent individual heterogeneity in probability of observation and the probability of remaining in or out of a marine sanctuary. Simulation studies demonstrated accurate estimation of multiple groups of latent individual heterogeneity. Simulations were also used to identify the limits of the Dirichlet process. The ability of the Dirichlet process to pick up unimodal heterogeneity was explored in order to avoid potential spurious multimodality. In application to a subset of the data from the North Atlantic humpback whales we were able to estimate annual population-level variation in usage of the marine sanctuary and three measures of individual-level variation. With the Dirichlet process prior we were able to detect multimodality in each parameter.

stat.AP↗