Accounting for missing data when assessing availability in animal population surveys: an application to ice-associated seals in the Bering Sea

Summary Population ecologists often use telemetry to estimate the probability that an animal is available for sampling (availability probability). Such estimates are paramount for generating reliable estimates of animal abundance and are sometimes of biological interest in their own right (e.g. for characterizing pinniped haul-out behaviour or landbird singing frequency). We consider the estimation of availability probability from telemetry data when there are missing records. When records are missing-completely-at-random (i.e. not influenced by whether an animal is available for sampling or other covariates), approaches that censor missing records result in unbiased estimates of availability probability. However, censoring such records can result in bias when data are not missing-completely-at-random. We present a novel Bayesian temporal availability model that can be used to explicitly account for the process by which missing data arise. Our approach couples an underlying model for the partially observed availability process together with an observation model that allows state-dependent probabilities of missingness. This approach provides relatively unbiased estimates of availability probability even in the face of violations of the missing-at-random assumption. A new R package, TempOcc, is introduced to automate estimation. We demonstrate the utility of our approach by analysing both simulated data and hourly satellite telemetry records for 157 ice-associated seals in the Bering Sea, for which 27% of records were missing. In this case, we were interested in generating unbiased estimates of haul-out probabilities. Such probabilities are necessary when estimating abundance from aerial survey counts and assessing possible changes in seal phenology. Our analysis indicates that the missing-at-random assumption is indeed violated in telemetry studies of phocid seals in the Bering Sea. However, estimates of availability probability do not depend critically on this assumption. We recommend that ecologists routinely consider the implications of violating the missing-at-random assumption when estimating availability probabilities. Where possible, the modelling framework articulated in this study can be used to diagnose and correct for violations of the missing-at-random assumption.

[1]  Erin M. Bayne,et al.  Dealing with detection error in site occupancy surveys: what can we do with a single survey? , 2012 .

[2]  C. Robert,et al.  Deviance information criteria for missing data models , 2006 .

[3]  Roderick J. A. Little,et al.  Statistical Analysis with Missing Data , 1988 .

[4]  Shinichi Nakagawa,et al.  Missing inaction: the dangers of ignoring missing data. , 2008, Trends in ecology & evolution.

[5]  Mevin B. Hooten,et al.  Spatial occupancy models for large data sets , 2013 .

[6]  J. Andrew Royle,et al.  Hierarchical Modeling and Inference in Ecology: The Analysis of Data from Populations, Metapopulations and Communities , 2008 .

[7]  Andrew Thomas,et al.  WinBUGS - A Bayesian modelling framework: Concepts, structure, and extensibility , 2000, Stat. Comput..

[8]  J. Besag,et al.  On conditional and intrinsic autoregressions , 1995 .

[9]  J. Andrew Royle,et al.  A Bayesian state-space formulation of dynamic occupancy models. , 2007, Ecology.

[10]  Kenneth H. Pollock,et al.  TIME-OF-DETECTION METHOD FOR ESTIMATING ABUNDANCE FROM POINT-COUNT SURVEYS , 2007 .

[11]  M. Conroy,et al.  Analysis and Management of Animal Populations , 2002 .

[12]  M. D. Samuel,et al.  Sightability adjustment methods for aerial surveys of wildlife populations , 1989 .

[13]  K. Pollock,et al.  Estimating Animal Abundance in Heterogeneous Environments: An Application to Aerial Surveys for Dugongs , 2006 .

[14]  Roger Pradel,et al.  Multievent: An Extension of Multistate Capture–Recapture Models to Uncertain States , 2005, Biometrics.

[15]  A. Brix.  Bayesian Data Analysis, 2nd edn , 2005 .

[16]  D. Heitjan,et al.  Distinguishing “Missing at Random” and “Missing Completely at Random” , 1996 .

[17]  Ian A. McLaren,et al.  METHODS OF DETERMINING THE NUMBERS AND AVAILABILITY OF RINGED SEALS IN THE , 1961 .

[18]  Jay M. Ver Hoef,et al.  Fast computing of some generalized linear mixed pseudo-models with temporal autocorrelation , 2010, Comput. Stat..

[19]  J. Besag,et al.  Bayesian image restoration, with two applications in spatial statistics , 1991 .

[20]  J. Gaillard,et al.  Statistical evaluation of parameters estimating autocorrelation and individual heterogeneity in longitudinal studies , 2012 .

[21]  William A. Link,et al.  Bayesian Inference: With Ecological Applications , 2009 .

[22]  Alan E. Gelfand,et al.  Model choice: A minimum posterior predictive loss approach , 1998, AISTATS.

[23]  H. Künsch.  Gaussian Markov random fields , 1979 .

[24]  J. Andrew Royle,et al.  ESTIMATING SITE OCCUPANCY RATES WHEN DETECTION PROBABILITIES ARE LESS THAN ONE , 2002, Ecology.

[25]  Bradley P. Carlin,et al.  Bayesian measures of model complexity and fit , 2002 .

[26]  S. Chib,et al.  Bayesian analysis of binary and polychotomous response data , 1993 .

[27]  D. Rubin.  INFERENCE AND MISSING DATA , 1975 .

[28]  Duane R. Diefenbach,et al.  INCORPORATING AVAILABILITY FOR DETECTION IN ESTIMATES OF BIRD ABUNDANCE , 2007 .

[29]  H. Marsh,et al.  Correcting for visibility bias in strip transect aerial surveys of aquatic fauna , 1989 .

[30]  Mevin B. Hooten,et al.  Predicting the spatial distribution of ground flora on large domains using a hierarchical Bayesian model , 2003, Landscape Ecology.

[31]  P. Boveng,et al.  Ringed and bearded seal densities in the eastern Chukchi Sea, 1999–2000 , 2005, Polar Biology.

[32]  John R Fieberg,et al.  Spending degrees of freedom in a poor economy: A case study of building a sightability model for moose in northeastern Minnesota , 2012 .

[33]  K. McCoy,et al.  Cryptic vector divergence masks vector-specific patterns of infection: an example from the marine cycle of Lyme borreliosis , 2010, Evolutionary applications.

[34]  Kevin G. Thompson.  Use of site occupancy models to estimate prevalence of Myxobolus cerebralis infection in trout. , 2007, Journal of aquatic animal health.

[35]  Evan G. Cooch,et al.  Multistate capture–recapture analysis under imperfect state observation: an application to disease models , 2009 .