Use of hidden Markov capture–recapture models to estimate abundance in the presence of uncertainty: Application to the estimation of prevalence of hybrids in animal populations

Abstract

Estimating the relative abundance (prevalence) of different population segments is a key step in addressing fundamental research questions in ecology, evolution, and conservation. The raw percentage of individuals in the sample (naive prevalence) is generally used for this purpose, but it is likely to be subject to two main sources of bias. First, the detectability of individuals is ignored; second, classification errors may occur due to some inherent limits of the diagnostic methods. We developed a hidden Markov (also known as multievent) capture–recapture model to estimate prevalence in free‐ranging populations accounting for imperfect detectability and uncertainty in individual's classification. We carried out a simulation study to compare naive and model‐based estimates of prevalence and assess the performance of our model under different sampling scenarios. We then illustrate our method with a real‐world case study of estimating the prevalence of wolf (Canis lupus) and dog (Canis lupus familiaris) hybrids in a wolf population in northern Italy. We showed that the prevalence of hybrids could be estimated while accounting for both detectability and classification uncertainty. Model‐based prevalence consistently had better performance than naive prevalence in the presence of differential detectability and assignment probability and was unbiased for sampling scenarios with high detectability. We also showed that ignoring detectability and uncertainty in the wolf case study would lead to underestimating the prevalence of hybrids. Our results underline the importance of a model‐based approach to obtain unbiased estimates of prevalence of different population segments. Our model can be adapted to any taxa, and it can be used to estimate absolute abundance and prevalence in a variety of cases involving imperfect detection and uncertainty in classification of individuals (e.g., sex ratio, proportion of breeders, and prevalence of infected individuals).

A standardized approach to empirically define reliable assignment thresholds and appropriate management categories in deeply introgressed populations

“…Nonetheless, these percentages of admixed individuals cannot be intended as estimates of prevalence of admixed individuals in the Italian wolf population because the analysed samples had not been randomly collected, but mostly derived from specific monitoring projects focused on hybrid detection and from heterogeneously monitored areas 26,31,33,52,56 . Conversely, reliable estimates of hybridization prevalence could be assessed through statistical multi-event models applied to capture-recapture data obtained from well-planned long-term genetic and camera-trapping monitoring projects carried out through the entire Italian wolf distribution range [79][80][81] .…”

“…On the other side, the active management of introgressed individuals might become a necessary option where they locally occur at a high prevalence (that can be sometimes much higher than region-or population-wide estimates), thus increasing the probability of interbreeding between hybrids and retaining domestic variants on the long term 81,82 .…”

Estimating Admixture at the Population Scale: Taking Imperfect Detectability and Uncertainty in Hybrid Classification Seriously

“…The multievent formulation of open population CR models explicitly handles uncertainty in individual classification by modeling the observed capture histories as 2 time series: the state process (i.e., the population dynamics during the study) and the event process (i.e., what we can observe through sampling; Pradel 2005). Following Pradel (2005) and Santostasi et al (2019), we modeled the state process as a Markov chain of 3 partially hidden states: alive in the study area as wolf, alive in the study area as admixed, and dead or permanently emigrated. The state process was described by the initial state probability (π w = the probability that an individual was in one or the other state when first encountered) and the apparent survival probability (φ = the probability that an individual survived and remained in the study area between sampling occasions).…”

“…Second, they should derive from estimation methods that formally account for imperfect detectability and other potential sources of bias (Anderson 2001, Yoccoz et al 2001); in particular, because prevalence is essentially a proportion measuring the relative abundance of admixed and parental individuals, the estimation process should account for a potentially different detectability of the 2 forms (i.e., admixed vs. parental). Third, the inherent uncertainty that generally afflicts the classification of individuals as parental or admixed, especially if based on poor‐quality DNA samples, should be formally accounted for within the estimation framework (Santostasi et al 2019). Specifically, even though genetic markers are considered at large more reliable than phenotypic cues of hybridization (Allendorf et al 2001), uncertainty in detecting admixed individuals still remains and depends on 2 interacting factors: the number and type of genetic markers used, and the statistical methods and options adopted to assign sampled individuals to the parental or admixed reference populations (Vähä and Primmer 2006, Bohling et al 2013).…”

Advances in modelling demographic processes: The Euring 2017 Analytical Meeting

“…Santostasi et al. () developed a multievent capture–recapture model to estimate hybrids’ prevalence in wolf populations accounting for both detection probabilities and error rates in hybridization assessment.…”