Using multiple datasets to account for misalignment between statistical and biological populations for abundance estimation

  1. Michelle Kissling
  2. Paul Lukacs
  3. Kelly Nesvacil
  4. Scott Gende
  5. Grey Pendleton

Reviewed by

Read the full article See related articles

Discuss this preprint

Start a discussion What are Sciety discussions?

Listed in

Log in to save this article

Abstract

A fundamental aspect of ecology is identifying and characterizing population processes. Because a complete census is rare, we almost always use sampling to make inference about the biological population, and the part of the population at risk of sampling is referred to as the statistical population. Ideally, the statistical population is the same as, or accurately represents its corresponding biological population. However, in practice, they rarely align in space and time, which can lead to biased inference. We often view a population misalignment as a temporary emigration process and resolve it with replicate and/or repeat sampling, though this approach is not feasible for all species and habitats. We developed a hierarchical modeling framework to estimate abundance of a biological population of the Kittlitz’s murrelet ( Brachyramphus brevirostris ), a highly mobile, non-territorial, ice-associated seabird of conservation concern in Alaska and eastern Russia. Our framework combines datasets from boat and telemetry surveys to account for all components of detection probability, specifically using telemetry locations to estimate probability of presence ( p p ) and line transect distance sampling to estimate probability of detection ( p d ). By estimating p p directly, we were able to account for temporary emigration from the sampled area, which changed with shifting icefloes between sampling occasions. Between 2007 and 2012, annual p p was highly variable, ranging from 0.33 to 0.75 (median=0.50, standard deviation=0.02), but was not predictable using five environmental covariates. In years when two boat surveys were conducted, our model reduced the coefficient of variation (CV) of abundance estimates for the biological population compared to the statistical population by 13–35%, yet in the year with only one boat survey (2009), the CV skyrocketed about 10-fold, emphasizing the importance of a second survey if p p varies. Although we increased the precision of annual abundance estimates by accounting for p p , we did not see the same improvement in the temporal trend estimate. This result indicates that while we reduced within-year variance, we failed to account for a source(s) of variation across years, which we suspect is related to the propensity for murrelets to skip breeding in some years. Our modeling framework to account for a population misalignment is simple, flexible, and scalable for generating unbiased and precise abundance estimates of highly mobile species that occupy dynamic habitats where other open population models are not possible. Importantly, it improves inference of the biological population, which is the population of interest. We urge ecologists to think critically about the population in which they want to draw inference, especially as tracking technology improves and model complexity increases.

Article activity feed

  1. Version published to 10.24072/pcjournal.528
  2. Peer Community in Ecology

    Estimating population parameters is critical for analysis and management of wildlife populations. Drawing inference at the population level requires a robust sampling scheme and information about the representativeness of the studied population (Williams et al. 2002). In their textbook, Williams et al. (see chapter 5, 2002) listed several sampling issues, including both temporal and spatial heterogeneity and especially imperfect detection. Several methods, either sampling-based or model-based can be used to circumvent these issues.

    In their paper, Kissling et al. (2024) addressed the case of the Kittlitz’s murrelet (Brachyramphus brevirostris), a cryptic ice-associated seabird, combining spatial variation in its distribution, temporal variation in breeding propensity, imperfect detection and logistical challenges to access the breeding area. The Kittlitz’s murrelet is thus the perfect species to illustrate common issues and logistical difficulties to implement a standard sampling scheme. 

    The authors proposed a modelling framework unifying several datasets from different surveys to extract information on each step of the detection process: the spatial match between the targeted population and the sampled population, the probability of presence in the sample area, the probability of availability given presence in the sample area and finally, the probability of detection given presence and availability. All these components were part of the framework to estimate abundance and trend for this species. 

    They took advantage of a radiotracking survey during several years to inform spatial match and probability of presence. They performed a behavioural experiment to assess the probability of availability of murrelets given it was present in sampling area, and they used a conventional distance-sampling boat survey to estimate detection of individuals. This is worth noting that the most variable components were the probability of presence in the sample area, with a global mean of 0.50, and the probability of detection given presence and availability ranging from 0.49 to 0.77. The estimated trend for Kittlitz’s murrelet was negative and all the information gathered in this study will be useful for future conservation plan. 

    Coupling a decomposition of the detection process with different data sources was the key to solve problems raised by such “difficult” species, and the paper of Kissling et al. (2024) is a good way to follow for other species, allowing to inform the detection components for the targeted species - and also for our global understanding of detection process, and to infer about the temporal trend of species of conservation concern. 

    References

    Williams, B. K., Nichols, J. D., and Conroy, M. J. (2002). Analysis and management of animal populations. Academic Press.

    Michelle L. Kissling, Paul M. Lukacs, Kelly Nesvacil, Scott M. Gende, Grey W. Pendleton (2024) Using multiple datasets to account for misalignment between statistical and biological populations for abundance estimation. EcoEvoRxiv, ver.3 peer-reviewed and recommended by PCI Ecology https://doi.org/10.32942/X2W03T

  3. Version published to 10.32942/x2w03t