Special Collection on Integrated Population Models

11 articles / 1 book cited, including 3 from The Auk: Ornithological Advances 

Improving Population-Level Inference through Study of Avian Life Histories with Integrated Population Models

Mitch D. Weegman, School of Natural Resources, University of Missouri, Columbia, Missouri, USA, [email protected]

Conservation and management policies and decisions about bird populations are commonly based on population models that rely on piecemeal analyses of abundance, survival, and reproductive success. With the advent of integrated population models (IPMs), researchers are able to combine datasets that directly and indirectly inform abundance and demographic parameters (e.g., reproduction, survival, and movements; Besbeas et al. 2002, Schaub et al. 2007, Kéry and Schaub 2012). The integration of datasets also enables cross-seasonal and spatial comparisons of animal population dynamics. As such, IPMs provide a more natural link to conservation planning than traditional population models (e.g., Koons et al. 2017). IPMs are often constructed in the Bayesian framework, which allows researchers to incorporate prior knowledge based on previous studies or expert opinion. Key benefits of IPMs include increased precision of parameter estimates and the ability to estimate unmeasured demographic rates (Weegman et al. 2016). These properties allow for more detailed insights into the processes that drive variation in key demographic rates that have hitherto been difficult to study (for an example IPM with linkage among datasets for parameter estimation, see Figure 1).

Figure 1. An example integrated population model (IPM) with two ages (juvenile [juv] and adult [ad]) formed using capture–recapture, count, and productivity data (data streams are in small boxes). There are three model processes for these data, each corresponding to a box with a different type of line. In this case, the capture–recapture data (m) are modeled in a multistate framework, where ?juv and ?ad represent juvenile and adult apparent survival rates, ψjuv and ψad represent juvenile and adult emigration rates, and pjuv and pad represent juvenile and adult recapture (resighting) rates. The count data (y) are modeled in a state-space framework, where immigration rate (ω) is inferred from estimation of all other demographic rates, and population size N (with observation error σy2) is estimated from all demographic rates. The productivity data are modeled using a Poisson model with the numbers of broods (R) and juveniles (J) to estimate recruitment rate (f). These datasets are modeled in a joint likelihood, which is the composite of single likelihood models, to holistically estimate parameters. Thus, parameter estimation is often informed by multiple datasets (i.e. in this figure, wherever model processes overlap; e.g., ϕjuv and ϕad are estimated using capture–recapture and count data). Adapted from Kéry and Schaub (2012).

This Special Collection is based on contributed talks from a symposium at the sixth North American Ornithological Conference, held in Washington, D.C., during August 2016, which highlighted the breadth and scope of recent developments in the use of IPMs to study avian life histories, ecology, and conservation. Researchers in avian ecology have used IPMs to address a variety of questions for a variety of species (e.g., Wilson et al. 2016, Arnold et al. 2018; for a review, see Zipkin and Saunders 2018). Our objectives in the symposium were threefold:

1. To demonstrate the diversity of IPMs to answer broad questions pertaining to avian life histories.
2. To examine how IPMs increase population-level inferences (e.g., quantifying demographic rates not previously identifiable using traditional population models).
3. To discuss the potential for IPMs to reform our understanding of population dynamics in avian ecology, and advance conservation and management in North America and globally.

To accomplish these objectives, we covered a variety of topics during the symposium, including

• informing management of endangered species;
• estimation of unmeasured demographic rates;
• continental-scale management of migratory game birds;
• quantifying the influence of environmental drivers on population dynamics of migratory and non-migratory birds;
• forecasting population dynamics under future environmental scenarios;
• better understanding migratory connectivity through population dynamics.

Specifically in this Special Collection, Coates et al. (2018) utilized the IPM framework to guide management and conservation efforts for Greater Sage-Grouse (Centrocercus urophasianus) in California and Nevada, USA, by integrating lek count and telemetry data collected for 6 subpopulations. While there was no substantial positive or negative trend in the metapopulation of Greater Sage-Grouse, Coates et al. identified that annual variation in population growth rates were most strongly associated with a lagged precipitation effect. Thus, this paper provides examples of tests of contributions of demographic rates to population growth rate and inclusion of covariates to better describe the variation observed.

McConnell et al. (2018) use capture-recapture and harvest data to develop an integrated population model to robustly quantify Northern Bobwhite (Colinus virginianus) demography in northern Florida and southern Georgia, USA. They used these data types to inform estimation of fecundity and chick survival, when no direct data for these two rates existed, which has been a hallmark of integrated population models. The authors also include hypothesis tests of population size, as well as summer and fall temperature and precipitation. Their results suggest that recruitment most strongly explained population growth rate and that breeding season maximum temperature and precipitation explained variation in recruitment. Thus, McConnell et al. provide a critical example of borrowing strength among data types to inform otherwise inestimable rates, with subsequent hypothesis tests of demographic drivers.

Ross et al. (2018) integrated lek count, nest monitoring, and telemetry data, as well as Palmer Drought Severity Index data, to quantify drivers of Lesser Prairie-Chicken (Tympanuchus pallidicinctus) population dynamics in western and southern Kansas, USA, which comprises the core range for these birds. They found that variability in population growth rate was best explained by variation in juvenile survival (even though juvenile survival was measured only in some years), while the Palmer Drought Severity Index had no substantive impact on the population dynamics of these birds, although the population growth rate declined after severe droughts. Thus, Ross et al. provide an important example of accounting for demographic and environmental sources of variability (including responses to climate forecasts) when designing conservation plans.

Chandler et al. (2018) illustrate integration of spatial capture–recapture data collected at few sites with count data collected at a broader scale for Canada Warblers (Cardellina canadensis) in North Carolina, USA. Although spatial mismatch in datasets is often problematic, Ahrestani et al. (2017) and Chandler et al. (2018) provide compelling examples of how researchers can utilize broad- and fine-scale data in IPMs. Chandler et al. extend their spatially explicit IPM by including a spatio-temporal point process model for recruitment rates. They modeled bird density, survival, and recruitment as a function of elevation for spatial comparisons. They showed that survival of male Canada Warblers increased with elevation, while no relationship existed with recruitment, and yet density varied greatly by elevation. Chandler et al. provide important foundational research for continued work at the interface of IPM development and spatial estimation of demographic rates.

In summary, the IPM framework offers great potential for avian ecologists. Our field is primed for researchers to utilize and advance IPMs, particularly because ornithologists have access to rich spatial and temporal datasets, and there is a broad need to answer population-level questions to better inform conservation and management of birds across landscapes. We consider that this Special Collection can serve as a reference for researchers with varied backgrounds to develop IPMs using a variety of datasets to answer pressing ecological questions. The works included here represent a portion of the symposium. We anticipate further populating this Special Collection as additional works are peer reviewed.

Literature Cited

Ahrestani, F. S., J. F. Saracco, J. R. Sauer, K. L. Pardieck, and J. A. Royle (2017). An integrated population model for bird monitoring in North America. Ecological Applications 27:916–924.

Arnold, T. W., R. G. Clark, D. N. Koons, and M. Schaub (2018). Integrated population models facilitate ecological understanding and improved management decisions. Journal of Wildlife Management 82:266–274.

Besbeas, P., S. N. Freeman, B. J. T. Morgan, and E. A. Catchpole (2002). Integrating mark–recapture–recovery and census data to estimate animal abundance and demographic parameters. Biometrics 58:540–547.

Chandler, R. B., J. Hepinstall-Cymerman, S. Merker, H. Abernathy-Conners, and R. J. Cooper (2018). Characterizing spatio-temporal variation in survival and recruitment with integrated population models. The Auk: Ornithological Advances 135:409–426.

Coates, P. S., B. G. Prochazka, M. A. Ricca, B. J. Halstead, M. L. Casazza, E. J. Blomberg, B. E. Brussee, L. Wiechman, J. Tebbenkamp, S. C. Gardner, and K. P. Reese (2018). The relative importance of intrinsic and extrinsic drivers to population growth vary among local populations of Greater Sage-Grouse: An integrated population modeling approach. The Auk: Ornithological Advances 135:240–261.

Kéry, M., and M. Schaub (2012). Bayesian Population Analysis Using WinBUGS: A Hierarchical Perspective. Academic Press, Waltham, MA, USA.

Koons, D. N., T. W. Arnold, and M. Schaub (2017). Understanding the demographic drivers of realized population growth rates. Ecological Applications 27:2102–2115.

McConnell, M. D., A. P. Monroe, R. Chandler, W. E. Palmer, S. D. Wellendorf, L. W. Burger, Jr., and J. A. Martin (2018). Factors influencing Northern Bobwhite recruitment, with implications for population growth. The Auk: Ornithological Advances 135:1087-1099.

Ross, B. E., D. A. Haukos, C. A. Hagen, and J. Pitman (2018). Combining multiple sources of data to inform conservation of Lesser Prairie-Chicken populations. The Auk: Ornithological Advances 135:228–239.

Schaub, M., O. Gimenez, A. Sierro, and R. Arlettaz (2007). Use of integrated modeling to enhance estimates of population dynamics obtained from limited data. Conservation Biology 21:945–955.

Weegman, M. D., S. Bearhop, A. D. Fox, G. M. Hilton, A. J. Walsh, J. L. McDonald, and D. J. Hodgson (2016). Integrated population modelling reveals a perceived source to be a cryptic sink. Journal of Animal Ecology 85:467–475.

Wilson, S., K. C. Gil-Weir, R. G. Clark, G. J. Robertson, and M. T. Bidwell (2016). Integrated population modeling to assess demographic variation and contributions to population growth for endangered Whooping Cranes. Biological Conservation 197:1–7.

Zipkin, E. F., and S. P. Saunders (2018). Synthesizing multiple data types for biological conservation using integrated population models. Biological Conservation 217:240–250.