Integrated modeling of waterfowl distribution in western Canada using aerial survey and citizen science (eBird) data

Adde, Antoine; Amat, Clara Casabona i; Mazerolle, Marc J.; Darveau, Marcel; Cumming, Steven G.; O’Hara, Robert B.

doi:10.1002/ecs2.3790

Cited by 12 publications

(18 citation statements)

References 57 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For each model, the spatial field shared among the three likelihoods captured the spatial structure not explained by the covariates, as well as any spatial autocorrelation among the sampled locations, with higher values where data were more clustered (Adde et al 2021). For the three models, the spatial field contributed greatly to the overall predicted distribution, with the covariates having in general small effects for all species.…”

Section: Discussionmentioning

confidence: 99%

“…This approach provides a computationally fast modelling environment for hierarchical Bayesian models, where complex spatially structured random effects can be added to models for a wide variety of response variables (e.g. binomial models for PA data or Poisson models for PO and count data, Bakka et al 2018), and where joint‐likelihood models can be developed to integrate different sources of data (Adde et al 2021, Cunningham et al 2021, Simmonds et al 2020).…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Bayesian species distribution models integrate presence‐only and presence–absence data to predict deer distribution and relative abundance

et al. 2022

View full text Add to dashboard Cite

Using geospatial data of wildlife presence to predict a species distribution across a geographic area is among the most common tools in management and conservation. The collection of high‐quality presence–absence (PA) data through structured surveys is, however, expensive, and managers usually have access to larger amounts of low‐quality presence‐only (PO) data collected by citizen scientists, opportunistic observations and culling returns for game species. Integrated species distribution models (ISDMs) have been developed to make the most of the data available by combining the higher‐quality, but usually scarcer and more spatially restricted, PA data with the lower‐quality, unstructured, but usually more extensive PO datasets. Joint‐likelihood ISDMs can be run in a Bayesian context using integrated nested laplace approximation methods that allow the addition of a spatially structured random effect to account for data spatial autocorrelation. Here, we apply this innovative approach to fit ISDMs to empirical data, using PA and PO data for the three prevalent deer species in Ireland: red, fallow and sika deer. We collated all deer data available for the past 15 years and fitted models predicting distribution and relative abundance at a 25 km2 resolution across the island. Model predictions were associated to spatial estimate of uncertainty, allowing us to assess the quality of the model and the effect that data scarcity has on the certainty of predictions. Furthermore, we checked the performance of the three species‐specific models using two datasets, independent deer hunting returns and deer densities based on faecal pellet counts. Our work clearly demonstrates the applicability of spatially explicit ISDMs to empirical data in a Bayesian context, providing a blueprint for managers to exploit unexplored and seemingly unusable data that can, when modelled with the proper tools, serve to inform management and conservation policies.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Bayesian species distribution models integrate presence‐only and presence–absence data to predict deer distribution and relative abundance

et al. 2022

View full text Add to dashboard Cite

show abstract

“…Assessing model fit when using multiple streams of data in an integrated framework is often not straightforward 21,59,66 . Our process for measuring model error was to use both out-of-sample data and crossvalidation approaches.…”

Section: Discussionmentioning

confidence: 99%

“…Plover breeding densities in the PPR showed high spatial and temporal autocorrelation. Accounting for this autocorrelation carries distinct advantages over non-explicit alternatives when modeling mobile species that select habitats based on behavioral or other cues of the environment that can be difficult to account for 31,59 . Avian species may shift their breeding distributions across time to take advantage of suitable habitats 71,72 , yet high site fidelity can lead to greater reproductive success 73 or population persistence 74 when site quality is temporally correlated.…”

Section: Discussionmentioning

confidence: 99%

“…We used SPDE to model a spatial random field with mean 0 and a Matérn covariance function 58 . To define the mesh design and triangle size, we started with a low-resolution mesh and iteratively fit smaller triangles to identify an optimal resolution that balanced computational cost and quality of approximation 59 . Modeling temporal variation in species distributions can be achieved by allowing environmental associations or residual spatial variation to fluctuate over time.…”

Section: Preliminary Covariate Selectionmentioning

confidence: 99%

See 1 more Smart Citation

Data integration reveals dynamic and systematic patterns of breeding habitat use by a threatened shorebird

Ellis

Anteau

MacDonald

et al. 2023

Sci Rep

View full text Add to dashboard Cite

Incorporating species distributions into conservation planning has traditionally involved long-term representations of habitat use where temporal variation is averaged to reveal habitats that are most suitable across time. Advances in remote sensing and analytical tools have allowed for the integration of dynamic processes into species distribution modeling. Our objective was to develop a spatiotemporal model of breeding habitat use for a federally threatened shorebird (piping plover, Charadrius melodus). Piping plovers are an ideal candidate species for dynamic habitat models because they depend on habitat created and maintained by variable hydrological processes and disturbance. We integrated a 20-year (2000–2019) nesting dataset with volunteer-collected sightings (eBird) using point process modeling. Our analysis incorporated spatiotemporal autocorrelation, differential observation processes within data streams, and dynamic environmental covariates. We evaluated the transferability of this model in space and time and the contribution of the eBird dataset. eBird data provided more complete spatial coverage in our study system than nest monitoring data. Patterns of observed breeding density depended on both dynamic (e.g., surface water levels) and long-term (e.g., proximity to permanent wetland basins) environmental processes. Our study provides a framework for quantifying dynamic spatiotemporal patterns of breeding density. This assessment can be iteratively updated with additional data to improve conservation and management efforts, because reducing temporal variability to average patterns of use may cause a loss in precision for such actions.

show abstract

Spatial modeling of the relative abundance of wading birds in peninsular Florida using citizen science data

Lee

Skripnikov

2023

Ecosphere

View full text Add to dashboard Cite

Recent literature has shown citizen science data that include reported counts, such as observations from the eBird database, can be used to model the relative abundance of selected species. The objective of our study was to extend on both the methodology and the range of species considered by previous research on relative abundance modeling; we selected the Floridian peninsula as our geographical area of interest. We confirmed that, much like for other bird species and geographical areas, modeling such data with a classic Poisson generalized linear model was not appropriate due to overdispersion. Moreover, assuming linearity for the effects of all environmental and observer effort covariates on relative abundance was shown to be overly simplistic. These concerns led us to consider a variety of potentially suitable modeling techniques; the quasi‐Poisson generalized additive model displayed the best combination of flexibility, strong predictive performance, and ease of interpretation. Expanding on previous research for eBird data, we chose to explicitly incorporate spatial dependence into the modeling task by performing hierarchical generalized additive modeling with a spatial conditional autoregressive structure for random effects. This practice allows us to account for spatial factors not represented by model covariates, such as species dispersal patterns or colonial breeding habits. After conducting appropriate testing procedures and fitting spatially explicit models, we found that our data contain moderate spatial dependence and that spatially explicit models have overall superior predictive performance to models that do not account for spatial dependence. Upon closer inspection, spatially explicit models perform especially well compared with their nonspatial counterparts in identifying situations with low abundance, with a negligible trade‐off in prediction accuracy for situations with high abundance. We conclude that quasi‐Poisson hierarchical generalized additive models with spatial random effects provide the best representation of the relative abundance of bird populations. Moreover, our spatially explicit models are more realistic based on domain knowledge when regarding the impact of environmental covariates, which is important when considering conservation implications and future projections.

show abstract

Integrated modeling of waterfowl distribution in western Canada using aerial survey and citizen science (eBird) data

Abstract: Integrated modeling of waterfowl distribution in western Canada using aerial survey and citizen science (eBird) data.

Cited by 12 publications

References 57 publications

Bayesian species distribution models integrate presence‐only and presence–absence data to predict deer distribution and relative abundance

Bayesian species distribution models integrate presence‐only and presence–absence data to predict deer distribution and relative abundance

Data integration reveals dynamic and systematic patterns of breeding habitat use by a threatened shorebird

Spatial modeling of the relative abundance of wading birds in peninsular Florida using citizen science data

Contact Info

Product

Resources

About