Satellite imagery can predict bird species occupancy and inform multispecies management in pine savannas

Cory R. Allred; Todd M. Schneider; Elizabeth Ann Hunter

doi:10.1093/ornithapp/duaf029

Home/Archive/USGS

USGS · 70269918

Satellite imagery can predict bird species occupancy and inform multispecies management in pine savannas

Cory R. Allred·Todd M. Schneider·Elizabeth Ann Hunter

View original source ↗Export metadata ↓

Abstract

Multispecies management can contribute to meeting growing challenges of preserving biodiversity, yet current game and threatened species management often focuses on individual species. Satellite imagery available at high spatial and temporal resolution provides a potential tool to overcome the challenge posed by multispecies management of linking patterns of habitat use among species. We sought to determine whether satellite imagery could be used to describe patterns of species occupancy and inform multispecies management in pine savannas in Georgia, USA. We conducted point-count surveys at 7 sites in 2022 for 3 bird species: Colinus virginianus (Northern Bobwhite), Dryobates borealis (Red-cockaded Woodpecker), and Peucaea aestivalis (Bachman’s Sparrow). We built single-season occupancy models comparing a set of models using covariates collected from field vegetation surveys and another set using covariates extracted from Sentinel-2 satellite imagery. We then used a multi-objective optimization algorithm to identify quasi-optimal management solutions (i.e., sets of covariate values from top satellite imagery metric models). We found that models created using satellite imagery performed well at predicting occupancy of all 3 species as measured by the area under the receiver operating characteristic curve (AUC > 0.8) and had higher AUC scores than field-derived habitat covariate-based models. We found combinations of metrics that could result in high rates of predicted probability of occupancy for all species (within 86% of highest possible occupancy probability), but these combinations did not exist at any of the sites. Our results demonstrate that (1) satellite imagery can allow users to build reliable occupancy models without intensive field-based vegetation surveys; and (2) C. virginianus , D. borealis, and P. aestivalis in pine savanna ecosystems could be simultaneously managed through more frequent burning, changes in canopy cover or by producing suitable heterogeneity of habitats after identifying an appropriate scale of management.

Explore related subjects

Ornithological Applications Article

Source-reported bounding extent: 30.357851° to 35.000768° latitude; -85.605165° to -80.840549° longitude. This indicates report coverage, not an exact sampling location. View area on OpenStreetMap.

Keep this discovery

Explore connections, maps & timelines

BibTeX RIS

Cory R. Allred, Todd M. Schneider, Elizabeth Ann Hunter. 2025-04-25. Satellite imagery can predict bird species occupancy and inform multispecies management in pine savannas. https://doi.org/10.1093/ornithapp%2Fduaf029

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related USGS reports

USGS2026

The cost of limited nocturnal roost availability on pre-migratory energy budgets in a central-place foraging shorebird, Numenius hudsonicus (Hudsonian Whimbrel)

Mobile species often travel long distances between the sites they use for different behaviors, such as central-place foragers that make daily commutes between a central roost and foraging sites. For migratory shorebird species that use coastal habitats, the limited availability of suitable nocturnal roost sites can lead individuals to commute long distances to their foraging territories, forcing them to balance the energetic cost of commuting with the need to refuel in between migratory flights. We investigated this potential trade-off in a population of Numenius hudsonicus (Hudsonian Whimbrel) that show wide variation in commute distance between a shared central nocturnal roost and individual foraging territories in coastal South Carolina, USA. We combined high-resolution movement data, behavioral state modeling, and empirical foraging behavior data to assess how increasing commute distances from a central nocturnal roost affect individual daily activity budgets, foraging habitat use, potential net energy gain, and breeding ground departure date. We found that N. hudsonicus that commute farther from their nocturnal roost spend a higher proportion of each day foraging and commuting, and that they delay their departure to the breeding grounds by 3–6 days. Our results indicate that roosting habitat close to foraging areas can reduce the energetic cost of commuting to help shorebirds avoid delays in migratory timing. Thus, protecting, restoring, and creating nocturnal roosting habitat close to foraging areas may enhance successful shorebird migration.

South Carolina↗

USGS2026

Modeling observational uncertainty improves estimation of breeding-season parameters that inform lek-based monitoring for Centrocercus urophasianus (Greater Sage-Grouse)

Counts of males on leks are widely used as abundance indices to monitor populations of lek-breeding species, but they can be influenced by variation in male attendance, inter-lek movement, and mortality. When there is imperfect observability of males on leks, estimates of lek attendance and inter-lek movement are likely to be biased low. We compared traditional multi-state mark–recapture (MSMR) and MSMR models with state uncertainty (MSMR-SU) to determine how accounting for uncertainty in GPS-derived lek attendance improves inference of rates underlying lek-based abundance indices for male Centrocercus urophasianus (Greater Sage-Grouse). We found that modeling observational uncertainty with MSMR-SU models consistently reduced bias in lek attendance and inter-lek movement estimates for adult and yearling males compared to MSMR models. Estimates of daily maximum male lek attendance (adults: 0.79; yearlings: 0.68) and inter-lek movement rates (adults: 0.10; yearlings: 0.07) from MSMR-SU models were higher than daily maximum lek attendance (adults: 0.74; yearlings: 0.55) and inter-lek movement rates (adults: 0.01; yearlings: 0.04) from MSMR models. Consistent with previous studies, adults exhibited higher lek attendance and mortality rates than yearlings. However, unlike earlier studies, inter-lek movement rates were similar between age classes, with adults tending to move earlier in the lekking season than yearlings. Wind speed and precipitation had a weak negative effect on lek attendance. Greater distance to the nearest known lek strongly reduced inter-lek movement. Accounting for lekking-season mortality shifted the estimated date of maximum lek attendance earlier in the season by 8–9 days for adults and 1–3 days for yearlings. Accounting for state uncertainty due to error in state assignment should reduce bias in demographic estimates for use in wildlife conservation and management decisions.

Colorado, Wyoming↗

USGS2026

Incorporating location uncertainty improves inference with stop-level North American Breeding Bird Survey data

Ecological models should account for uncertainty to be most effective and useful. Yet, uncertainty from model covariates—unlike that from other sources, such as sampling error or process variability—is seldom explicitly incorporated. This can cause underestimates of uncertainty to cascade through model parameter estimates, predictions, and downstream uses. Burner et al. proposed a method for quantifying uncertainty in covariates and incorporating it into models using informative Bayesian priors. This method was applied to stop-level Breeding Bird Survey (BBS) analyses, where land cover uncertainty at each stop arises from substantial stop location uncertainty. A limited validation of model-estimated land cover, using stops with known locations, indicated the method’s potential effectiveness, but it was not rigorously evaluated. We conduct a robust simulation-based test, generating stop locations, extracting land cover, and simulating bird communities across 210 BBS routes in the upper Midwest. We compare 3 models: a “known” model with true land cover, a “naive” model assuming consistent 800-m stop spacing, and a “full” model using informative priors to estimate land cover. Species parameter estimates and predicted prevalence patterns across gradients in land cover from the full model approached those of the known model and were substantially closer to the true values used in simulations relative to those from the naive model. Naive model parameters were more biased relative to the other models, and credible intervals of predicted species prevalence rarely included the true simulated values. The full model also produced land cover covariate estimates closer to true simulation values relative to the mean informative priors. Our results show that, for the BBS, informative priors enable more accurate stop-level analyses despite location uncertainty. In contrast, naive models that ignore this uncertainty yield poor inferences. More broadly, we demonstrate empirically the utility of informative priors to account for covariate uncertainty in ecological models.

Michigan, Minnesota, Wisconson↗