Sample size considerations for satellite telemetry and animal distributions

Roberts, Anthony J.; Silverman, Emily; Gifford, Suzanne J.

doi:10.1002/jwmg.21504

Cited by 4 publications

(4 citation statements)

References 19 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In addition, generating a representative network requires distributing transmitters throughout the population(s) of interest. This study benefited from a coordinated effort to deploy transmitters throughout the populations of interest, resulting in comprehensive coverage of scoter species' ranges (Pearce et al 2019) with sample sizes sufficient to characterize seasonal distributions (ie at least 80-130 individuals per species; Roberts et al 2018). However, low-density wintering areas (eg the Gulf of Mexico for black scoters) were unsampled and are underrepresented in the network.…”

Section: Discussionmentioning

confidence: 99%

See 1 more Smart Citation

Evaluating conservation units using network analysis: a sea duck case study

Lamb,

Cooper‐Mullin,

Gilliland

et al. 2024

Frontiers in Ecol & Environ

View full text Add to dashboard Cite

Conserving migratory wildlife requires understanding how groups of individuals interact across seasons and landscapes. Telemetry reveals individual movements at large spatiotemporal scales; however, using movement data to define conservation units requires scaling up from individual movements to species‐ and community‐level patterns. We developed a framework to define flyways and identify important sites from telemetry data and applied it to long‐term, range‐wide tracking data from three species (640 individuals) of sea ducks: namely, North American scoters (Melanitta spp). Our network of 88 nodes included both multispecies hotspots and areas uniquely important to individual species. We found limited spatial overlap between scoters wintering on the Atlantic and Pacific coasts of North America, with differing connectivity patterns between coasts. Finally, we identified four multispecies conservation units that did not correspond to traditional management flyways. From this approach, we show how individual movements can be used to quantify range‐wide connectivity of migratory species and reveal gaps in conservation strategies.

show abstract

Section: Discussionmentioning

confidence: 99%

“…2019) with sample sizes sufficient to characterize seasonal distributions (ie at least 80–130 individuals per species; Roberts et al . 2018). However, low‐density wintering areas (eg the Gulf of Mexico for black scoters) were unsampled and are underrepresented in the network.…”

Section: Discussionmentioning

confidence: 99%

Evaluating conservation units using network analysis: a sea duck case study

Lamb,

Cooper‐Mullin,

Gilliland

et al. 2024

Frontiers in Ecol & Environ

View full text Add to dashboard Cite

show abstract

“…Sampling parameters are dictated by the cost, size, and technological limitations of tracking devices, rather than designed to maximize statistical power of the results to detect target patterns of movement [ 1 , 2 ]. Nevertheless, accumulation of tracking data over time has created opportunities to synthesize existing data and assess how the number and distribution of sampled individuals affects detection of variation and population-level inferences [e.g., 3 , 4 ]. Such retrospective studies have generally focused on active transmitters that actively collect, store, and/or upload location data via satellite networks.…”

Section: Introductionmentioning

confidence: 99%

Distributing transmitters to maximize population-level representativeness in automated radio telemetry studies of animal movement

2023

View full text Add to dashboard Cite

Telemetry is a powerful and indispensable tool for evaluating wildlife movement and distribution patterns, particularly in systems where opportunities for direct observation are limited. However, the effort and expense required to track individuals often results in small sample sizes, which can lead to biased results if the sample of tracked individuals does not fully capture spatial, temporal, and individual variability within the target population. To better understand the influence of sampling design on results of automated radio telemetry studies, we conducted a retrospective power analysis of very high frequency (VHF) radio telemetry data from the Motus Wildlife Tracking System for two species of birds along the United States Atlantic coast: a shorebird, the piping plover (Charadrius melodus), and a nearshore seabird, the common tern (Sterna hirundo). We found that ~ 100–150 tracked individuals were required to identify 90% of locations known to be used by the tracked population, with 40–50 additional individuals required to include 95% of used locations. For any number of individuals, the percentage of stations included in the sample was higher for common terns than for piping plovers when tags were deployed within a single site and year. Percentages of stations included increased for piping plovers when birds were tagged over multiple sites and, to a lesser extent, years, and increased with average length of the tracking period. The probability that any given receiver station used by the population would be included in a subsample increased with the number of birds tracked, station proximity to a migratory stopover or staging site, number of receiving antennas per station, and percentage of the tracked population present. Our results provide general guidance for the number and distribution of tagged birds required to obtain representative VHF telemetry data, while also highlighting the importance of accounting for station network configuration and species-specific differences in behavior when designing automated radio telemetry studies to address specific research questions. Our results have broad applications to remotely track movements of small-bodied migratory wildlife in inaccessible habitats, including predicting and monitoring effects of offshore wind energy development.

show abstract

“…The number collared may be too low to get accurate estimates of the UD. Seaman et al (1999) estimated that at least 30 observations (but preferably ≥ 50) are necessary using KDE with least-squares cross validation to estimate the smoothing bandwidth and other studies have suggested much larger samples are necessary (Arthur & Schwartz, 1998;Lindberg & Walker, 2007;Roberts et al, 2018). In practice, this means that achieving an adequate sample size from collared individuals may require combining multiple years of location data or may not be possible.…”

Section: Introductionmentioning

confidence: 99%

Aerial survey and telemetry data analysis of a peripheral caribou calving area in northwestern Alaska

Prichard

Klimstra

Person

et al. 2019

Ran

View full text Add to dashboard Cite

With industrial development expanding in the Arctic, there is increasing interest in quantifying the impacts of development projects on barren ground caribou (Rangifer tarandus granti). The primary data source to assess caribou distribution and predict impacts in remote areas of Alaska has shifted in recent decades from aerial survey data to telemetry data, but these techniques have different strengths and weaknesses. The ranges of two caribou herds, the Western Arctic Herd and the Teshekpuk Herd, overlap in northwest Alaska between Wainwright and Atqasuk, Alaska. Based on long-term telemetry data sets, this region was thought to be outside of the core calving ranges of both herds. Calving has long been reported to occur in this general area, but early reports assumed caribou were from the Western Arctic Herd and only one systematic aerial survey of caribou density and distribution during calving has been conducted in this area in recent decades. Following interest in industrial development in this area, we conducted aerial strip-transect surveys during early to mid-June 2013–2015 to directly assess the density and distribution of caribou in the area and we used existing telemetry data to compare our results to the seasonal distribution of both herds. Total caribou densities varied between 0.36 and 1.06 caribou/km² among years, and calf densities varied 0.04 and 0.25 calves/km² among years. Contrary to assumptions by early researchers in the area, telemetry data indicated that caribou in this area during early to mid-June were from the Teshekpuk Herd. The use of telemetry data alone underestimated the importance of this area for calving, but the combination of aerial surveys and telemetry data provided complementary information on caribou use of this area showing the importance of collecting the appropriate types of data for assessing potential impacts of development on caribou.

show abstract

Sample size considerations for satellite telemetry and animal distributions

Cited by 4 publications

References 19 publications

Evaluating conservation units using network analysis: a sea duck case study

Evaluating conservation units using network analysis: a sea duck case study

Distributing transmitters to maximize population-level representativeness in automated radio telemetry studies of animal movement

Aerial survey and telemetry data analysis of a peripheral caribou calving area in northwestern Alaska

Contact Info

Product

Resources

About