Leveraging the strengths of citizen science and structured surveys to achieve scalable inference on population size
Data files
Sep 22, 2023 version files 248.40 KB
-
jags_data.rds
135.51 KB
-
LinearModel_Data.csv
108.83 KB
-
README.md
4.06 KB
Abstract
- Population size is a key metric for management and policy decisions, yet wildlife monitoring programs are often limited by the spatial and temporal scope of surveys. In these cases, citizen science data may provide complementary information at higher resolution and greater extent.
- We present a case study demonstrating how data from the eBird citizen science program can be combined with regional monitoring efforts by the U.S. Fish and Wildlife Service to produce high-resolution estimates of golden eagle abundance. We developed a model that uses aerial survey data from the western United States to calibrate high-resolution annual estimates of relative abundance from eBird. Using this model, we compared regional population size estimates based on the calibrated eBird information to those based on aerial survey data alone.
- Population size estimates based on the calibrated eBird information had strong correspondence to estimates from aerial survey data in two out of four regions, and population trajectories based on the two approaches showed high correlations.
- We demonstrate how the combination of citizen science data and targeted surveys can be used to (a) increase the spatial resolution of population size estimates, (b) extend the spatial extent of inference, and (c) predict population size beyond the temporal period of surveys. Findings based on this case study can be used to refine policy metrics used by the U.S. Fish and Wildlife Service and inform permitting regulations (e.g., mortality/harm associated with wind energy development).
- Policy implications. Our results demonstrate the ability of citizen science data to complement targeted monitoring programs and improve the efficacy of decision frameworks that require information on population size or trajectory. After validating citizen science data against survey-based benchmarks, agencies can harness strengths of citizen science data to supplement information needs and increase the resolution and extent of population size predictions.
https://doi.org/10.5061/dryad.dfn2z357x
Description of the data and file structure
This file contains information and explanation for the data and code that accompany the following project:
Stillman, A.N., P.E. Howell, G.S. Zimmerman, E.R. Bjerre, B.A. Millsap, O.J. Robinson, D. Fink, E.F. Stuber, and V. Ruiz-Gutierrez. 2023. Leveraging
the strengths of citizen science and structured surveys to achieve scalable inference on population size. Journal of Applied Ecology.
This .README file accompanies the archived data for this project. Two files marked with the word “Script:” provide R code for two Bayesian models described in the main text. Two files marked with the word “Dataset:” provide the necessary data to run the models. Data from the eBird Status and Trends program are freely available online from the Cornell Lab of Ornithology. Software versions are recorded in the references section of the manuscript.
Sharing/Access information
Links to other publicly accessible locations of the data:
Code/Software
Dataset: jags_data.rds
This file contains the data to run the model in PopEstSide.mdl. Reading the file into R using the line “mod.data <- readRDS(“jags_data.rds”)” will
return a large list with the following components:
- Detection rates used in FWS Golden Eagle population size updates, developed by Western EcoSystems Technology, Inc., Cheyenne, Wyoming, USA.
mndet = Mean for normal distribution (from Western EcoSystems Technology)
sedet = SD for normal distribution (from Western EcoSystems Technology)
ndetcls = Number of detection classes
fixedyear_W = Median Year (for centering), same for both perched and flying birds
nOVstrata = BCR (strata)
nyears_W = Number of years
A = Area from NABCI layers
n.YrBCRTrn = Number of unique Year-BCR-Transect combinations
- Indexing for summing subtransect counts to the transect level for perched (P) and flying (F) birds
startP
stopP
startF
stopF
- Data for perched birds
ncounts_P = Number of observations
areaP = Transect area (one side) in km^2
detcls_P = Detection classes
strat_P = BCRs (strata; 9, 10, 16, 17)
year_P = Year
obser_P = BCR-Transect
count_P = Number of perched eagles observed along a Year-BCR-Transect-Elev-Side
nobservers_P = Number of BCR-Transects
Trn_AP = Transect area (both sides) in km^2
- Data for flying birds
ncounts_F = Number of observations
areaF = Transect area (one side) in km^2
detcls_F = Detection classes
strat_F = BCRs (strata)
year_F = Year
obser_F = BCR-Transect
count_F = Number of eagles observed along a Year-BCR-Transect-Elev-Side
nobservers_F = Number of BCR-Transects
Trn_AF = Transect area (both sides) in km^2
Dataset: LinearModel_Data.csv
BCRTrn_year = Unique ID for each transect flown during the aerial survey, in the format: [BCR number].[transect number].[year].
BCR = Bird Conservation Region (BCR) where each transect is located.
year = Year in which the aerial transect survey occured.
mean.D = Estimated density of Golden Eagles within the surveyed area, in eagles/km^2. These numbers come from the model of aerial survey data.
eBirdMean = The mean relative abundance value for all ~3x3 gridcells that overlap the flight track during a single transect survey. Relative abundance values are annualized to match the year of the survey.
Script: PopEstSide.mdl
This is the model code for the hierarchical Poisson model in Methods Section 2.2. The input data are contained in jags_data.rds.
Outputs from this model include (1) population size estimates for each BCR in every year, and (2) density estimates (eagles/km^2) for each transect in every year.
Script: Run_LinearModel.txt
This is the model code for the hierarchical linear model in Methods Section 2.3. The input data are contained in LinearModel_Data.csv.