Supporting data for assessing impacts of satellite GPS transmitters on survival, nesting propensity, and nest success of greater sage-grouse

Stevens, Bryan1 ; Conway, Courtney2; Tisdale, Cody1; Denny, Kylie1; Meyers, Andrew1; Makela, Paul3

Published Dec 08, 2023 on Dryad. https://doi.org/10.5061/dryad.573n5tbf3

Data files

Dec 08, 2023 version files 265.48 KB

Abstract

Telemetry technology and data are commonly used to study behavior and demography of wildlife. Satellite-based, global positioning system (GPS) telemetry allows researchers to remotely collect a high volume of fine-resolution animal location data but may also come with hidden costs. For example, recent studies suggested GPS transmitters attached via backpacks may reduce survival of greater sage-grouse (Centrocercus urophasianus) relative to very high frequency (VHF) telemetry transmitters attached via collars. While some evidence suggests GPS backpacks can reduce survival, no studies examined their effects on sage-grouse breeding behavior and success. We compared survival, breeding behavior, and nest success for sage-grouse hens marked with either VHF collars or GPS backpack transmitters in central Idaho, USA. GPS backpacks reduced spring-summer survival relative to VHF collars, yet GPS backpacks did not consistently affect nest success or the likelihood or timing of nest initiation relative to VHF collars. Daily nest survival varied annually and with timing of nest initiation and nest age, but marginal effects of transmitter type were statistically insignificant, and interactions between transmitter type and study year were inconsistent. These results demonstrate the effect of GPS backpacks on sage-grouse survival but also suggest GPS backpacks do not appear to affect components of fecundity.

README: Supporting data for assessing impacts of satellite GPS transmitters on survival, nesting propensity, and nest success of greater sage-grouse

https://doi.org/10.5061/dryad.573n5tbf3

The data files contain data used to conduct the following analyses: 1) multi-state modeling of daily mortality probabilities for hen sage-grouse in program MARK (MultiState.inp), 2) nest survival modeling to estimate daily nest survival probabilities for greater sage grouse in program MARK (NestSurvival.inp), 3) mixed effects logistic regression modeling of nesting propensity for hen sage-grouse (i.e., proportion of females initiating a nest) in program R (nest_propensity.csv), and 4) mixed effects Gaussian regression modeling of timing of nest initiation for hen sage-grouse in program R (NestTiming.R). In addition, analysis scripts, written in R language, are provided for analyses 3 (LogisticReg.R) and 4 (GaussianReg.R) above. These R scripts are self contained and annotated within to explain each piece of code.

Description of the data and file structure

Description of the data and file structure for file MultiState.inp for multi-state model analyses in MARK

Each row of data represents a single encounter history observation and the corresponding covariates.

The first column of data is a string of data for n = 154 encounter occasions, with each observation consisting of either: "A" (observed alive), "D" (observed dead), "." (no attempt to monitor and therefore no data), "0" (attempted to monitor but failed to detect).

Columns 2-4 represent animal group number, transmitter type (0 = VHF transmitter, 1 = PTT transmitter), and age of hen (1 = yearling, 0 = adult). Columns 5-9 represent individual covariates (binary indicator variables) for each corresponding year of study (year 2016 is indicated by all other year columns being set to 0). All remaining columns contain time-specific individual covariate data for each observation day (days 1-153), which were used to fit the linear and quadratic time effects to model changes in daily mortality risk across the spring-summer period.

Description of the data and file structure for file NestSurvival.inp for nest survival model analyses in MARK

Nesting data are organized for analysis according to two groups: 1) hens marked with VHF transmitters, and 2) hens marked with PTT transmitters.

Each row of data represents a single nest and its corresponding nest check data and covariates. The first column is a nest identification code that does not get used at all in the statistical analysis; it's there simply to indicate which nest is being observed.

Columns 2-5 contain the standard nest survival data (going left to right): day of the nesting season on which the nest was found, last day the nest was checked when alive, last day the nest was checked, and nest fate (0 = successful, 1 = depredated). Column 6 is all 1s and indicated the number of nests each row represents. Columns 7-13 are individual covariates for each corresponding year of study. Column 14 is ordinal day of nest initiation, and column 15 is nest age on the first observation day of each season (used to model nest age as an individual covariate).

Description of the data and file structure for file nest_propensity.csv for mixed effects logistic regression analyses in R

This csv file contains 6 columns: 1) transmitter type, 2) Year, 3) number of nests initiated (No_nests), and 4-5) total number of hens available for nesting under the conservative (Total_method_1) and liberal (Total_method_2) definitions of hens available for nesting (see methods section in manuscript text).

Description of the data and file structure for file nest_initiation_data.csv for mixed effects Gaussian regression analyses in R

This csv file contains 5 columns of data: 1) individual nest ID, 2) year, 3) ordinal day of nest initiation, 4) transmitter type (VHF or PTT), and 5) age of hen (adult [A], yearling [Y], or not available [NA]).

Sharing/Access information

Code/Software

Code scripts described above are written and executed in the R statistical computing language.