https://doi.org/10.5061/dryad.ffbg79d4n

Description of the data and file structure

We conducted a comparative study of northern myotis populations in the United States to explore the mechanisms that may be supporting species survival following the arrival of white-nose syndrome (WNS). Here, the core of the species range is represented by mainland populations and the periphery is represented by the coastal island populations of Long Island, New York, and Martha’s Vineyard and Nantucket, Massachusetts. First, we presented mainland colony declines from hibernacula across the species range and we compared capture probabilities of island and mainland bats to assess the severity of population decline in post-WNS years. We quantified seasonal patterns of Pseudogymnoascus destructans prevalence and infection intensity between the islands and mainland. We compared spring body weights of island bats to evaluate differences in body condition between pre- and post-WNS years and assessed the degree of post-WNS wing damage due to prior infection. We modeled the predicted length of hibernation for the islands and mainland in relation to local climate conditions and measured seasonal bat activity on the islands to document the periods of hibernation entry and emergence. In addition, we compared the temperature and humidity of island and mainland hibernacula over the winter season to assess if microclimates differ by location. Our data is split into 12 .csv files (described below) and Program R is required to run the 2 .R files containing code to conduct analyses. 

Files and variables

File: CapProb_forpub.csv

Description: This file contains mist-netting survey results for northern myotis on the U.S. mainland and islands of Long Island, New York, and Martha’s Vineyard and Nantucket, Massachusetts. The dataset was used to calculate capture probabilities for each location (mainland vs. islands) in years since WNS arrival. Each row represents a specific survey night by year.

Variables

• Location: categorical variable, representing each island or mainland state (LI=Long Island, MV=Martha’s Vineyard, N=Nantucket, NY=New York, WI=Wisconsin)
• Loc2: categorical variable, representing general location (Island or Mainland)
• YSW: years since WNS arrival for each survey
• Year: survey year
• Date: date of survey (day-month)
• Cap: binomial (0/1) presence/absence of MYSE for each survey night
• MYSE: number of northern myotis captured each survey night
• Net_events: number of surveys conducted on each date
• Nets: number of nets used per survey

File: DiseaseSeverity.csv

Description: Data collected from captured northern myotis during spring between 2017–2021 on the islands of Long Island, New York, and Martha’s Vineyard and Nantucket, Massachusetts. This dataset was used to compare two disease severity metrics, wing-damage index (WDI) and UVscore, and each row represents one individual bat.

Variables

• Date: the date of capture (month/day/year)
• Island: categorical variable, location of capture (LI=Long Island, N=Nantucket, MV=Marthas Vineyard)
• WingScore: a standardized 4-point wing-damage index (WDI; Reichard and Kunz 2009) to assess incidence and severity of wind damage after exposure to WNS (0 to 3)
• Uvscore: ultraviolet (UV) light score, measured as the area of the wing with orange fluorescence, to document degree of infection (0 = none, 1 = 1–10%, 2 = 10–50%, 3 = >50% of wing area; Turner et al. 2014)
• Band: band identification number of captured bats
• Wyear: categorical variable, winter year (essentially the year of collection as all individuals used in this analysis were captured during spring)
• Recap: categorical variable representing whether each bat was a recapture, indicated by prior banding (N=no, Y=yes)
• OrigCap: year of original capture for recaptured individuals
• Load: Pseudogymnoascus destructans fungal load (in ng of DNA, und=represents negative sample); presence of *P. destructans *was tested for using qPCR, and a standard curve was used to determine the quantity of fungus
• Uvprop: proportion of individuals wing that fluoresced orange under UV light, wing images were analyzed using ImageJ software
• Notes: any additional notes regarding the captured individual

File: EnvaDat_forpub.csv

Description: This data file represents samples from anthropogenic hibernacula collected on the islands of Long Island, New York, and Martha’s Vineyard and Nantucket, Massachusetts. We used this dataset to compare prevalence of P. destructans and the extent of the environmental reservoir (measured by fungal load) for one island hibernaculum that had the largest sample size.

Variables

• Date: date of sampling (month/day/year)
• Season: categorical variable, season of sampling (Hibernation, Emergence)
• Month: month sampling was conducted
• YrMonth: combination of month and year of sampling
• Year: winter year of sampling
• SeasonYear: combination of season and winter sampling period
• Location: categorical variable of island where samples were collected (Marthas Vineyard MA, Nantucket MA, Long Island NY)
• Island: island name abbreviation (Main=Mainland, LI=Long Island, N=Nantucket, MV=Marthas Vineyard)
• Site: sampling site name
• Prev: binary (0/1) prevalence of P. destructans
• Pdload: P. destructans fungal load (ng of DNA); NA indicates negative sample
• Log10Load: the log10 P. destructans fungal load
• Details: any additional notes on the sampling locations

File: epi_dec_forpub.csv

Description: This file contains data for northern myotis hibernating colonies with counts before and after WNS detection at 74 sites across 8 US states (data obtained from state government reports or publicly available from the USGC North American Bat Program). To document population declines at these mainland hibernacula, we estimated the population growth rate ( using the single most recent pre-WNS census as a proxy for colony size prior to WNS arrival.

Variables

• county: categorical variable, US county where site is located
• site: categorical variable, anonymized hibernacula site names
• species: species scientific name abbreviation (MYSE=Myotis septentrionalis)
• clustersize: the colony count for that associated survey year
• state: categorical variable, US state where site is located
• wyear: categorical variable, year of winter sampling
• yoa: categorical variable, year of pathogen arrival
• ysw: categorical variable, year since WNS arrival
• preN: the most recent pre-WNS arrival colony count
• preNyear: categorical variable, the year of the most recent pre-WNS arrival colony count
• surveydiff: number of years between WNS arrival and the most recent pre-WNS arrival colony count
• lambda.post: population growth rate post-WNS arrival
• wtemp: winter surface temperature at each hibernacula (°C)
• postwns_surv: the number of surveys conducted for each site in post-WNS years           
• surv_including yoa: the number of surveys conducted for each site in post-WNS years including the year of pathogen arrival

File: MYSEweights_forpub.csv

Description: Data collected from captured northern myotis on the islands of Long Island, New York, and Martha’s Vineyard and Nantucket, Massachusetts. Pre-WNS data was collected on Martha’s Vineyard (1997–1998; Buesch 1999) and Nantucket (2017–2020) and post-WNS data was collected on Long Island (2017–2019) and Martha’s Vineyard (2018–2021). This dataset was used to compare individual weights from bats captured prior to the arrival of WNS on each island and post-WNS arrival, and each row represents one individual bat.

Variables

• WNS: categorical variable representing the time period of data collection, either pre- or post-WNS arrival (Pre, Post)
• DateBanded: the date of capture
• Species: species scientific name abbreviation (MYSE=Myotis septentrionalis)
• Age: categorical variable, age class (A=adult, J=juvenile, unk=unknown)
• Sex: categorical variable (M=male, F=female)
• Weight: weight of bats captured (grams)
• Capture Method: all data were collected from bats captured in mist-nets (Mist Netting)
• Season: categorical, all data used in this analysis were from bats captured during Spring
• Loc2: categorical variable, represents the island of capture (LI=Long Island, N=Nantucket, MV=Marthas Vineyard)
• Band #: band identification number of captured bats
• ReCap: categorical variable representing whether each bat was a recapture, indicated by prior banding (N=no, Y=yes)

File: MYSEactivity.csv

Description: This data file contains pooled northern myotis activity (measured by the number of confirmed acoustic files) by day of year for each survey period.

Variables

• Jdate: julian date
• period: survey period (e.g. 1718=data collected Sept 1, 2017 – May 31, 2018)
• total_act: pooled northern myotis files by day of year
• total_det: the number of detectors operational on that survey day
• act_det: the pooled northern myotis files/detectors operational for each survey day

File: temp_compare.csv

Description: This file contains internal temperature data for northern myotis island hibernacula and winter surface temperatures. Each row represents a temperature record for each day during the survey period. Island hibernacula were monitored with iButtons (model DS1923 Hygrochron), and data were collected every two hours. Daily maximum and nightly minimum surface temperatures were collected at weather stations nearest to island hibernacula (National Oceanic and Atmospheric Administration (NOAA); NCEI DSI 3505) to assess the relationship between internal hibernacula and external ambient temperature.

Variables

• Site: hibernacula site name
• Site1: unique site and location abbreviations
• Island: categorical variable, island name abbreviation (LI= Long Island, MV=Martha’s Vineyard, N=Nantucket)
• Year: winter year of data collection (e.g. year 1819 means data were collected for autumn 2018 through spring 2019)
• DOY: day of year
• Month: month of data collection
• Temp: daily average internal temperature for island hibernacula (°C)
• RH: daily average percent relative humidity for island hibernacula
• TaAvg: the daily average ambient (surface) temperature (°C)
• Tdiff: difference between internal and ambient temperature (positive values indicate that internal temperature is greater than ambient)

File: LocSeasonalTemps.csv

Description: This data file contains ambient temperature records obtained from weather stations nearest to mainland or island hibernacula (downloaded from the National Climatic Data Center, NOAA, 2023). Temperatures have been averaged across survey periods, and each row represents the average temperature for each day of the year during the sampling period (Sept 1 – May 31). The nightly minimum temperature is used to estimate the days that fall below 0°C, indicating the approximate hibernation duration.

Variables

• Jdate: julian date
• DOY: day of year (beginning first day of survey period, Sept 1)
• Location: categorical variable (Main=mainland, Island=island)
• Site: site name
• DayMin: daily minimum temperature, °C
• NightMin: nightly minimum temperature, °C

File: MYSE_PdData_forpub.csv

Description: Data collected from captured northern myotis on the mainland U.S. and islands of Long Island, New York, and Martha’s Vineyard and Nantucket, Massachusetts to assess prevalence and infection intensity of WNS. This dataset was used to compare disease dynamics between the mainland and islands, and each row represents data collected from one individual bat.

Variables

• Date: date of data collection(month/day/year)
• DOY: day of year
• Month: month of data collection
• Pdate: modified time variable beginning the first day of autumn sampling and expressed in units of partial months
• Year: year of data collection
• Wyear: winter year of data collection
• YOA: year of WNS arrival
• YSW: year since pathogen arrival
• Group: categorical variable indicating whether the data was collected during the Invasion (years 0–3) or Post-Invasion (4+) disease phase
• Season: categorical variable, season of data collection (fall, spring, summer)
• Location: categorical variable (Island, Mainland)
• Site: categorical variable, separates islands (Main=Mainland, LI=Long Island, N=Nantucket, MV=Marthas Vineyard)
• SurveySite: anonymized site names where bats were captured
• State: U.S. state of data collection
• Sex: categorical variable (M=male, F=female)
• Age: categorical variable, age class (A=adult, J=juvenile, unk=unknown)
• Weight: individual weights of bats captured (grams)
• Prev: binary (0/1) *P. destructans *prevalence (0=negative sample, 1=positive sample)
• Pdload: P. destructans fungal load (ng of DNA); NA indicates negative sample
• Log10Load: the log10 P. destructans fungal load

File: Seasonal_long.csv

Description: This file contains northern myotis acoustic activity by day of the year and weather variables (downloaded from the National Climatic Data Center, NOAA, 2023). This dataset is used to model the probability of northern myotis presence in relation to ambient conditions, and show northern myotis activity during the winter months on the islands.

Variables

• Jdate: julian date
• Date: sampling date (month/day/year)
• Month: sampling month
• Day: day of sampling
• Year: sampling year
• Island: categorical variable, island name abbreviation (LI= Long Island, MV=Martha’s Vineyard, N=Nantucket)
• Activity: total number of northern myotis calls recorded for each day of the year
• Pres: binary (0/1) variable, presence/absence of northern myotis
• T_Min_N: nightly minimum temperature, °C
• T_Max_D: maximum day temperature, °C
• WindAvg_N: nightly average wind speed (mph)
• %MFI: percent of moon face illuminated
• Precip_Hours: total hours of nightly precipitation
• Period: survey period (e.g. 1718=Sept 1, 2017 – May 31, 2018)

File: SiteCompare.csv

Description: This file contains microclimate data for island and mainland northern myotis hibernacula (temperature and relative humidity records). This dataset was used to compare island vs. mainland microclimates, and each row represents a daily record for each site.

Variables

• Site: hibernacula site name
• Site1: unique site and location abbreviations
• Location: categorical variable (Island or Mainland)
• Location1: specific location (Islands: LI= Long Island, MV=Martha’s Vineyard, N=Nantucket, or mainland: New York=New York State)
• Date: date (day-month)
• Year: year of data collection
• DOY: the day of year
• Month: the month of data collection
• Temp: temperature recordings from microclimate loggers (°C daily averages; mainland data were recorded every 3 hours using HOBO H08 Pro Series, island data were recorded every 2 hours using iButton model DS1923 Hygrochron)
• RH: percent relative humidity recordings from microclimate loggers (daily averages; mainland data were recorded every 3 hours using HOBO H08 Pro Series, island data were recorded every 2 hours using iButton model DS1923 Hygrochron)
• RHscaled: scaled relative humidity values that range from 0–100%

File: ActivityPercent.csv

Description: This file contains summarized acoustic data across all sampling years for northern myotis. Data were pooled from all operational detectors each night and the total number of calls by season were summed for each survey year.  We then calculated the total nightly activity/detectors for each season with all years combined. Relative activity was converted to a percentage of total activity by season for autumn (Sept 1 – Dec 31) and spring (Jan 1 – May 31) and subtracted from 100%.

Variables

• Jdate: julian date
• 1718total: sum of northern myotis calls by date during the 2017-2018 survey period
• 1819total: sum of northern myotis calls by date during the 2018-2019 survey period
• 1920total: sum of northern myotis calls by date during the 2019-2020 survey period
• Sum: the sum of northern myotis calls by date across all survey periods
• PercentT: the percent of total northern myotis calls by date out of 100
• CumTotal: the cumulative percent of northern myotis calls subtracted from the total (100%)

Code/software

Program R is required to run the following files:

R code file 1: IslandMYSE_WNS_forpub.R

This code runs analyses to compare prevalence and infection intensity of northern myotis between island and mainland populations, assess disease severity of island bats, compare hibernacula microclimates between the islands and mainland, document environmental prevalence and the extent of the environmental reservoir in island hibernacula, and calculate the population decline on the mainland and capture probabilities for the mainland and islands following WNS arrival.

Session Info

R version 4.3.1 (2023-06-16)

Platform: aarch64-apple-darwin20 (64-bit)

Running under: macOS Ventura 13.2

Packages

attached base packages: grid; stats; graphics; grDevices; utils; datasets; methods; base    

other attached packages:

 [1] EnvStats_2.8.1    gridExtra_2.3     cowplot_1.1.1     DHARMa_0.4.6      glmmTMB_1.1.7   

 [6] knitr_1.43        AICcmodavg_2.3-2  rcompanion_2.4.30 car_3.1-2         sjPlot_2.8.14   

[11] effects_4.2-2     carData_3.0-5     scales_1.2.1      ggpubr_0.6.0      lme4_1.1-33     

[16] Matrix_1.5-4      multcomp_1.4-25   TH.data_1.1-2     MASS_7.3-60       survival_3.5-5  

[21] mvtnorm_1.2-2     forcats_1.0.0     stringr_1.5.0     dplyr_1.1.3       purrr_1.0.1     

[26] readr_2.1.4       tidyr_1.3.0       tibble_3.2.1      ggplot2_3.4.4     tidyverse_2.0.0 

[31] lubridate_1.9.2   emmeans_1.8.7   

R code file 2: hiber_duration.R

This code runs analyses to estimate the hibernation duration for island northern myotis populations on Long Island, New York, and Martha’s Vineyard and Nantucket, Massachusetts. The first analysis uses landscape bat acoustic data collected Sept 1 – May 31, 2017–2020 to determine the window of entry into and emergence from hibernation. The second analysis uses ambient conditions (minimum nightly temperature) to estimate the days below 0°C for each location (mainland vs island hibernacula) and quantify the approximate hibernation length. Additionally, this file contains code to run analyses that assess the effects of ambient conditions on seasonal northern myotis presence.

Session Info

R version 4.3.1 (2023-06-16)

Platform: aarch64-apple-darwin20 (64-bit)

Running under: macOS Ventura 13.2

Packages

attached base packages: grid; stats; graphics; grDevices; utils; datasets; methods; base    

other attached packages:

 [1] tidymv_3.4.2      mgcv_1.8-42       nlme_3.1-162      gtable_0.3.4      ggeffects_1.2.3 

 [6] BBmisc_1.13       EnvStats_2.8.1    gridExtra_2.3     cowplot_1.1.1     DHARMa_0.4.6    

[11] glmmTMB_1.1.7     knitr_1.43        AICcmodavg_2.3-2  rcompanion_2.4.30 car_3.1-2       

[16] sjPlot_2.8.14     effects_4.2-2     carData_3.0-5     scales_1.2.1      ggpubr_0.6.0    

[21] lme4_1.1-33       Matrix_1.5-4      multcomp_1.4-25   TH.data_1.1-2     MASS_7.3-60     

[26] survival_3.5-5    mvtnorm_1.2-2     forcats_1.0.0     stringr_1.5.0     dplyr_1.1.3     

[31] purrr_1.0.1       readr_2.1.4       tidyr_1.3.0       tibble_3.2.1      ggplot2_3.4.4   

[36] tidyverse_2.0.0   lubridate_1.9.2   emmeans_1.8.7