https://doi.org/10.5061/dryad.gqnk98svz

Description of the data and file structure

Data include:

ticks.hind: counts of individual winter ticks, Dermacentor albipictus, found along transects through the hair on the rumps of 750 live-captured moose, Alces alces, during winter (December 13%, January 52%, February 20%, March 16%) in 5 western US states (state).

transect.cm.hind: length of the transects along which hair was parted and searches for ticks were conducted

moose ID: individual moose that were captured

capture date: date of the capture events

density.category: We used expert opinion to characterize individual moose as being located within areas of low (< 0.21), medium (between 0.21 and 0.41), or high (> 0.41) adult females per km2, following quantitative delineation of moose densities with focus on adult females by Schmidt et al. (2007). We imagined overlaying a 5 x 5 km neighborhood surrounding each winter capture location and the biologist with field experience in each area estimated how many adult female moose would be found in that neighborhood; our low density category corresponded to ≤5 cows sharing that 25-km2 neighborhood, medium density to 6–10 cows sharing that neighborhood, and high density to >10.

migratory: We treated seasonal migration as a proxy for spatial overlap of individual moose during the tick drop-off and questing seasons, and assigned ‘resident’, ‘migratory’, or ‘unknown’ categorical depictions of migration behavior according to project-specific spatial analyses or subjective assessments of seasonal movements. We assumed that migration behavior subsequent to capture was an adequate proxy for behavior during the year prior to capture, when measured tick loads would have accumulated on each moose.

shared.winter.range: presence of other ungulates on shared winter range; We included a binary categorical variable concerning whether moose shared their winter range with deer or elk populations of equal or greater abundance based on field experience in each study area

Climate metrics: A suite of climate metrics are also included for each record specific to that capture location and to the specific year and season prior the capture event. These are labelled according to the biological season prior to moose captures during which they were summarized (q=questing [fall, 15-Sep to 30-Nov]; d=dropoff [spring, 15-Mar to 30-Apr]; s=summer [1-Aug to 14-Sep]) as described in the manuscript. 

qSnowProp & dSnowProp: proportion of days with measurable snow across questing and drop-off periods according to daily gridded measures of normalized difference snow index (NDSI) detections of snow cover from MODIS version 6 MOD10A1 (Terra) and MYD10A1 (Aqua) data at a 500 m x 500 m resolution (Hall et al. 2006). We merged Terra and Aqua NDSI data according to the maximum snow cover value measured between each platform per day (Tran et al. 2019) and used a threshold NDSI value of 0.1 to distinguish snow from non-snow, following Hussainzada et al. (2021).

qSWEmid, qSWEend, dSWEmid, dSWEend: the amount of snow using estimates of snow water equivalent (SWE) during the mid-point and end-point of the both the drop-off and questing periods using daily SWE data from Daymet V4R1 at 1 km x 1 km resolution.

qTempMean, dTempMean, sTempMean: mean temperature; We quantified temperature as the mean of daily low temperatures across each of the late summer, questing, and drop-off periods using Daymet version 4 revision 1 data at a 1 x 1 km resolution (Thornton et al. 2022), accessed via the R package (Hufkens et al. 2018, R Core Team 2018).

qTempDays0, qTempDaysm20, dTempDays0, dTempDaysm20: To capture potential threshold effects of cold temperatures during the questing and drop-off seasons, we also used the same Daymet data to measure the number of days with low temperatures below thresholds of 0˚ (Powers and Pekins 2020) and -20C˚ (Holmes et al. 2018).

qRHmin, qRHmax, sRHmin, sRHmax: We quantified daily minimum and maximum relative humidity during the late summer and questing seasons using daily minimum and maximum estimates of vapor pressure deficit and temperature produced by the PRISM (parameter-elevation regressions on independent slopes model) Climate Group at 4 km x 4 km resolution (PRISM Climate Group 2023) and following equations of Daly et al. (2015).

sPrecip: cumulative precipitation; we used daily precipitation data from PRISM, also at 4 km x 4 km resolution, to quantify the total precipitation during the late summer period.

landcover: We characterized vegetation type at each capture site according to year-specific land cover data using the International Geosphere Biosphere Programme (IGBP) global vegetation classification of the moderate resolution imaging spectroradiometer (MODIS) land cover product (MCD12Q1) at a 500 m x 500 m resolution, following Liang et al. (2015). We simplified the 6 land cover types identified in the study areas (evergreen forest, woody savanna, savanna, grassland, cropland, and urban) into 2 land cover categories by aggregating together the first three (“forest”) and the latter three (“open”).

***_2*: Varibles with a "2" after the name are simply quadratic (squared) terms for some variables (_2)

qSnowPropavg, dSnowPropavg, qSWEmidavg, qSWEendavg, dSWEmidavg, dSWEendavg, qTempMeanavg, dTempMeanavg, sTempMeanavg, qTempDays0avg, dTempDays0avg, qRHmaxavg, qRHminavg, sRHmaxavg, sRHminavg, sPrecipavg: These variables are 10-year averages of the same metrics described above, averaged across the entire study period, 2013-2022, as opposed to the variables above which were estimated specific to the year prior to data collection for each tick count measurement.