Data from: Female moose pregnancy and survival in Colorado
Data files
Oct 02, 2024 version files 86.62 KB
-
MooseMatrixCode.txt
1.13 KB
-
MoosePregnancy.txt
9.88 KB
-
MooseSurvival_AllMortality.txt
36.23 KB
-
MooseSurvival_Natural.txt
36.23 KB
-
README.md
3.14 KB
Abstract
Wildlife management agencies are obliged to provide evidence-based management recommendations to stakeholders. However, the allocation of resources is rarely uniform among species. The consideration of life-history characteristics of moose offers wildlife managers a more robust understanding of population ecology, while also providing insight into potential limiting factors for long-term management. We measured the survival of adult female moose, nutritional condition and pregnancy status of adult female moose, and winter calf-at-heel ratios as a means to document productivity and life-history characteristics of moose in Colorado. We measured these parameters from three distinct geographical regions in Colorado, from 2014-2020. There was very little geographical or annual variation in survival rates, although variation in monthly survival rates was observed. The primary source of mortality of moose was in fall during hunting seasons, whereas natural mortality events typically occurred between January and May. Pregnancy status of moose was best modeled using body condition score, year, and presence of a 6-month-old calf from the previous breeding season. A positive relationship between the percent ingesta-free body fat and moose pregnancy probability was evident.
README: Data from: Female moose pregnancy and survival in Colorado
https://doi.org/10.5061/dryad.gtht76hwg
Description of the data and file structure
Data files to be associated with recently published work on moose in Colorado.
Four files are included.
The data in the first file, "MooseSurvival_AllMortality.txt", were collected from radio-collared moose in Colorado. These data are prepared for analysis using a logistic regression, known-fate, modelling framework in Program MARK. These data account for all known mortalities, attributed to the month during which mortality occurred. Survival is estimated at monthly intervals, corresponding to June from one year through May of the following year. The year of survival is determined by the year in which June occurs. Data are broken into 7 groups, corresponding to the years 2013-2019. The final two covariates, coded as "dummy variables", correspond to spatial regions in Colorado, 0 0 = Northwest Colorado, 1 0 = Northeast Colorado, and 0 1 = Southwest Colorado.
The data in the second file, "MooseSurvival_Natural.txt", were collected from radio-collared moose in Colorado. This dataset deviates from the first survival dataset in that only non-human mortality sources are included for survival estimation purposes. Animals that were harvested by hunters are censored from the dataset at the time of harvest. These data are prepared for analysis using a logistic regression, known-fate, modelling framework in Program MARK. These data account for all known mortalities, attributed to the month during which mortality occurred. Survival is estimated at monthly intervals, corresponding to June from one year through May of the following year. The year of survival is determined by the year in which June occurs. Data are broken into 7 groups, corresponding to the years 2013-2019. The final two covariates, coded as "dummy variables", correspond to spatial regions in Colorado, 0 0 = Northwest Colorado, 1 0 = Northeast Colorado, and 0 1 = Southwest Colorado.
The third dataset, "MoosePregnancy.txt", pertains to moose pregnancy data. These data were derived from captured moose each winter, from hormonal assays from blood samples. This dataset is structured to estimate pregnancy using logistic regression. Data (and associated headers) include a unique animal ID (AnimalID), capture date (CaptureDate), and pregnancy status (Pregnant, 1= pregnant, 2 = not pregnant). Other covariates (and associated headers) include two factors: temporal variation (YrFactor) and regional variation (RgnFactor). Individual covariates (and associated headers) include the presence/absences of a previous year's calf at the time of capture (Calf Presence, 1 = yes, 2 = no), hand palpated body condition score (BCS), and ultrasonic measurements of loin muscle depth (LoinDepth, measured in millimeters) and loin fat (LoinFat, measured in millimeters).
The final file, "MooseMatrixCode.txt", is R code pertaining to the population projection model presented in the associated paper.
Code/software
These data were analyzed using R and Program MARK