Presenter Information

Kylie BrunetteFollow

Presentation Type

Poster

Faculty Mentor’s Full Name

Chad Bishop

Faculty Mentor’s Department

Wildlife Biology

Abstract

Declining mule deer (Odocoileus hemionus) populations across much of western North America in the past 25 years have inspired many studies examining population dynamics and possible causes of this decline. Here I analyze data retrieved from GPS collars on does of the Piceance basin mule deer herd in northwestern Colorado to test if GPS collar point distributions can be used to infer specific causes of mortality. I hypothesize that spatial movements of a carcass post-mortality differ between different mortality causes. I predict that coyote-caused mortalities will show significantly more movement post-mortem than mountain lion or malnutrition-caused mortalities. I predict that mountain lion and malnutrition-caused mortalities will not differ much from each other due to a lack of movement post-mortem in both cases. To test this, I will analyze GPS data in ArcMap to calculate average distance moved over time pre- and post-mortem, as well as average area moved pre- and post-mortem. I will randomly select one-half of the known mortality dataset to develop a model for determining cause of mortality based on spatial array of points and total distance moved. I will also include habitat type, slope, and timing of mortality as additional covariates in the model. I will then use the other half of the mortality dataset to validate the model. If my hypothesis is correct, studies using GPS collars may be able to more accurately determine cause of death based on the GPS data collected. If, on the other hand, GPS data alone cannot differentiate between mortality causes, then this means any study desiring accurate data on cause of death needs to prioritize reaching mortalities as soon as possible.

Category

Life Sciences

Share

COinS
 
Apr 17th, 11:00 AM Apr 17th, 12:00 PM

Dead Deer Do Tell Tales: Use of GPS Data to Infer Cause of Mortality in Mule Deer

UC South Ballroom

Declining mule deer (Odocoileus hemionus) populations across much of western North America in the past 25 years have inspired many studies examining population dynamics and possible causes of this decline. Here I analyze data retrieved from GPS collars on does of the Piceance basin mule deer herd in northwestern Colorado to test if GPS collar point distributions can be used to infer specific causes of mortality. I hypothesize that spatial movements of a carcass post-mortality differ between different mortality causes. I predict that coyote-caused mortalities will show significantly more movement post-mortem than mountain lion or malnutrition-caused mortalities. I predict that mountain lion and malnutrition-caused mortalities will not differ much from each other due to a lack of movement post-mortem in both cases. To test this, I will analyze GPS data in ArcMap to calculate average distance moved over time pre- and post-mortem, as well as average area moved pre- and post-mortem. I will randomly select one-half of the known mortality dataset to develop a model for determining cause of mortality based on spatial array of points and total distance moved. I will also include habitat type, slope, and timing of mortality as additional covariates in the model. I will then use the other half of the mortality dataset to validate the model. If my hypothesis is correct, studies using GPS collars may be able to more accurately determine cause of death based on the GPS data collected. If, on the other hand, GPS data alone cannot differentiate between mortality causes, then this means any study desiring accurate data on cause of death needs to prioritize reaching mortalities as soon as possible.