There have been a number of blogs on how deer respond to the rifle season. There is my favorite, Hillside Doe, and our most famous, Buck 8917.
All show changes in deer movements in response to the hunter pressure. Deer knew there was something amiss and altered their movements right before and during the rifle season. It’s fun making movies of their movements, and the movies are even more fun to watch. Those posts generated a lot of discussion, but there isn’t much that can be said from a scientific perspective. A single deer and an opinion do not research make.
But many deer and a complex statistical model do! Are there larger patterns at play? Katie Gundermann, a graduate student at Penn State University, dug deep looking for those larger patterns in her research for her M.S. degree.
Katie had two predictions about how deer might respond to the hunting season. One – they could shift their home range to a different location, as illustrated below (the brown track represents its home range prior to the hunting season and the blue is the result of hunting activity).
Or two – they could contract their home range. That is, only use a portion of their home range like this.
To assess whether deer shift or contract their home range (or don’t!) she used a relatively new tool in ecology called Hidden Markov Models, or HMMs. The models are named after Russian mathematician Andrey Andreyevich Markov (1856–1922) who studied stochastic processes and what is now known as a Markov chain.
Markov models are used whenever the condition at time t depends on time t-1. For example, the probability I will have a scrambled egg for breakfast might depend on whether I had one the day before. If so, a HMM might be able to predict when I will have an egg for breakfast!
Katie decided to use HMMs to see if she could predict how deer respond to hunting pressure in the rifle season. She used a lot of math and computer power to develop and process her models, but let’s just say she took into account factors like distance to road and slope as measures of hunting pressure (we know hunting pressure is less farther from roads and on steeper slopes). Her models also included time of day, hunting season dates, etc. to see if she could detect a change in deer movements.
She used data from 38 deer monitored for one or more hunting seasons. In total, she had 54 deer-years of data.
So what was your wager? Shift or Shrink?
The winning bet was shrink! Most deer exhibited some kind of home-range contraction, like the deer on the left below (the red locations are related to the rifle season). Some deer did shift their home range, but much less likely. However, less than half (43%) of deer exhibited evidence of home-range contraction due to hunting pressure.
Predicting deer movement is the holy grail of hunting. It is something every hunter seeks. However, I wonder if they would be happy if we actually found it. Would we still call it hunting?