Estimating Detection for Gray Wolf (Canis lupus) Pups in Yellowstone National Park
Presentation Type
Oral Presentation
Abstract/Artist Statement
Carnivore restoration has gained attention all over the world for being a solution for bolstering not only individual species, but ecosystem processes such as predation at a level that sustains the structure and function of an ecosystem. Reintroduction of gray wolves (Canis lupus) to Yellowstone National Park (YNP) has been amongst the more innovative ‘natural’ experiments in the world, and successful in bringing these large carnivores back to their native range. While wolf packs are visible traveling in many areas of YNP, estimating detection for young wolves at den sites remains a challenge due to location, cover, and access. Recruitment of pups into the population remains an important factor in population growth and stability, although there has been little work on estimating detection of pups at den sites. This project’s aim is to estimate the validity of pup detection at individual den sites in YNP.
Before beginning analysis of real data, simulated data has been used as the first step to test the feasibility and scope of the project. Simulated data for this project consists of individual wolf packs, months of a biological year (April-May), and particular visits to den sites throughout a month in a three-dimensional array. This arrangement is very similar to data collection methods in YNP. Initial average litter sizes, 4.8, populated the array from previous research (Stahler et al 2013), and survival is fixed at 90% per month.
Results from the N-mixture model with simulated data include: maximum and average litter size of any pack per month, maximum litter size at any point in time, the total pups per month in the entire population, and detection per month. For the total pups in the population by month, the starting value was 26.275 and final value was 10.821, peaking in month three at 38.992. For the average litter size per month, the starting value was 5.254 and final value was 2.169, peaking in month three at 7.810. Detection, varied from a low of 0.583 in month three to a high of 0.925 in month six. By examining total population size (of pups) in the first and last month, survival throughout the year (recruitment) was 41% which is in the realm of biological probability. This model is not overly complicated, it was a great exercise to pick apart what each piece was doing to understand how it will work when more complexity is added.
This work is the first step in understanding how detection changes the perception of survival of young animals, important drivers of population growth and stability. Additionally, this work can be extrapolated to other species that are not as heavily studied as gray wolves in North America.
Mentor Name
Mark Hebblewhite
Estimating Detection for Gray Wolf (Canis lupus) Pups in Yellowstone National Park
UC 332
Carnivore restoration has gained attention all over the world for being a solution for bolstering not only individual species, but ecosystem processes such as predation at a level that sustains the structure and function of an ecosystem. Reintroduction of gray wolves (Canis lupus) to Yellowstone National Park (YNP) has been amongst the more innovative ‘natural’ experiments in the world, and successful in bringing these large carnivores back to their native range. While wolf packs are visible traveling in many areas of YNP, estimating detection for young wolves at den sites remains a challenge due to location, cover, and access. Recruitment of pups into the population remains an important factor in population growth and stability, although there has been little work on estimating detection of pups at den sites. This project’s aim is to estimate the validity of pup detection at individual den sites in YNP.
Before beginning analysis of real data, simulated data has been used as the first step to test the feasibility and scope of the project. Simulated data for this project consists of individual wolf packs, months of a biological year (April-May), and particular visits to den sites throughout a month in a three-dimensional array. This arrangement is very similar to data collection methods in YNP. Initial average litter sizes, 4.8, populated the array from previous research (Stahler et al 2013), and survival is fixed at 90% per month.
Results from the N-mixture model with simulated data include: maximum and average litter size of any pack per month, maximum litter size at any point in time, the total pups per month in the entire population, and detection per month. For the total pups in the population by month, the starting value was 26.275 and final value was 10.821, peaking in month three at 38.992. For the average litter size per month, the starting value was 5.254 and final value was 2.169, peaking in month three at 7.810. Detection, varied from a low of 0.583 in month three to a high of 0.925 in month six. By examining total population size (of pups) in the first and last month, survival throughout the year (recruitment) was 41% which is in the realm of biological probability. This model is not overly complicated, it was a great exercise to pick apart what each piece was doing to understand how it will work when more complexity is added.
This work is the first step in understanding how detection changes the perception of survival of young animals, important drivers of population growth and stability. Additionally, this work can be extrapolated to other species that are not as heavily studied as gray wolves in North America.