Michelle Gooch

The Effects of Urbanization on Frog Populations

Acris crepitans (Northern Cricket Frog) calling at a pond

Golf-course pond in a local subdivision

I’m currently studying the effects of urbanization on summer-breeding frog populations. During the summer of 2004, I worked with Aubrey Heupel of Iowa State University to conduct frog-calling surveys at 35 ponds in Mecklenburg County. We located the ponds using digital aerial photos of Mecklenburg County in ArcView. We tried to find ponds that represented a wide variety of urbanization-levels. For example, some of our ponds were completely surrounded by forest, others had some surrounding forest, agricultural land, or development, while others were in the middle of subdivisions or golf courses.
We divided our ponds up into routes of 6-11 ponds per route. We would run one route per night, driving to all of the ponds between the hours of 9 p.m. and 1:30 a.m. At each pond, we would listen for 10 minutes and record each species of frog that we heard calling. We would also rank the abundance of each species on a scale of 1-3. We recorded other variables such as weather, air temperature, water temperature, and humidity at each site.
We also performed local assessments at each pond; this included visually estimating the percentage of the pond edge that was mowed, grassland, shrub, forest, or mud. We also recorded habitat characteristics such as the presence of algae, livestock, fountains, or emergent vegetation. Using ArcView, we created 200 m buffer zones around each pond. We could then measure the percentage of these buffer zones taken up by development, grassland, agriculture, or forest. In this way we characterized the amount of urbanization surrounding the ponds.
We went on to analyze our data using logistic regression models. We are currently in the process of further analyzing the data and writing up a paper to be submitted for publication. While Aubrey heads up this paper back in Iowa, I am currently working on another paper involving detection probability, using our summer data. This paper will be the final outcome of my class in Investigations in Herpetology with Dr. Dorcas.
While numerous studies have been done that utilize frog-calling surveys, very few of these studies have taken into account the probability of detection for certain species. Detection probability data is important because oftentimes a species will be recorded as “absent” at a pond, when it is in fact present. This can happen for many reasons- sometimes a species will not be calling on a specific night due to weather conditions. The amount of time spent listening at each pond and number of surveys done per season can also drastically affect the results of a study. Using the computer program PRESENCE (created by Darryl MacKenzie of Proteus Wildlife Research Consultants), I plan to determine the detection probabilities for each frog species using calling surveys that last 10 minutes, 5 minutes and 3 minutes. I have been able to manipulate our summer data in an Excel spreadsheet to determine how it would have been affected if we had used shorter surveys.
This is a study-in-progress and should be completed by the end of fall semester, 2004. Hopefully I will be submitting my paper to a peer-reviewed journal for publication. I also hope to present my results at the meeting of the Association of Southeastern Biologists in November

Aubrey & I having way too much fun listening for frogs!
Aerial map of Mecklenburg County with ponds shown
Pond with 200 m buffer and habitat outlines as done in ArcView 3.2
Taking water temperature


E-mail me with questions