Feral swine (Sus scrofa) are present in every county in Florida with an estimated population of over 500,000. They are considered an invasive species and have negative impacts upon native vegetation, agricultural crops, wildlife, and increase the risk of livestock disease transmission. Ongoing collaborative projects have been established to quantify impacts of feral swine on rangelands, and understand economic costs of feral swine to the ranching community. The following questions are being asked:
How dense are feral swine populations?
What are their movement patterns and habitat preferences?
What is their contact rate with livestock or point sources?
How does rooting impact forage production and how are native plant communities shifted from feral swine impacts?
What diseases are shared among feral swine, wildlife and livestock and what is the risk of transmission?
How often do feral swine depredate ground nesting birds (Turkey, Bob-white Quail and Sand-hill Cranes)?
Do wildlife prefer areas where feral swine have been controlled and excluded?
What are the best control and exclusion methods?
Feral Swine Population Dynamics, Response to culling, Space-use patterns and Behavioral Interactions in Florida
Boughton Lab in Collaboration with USDA-APHIS
The overall objective of this research is to estimate feral swine (Sus scrofa) population sizes and densities, habitat use, behavioral interactions and determine effort, costs and effectiveness of removal efforts. Specific objectives are:
Estimate population size and densities of feral swine using a mark re-sight method with an array of game cameras.
Estimate space-use patterns, habitat selection and home range size using GPS collars.
Collect biological samples to examine disease and genetic relationships.
This is a 5 year study being conducted on Buck Island Ranch in Highlands County.
As of October 2018, we have tagged and sampled over 736 feral swine, deployed and recovered 94 GPS collars composed of over 580,000 GPS data location points.
An array of 44 motion-activated game cameras have been established to re-sight pigs, one camera every 1 km2.
Pigs are trapped and individually marked with ear tags and/or collar bands and some with GPS collars.
Pigs are removed alternately from the north or south of the ranch by trapping.
Biological samples are collected from all pigs, including several repeated samples to examine diseases and bacterial pathogens.
This study will fill data gaps on the biology, habitat use, and behavior of wild pigs in Florida, as well as provide information on how wild pigs respond to management activities.
Computers Used to Identify Wildlife in Game Camera Pictures
By: Dr. Raoul Boughton
The artificial-intelligence breakthrough, is now detailed in a paper published in the scientific journal Methods in Ecology and Evolution, and is described as a significant advancement in the study and conservation of wildlife. The computer model is now available in a software package for Program R, a widely used programming language and free software environment for statistical computing. A computer model developed at the University of Wyoming by researchers, in collaboration with USDA’s National Wildlife Research Center, Arizona State University, California’s Tejon Ranch Conservancy, the University of Georgia, the University of Florida, Colorado Parks and Wildlife, the University of Saskatchewan, and the University of Montan, has demonstrated remarkable accuracy and efficiency in identifying images of wild animals from camera-trap photographs in North America.
“The ability to rapidly identify millions of images from camera traps can fundamentally change the way ecologists design and implement wildlife studies,” says the papers lead authors Michael Tabak and Ryan Miller, both of the U.S. Department of Agriculture’s Center for Epidemiology and Animal Health in Fort Collins, Colo.
The study builds on UW research published earlier this year in the Proceedings of the National Academy of Sciences (PNAS) in which a computer model analyzed 3.2 million images captured by camera traps in Africa by a citizen science project called Snapshot Serengeti. The artificial-intelligence technique called deep learning categorized animal images at a 96.6 percent accuracy rate, the same as teams of human volunteers achieved, but at a much more rapid pace. In the latest study, the researchers trained a deep neural network on Mount Moran, UW’s high-performance computer cluster, to classify wildlife species using 3.37 million camera-trap images of 27 species of animals obtained from five states across the United States. The model then was tested on nearly 375,000 animal images at a rate of about 2,000 images per minute on a laptop computer, achieving 97.6 percent accuracy — likely the highest accuracy to date in using machine learning for wildlife image classification.
In our December Ona Highlight https://www.youtube.com/watch?v=u2kEoOoNeJU, I explain how one set of millions of images taken on Buck Island Ranch was included in the computer model development to specifically allow increased understanding of feral swine populations, and rapid identification of pictures containing swine.
Tabak, MA, Norouzzadeh, MS, Wolfson, DW, et al. Machine learning to classify animal species in camera trap images: Applications in ecology. Methods Ecol Evol. 2019; 10: 585– 590. https://doi.org/10.1111/2041-210X.13120
Along with photographs of marked and unmarked swine, the motion-activated cameras record photographs of other wildlife species. Although feral swine are the focus of this study, the array could very easily be used to study other wildlife in a similar manner, for example, estimating white-tailed deer abundance.
Impacts of Feral Swine on Florida Rangeland Amphibian Communities
Wesley Anderson, PhD Student
The overall objective of this research is to determine direct and indirect impacts wild pigs are having upon amphibians and seasonal wetlands on Florida rangelands. Specific objectives are:
Quantify wild pig habitat disturbance in seasonal wetlands using a UAV (quadcopter drone)
Compare tadpole species richness, growth, and stress hormone levels between seasonal wetlands of varying disturbance levels and correlate with environmental data
Estimate annual survival and/or occupancy of large aquatic salamanders in seasonal wetlands of varying disturbance levels to examine impacts of direct predation by pigs
Conduct DNA metabarcoding analyses on pig fecal samples to further investigate predation
Study is being conducted in 36 seasonal wetlands on Buck Island Ranch in improved and semi-native pastures.
UAV flights began in 2016 and are conducted four times a year from Nov.- May.
Preliminary data suggest tadpoles occur at lower relative abundances in rooted areas.
Greater sirens are the most common salamander and found in wetlands within both pasture types.
Lesser sirens appear restricted to wetlands in semi-native pastures.
Pig fecal samples for diet analyses collected across an entire year showed amphibians did not occur frequently, however, sirens were consumed during the dry season which may suggest pigs are unearthing them while rooting in wetlands that have dried down.
When complete these complementary studies will provide a clearer picture of the impacts wild pigs are having on amphibians and wetlands across this subtropical ecosystem.
Potential Resource Competition between Feral Swine and White-tailed Deer for Food Resources on Florida Rangelands
The overall objective of this research is to test for the following three types of behavioral changes indicating niche differentiation of deer caused by competition with swine:
Avoidance of swine by deer through time (Temporal shift)
Avoidance of swine by deer through space (Spatial shift)
Reduction in limited resources available to deer due to higher swine activity and feeding ability (Differential resource use efficiencies)
Camera traps were placed over an area of ~20 km2 in two studies, the first at naturally occurring acorn producing oak trees and the second using human-placed supplemental food.
Results of Study at Acorn Producing Oak Trees:
Deer increased their diurnal activity and reduced their crepuscular activity to avoid highly nocturnal feral swine.
Average daily white-tailed deer site use was negatively correlated with average daily feral swine use of same site (p<0.001), suggesting that the two species utilized spatially separate trees.
Feral swine dominated the highest producing trees, and that white-tailed deer utilized the lowest producing trees.
These relationships suggest that white-tailed deer spatially and temporally avoid feral swine, and that increased feral swine activity at the best oak trees may limit mast availability for white-tailed deer.
Results of Study at Supplemental Food Sources:
Feral swine dominated supplemental feeding stations and decreased bait availability for white-tailed deer, as evidenced by the significantly higher (p<0.001) visitation rates of swine to baited swine-only treatments than deer to baited deer-only treatments.
Through these two studies using food resources we have provided evidence that feral swine limit both food availability and space for white-tailed deer, leading to a reduction in the realized niche of white-tailed deer on Florida rangelands caused by competition with invasive feral swine. The reduced availability of food and spatial resources for white-tailed deer caused by feral swine competition may have cascading negative effects for deer populations in the long-term if populations of feral swine continue to increase.
These projects were made possible by funding and collaboration with the USDA Veterinarian Services and Renewable Resources Extension Act. Collaborators: Ryan Miller, Dan Grear, USDA. Matt Farnsworth, Jesse Lewis, CSP. Samantha Wisely, Marty Main, UF-WEC. Contributors: The MacArthur Agro-ecology Research Center.