Contact Rates Among Feral Swine and Domestic Cattle in Texas: Addressing Vulnerability to Foreign-Animal Diseases

NON-TECHNICAL SUMMARY: Very little is presently known of the potential for `re-emerging disease and pest threats' to be spread among feral swine and domestic livestock. We argue that feral swine in the southern United States represent a potentially intractable reservoir of infection should foreign animal diseases (e.g., foot-and-mouth disease, classical swine fever) ever be introduced to the population. To address this gap in knowledge we will intensively study real-time contact rates among feral swine and cattle in a defined region of south Texas. The results of this study will be used to better prepare for an outbreak of exotic diseases and to better model or predict the scope and scale of an outbreak and the best way to contain it.

<P>

APPROACH: <BR> Objective #1: We will conduct an intensive 1.5 year-long study tracking concurrent movements of cattle and feral swine on a single large ranch in South Texas to better estimate direct and indirect inter-species contact across a range of landscapes within a clearly defined study area. The relation of landscape features, vegetation cover, water availability and seasonal meteorological information will be determined. <BR>Objective #2: This research objective will be met by conducting an additional 0.5 year study on an 3-4 more ranches in the south Texas region. We will shift GPS collar application to animals on other ranches in order to replicate and generalize the findings of Objective 1 to additional landscapes. <BR>Objective #3: This objective will be met by conducting a concurrent survey of local residents to determine if a survey instrument might be useful for rapid assessment of the level of inter-specific contact between cattle and feral hogs in any given region. While retaining the simplicity necessary for rapid deployment and analysis, we will develop and test a somewhat sophisticated instrument, to be delivered to a wider audience in order to determine sightings of feral swine, visible interaction of swine (direct and indirect) with cattle and deer, and simply evidence of swine presence (e.g., through damage). The results of these surveys will be compared to the actual animal contact data to determine correlation (agreement) and any adjustments that might be applied to improve estimation.
<BR><BR>
PROGRESS: 2003/07 TO 2008/06<BR>
OUTPUTS: Results of our work remain of great interest to epidemiologic modelers, veterinarians, cattle producers and other landowners, and animal health policy makers. In particular, there is growing demand for epidemiological modelers to more accurately quantify the rates at which infectious disease agents might transmit among domestic livestock and wildlife species. Reduction of contact rates between animals is paramount to limiting the spread of disease. At low population density disease tends to die out , but at high density, as is the current situation with feral swine in Texas, disease is likely to become well established and possibly endemic Eradication of feral swine is not possible, nor would it be popular with sport hunters, but reduction in numbers would help limit disease incursions. This study identified water and cultivated fields as high risk areas for disease transfer between feral swine and cattle. Water was the primary zone of contact both within and between species, shared use of this resource was particularly high in droughty conditions and during the hot summers. While it is often convenient to provide cattle with mineral and feed supplements at water points, this increases the animals' attraction to high risk areas. A better plan would be to move the supplements to the more open areas of the rangeland, or ranch roads which are extensively used by cattle but generally are not favored by feral swine. Concentrated food resources are another strong attractant for livestock and wildlife. Cultivated food plots were zones of high contact between animals. Fencing off these food sources easily deters hogs but still allows the deer access. On many ranches shelled corn is used to bait deer for survey and harvest. One practice is corning the roads where by shelled corn is trickled along ranch roads to lure deer out of the brush for better viewing. Even though feral swine did not normally use roads for travel as cattle do, they congregated on the roads to feed on the corn. In the event of a disease problem this practice should be curtailed. In 2008, we held a final expert workshop on methods for estimating feraal hog densities in the Texas landscape. This remains the single most important deficit in data needed to model a potential CSF or FMD epidemic. Experts from throughout the US were brought together in Uvalde in early June 2008 and a workshop report has been printed (see below). In addition, a PhD student and modeling group at UC davis is now utilizing these data to improve their models and in particular is developing new methods to predict animal movements and landuse on varied landscapes. <BR>PARTICIPANTS: Study participants: Individuals (all at Texas A&M University unless otherwise noted): PI: H. Morgan Scott DVM, PhD Co-PIs: Susan Cooper, PhD; James Cathey, PhD Collaborators: Roel Lopez, PhD; MS Students: Aubrey Deck; Walley de la Garza III PhD Student: Lindsey Holmstrom DVM Technical staff: Lang Alford; Andrea Wappell; Shane Sieckenius Partners: Texas Cooperative Extension; Texas Agricultural Experiment Station; Private (Ranch) Landowners Workshop participants: H. Morgan Scott DVM, PhD, Susan Cooper, PhD, Jim Gallagher PhD, Lindsey Holmstrom DVM, Richard B. Minnis PhD, Rick Taylor, Matt Reidy MS, Shane Sieckenius, BS, Heather Simmons DVM, MSVPH (Rapporteur), Andrea Wappel BS TARGET AUDIENCES: Epidemiologic modelers: academic and government Government agencies: Texas Animal Health Commission; USDA-APHIS-VS; Ranchers and other landowners; Extension personnel; Law enforcement and otheer first responders<BR> PROJECT MODIFICATIONS: Because of difficulties we encountered with determining feral hog densities on varying landscapes within the study area (including via flyover transects, querying landowners and locals, and the vast scale of the ranches themselves (up to 85,000 acres)), additional approaches to achieving estimated distributions of hog density by landscape features were based on a workshop panel held in June 2008. These estimated distributions (using GIS and vegetative indices) are being employed within a stochastic modeling framework to better provide a full range of probable outcomes in epidemiologic models to complement the empirical nature of our GPS contact data. This approach replaces our initial plans to rely on a survey instrument provided to local landowners, police, extension agents, etc. In addition, we are employing movement analysis to better populate the epidemiologic models.
<BR><BR>
IMPACT: 2003/07 TO 2008/06<BR>
A workshop on estimating swine densities was held in Uvalde Texas in June 2008. The primary objective of the Workshop was the following: 1. To provide an understanding for all of the known and/or potential methods for estimates of feral swine densities or probabilities across a wide variety of Texas landscapes. The secondary objectives of this Workshop included the following: 1. To conduct a matrix analysis describing the advantages/ disadvantages, such as cost, speed, and validity for various methods used to estimate feral swine density, 2. To explore if the "interpreted methodologies", using a blend of empirical data, satellite imagery, and eco-region experts is actually achievable. 3. To recognize what critical data and/or research are needed, focusing on disease modeling and management.4. To provide recommendations to animal disease managers and decision- makers regarding the best empirical approaches to estimate feral swine densities across a variety of landscapes and scales (area) during the prevention, detection, response, and recovery phases of an outbreak. Various techniques exist to estimate or index feral swine abundance. That said, feral swine density estimation methods developed must be efficient, economical, accurate, and valid for emergency response managers to compare feral swine density abundance between different locations, habitats, and times. Feral swine abundance can be measured in three ways: as the number of animals in a population, the number of animals per unit of area (absolute density), and as the density of one population relative to that of another (relative density). Abundance estimates and indices can be complex, requiring systematic measurement, repeated sampling and usually some form of numerical analysis; or simple, requiring less complex measurement and no numerical analysis. New modeling efforts are underway that rely on movement data we achieved in the study. We analyzed differences in path tortuosity by calculating fractal dimensions and path lengths. There were no statistically significant differences of fractal dimensions between hog type, sex, age or social group (p>0.05). The average (95% confidence interval, CI) fractal dimension across all groups was 1.39 (1.34, 1.44). There were differences in movement pattern based on proximity to landscape features. Movement patterns were directed between and convoluted within riparian areas, cultivated fields and water areas, identifying these landscape features as areas of increased contact and potential disease spread. The average (95% CI) path length traveled each night was 6.45 (5.44, 7.46) km for boars, 4.43 (3.71, 5.14) km for sows, and 3.70 (3.16, 4.23) km for shoats. The path lengths were statistically different between boars and sows, boars and sounders, male and female shoats, and between adults and juveniles (p<0.001 for all comparisons). These results will be used to set movement rules for feral swine in the development of a wildlife disease model to simulate disease spread over various landscapes.