Development of a Differential Global Positioning System to Determine Levels of Landscape Resource Utilization by Free-Ranging Cattle.
P.D. Teel, B.R. Bush, O.F. Strey, and M.T. Longnecker
Cattle influence the dispersal of ticks through their patterns of resource utilization across rangeland landscapes. Ticks reside off-host in the litter of canopied brush mottes, where environmental conditions are more favorable for their survival. Distribution and maintenance of tick populations within heterogeneous brush habitats is known to be dependent upon the probability of host contact and rate of successful tick attachment and feeding. Mesquite invades herbaceous areas through seeds in manure pats that are deposited within the cooler micro environments of canopied brush habitats. Once established, mesquite canopy provides favorable conditions for the establishment and expansion of additional brush species, further reducing the grass available for cattle to graze.
Previous studies at LCRA had focused on each of these pervasive rangeland problems as separate and discrete entities. It was increasingly apparent however, that these dynamic pest systems are interrelated, and synergistic to the movements of grazing cattle. Despite this, little was known about the interactions of cattle within landscape environments, and how they correlated with resident tick populations and the establishment of invasive woody plants such as mesquite. These associations were known to exist but the actual interactions that take place over a large landscape with several animals was unknown. Previous studies evaluating landscape utilization by cattle had relied upon human observers to visually track and record the movements of a few cattle confined within a small fenced in area. While these methods were adequate for monitoring daytime movements of a few cows across small scale landscapes, they were impractical for accurately assessing the landscape utilization of a large cattle herd distributed over a pasture sized landscape.
Global Positioning System (GPS) technology is vastly superior to other methods of animal tracking, including radio telemetry, LORAN, and human observation. GPS surveillance provides a continuous (24 h/day) real-time record for tracking cattle movements and calculating residence times. GPS signals are transmitted from a series of more than 20 satellites. At any time, signals from more than eight satellites can be picked up at a ground site. GPS Commercial Availability codes are generally available to the public worldwide for positioning information. The Department of Defense introduces intentional, slow-moving, pseudo-random clock “misinformation”. While time data are not affected, pseudo-range data between the satellite and receiver appear to wander randomly about a central point, resulting in positions with accuracy measurements within a 10 meter radius. A differential receiver /transmitter which is placed at a georeferenced point in the pasture provides site specific correction adjustments to the satellite signals. This Differential Global Positioning System (DGPS) is even more accurate than standard GPS, providing accurate positions within about a 1 meter radius on each position tracked. DGPS is an automated data capture process which maximizes the data to be obtained from a cattle herd without interruption by human observers, and enables the computer-logged data to be readily integrated into a computer based Geographic Information System (GIS) for complete spatial analysis. Only periodic monitoring is necessary from the base station to insure that each cow’s GPS device is logging properly, and an individual cow’s position is obtainable at any given time during the monitoring session.
The objective of our study was to evaluate landscape-level distributions of ticks and manure pats relative to vegetation heterogeneity and cattle grazing patterns. To accomplish this we: 1) mapped out the vegetation areas in the study pastures in accordance with a predetermined key by utilizing ARC/INFO, GIS software, 2) using DGPS technology, established a database of cattle movement within the study pastures during the recent drought of 1996, and 3) conducted sample CO2 surveys on resident ticks in brush areas utilized by DGPS monitored cattle.
The study area consisted of two 80 hectare pastures (Deferred pastures 1 & 4) located on the La Copita Research Area (LCRA). The vegetation zones of each pasture were predetermined from aerial infrared images and from ground coordinates using a hand held GPS unit on site to insure the accuracy of the interpretations. Once the zones had been determined, a digitized map of the pastures was produced. The map includes all relevant structures, fences, watering sites, and vegetation zones. Three fixed position control points forming a triangulation along the perimeter of the two study pastures was selected for stationary GPS receivers, providing additional constant reference points for DGPS corrections of cow positions. Satellite signals received by each cow harness and corrected by the differential receiverwere transmitted to the final component, a PC computer datalogging station located in the ranch headquarters facilities. The signal containing time, latitude, longitude, altitude, and animal identification numbers was recorded and stored in a LOG file on the base station computer (Fig. 1). The data was downloaded to 1/4 inch tape for easy transportation and uploading to ARC/INFO GIS software on a Sun computer workstation.
Each pasture was grazed by one herd of 16 acclimated cows for 2 days to yield 32 cow days/pasture. Cattle were inspected over for ticks at the beginning and end of each 2 day session. During each inspection, removal of all ticks was necessary in order to give us an accurate record of ticks acquired by each cow during the monitoring period. Cattle were restrained in a standard commercial working chute, enabling tick inspections and harness fitting to be conducted efficiently and safely. Following tick inspections, each cow was fitted with a customized GPS tracking harness and released into one of the two pastures. Each harness was equipped with a GPS unit and antenna, a terminal node controller (TNC) that translated the GPS position data into a broadcastable frequency, and a radio transceiver and antenna that transmitted the cow’s position to the base station at LCRA headquarters. Once the data reached the base station, it was then logged onto a computer running Automatic Packet Reporting System (APRS) software. A computerized map ran as a backdrop along with the APRS surveillance software. This gave us an accurate real-time, visual record of updated cattle herd positions, individual cow positions, and herd distributions across both pastures throughout the DGPS monitoring sessions. The data from the tick inspection information was compared with the position data looking specifically at the time spent occupying the different vegetation zones. Landscape utilization from each pasture was analyzed and compared through statistical analyses to determine significant differences in landscape usage. Data was pooled where no significant differences were found. Samples of manure were collected from study animals at the end of each two day DGPS monitoring session. Samples were returned to the Grazing Animal Nutrition Lab at TAMU and analyzed for estimates of cattle diet quality and % dicots (brush utilization) using near-infrared spectroscopy (NIRS) technology.
Using the cattle position data and the map of the pasture it was possible to determine where to census for ticks. Landscape level distribution of the resident free-living population among habitat-types was quantified by conducting surveys using CO2 tick traps . The surveys were conducted approximately 5 days following the DGPS data capture sessions to allow time for data processing and to identify vegetation zones in each pasture for sampling. Trapping was conducted along transects in brush habitat types utilized by cattle (lowland drainage, upland coalesced, & upland mature brush mottes) during the prior monitoring session. The location of each trap was recorded with a hand held GPS unit. Trap location and tick numbers were recorded as a thematic map in the GIS for evaluation of landscape-level distributions of ticks relative to vegetation heterogeneity and cattle grazing patterns (Fig. 2). Hierarchical analyses of variance were conducted to compare tick density across habitat type using the General Linear Models (GLM) Procedures of the Statistical Analysis System (SAS). A climate monitoring station positioned near the fence line dividing the two pastures recorded ground and air temperature/RH in open and canopied habitats, as well as rainfall and solar radiation. These data were used as covariates in tick distribution analyses.
Following analysis of DGPS cattle position information and supportive field data, a computer simulation model of cow movement was evaluated for its potential use as a decision-making tool. This is a physiologically based landscape-use model for cattle developed by researchers at TAMU. The model is driven by diet quality, forage availability, and by climatic stressors such as temperature, relative humidity, and solar radiation, which are recorded by the climate monitoring station at the research site. The model was run concurrent with the GIS of cattle landscape use patterns. Model simulations of time spent each day in each habitat type was compared to the observed DGPS data to determine its validity in simulating landscape resource utilization by cattle.