The 2008 Joint Meeting of the Society for Range Management and the America Forage and Grassland Council. |
Wednesday, January 30, 2008
30
Determining the Optimum Number of GPS Collars for Livestock Distribution Studies or How Many Tracking Collars Does It Take?
Brandee Williams1, Vibhuti Pandey2, Ken Lackmann1, and Mimi Williams1. (1) Natural Resource Conservation Service, USDA, 324 8th Ave. West, Suite 104, Palmetto, FL 34221, (2) CH2M Hill, 3001 PGA Blvd, Suite 300, Palm Beach Gardens, FL 33410
Practices such as water placement, mineral and supplement feeders, and burning are recommended to enhance sustainable distribution of livestock in South Florida rangelands. The effectiveness of such practices, particularly water point distribution, although well demonstrated under western rangeland conditions, are less understood for the subtropical rangelands of Florida . GPS tracking collars have been effectively used to determine animal distribution across the landscape, particularly when the environment did not permit direct observation. But these studies have shown that individual animals can differ widely in their use of the landscape, particularly in extensive rangeland systems. The question arises, that if resource managers want to determine the effectiveness of a given management recommendation using GPS technology, how many tracking collars are necessary to obtain representative grazing data. Florida NRCS placed seven and nine tracking collars, for a total of 16, on two separate herds of cattle for two grazing seasons to determine the minimum number of tracking collars it takes to establish representative data. Analysis of the first season’s grazing showed similar grazing distribution for most of the animals collared although distribution differed from anticipated. Cattle preferred tame pasture sites compared to native range areas in the pasture unit, which could result in overgrazing and potentially reduced animal performance. Outlier animals were observed and reliability of the units was a problem. Preliminary data from the first year demonstrates that one would need more than one animal to be collared to obtain accurate grazing data.