Printable Format (.pdf)

Executive Summary

Background

Features of Unit Tested

Setting for the Test

Results

Data Summary

Conclusions

Acknowledgements &
Further Information

Preparation
In anticipation of the test, one of the collars was tested for a 3-week period in March and April 2001 at the U.S. Sheep Experiment Station in Dubois, Idaho. The purpose of the test was to compare the GPS_2200L(R) collar with a prototype backpack harness unit. Concerns had been raised by experiment station staff regarding the durability of the collars, given the head-down grazing habits of sheep. Following this brief test, concerns were reduced and the station staff recommended the Sawtooth NRA proceed with the full test as planned, using the collars. The backpack prototype had its own particular challenges (G. Lewis, station manager, personal communication).
 

Before deploying the collars, the GPS_2000 Host software had to be loaded onto a standard PC, and configured for the particular computer and geographic location. A current satellite almanac was downloaded from Lotek’s Web site just before deploying the collars. The almanac is required to inform the collar of the approximate locations of available satellites upon its first fix attempt. After this first fix, this information is communicated automatically from the satellites to the collars. Finally, the approximate geographic location of the deployment site was entered—again, to assure a quick and successful first fix.

A link with the collar was established and the position and almanac information, as well as the schedule for position fixes, was sent to the collar. The first fix must be obtained within 10 minutes or the VHF beacon will switch into “Recovery” mode.

Deployment
 

The Lotek GPS_2200L(R) collars were fitted on two randomly selected sheep as they were off-loaded from the livestock shipping trucks. The permittee used a shepherd’s crook to select and control the sheep. The collar strap was quickly looped around the animals’ necks and secured with holding plates and lock nuts tightened to the specified 1.7 N×m torque. The collars were adjusted on both sides to center the antenna on top. The sheep were marked with colored chalk to distinguish them within the band, and then were released.

Active Use

The  Sawtooth NRA did not possess a VHF receiving antenna and assumed it could monitor the sheep by visual observation during allotment administration. The agency arranged to borrow an antenna from a sister agency if needed. Within a week, one of the collared sheep was missing. An antenna was obtained and the ewe was shortly found with a broken leg. The “recovery” beacon pulse was never heard, suggesting she continued sufficient movement to maintain the “normal” beacon.
 

No other problem was encountered. The chalk quickly wore off, and the VHF antenna was the only means to monitor the status of the sheep and their collars. The VHF antenna also proved to be a very effective tool in monitoring the band location, whether nearby and visible, or more distant. If visual identification is important, a more durable product, such as marking paint, may be required.

The effect of fleece growth through the season on the fit of the collar was a speculated concern initially, but proved to be of no consequence. The collar remained snug on the neck, and the fleece accumulated around it. No interference with the satellite signal was apparent.
Incidentally, the sheep’s leg was likely broken during the permittee’s scuffle to capture and constrain the sheep initially to attach the collar. The herder had noted the sheep limping as she walked away after the collar was attached. Care in such occasions is obviously important to avoid injury to the animal as well as to the handlers.
 

Retrieval The collars were collected without difficulty on October 12 by the permittee. Close inspection of the collar straps and unit housings found no compromised conditions beyond normal wear. In contrast to some pretest concerns, the collars appeared to be very durable. The process to retrieve the data from the collars was a reverse of the initial launching. Once connected to the PC through the download link unit, the data was collected from both collars without incident. Post Processing The data was exported from the Lotek GPS_2000 Host software to a Microsoft Excel® spreadsheet. Once in Excel, the data was corrected, analyzed, and prepared for further analysis in the Geographic Information System (GIS).

The Lotek collar stores latitude and longitude positions in degrees, minutes, and decimal seconds (to thousandths). When exported from the host software via the spreadsheet option, the values are converted to integer milliseconds. (See table 3.)

To prepare the data for analysis within the GIS the millisecond latitude and longitude values were converted to decimal degrees by dividing the milliseconds values by 3,600,000 (60 minutes x 60 seconds x 1000). (See table 4.)

Simple formatting and heading changes were then also applied to a few of the remaining fields for compatibility with the GIS. The result was saved in dBase IV® format.
ArcView GIS® version 3.1 was used to create a geographic theme for analysis from the table of position values (i.e., an “Event Theme”). For compatibility with the Sawtooth National Forest corporate GIS data themes (e.g., roads, streams, lakes), the new collar themes were reprojected. Once the export position format of the GPS_2000 Host software was understood, this relatively straightforward export process was completed and field maps produced for grazing use correlation within a day.