Testing virtual fence systems for fire fuel management

Research objectives:

Test the viability of GPS controlled shock collars for use in a fire fuel management context. Collars must achieve:

Effective Containment. Animal movement will be compared to the containment level of sheep in electric fence with the potential for equipment failure or animal breakout in both instances. Ecological goals will be set for both the virtual fence and the control group and compared.

Reduced Labor. Labor of monitoring, applying the collars to the animals, changing out defective equipment, charging, or unpredicted labor needs will be compared with the daily needs of deploying electric fence in similar contexts.

Reduced Projected Costs. All costs of associated equipment and subscription fees in the GPS collar system will be compared to the costs in Labor, and electric fence systems including damaged or replaced equipment projected over a 3 year term.

We will purchase 100 GPS controlled collars designed for livestock from No Fence, who has agreed to enroll us in their pilot program for $120 a unit. Experimentation will take place in three locations, all of which are contracted for vegetation removal by the Fire Safe Council of Santa Barbara County.

The team will identify three locations for side-by side trial of the system in different vegetation types. We will choose the sites to reflect a diversity and spectrum of what a contract grazer may be asked to do in a fire fuel management context. Sites will include, annual dominate grasslands with moderate grade, diverse mixture of shrub and weed types of moderate to extreme grade, and dense mature shrub types  with moderate to extreme grade. A control  group will be managed under electric fence or herding methods. 

Containment will be measured in both systems through documented instances of escape. Notes will be made as to whether escapes are single animals or whole herd escape events and document the time and lengths it takes to return animals to containment. Should a majority of animals continuously escape designated areas, they will be removed and the test will be considered a failure. No system is perfect so expected breakouts in electric fence will be an important comparison to realistic gains from the virtual fence system.

To measure effectiveness of the treatment and ecological goals, residual dry matter measurements will be collected before and after both treatments to measure effective percent of dry matter removed. Photo documentation will also supplement measured findings with visual confirmation of fuel removal and space in shrub layers.

All labor hours of both system will be counted and compared using a time clock app, including any management of collars, pulling on or off, maintaining, trouble shooting of individual units, collars catching on brush or dealing with escapes. Comparisons will be made in the training of animals to the virtual fence system and the labor it takes to train animals to electric fence.

Cost of both systems will be measured based on cost of equipment and labor for implementation of grazing management goals in each system. Costs for electric fence will be for a standard netting system and electric fence chargers. Costs for the virtual fence system will include cost of each unit as well as shipping and processing. If the units fail or persist after three years we will note their success or failure, and take into account longevity of equipment in both systems. Labor costs for the GPS units will include putting the collars on and any troubleshooting or escapes. A yearly fee will also be compared to the labor of stretching fence out and retrieving it. Each trial will last one to two weeks, and expenses will be projected over a three year term for both groups and categorized both in whole expenses and in a per animal unit expenses.

Other grazing management strategies such as electric netting and herding will be examined in comparison to asses viability of collars compared to realistic alternative approaches to fire fuel management with prescribed herbivory. Previous studies on virtual fence systems has preliminary success in pasture settings, but more research is needed to demonstrate effectiveness in harder vegetation and slope settings. There has been no study the author could find that referenced predator impact on the new technology. We will document the predator pressure in wild land environments and will whether a dog on a collar is sufficient to protect from basic coyote and lion pressure.

In order to do this work we will utilize 100 collar units, including one guardian dog, as well as a tablet with software that is capable of effectively mapping and controlling the units. We will also use electric net fence systems that we already have and herding dogs for moving outside of fence for the animals in the control. Each fence paddock will have one 3 joule charger run on a 12volt battery with a 40 watt solar panel. Each herder will have a time clock software on their personal phones to accurately measure when they are managing sheep or issues in one system vs. the other. They will also document the progress before and after moving animals. We will use a one meter square PVC unit to measure RDM and clip forage with a pair of pruners to avoid removal of roots or un-grazeable material.

Comparing all this data we will asses over a three year period if the Virtual Fence systems are a viable solution for fire fuel mitigation projects in the West. We will also utilize the collars, where applicable, outside of the direct side-by-side analysis to have robust experience of actually operating with the technology in that three years to share with other producers and contractors. We will carefully compare benefits of traditional management and draw a conclusion that includes, effectiveness, cost and labor analysis, appropriate application and the quality of life improvements of the owners and operators.