Effects of conditions of an experimental model to evaluate methods of electric-fence strand addition to barbed-wire fence to contain goats.

Growing meat goats of 4 types (Boer and Spanish of both wethers and doelings) were used to evaluate conditions for a method of testing efficacy of electric-fence strand additions to barbed-wire fence used for cattle to also contain goats. Animals were allocated to 8 sets, with each set consisting of 5 groups. There was 1 goat of each of the 4 types in a group. One side of five 2.4- × 3.7-m evaluation pens consisted of barbed-wire strands at 30, 56, 81, 107, and 132 cm from the ground. Evaluation pens were adjacent to a pasture with abundant vegetation. Fence treatments (FT) were electrified strands (6 kV) at 15- and 43- (LowHigh), 15- and 23- (LowMed), 15- (Low), 23- (Med), and 43-cm (High), where Low, Med, and High abbreviations are for low, medium, and high heights from the ground, respectively. For adaptation, there were 4-wk and sequential exposures to evaluation pens: wk 1, no electric strands; wk 2, 1 strand at 0 kV; wk 3, LowHigh; and wk 4, LowHigh. There were 6 periods for measurements, each separated by 1 wk. During the 1-wk intervals on pasture, sets were exposed to 1 interval treatment without and another with 2 electric strands (6 kV) positioned next to supplement troughs, to potentially affect familiarity with electrified strands and influence subsequent behavior. All animal sets were used for measurements in period 1 in a completely randomized design (CRD). Four sets were also used in 4-wk subsequent measurement periods for a 5 × 5 Latin square (LS). All animal sets were exposed to the same FT in period 6 as in period 1. Behavior in evaluation pens was observed for 1 h with a video surveillance system in the 6 periods. There were no effects of gender and few and minor effects of preliminary and interval treatments. The percentage of animals that exited evaluation pens differed (P < 0.05) among FT, with the CRD approach in period 1 (25%, 47%, 38%, 66%, and 84%; SEM = 8.0) and with repeated measures in periods 1 and 6 (6%, 22%, 22%, 63%, and 81% for LowHigh, LowMed, Low, High, and Med, respectively; SEM = 4.9), and between breeds in periods 1 (34% and 70%) and 1 and 6 (28% and 50% for Boer and Spanish, respectively). For the LS approach, FT affected exit (31%, 23%, 16%, 35%, and 30%; SEM = 5.3) and breeds differed (P < 0.05), as well (12% and 43%). Exit decreased as period advanced (60%, 35%, 23%, 10%, and 8%, for 1, 2, 3, 4, and 5, respectively; SEM = 5.3). In conclusion, breed should be considered in the model being developed. A LS approach was not suitable, but a CRD experiment after these adaptation procedures appears promising.