1207
Development of an objective on-farm equine temperament scoring system
To evaluate interest in development of an objective equine temperament scoring system (TSS), an online survey was distributed via hyperlink to faculty at 3 public universities and their equine interest groups and social media pages. Of 123 respondents, 71.4% were familiar with existing TSS, 81.1% of those familiar with existing TSS felt they did not accurately represent the horse’s temperament, and 70.5% were interested in an improved TSS to replace subjective scales currently used on equine websites (equinenow.com and horsetopia.com most frequently cited). Existing TSS rank the horse from 1 to 10 (1 = calm, 10 = spirited), however no further clarification is provided. In order to assess equine temperament more objectively, 27 horses were utilized in a two phase test. Phase 1 incorporated ten minutes of isolation while tied. Frequency of vocalization, defecation, urination, lateral movements, forelimb pawing, and blatant pulling on lead were recorded and used to develop a 0 to 6 scoring system (score assigned based on frequency of aforementioned behaviors). Phase 2 tested willingness to cross an unfamiliar obstacle. Horses were allowed 1 min to cross the obstacle. If the obstacle was refused, horses were trotted in a 5 m circle for 1 min. This procedure was repeated up to 5 times, and horses were assigned a score from 0 to 6 based on time required to cross with little resistance or fear. For both phases, vital signs of heart rate (HR), respiration rate (RR), and rectal temperature (RT) were evaluated immediately before and after each test. Blood samples were taken via jugular venipuncture and later analyzed for cortisol concentration utilizing a commercially available ELISA kit. Pre-test values of HR, RR, RT and cortisol concentrations were subtracted from post-test values to calculate the difference. Pearson’s correlations were analyzed for all variables within phase, and Spearman’s rank test performed to compare phases. Strong positive correlation existed between difference in cortisol and assigned score in both phase 1 and 2 (r≥ 0.61; p < 0.001). A strong positive correlation between assigned score in both phases and difference in vital signs (HR, RR, RT; r ≥ 0.4387, p ≤ 0.02) was noted. However, there was no correlation between phase 1 and phase 2 scores (r = 0.25, p= 0.20). This suggests each phase successfully evaluated different aspects of equine behavior using clearly defined 0 to 6 point scales, and both phases could be used together to evaluate equine temperament.
Keywords: behavior evaluation, equine, temperament scoring