Evaluating accurate and efficient sampling strategies designed to measure social behavior and brush use in drylot housed cattle

Abstract

Efficient sampling strategies expedite behavioral data collection. While multiple studies have evaluated sampling strategies for core behaviors in cattle, few have focused on social interactions. To identify sampling strategies that accurately captured cattle social behaviors and brush use feedlot steers (n = 3 pens; 9 steers/pen) were observed from 8:00 to 17:00. Average bout duration (sec), total duration per day (sec), and bout frequency were recorded for allogrooming, bar licking, tongue rolling, and brush utilization. Frequency was recorded for headbutting and mounting. Data was extracted from continuous observation datasets using eight different sampling strategies and the results subsequently compared. Differences among sampling strategies were evaluated using a non-parametric One-Way ANOVA Kruskal-Wallis Test. Pearson correlation evaluated the strength of association between a specific sampling strategy and continuous observations. Bout duration for allogrooming (P > 0.65), bar licking (P > 0.60), tongue rolling (P > 0.99), brush use (P > 0.99), and mounting frequency (P > 0.70) did not differ from continuous observations. Tongue rolling (r2 > 0.95, P <0.0001) and brush use (r2 > 0.70, P < 0.0003) were best captured when cattle were observed from 08:00 to 14:00. When cattle were continuously observed from 08:00 to 14:00 or for 15 minutes every 30 minutes, allogrooming (P > 0.2) (frequency, duration), bar licking (P > 0.95) (frequency, duration), brush use (P > 0.1) (frequency, duration), heat butt (P > 0.30) (frequency), or tongue rolling (P > 0.30) (frequency, duration) did not differ from continuous observations. Observing cattle for 15 minutes every 30 minutes yielded the highest accuracy for all behavioral metrics and was considered the most effective strategy for comprehensively evaluating cattle social behavior (r2 > 75; P < 0.05). These results provide insight into accurate and efficient sampling strategies that expedite social behavior data collection in cattle and will facilitate efficient generation of new knowledge regarding cattle social behaviors.