Subjective methods of quantifying temperament in heifers are indicative of physiological stress

Abstract To effectively select for temperament, selection criteria must reflect an animal's response to stress. Subjective measures such as chute (CS) and exit (ES) score provide fast and easy assessments of temperament but may be subject to observer bias and inconsistency. Objective measures such as cortisol and other metabolites have been associated with increased stress in cattle, especially regarding transportation. Our objectives were to identify relationships between objective and subjective methods of evaluating temperament, and to determine whether objective measures of stress change under repeated, routine handling. Over three consecutive years, a factorial design of two measurement protocols [frequent (F), infrequent (IN)], and three recording periods (event) was used. The F measurements were collected over three consecutive days, and IN measurements only on day one, within all three events every year. Each year, 40 commercial Bos taurus heifers were randomly split into the two protocols. While restrained in a chute, behavior was quantified using CS. Thereafter body temperature and heart rate were measured, and a fecal and blood sample collected. These samples were analyzed for levels of fecal and serum cortisol, glucose, creatine kinase, non-esterified fatty acids, and blood urea nitrogen. Upon release, ES was recorded. For all heifers, the effects of protocol, event, and their interaction, were compared on day one of each of the three events. For heifers assigned to F, event and day within event also were evaluated, with as many as nine observations on each heifer. For both analyses, body weight, ambient temperature, and wind speed were included as covariates, with sire and year fitted as random effects. Pearson correlations between subjective and objective measures were calculated for the first day of all events combined. Chute (r = 0.24 to 0.32) and ES (r = 0.26 to 0.33) were positively correlated (P