Abstract: Two herds of Przewalski horses at the Minnesota Zoological Garden were observed during 1980 in each of 2 enclosures that differed in size. The larger enclosure was a 3.4-ha pasture; the smaller enclosure was a 17 x 30-m grass-less pen. One herd was composed of a stallion, 3 adult mares and 2 foals. The other consisted of a stallion and 2 mares. All occurrences of aggression, mutual grooming and snapping were recorded, and 5-min scan-samples of the activity state of each horse were taken. The time budgets, frequency of aggression and frequency of mutual grooming differed significantly with enclosure size for both herds. More time was spent pacing and milling in the smaller enclosure, and the frequency of aggressions and of mutual grooming was also higher. Only the foals exhibited snapping; frequency of snapping did not vary with enclosure size. More time was spent feeding in the larger enclosure. Provision of hay in the smaller enclosure eliminated the differences in time spent feeding. A second study was conducted during the spring of 1984 in an intermediate-sized enclosure, 0.4 ha, a sub-division of the pasture on which the horses were kept in 1980. One herd consisted of a stallion, 2 mares and 2 yearlings; the other consisted of a stallion, 3 mares and a foal. One of the stallions and all of the mares were those studied in 1980, but that stallion and one of the mares were in different herds than they had been in 1980. The frequency of aggression was similar to that observed in 1980.

Abstract: Twelve ponies were fed their total daily ration either as one large meal or divided into six small meals. Pre- and post-feeding behavior was recorded six times a day. Blood samples were taken for 30 min before and two hr after the meal. Plasma protein increased from 7.0 to a peak of 7.3 g/dl with small meals and from 7.3 to 8.1 g/dl with large meals, and returned to pre-feeding levels by 90 min post-feeding. Hematocrit rose from 33.3 to 34.1% with small meals and from 33.0 to 36.0% with large meals. These rapid and short-lived increases indicate a decrease in plasma volume. Plasma osmolality rose with feeding from 283 to 285 mosmoles/kg with small meals and from 281 to 288 mosmoles/kg with large meals. Water availability had no significant effect on blood changes. Digestibility and rate of passage were measured with chromic oxide, but there were no differences. Vocalizing (neighing) and walking occurred more often before than after feeding, while eating bedding and engaging in other oral behaviors were more frequent after feeding.

Abstract: Three hundred and seventy two horses of varying breeds, height and fatness were weighed and measured for height at the withers. They were assessed for condition score by adaptation of a previously published method. The heart girth and length of 281 of the horses were also measured. Weight of horses was highly correlated (P less than 0.001) with height (r2 = 0.62), condition score (r2 = 0.22) and girth2 x length (r2 = 0.90). Nomograms were constructed to predict weight from height and condition score, and girth and length measurements. Weight can also be accurately estimated from the formula: (formula, see text) The average value of 'Y' in this experiment was 11900 and this estimated weight with more accuracy than some previously published values of 'Y'. Racing Thoroughbred horses were found to be significantly lighter than non-racing Thoroughbreds of the same height and condition score. The method of assessment of condition score was shown to be repeatable between different operators with varying degrees of experience.

Abstract: Thirty-two Quarter horse yearlings were utilized in a split-plot experiment to establish behavioral tendencies for two levels of emotionality; normal and a more highly reactive level of emotionality termed nervous. Four observers who were experienced with horses scored the yearlings on an emotionality scale from 1 to 4, with 1 representing the highly nervous yearlings and 4 representing the quiet yearlings. Emotionality evaluations were based upon the response of the yearlings to a standard regimen of standing in a chute, being identified and being released from the chute. The inter-rater reliability coefficients were 0.90 for the normal yearlings and 0.65 for the nervous yearlings, indicating that the raters agreed less when evaluating the nervous yearlings. Yearling heart rates in the chute were correlated (r = -0.54, P<0.002) with the average emotionality score. Observational data on behavior collected 2 days following the emotionality scoring procedure indicated that the normal yearlings maintained a greater (P<0.001) individual distance than the nervous yearlings. From an activity summary, the normal yearlings spent 10.9% of the time lying down (LD), 79.1% standing (S), 9.6% walking or trotting (WT) and 0.4% of the time cantering or galloping (CG). The nervous yearlings spent 5.7% of the observational periods LD, 79.2% S, 11.7% WT and 3.4% CG. The nervous yearlings tended to have a higher overall activity index level than did the normal yearlings. Results indicate horses of different emotionality levels exhibited different behavioral patterns.

Abstract: The relationships between 4-5 adult zebra stallions, living in a safari park, were investigated over a period of 5 years. Asymmetries in the distributions of a number of behaviours could be explained by adopting dominance as an intervening variable. Dominance in stallions was of a bipolar nature with on the one hand behaviours representing subordinance and defence, and on the other hand behaviours reinforcing and confirming dominance. Expression of formal dominance seems to play a minor role. The dyadic relationships of stallions differed as to the number of behaviours reflecting dominance relationships. Although often linear rank-orders could be constructed, these rank-orders were not necessarily identical. This means that the concept of dominance is of only limited value for describing relationships between zebra stallions.