Abstract: Six pony mares deprived of water for 24 hours showed significant increases in plasma vasopressin (2.8 pg/ml) and osmolality (9 mosmol/kg). When water was made available the ponies drank rapidly (5 of 6 drank to satiety within 90 seconds) and corrected their fluid deficits precisely. Vasopressin did not return to predehydration levels until osmolality did after 15 minutes of access to water. The horse differs from rodents and humans, but is similar to pigs in that vasopressin levels do not fall before osmolality returns to normal. Oropharyngeal factors, therefore, may not be as important in vasopressin release in horses as in other species.

Abstract: This study investigated drinking behaviour of unhandled, semi-feral Dartmoor ponies. Aspects studied were drinking behaviour, latency to drink from novel unfamiliar drinkers after transport, preferences for different types of artificial water drinkers, effects of mixing with unfamiliar ponies and group size, on drinking behaviour, and the effect of a simulated market on the latency to drink. Ponies were tested in groups of three or six animals, or as individuals in test pens that were equipped with three water drinkers: bucket, automatic drinking bowl, flowing water trough. Behaviour was recorded using time-lapse video. An individual pony drank on average 10 l per day. Ponies also drank, but at a lower rate, during the night. The latencies to drink after 4.5 h of transport showed large variation, but most ponies drank within the first hour after being transported (all groups 80.5 ± 32.94 min, mean ± SEM). In the individual choice tests, the preferred drinkers were the bucket and the flowing water trough, but not the automatic drinking bowl (drinking time 25.2 ± 4.66, 11.5 ± 4.26, 2.4 ± 2.23 min for bucket, trough and bowl respectively, mean ± SEM; paired t-tests, bowl versus other drinkers, all tests p < 0.02). A possible reason for the avoidance of the automatic bowl was the noise it made when filling. After mixing a group of three ponies with a group of three unfamiliar animals, the ponies did not express their individual drinker preferences anymore. The use of the previously preferred bucket decreased significantly and the use of the initially, non-preferred, bowl increased significantly. This was likely caused by the fact that ponies were either intentionally or accidentally obstructing drinkers in certain areas of the pen and unfamiliar ponies did not want to push past them. In the simulated market, the differences in latencies to drink between ponies in the home pen and market groups did not reach significance. No significant effect of group size (groups of three versus six ponies) on drinking behaviour was detected. The results have implications for situations where only automatic water bowls are provided, such as during pony sales at livestock markets. Preventing ponies from expressing their drinking choice, either by offering non-preferred drinkers or by mixing with unfamiliar animals, could adversely affect their welfare especially if this happens in conjunction with other stressful events such as transport and markets, and potentially weaning.

Abstract: In animal groups, collective movements emerge from individual interactions. Biologists seek to identify how characteristics of actors in these groups, and their relationships, influence the decision-making process. We distinguished two basic factors determining leadership in group choices: identity and state. We hypothesized that identity is more important to leadership in groups with stable relationships, which permit the development of habitual roles. In groups with fluid membership, particular individuals or subgroups are less likely to emerge as consistent leaders. Instead, we predicted that movement initiation in unstable groups depends on individual state at the time of the decision. We characterized how identity and reproductive state influenced leadership patterns in the movements of plains zebra. As in many other mammals, lactation in this species significantly alters water and energy needs. We investigated leadership in tightly knit harems and loosely bonded herds of multiple harems. Harem females tended to have habitual roles in the initiation of harem movement. In herds, however, we found no consistent leaders among harems. At both levels of social organization, lactation was a key determinant of leadership. In harems, lactating females were more likely to initiate movement than nonlactating females. In turn, harems containing lactating females were more likely to lead herd movements. Thus, we conclude that social relationships and reproductive state together shape the interactions that produce group behaviours. One benefit to lactating females of leading herd movements is preferential access to scarce water. Thus, leadership roles in group decisions may have fitness consequences.

Abstract: The stimuli that elicit thirst were studied in four ponies. Nineteen hours of water deprivation produced an increase in plasma protein from 67 +/- 0.1 g/litre to 72 +/- 2 g/litre, a mean (+/- se) increase in plasma sodium from 139 +/- 3 to 145 +/- 2 mmol/litre and an increase in plasma osmolality from 297 +/- 1 to 306 +/- 2 mosmol/litre. Undeprived ponies drank 1.5 +/- 0.9 kg/30 mins; 19 h deprived ponies drank 10.2 +/- 2.5 kg/30 mins and corrected the deficits in plasma protein, plasma sodium and plasma osmolality as well as compensating for the water they would have drunk during the deprivation period. In order to determine if an increase in plasma osmolality would stimulate thirst, 250 ml of 15 per cent sodium chloride was infused intravenously. The ponies drank when osmolality increased 3 per cent and when plasma sodium rose from 136 +/- 3 mmol/litre to 143 +/- 3 mmol/litre. Ponies infused with 15 per cent sodium chloride drank 2.9 +/- 0.7 kg; those infused with 0.9 per cent sodium chloride drank 0.7 +/- 0.5 kg. In order to determine if a decrease in plasma volume would stimulate thirst, ponies were injected with 1 or 2 mg/kg bodyweight (bwt) frusemide. Plasma protein rose from 68 +/- 2 g/litre pre-injection to 75 +/- 2 g/litre 1 h after 1 mg/kg bwt frusemide and to 81 +/- 1 g/litre 1 h after 2 mg/kg bwt frusemide.(ABSTRACT TRUNCATED AT 250 WORDS)