Measures related to the physical development

In the present work only few differences were found between single and group housed horses concerning physiological measures (Paper I). Both treatment groups followed a normal development pattern for young horses concerning feed intake, growth and bone development.

4.2.1. Bone development

Group housed horses were more active than single housed horses when in paddocks and it was expected that the higher energy expenditure would give a difference in feed intake and/or growth and thus also in bone development. However, this was not the case. Activity or exercise has been shown to have beneficial effects on bone development in horses (Barneveld and van Weeren, 1999) and thus also a difference in bone quality as assessed by radiology was expected but not confirmed in the present work (Paper I). However, group housed horses showed higher levels of calcium, magnesium and the enzyme alkaline phosphatase in plasma, which indicate a higher activity level in bone development. Still, levels for single housed horses were within the normal range, and thereby not considered to affect the bone development negatively. The reason for the lack of effect may be that both treatment groups actually exercised too little in order to achieve the beneficial effects of a higher physical activity. In most studies on bone development young horses have either been exercised or the time spent on pasture has been much longer than in the present work. Bell et al. (2001) found that pasture rearing or 12-h daily turnout was beneficial to maintain and increase the bone mineral content in comparison to horses kept in stalls. In their study horses were housed singly in stalls whereas the pastured horses were in a group. The differences found by Bell et al. (2001) was not an effect of the social environment but of the level of activity which, as indicated in the present work (Paper I), might be a secondary effect of the social environment. Bell et al. (2001) did not record behaviour but reported that horses on pasture were engaged in running, chasing and play-fighting. The same types of behaviour were observed in the present work and at a much higher level for group housed than for single housed horses (Paper I).

4.2.2. Feed intake and growth

In the present work no difference was found in feed intake despite the higher level of activity for group housed horses (Paper I). Size differed between housing treatments but contradicting for the two replicates, and there was no difference in daily gain or body condition scores. As the feed intake was not recorded on an individual level for horses housed in groups it is possible that there were large variations within the groups even though the overall mean did not differ from the horses housed singly. Even in groups of three horses as in the present study a hierarchy will exist and this is likely to influence the feeding behaviour and perhaps the feed intake of individuals in the group. For instance, Houpt and Wolski (1980) found that subordinate ponies spent less time eating when in the same paddock as a dominant pony compared to when the ponies were separated by a fence. Also, the dominant pony was less aggressive when the pony-pair was separated by a fence than when they were in one paddock. Feeding behaviour may be influenced not only by the rank but also by the

Effects of social environment

30

tolerance between animals in a group meaning that some horses might be prevented from feeding (Kolter, 1984). In the study by Kolter (1984) two horses in a group of 6 horses had a lower feed intake and had access to forage of lower quality because they were not tolerated by the other horses. In the present study horses were fed ad libitum and the quality of the Total Mixed Ration supposedly remained the same throughout the day; thus no horses were prevented from consuming the feed they needed/wanted.

4.2.3. Physiological measures

Heart rate – when measured before and in an arena test – was not affected by social environment in the present study (Paper II). This is in accordance with Kusunose et al.

(1986) who found that there was no relationship between heart rate measures and group size.

Also Heleski et al. (2002) measured faecal glucocorticoid metabolite concentration in weanling horses but found no differences between horses housed singly in stalls and horses pastured in groups. Lensink et al. (2001) found no difference between calves housed singly or in pairs before and during loading onto a truck but during transport pair housed calves had a lower heart rate than single housed calves. Pair housed calves were transported together which may have had a calming effect. Isolation in an unfamiliar environment gave a rise in heart rate independent of the social environment (Paper II), which implies that in relation to the later use of horses it is essential that horses learn to be on their own. It requires only a little training as indicated in the study by Jezierski and Górecka (1999) who also found an increased heart rate when isolating horses but the horses habituated to the situation within a few days.

Photo by Eva Søndergaard