Research Paper PrePrint – An investigation into the daily level of voluntary activity of stabled riding school horses.
Authors: H. Browna*, D. Marlinb and P. Harrisc
Affiliations: Hartpury College, University of the West of England, Hartpury, Gloucestershire, UK. Animal Web Ltd, Cambridge, UK
Equine Studies Group, WALTHAM Centre for Pet Nutrition, Leicestershire, UK.
The importance of correct feeding practice has been highlighted by the increasing prevalence of obesity in horses. Research in human subjects has suggested that voluntary activity (VA) levels may influence digestive energy (DE) requirements, accounting for 15 – 50% of human daily energy expenditure. Therefore, the aim of this study was to investigate whether levels of VA (classified as non-structured exercise) differ between stabled horses with similar bodyweight (BW) and similar structured workloads, but with different estimated DE intakes to maintain their BW. Twelve mature horses were selected based on their estimated DE intake and BW and were paired according to their BW, breed, estimated DE intake and structured exercise. Within each pair, one horse (L) had a relatively lower estimated DE intake than the other horse (H) to maintain a similar, constant BW. Estimated DE intake was significantly (P<0.01) different between Group L and Group H. Each pair was observed for 72 hours during which structured exercise and non-structured exercise were measured. HR was used as a measure of workload during the structured exercise. Two accelerometers, located on a roller (RT3-R) and head collar (RT3-H), were used to measure VA levels when stabled in addition to visual observations using focal sampling between 07:00h and 18:00h. RT3-R and RT3-H activity levels were not significantly (P>0.05) different between individual horses. Median activity counts were significantly (P≤0.001) higher during the daytime (06:30h – 18:29h) compared to the night time (18:30h – 06:29h). Both RT3-R and RT3-H were significantly (P≤0.001) higher around feeding time. Activity measured using RT3 accelerometers suggested higher activity levels for horses in Group H over the 24 hour periods, although this was not significant (P>0.05). It was therefore concluded that differences in VA levels during stabling could not explain the difference in estimated DE requirements between horses with a similar BW and workload.
The prevalence of obesity in horses used for leisure has been increasing, potentially as a direct result of domestication and current management practices (Argo, 2009; Sillence et al., 2002; Wyse et al., 2008). Weight gain is likely to occur when there is an imbalance between energy intake via the diet and energy requirement (Geor & Harris, 2005; Harris and Kronfeld, 2003). Various intrinsic and extrinsic factors have the potential to affect the digestible energy (DE) requirements of an individual horse being exercised including the horse’s body weight (BW) and age, differences in gastrointestinal function, metabolic factors, stage of training, workload, ability and weight of the rider n (Anon, 2007; Harris, 1997; Hintz and Cymbaluk, 1994). However, it has been suggested that the small individual differences in workload between individuals of a similar bodyweight, age and fitness undertaking similar types of structured exercise under the same environmental conditions cannot always explain the variability in estimated DE intake required to maintain a constant BW (Dekker et al., 2007; Dekker, 2009). Individual differences in feed digestibility, gastro-intestinal absorptive capacity or metabolic efficiency are a possible explanation (Anon, 2007), but other less obvious differences may exist between individuals including the amount of non-structured exercise they undertake.
Within current management systems, horses are often housed in individual stables with limited or no turn-out into a paddock or field, which limits their ability to undertake free movement or non-structured exercise (Brehme and Rose, 2007; Harris, 1999; Petersen et al., 2005). In fact very few studies have focussed on the levels of non-structured activity or voluntary activity (VA: any movement within the stable or during turn-out that is not a result of human encouragement) of horses during stabling or turn-out (Rose-Meierhofer et al., 2010; Werhahn et al., 2012) and the potential influence of this on individual DE requirements. Human studies have suggested that VA may play an important role in the regulation of energy balance, where VA is associated with posture maintenance, fidgeting and all other daily activities which are not classified as exercise (Levine et al., 1999; Levine et al., 2000). Reported data in humans has suggested that VA can account for ~15% of the daily energy expenditure in sedentary individuals to more than 50% of the daily energy expenditure in extremely active individuals (Deriaz et al., 1992; Levine, 2003; Livingstone et al., 1991; Ravussin et al., 1986), and may contribute to the inter-individual variation in susceptibility to weight gain as a response to overfeeding (Levine et al, 1999).
The aim of this study was therefore to investigate whether VA levels differed between stabled horses with similar BW, undertaking similar levels of structured exercise but requiring different estimated DE intakes to maintain their BW. It was hypothesised that VA levels would differ during a 24 hour period and that horses of a similar bodyweight but with a higher estimated DE intake would have higher levels of VA.
Materials & Methods
Horses and diets
Horses performing Elementary and Medium level dressage were monitored at six different competition venues. Twelve mature horses (8 geldings and 4 mares) with mean age of 10.9 ± 4.0 years (range: 5 – 18 years) of mixed breeds were selected from a large population of experienced riding school horses (Table 1). All horses were on a routine anthelmintic and dental care program and only animals in good health and dental status, which did not exhibit any stereotypic behaviour, were recruited. All horses had similar levels of structured exercise. This exercise consisted of on average ten student training sessions per week (range: 8 – 12 sessions per week), which included flat work and low level jumping. During the weekend horses were hacked out or walked in hand. None of the horses were turned out during the study period. All horses received a standard commercial diet during the study. The compound feeds were provided twice daily at 07:00h and 16:30h and the hay or haylage was provided three times per day at 07:30h, 12:30h, and 17:00h.
Table 1. Mean (± SD) age, height, bodyweight (BW), body condition score (BCS), estimated DE intake and relative workload (RW) for horses in group L (n = 6) and horses in group H (n = 6). Different letters (a, b, c etc.) indicate significant differences (P<0.05) between the two groups, i.e. ‘a’ is significantly different from ‘b’, etc.
|Variable||Group L (n = 6)||Group H (n = 6)|
|Age (years)||10.3 ± 4.9 a||11.5 ± 3.3 a|
|Height (cm)||158 ± 1 a||162 ± 5 a|
|BW (kg)||519 ± 26 a||530 ± 31 a|
|BCS||4.32 ± 0.79 a||4.32 ± 0.81 a|
|DE intake (MJ/day)||102 ± 14 a||125 ± 9 b|
|RW (%HRmax)||42.0 ± 4.2 a||41.6 ± 3.8 a|
Horses were selected based on their estimated DE intake and their BW, which was monitored weekly (Eziweigh 2, Tru-Test Ltd, Aukland, New Zealand) for a five week period prior to the start to confirm that the chosen horses were maintaining their BW on their individual diets. A BW variation of 1% BW between weekly weighing sessions was considered insignificant and the result of variation due to last time of defecation or urination. In addition, body condition score (BCS) of the horses was determined using either Henneke et al. (1983) or Kienzle and Schramme (2004) dependent on the horse’s breed. The individual DE intakes were estimated based on the DE values (MJ/kg) of commercially available feeds as provided by the feed manufacturers, and the DE values of hay or haylage as determined by an independent laboratory (Direct Laboratories, Wolverhampton, UK). I think all the feed details need to be provided, including names and estimated DE values and values for hay/haylage and amounts fed.
Horses were paired based on BW, breed, structured exercise load and estimated DE intake. A maximum difference of 6% BW between paired horses was allowed. Each pair consisted of one horse (L) which received a relatively lower amount of estimated DE (MJ/kg) (mean 18 ± 8% lower) than the other horse (H) in order to maintain a similar constant BW.
Two accelerometers (RT3, StayHealthy Inc, California, USA), located on an anti-cast roller (RT3-R) and on an easy-snap head collar (RT3-H), were used to determine VA levels. Previous work (Dekker, 2009) had demonstrated that these positions provided the most reliable measurements. Both accelerometers were set to record activity over one-minute intervals.
Each pair of horses was monitored? observed over a period of 72 hours. During this period, horses were used for their normal riding sessions during which heart rate (HR) was monitored (Polar S610i, Polar, Kempele, Finland), but RT3 accelerometers were removed. Relative workload (RW) was calculated as the mean HR during work as a percentage of the horse’s estimated maximum HR ((HRmax). HRmax was estimated using the equation: HRmax = 224 – age (Vincent et al., 2006).
In addition to data collected by the RT3 accelerometers, horses were observed visually between 07:00h and 18:00h during the 72h study period. Focal sampling with continuous recording was used for 30 minutes, with one horse being the focus at any one time. Observation periods were allocated evenly between paired horses, enabling a representation of their activity pattern to be evaluated. Horses were observed before, during, and after concentrate meals as well as between meals and training sessions. The observer was situated outside the stable with a clear view of the horse. Different stable activities and behaviours were recorded during one minute time periods, which coincided with the RT3 accelerometers’ intervals. These activities were further categorised as leg movement, head movement, and total movement per minute during data analysis. probably need an example record sheet?
Data were analysed using SPSS version 15.0 for Windows. Using the Shapiro-Wilk W test it was determined that raw data was not normally distributed (P>0.05). Therefore, the median of the data was used when aggregating the rough data to one hour intervals (n = 288), time periods (day: 06:30h – 18:29h and night: 18:30h – 06:29h, n = 24), day level (n = 36) and horse level (n = 12). Spearman’s correlation coefficient (rs) with a significance level of P≤0.05 was used to determine correlations between monitored activity and observed activity using the raw data.
Aggregated data for hourly intervals, day periods, day level, and horse level showed a normal distribution established using the Shapiro-Wilk W test with a significance level of P≤0.05, allowing for the use of further parametric tests. Pearson’s correlation coefficient (r) was used to determine correlations between horse characteristics and aggregated data at horse level. Secondly, a one-way ANOVA with a Bonferroni post-hoc test with a significance level of P≤0.05 was used to determine whether VA levels were consistent during 72h observation period using aggregated data at day level, as well as whether the mean daily activity measured by RT3-R and RT3-H were significantly different between horses. Thirdly, an independent t-test with significance level of P≤0.05 was used to determine significant differences for RT3-R and RT3-H between day time and night time using the aggregated data for time periods. Finally, aggregated data at hour level was used to determine patterns of daily activity in addition to observed activity through visual observations
Throughout the study the horses maintained a consistent BW (mean variation 2.2 ± 2.3 kg BW) and BCS and remained clinically healthy. All horses remained sound and maintained their normal weekly workload. Within pairs the average difference in BW was 3.3 ± 1.7 %BW.
On average each individual horse was observed for 516 ± 94 minutes (range: 451 – 627 minutes) between 07:00h and 18:00h over the three consecutive days. The number of observed minutes was not significantly different (P>0.05) between horses in Group L and Group H (507 ± 94 minutes and 524 ± 103 minutes, respectively). The main observed activities included standing (52.4 ± 10.1%), eating hay (41.7 ± 9.6%) and eating concentrate feed (5.8% ± 2.6%). Observed activity showed a correlation with monitored activity (P<0.05). Activity measured by RT3-R showed a weak correlation (r = 0.22) with the amount of time spent walking and moving around the stable. The activity recorded by the RT3-H showed a moderate correlation with eating (r = 0.40 r2 and p value) and vertical head movement (r = 0.24 r2 and p value).
Monitored daily activity
Data from RT3-R (median counts = 7, range: 0 – 144) was significantly (P<0.0001) lower compared to data from RT3-H (median counts = 144, range: 0 – 2074). There was no significant difference between individual horses for median daily activity as measured by the RT3-R (P>0.05) or RT3-H (P>0.05). No significant correlations (P>0.05) were found between median activity counts of the individual horses and the horses’ characteristics, including age, BW, BCS, DE intake, and RW.
Figure 1. Median RT3 roller and median RT3 head activity during the day time (06:30h – 18:29h) and the night time (18:30h – 06:29h) (N = 12). N equals the number of horses. Bars with different letters are significantly different (P<0.05), i.e. ‘a’ is significantly different from ‘b’, etc.
Median activity counts of the horses was significantly (P≤0.001) higher during the day time (06:30h – 18:29h) for both RT3-R and RT3-H compared to night time (18:30h – 06:29h) (Figure 1). In addition, RT3-R and RT3-H differed significantly between individual hours (Figure 2), with increased activity being measured around 07:00h, 12:00h, and 16:30h. This was associated with the time of feeding. Both RT3-R and RT3-H were significantly (P≤0.001) higher during the eating of compound feeds compared to the eating of hay.
Figure 2. Activity pattern of mean RT3 roller activity and mean RT3 head activity over a 24h period (N = 12). N equals the number of horses. Arrows indicate occurrence of feeding.
Activity within pairs
Group L and Group H did not differ significantly (P>0.05) in mean age, mean height, mean BW, mean BCSor mean RW. Although the use of compound feeds did not differ between the two groups, the estimated DE intake (MJ/day) was significantly (P<0.01) different (see table 1). Horses in group L were visually observed to move significantly (P<0.05) more per minute than horses in group H (12.9 ± 7.8 movements per minute and 11.3 ± 7.2 movements per minute, respectively). However, activity measured using RT3-R and RT3-H indicated a trend higher overall activity counts for horses in group H compared to group L during the daytime and night time periods as well as over 24 hour periods, although these differences were not significant (give the actual P value5) (Table 2).
Table 2. Mean (± SD) counts of the RT3 roller and RT3 head activity for horses in Group L (n = 6) and Group H (n = 6). Mean values for daytime (06:30h – 18:29h), night time (18:30 – 06:29h), and 24 hour period derived from data collected over a period of 72 hours. Different letter (a, b, c, etc) indicate significant differences (P<0.05) between the two groups, i.e. ‘a’ is significantly different from ‘b’, etc.
|Time||RT3||Group L (n = 6)||Group H (n = 6)||All horses (n = 12)|
|Day time||Roller||20 ± 13 a||21 ± 17 a||20 ± 14|
|(06:30h – 18:29h)||Head||376 ± 148 a||322 ± 109 a||349 ± 127|
|Night time||Roller||2.1 ± 3.1 a||5.8 ± 7.0 a||3.9 ± 5.5|
|(18:30h – 06:29h)||Head||36.8 ± 30.7 a||69.8 ± 47.7 a||53.3 ± 42.0|
|24 hour period||Roller||9.7 ± 6.3 a||11.3 ± 9.7 a||12.1 ± 13.4|
|Head||117.4 ± 113.5 a||184.0 ± 88.9 a||201.3 ± 177.2|
The use of accelerometers to objectively measure activity in horses has been reported in various different contexts (Berger et al., 1999; Burla et al., 2014; Piccione et al., 2008; Scheibe et al., 1998), although to the authors’ knowledge no link has been made to energy expenditure and DE requirements of the horse. With the knowledge that VA levels can account for more than 50% of the daily energy expenditure in extremely active human individuals (Deriaz et al., 1992; Levine, 2003; Livingstone et al., 1991; Ravussin et al., 1986), it could be presumed that similar principles could be underlying the anecdotal reports that horses of similar BW and structured workload have significantly different DE requirements.
Because of the different variables known to influence DE requirements in horses, the horses in this study had been paired according to a similar BW, age, breed and estimated RW, but with significantly (P<0.05) different DE requirements (MJ/day) to maintain their BW and BCS?. Breed was considered an important factor, as previous studies have indicated that reactivity, temperament and VA levels tend to vary between different breeds (Hausberger and Muller, 2002; Lloyd et al., 2008; Søndergaard and Ladewig, 2004). Although horses in Group L were visually observed to move significantly (P<0.05) more than horses in Group H, this was not mirrored in the accelerometer recordings of activity. Therefore, it can be presumed that other variables have a larger influence on the daily DE requirements of horses than VA levels in the stable. Alternatively, the accelerometers may not be sensitive enough to characterise differences in VA between the two groups.
Although VA (by accelerometer recordings)was similar over the 72h observation periods, day time activity (06:30h – 18:29h) was found to be significantly higher 1 compared to night time activity (18:30h – 06:29h) for all horses. The higher activity levels recorded during the day time could be attributed especially to behaviour around feeding time (Figure 2). Increased activity levels have previously been recorded as anticipation behaviour (Cooper and Mason, 1998; Cooper et al., 2000; Cooper et al., 2005). However, in contrast to these studies, it is important to note that none of the horses in the current study performed any stereotypical behaviours. It can therefore be suggested that anticipation of feeding can result in an increase in activity in horses in general, not just those with stereotypies. Day time activity did not differ significantly between the two groups (Table 2), suggesting further that activity on the yard and the horses’ exercise and feeding regimen strongly influenced day time VA levels. Indeed, horses have previously been shown to spend significantly less time resting after exercise and more time feeding and drinking (Caanitz et al., 1991).
Although no visual observations were undertaken during the night time, it may be presumed that horses spent the majority of their time resting and ingesting hay, as reported by Greening et al. (2013). It is of interest to note that there was a trend for VA levels to be higher for horses in Group H. Horses were all stabled on a combination of rubber matting and shavings, limiting the variable of bedding on behaviour and activity patterns during the night time period (Greening et al., 2013). The difference between the two groups may potentially be the result of large individual differences within a relatively small sample size. Therefore, it is suggested that a larger sample size would be required to ascertain this. Although VA levels during turn-out were not evaluated during this study, this should also be considered as an influencing factor for differences in daily DE requirements between similar horses.
The present study objectively measured the voluntary non-structured activity levels of stabled, general purpose horses with no stereotypical behaviours. Differences in VA between horses could not explain the difference in estimated DE requirements between pairs of animals of a similar age with similar BW, workload, etc., although the trend for differences over night might suggest a small role.
We would like to acknowledge Hartpury College and SPILLERS for the funding of this project and would like to thank the yard staff, riding instructors and students at Hartpury College for their help during this study.
Anon, 2007. Nutrient requirements of horses, 6th edition. National. Academic Press, Washington DC, USA.
Argo, C. McG., 2009. Appraising the portly pony: Body condition scores and adiposity. The Veterinary Journal 179: 158-160
Berger, A., Scheibe, K.M., Eichhorn, K., Scheibe, A. and Streich, J., 1999. Diurnal and ultradian rhythms of behaviour in a mare group of Przewalski horse (Equus ferus przewalskii), measured through one year under semi-reserve conditions. Applied Animal Behaviour Science 64: 1-17
Brehme, U. and Rose, S., 2007. Effect of different activity and space offers on the activity behaviour of stallions. Landtechnik. 62: 408-409
Burla, J-B., Ostertag, A., Schulze Westerath, H. and Hillmann, E., 2014. Gait determination and activity measurement in horses using an accelerometer. Computers and Electronics in Agriculture 102: 127-133
Caanitz, H., O’Leory, L., Houpt, K., Petersson, K. and Hintz, H., 1991. Effect of exercise on equine behavior. Applied Animal Behaviour Science 31: 1-12
Cooper, J.J. and Mason, G.J., 1998. The identification of abnormal behaviour and behavioural problems in stabled horses and their relationship to horse welfare: a comparative review. Equine Veterinary Journal Supplement27: 5-9
Cooper, J.J., McDonald, L. and Mills, D.S., 2000. The effect of increasing visual horizons on stereotypic weaving: implications for the social housing of stabled horses. Applied Animal Behaviour Science 69: 67-83
Cooper, J.J., McCall, N., Johnson, S. and Davidson, S., 2005. The short-term effects of increasing meal frequency on stereotypic behaviour of stabled horses. Applied Animal Behaviour Science90: 351-364
Dekker, H., 2009. A study of the effects of diet and management on the activity and body condition of stabled horses. PhD thesis, University of the West of England
Dekker, H., Marlin, D., Alexander, L., Bishop, R., and Harris, P., 2007. A pilot study investigating the relationship between perceived and actual workload and estimated energy intake in riding centre horses. Equine and Comparative Exercise Physiology 4: 7-14
Deriaz, O., Fournier, G., Tremblay, A., Despres, J.P. and Bouchard, C., 1992. Lean-body-mass composition and resting energy expenditure before and after long-term overfeeding. The American Journal of Clinical Nutrition 56: 840-847
Geor, R.J. and Harris, P.A., 2005. Nutritional Management of Endurance Horses. In: Harris, P.A., Mair, T.S., Slater, J.D. and Green, R.E. (Eds.), Equine Nutrition for All. The 1st BEVA and WALTHAM Nutrition Symposia. Equine Veterinary Journal, Newmarket, UK, pp. 71 – 78
Greening, L., Shenton, V., Wilcockson, K. and Swanson, J., 2013. Investigating duration of nocturnal ingestive and sleep behaviors of horses bedded on straw versus shavings. Journal of Veterinary Behaviour 8: 82-86
Harris, P.A., 1997. Energy sources and requirements of the exercising horse. Annual Review of Nutrition 17: 185-210
Harris, P.A., 1999. Review of equine feeding and stable management practices in the UK concentrating on the last decade of the 20th century. Equine Veterinary Journal Supplement 28: 46-54
Harris, P.A. and Kronfeld, D.S., 2003. Influence of dietary energy sources on health and performance. In: Robinson, N.E. (Ed.), Current Therapy in Equine Medicine 5. WB Saunders, Philadelphia, USA, pp.698-704
Hausberger, M. and Muller, C., 2002. A brief note on some possible factors involved in the reactions of horses to humans. Applied Animal Behaviour Science76: 330-344
Henneke, D.R., Potter, G.D., Kreider, J.L. and Yeates, B.F., 1983. Relationship between condition score, physical measurement and body fat percentage in mares. Equine Veterinary Journal 15: 371-372
Hintz, H.F. and Cymbaluk, N.F., 1994. Nutrition of the horse. Annual Review of Nutrition 14: 243-267
Kienzle, E. and Schramme S.C., 2004. Beurteilung des Ernährungszustandes mittels Body Condition Scores und Gewichtsschätzung beim adulten Warmblutpferd. Pferdeheilkunde 20: 517-524
Levine, J.A., Eberhardt, N.L. and Jensen, M.D., 1999. Role of nonexercise activity thermogenesis in resistance to fat gain in humans. Science 283: 212-214
Levine, J.A., Schleusner, S.J. and Jensen, M.D., 2000. Energy expenditure of nonexercise activity. The American Journal of Clinical Nutrition 72:1451-1454
Levine, J.A., 2003. Non-exercise activity thermogenesis. Proceedings of the Nutrition Society 62: 667-679
Livingstone, M.B.E., Strain, J.J., Prentice, A.M., Coward, W.A., Nevin, G.B., Barker, M.E., Hickery, R.J., McKenna, P.G. and Whitehead, R.G., 1991. Potential contribution of leisure activity to the energy expenditure patterns of sedentary populations. British Journal of Nutrition65: 145-155
Lloyd, A.S., Martin, J.E., Bornett-Gauci, H.L.I. and Wilkinson, R.G., 2008. Horse personality: variation between breeds. Applied Animal Behaviour Science 112: 369-383
Petersen, S., Tolle, K.H., Blobel, K.J., Grabner, A. and Krieter, J., 2005. Evaluation of horse keeping in Schleswig-Holstein. In: 56th Annual Meeting of the European Association of Animal Production, Uppsala, Sweden 33. EAAP – Book of Abstracts No 11
Piccione, G., Costa, A., Giannetto, C, and Caola, G., 2008. Daily rhythms of activity in horses housed in different stabling conditions. Biological Rhythm Research 39: 79-84
Ravussin, E., Lillioja, S., Anderson, T.E., Christin, L. and Bogardus, C., 1986. Determinants of 24-hour energy expenditure in man. Methods and results using a respiratory chamber. The Journal of Clinical Investigation 78: 1568-1578
Rose-Meierhofer, S., Klaer, S., Ammon, C., Brunsch, R. and Hoffmann, G., 2010. Activity behaviour of horses housed in different open barn systems. Journal of Equine Veterinary Science30: 624-634
Scheibe, K.M., Schleusner, Th., Berger, A., Eichhorn, K., Langbein, J., Dal Zotto, L. and Streich, W.J., 1998. ETHOSYS ® -new system for recording and analysis of behaviour of free-ranging domestic animals and wildlife. Applied Animal Behaviour Science 55: 195-211
Sillence, M., Noble, G. and McGowan, C., 2002. Fast food and fat fillies: the ills of western civilization. The Veterinary Journal 172: 396-397
Søndergaard, E. and Ladewig, J., 2004. Group housing exerts a positive effect on the behaviour of young horses during training. Applied Animal Behaviour Science 87: 105-118
Vincent, T.L., Newton, J.R., Deaton, C.M., Franklin, S.H., Biddick, T., McKeever, K.H., McDonough, P., Young, L.E., Hodgson, D.R. and Marlin, D.J., 2006. Retrospective study of predictive variables for maximal heart rate (HRmax) in horses undergoing strenuous treadmill exercise. Equine Veterinary Journal Supplement 36: 146-152
Werhahn, H., Hessel, E.F., Herman and Van de Weghe, H.F.A., 2012. Competition horses housed in single stalls (I): Behavior and activity patterns during free exercise according to its configuration. Journal of Equine Veterinary Science32: 45-52
Wyse, C.A., McNie, K.A., Tannahil, V.J., Murray, J.K. and Love, S., 2008. Prevalence of obesity in riding horses in Scotland. The Veterinary Record162: 590-591