Exposure is one of the greatest contributors to neonatal lamb mortality on outdoor lambing systems in the UK. Effective shelter has been predicted to reduce neonatal lamb mortality rates by up to half during periods of inclement weather. However, it is unclear how shelter provision influences the prevalence of other shepherding problems such as dystocia, ewe mismothering behaviour and poor lamb vigour.Periods of high wind speed and rainfall have been shown to influence shelter usage, however, it is not yet known how ewe factors such as breed, age and body condition score influence shelter-seeking behaviour. This study conducted on a working upland farmin the UK, was the first trial to quantify these variables. Firstly, it assessed the impact of shelter provision on the prevalence of shepherding problems. Secondly, it examinedthe and biological factors that influence individual ewe behaviour around natural and artificial shelter.Lambing ewes (n=147) were allocated on the basis of breed then agebetween two adjacent fields(total 6.3 hectares/stocking density 23 sheep/hectare); one field had existing natural shelter that was reinforced with three artificial shelters; the other exposed field was used as a control. Individual ewes were observed every 2 hours between 0800-1600 for 14 continuous days to monitor their location relative to shelter. Ewe breed (Aberfield and Highlander), age (2 to 8 years) and body condition score were considered as explanatory variables to explain flock and individual variance in shelter-seeking behaviour and the prevalence of shepherding problems. Any ewe observed with dystocia, a dead or poor vigour lamb or who exhibited mismothering behaviour was recorded as a shepherding problem.Mood’s median and Kruskal-Wallis testswere used to see how variables such as field allocation, the presence of a lamb and ewe breed, age and BCS influenced a ewe’s preference for shelter.Chi-square tests were used to see how the prevalence of shepherding problems varied between fields, breeds, age categories and ewe BCS. Windspeed (P=0.007) and the presence of a lamb (P=0.003) were highly significant in influencing ewe behaviour. Overall, ewesacross both breedsthat were given access to shelter experienced fewer shepherding problems than those without (P=0.048). This was reflected in a breed effect, with Highlander ewes significantly more likely to seek shelter than Aberfield ewes(P=0.001),and presentingsignificantly fewershepherding problems (P=0.035).
This was a relatively low power study with a narrow range of climatic variables, therefore, for shelter to still have a significant influence in reducing the prevalence of shepherding problems is an important finding. A proven reduction in shepherding workloadthrough the provision of shelter provides an industry incentive to adopt similar shelter interventions on upland farms in the UK.