Mortality data on 5163 Romney X Suffolk lambs were collected at Winchmore Irrigation Research Station from 1976 to 1.981. M?rtality from birth to tailing (25 days) for single, twm, and triplet births was 14.1, 14.7, and 33.0% with mean birthweights of 5.2, 4.1, and 3.3 kg respectively. Ninety-four percent of all deaths occurred within 72 h ofbirth. Dystocia accounted for 50.4% of all single deaths and starvation/exposure for 45.0% ofall multiple deaths. The proportion ofewes assisted at lambing increased with birthweight above 4.5 kg to reach almost 19 and 8% for single and twin births respectively. If lambs that were assisted at birth, but survived to weaning, are regarded as potential deaths, the overall potential mortality of single and twin lambs was 27.9 and 20.2% respectively. Ewe liveweight gain during the last 6 weeks of pregnancy was associated with increased birthweights of single and twin lambs (P< 0.05). A liveweight change of approxi.mately 10 kg in ewes was required to alter birthweight by 0.5 kg in both single and twin lambs. Single l~mb mortality was not significantly affected by liveweight gain or ewe livewei~ht at lamb~ng on.either a between-mob or a within-mob basis. ThIS suggests that with single bearing ewes it may be more economical to save feed for lactation when feed requirements of ewes are high, rather tha~to offer it in late pregnancy. However, when multiple bearing ewes were offered additional feed in la~e pregnancy to increase liveweight, lamb mortality was reduced, suggesting merit in improved pre-lamb feeding for these ewes.

No abstract

The usefulness of udder score, liveweight, and condition score measurements and their changes during pregnancy for predicting ewe litter size was tested using discriminant analysis on records of 7 058 ewes from 13 flocks covering a range of breeds. When each variable was used individually to discriminate between ewes bearing singles and those bearing twins, udder score was the best predictor variable with 65% being correctly classified Prelambing liveweight was the next best predictor with 62% correctly classified, and mating liveweight, liveweight change during gestation, and condition score were similar,averaging 58%correctly classified. When all variables were used together in a combined index, the accuracy of prediction was only slightly improved (68%) over that achieved by udder scoring alone. For on-farm use, udder scores or prelambing liveweights could potentially be used to correctly classify about two-thirds of ewes.