In three successive years, sward height was maintained at 3, 5, 7 or 9 cm on grass swards receiving a total of 300 kg N ha " ' in six equal monthly dressings from April, and on grass/ clover swards receiving 50 kg N ha "' as a single dressing in early spring. From turnout in April until weaning in July, 64 ewes and their lambs (mean litter size 1 -5) were continuously grazed at the four sward surface heights on the two sward types. White clover content of grass/ clover swards remained low throughout the experiment ranging from 0-2 to 7-4% of the herbage mass. During the first two years, lamb gains averaged over sward types were 204, 260, 285 and 308 g d"' up to weaning, while in the third year gains were 238, 296, 296 and 260 g d"' on 3, 5, 7 and 9 cm swards respectively. Ewes lost Hve weight on 3 cm swards but apart from this sward height had little effect on performance. During the autumn, weaned lambs gained -27, 87, 147 and 167 g d " ' on 3, 5, 7 and 9 cm swards respectively. Sward type had only a small effect on the performance of lambs up to weaning but in the autumn, mean gains of weaned lambs were lower on grass/N swards (73 g d "') than on grass/clover swards (115 gd"'). Relative to 3 cm swards, carrying capacities of 5, 7 and 9 cm swards were 0-76, 0-57 and 0-52 respectively from turnout to Correspondence: D. M. B. Chestnut!, Agricultural Research Institute of Northern Ireland, Hillsborough, County Down, BT26 6DR, UK. t Also a member of staff of the Department of Agriculture for Northern Ireland and the Queen's University, Belfast. 70weaning and 0-66, 0-52 and 0-44 respectively during autumn. Grass/clover swards carried 0 • 67 of the ewes carried by grass/N swards from turnout to weaning and 0-51 of the live weight carried by grass/N swards during autumn.The reaction of the two sward types to sward height did not appear to differ but in the third year there was evidence of a reduction in white clover content when swards were grazed at 9 cm. The data suggest that lamb growth rates will increase as sward height increases up to 9 cm and the evidence for this was stronger with weaned lambs in autumn than with suckling lambs in spring.

A small-plot experiment was made to assess the influence on dry-matter output from grass of a wide range of fertilizer nitrogen and defoliation interval treatments. There were five defoliation treatments, 22, 28, 45, 75 and 112-day regrowth intervals throughout the growing season each at six levels of nitrogen application, ranging by 300 kg increments from 0 to 1500 kg/ha/year.There was a marked interaction effect between treatments; a positive dry-matter response was maintained to a higher level of applied nitrogen with more frequent defoliation. In 2 years out of 3 maximum dry-matter yield was produced under a 75-day defoliation interval although the mean yield advantage over a 45-day defoliation system was only 11 %. Mean yield of digestible dry matter appeared to reach a maximum under a 45-day defoliation interval at 600 kg N/ha but at the lower levels of N the maximum yield was reached at the longest growth interval.Seasonal response to nitrogen under the 22-and 28-day defoliation systems measured as the increase in yield resulting from increased N at each cutting date reached its peak in July-August. Application for these short growth periods early and late in the growing season appeared to be a relatively inefficient use of nitrogen.The less frequently the sward was harvested and the higher the nitrogen application the greater was the reduction in ground cover as estimated by eye at the end of the growing season, this reached an estimated 25% reduction under 112-day defoliation at 300 kg N/ha/year.In relation to published figures nitrate content of herbage did not reach dangerous levels until nitrogen application reached levels beyond those at which maximum dry-matter yield was achieved.

Texel (T) and Suffolk (S) sired ram, wether and ewe lambs (no = 180) were kept at grass with their dams until weaned at 20 weeks and then taken to slaughter weights of 40, 44 and 48 kg, also at grass. Sex type, but not sire breed, affected lamb growth rate (rams 293, wethers 253, ewes 224 g/day; P < 0·001). Both T and S ewe lambs were fatter at slaughter in all fat depots compared with their ram and wether siblings such that rams could be slaughtered at a calculated 7·7 kg greater live weight than ewes at equal carcass fat cover. T carcasses were greater than S carcasses in eye-muscle area (793 v. 732 mm2; P < 0·001) and killing-out proportion (481 v. 476 g/kg; P < 0·05) but not in any fat measurement. Slaughter weight influenced killing-out proportion (P < 0·001) and all fat measurements (P < 0·01) but did not significantly affect eye-muscle area.Mean serum insulin-like growth factor-1 (IGF-1) concentrations (μig/l), determined during weeks 8,11, 14, 17 and 20 in a balanced subset of 84 lambs, were higher in T than in S lambs (P < 0·05), in singles than in twins (P < 0·01; rams and wethers only) and in rams than in either wethers or ewes (P < 0·001). Mean serum IGF-1 concentration decreased between week 8 and week 20 with a greater rate of decline in singles than in twins (P < 0·001). IGF-1 was more strongly correlated with live weight at 8 weeks (r = 0·629; P< 0·001) than at 20 weeks (r = 0·293; P < 0·05). Mean IGF-1 (weeks 8 to 20) was correlated with rate of live-weight change (r = 0·576; P < 0·001). Significant differences for T and S lambs were found in the relationships between mean serum IGF-1 (weeks 8 to 20) and daily live-weight gain (weeks 8 to 20) and between mean IGF-1 and eye-muscle area at slaughter. Metabolite concentrations differed little between sire breeds and not at all between sex types.

The effects of continuous stocking by sheep at sward surface heights (SSH) of 3, 5, 7 and 9 cm in grass/clover (GC) and nitrogen‐fertilized grass (GN) swards were examined in relation to herbage mass and quality, clover content, tiller density and rates of herbage production and senescence in two periods in each of three grazing seasons (1987‐89). The GN swards received a total of 300 kg N ha−1 each year in six equal dressings from March; GC swards received a single dressing of 50 kg N ha−1 in March each year. Herbage mass measured from ground level increased linearly with SSH with overall mean herbage masses of 0·89, 1·38, 1·78 and 2·12 t OM ha−1 (s.e.m.0·024, P < 0·001) at SSH of 3, 5, 7 and 9 cm respectively. GN and GC swards had mean herbage masses of 1·58 and 1·51 t OM ha−1 (s.e.m. 0·051, NS) respectively. Mean N content of herbage on GN swards was greater than that on GC swards and declined with increasing SSH. Crude, fibre (CF) content of herbage was similar for both sward types and increased with increasing SSH. Clover content of GC swards remained low throughout the experiment, ranging from 0·002 to 0·074 of herbage mass. However, from tissue turnover rates it was estimated that its contribution to herbage production was in the range of 0·049–0·219 of net herbage growth. Total growth increased with increasing SSH in both sward types, with maximum growth rates in GN swards of 143 and 130 kg DM ha−1 d−1 and in GC swards of 88·2 and 85·4 kg DM ha−1 d−1 in Periods 1 (up to early July) and 2 (after July) respectively. Senescence rates ranged between 13·3 and 50·1 kg DM ha−1 d−1 and tended to be higher in Period 2 than in Period 1. Net production increased with increasing SSH in Period 1, while in Period 2 net production declined at SSH above 6·5 cm. The increased net herbage production in taller swards was not associated with greater utilized metabolizable energy production at sward heights above 5 cm.

Over three grazing seasons (1984)(1985)(1986)) a sward of perennial ryegrass, cv. Talbot, which received a total of 336 kg N ha" ' each season, was cut or grazed with ewes at 3-or 4-week intervals on a rotational basis.Sward productivity was higher under cutting than under grazing irrespective of the interval between defoliations. Under cutting, mean herbage organic matter (OM) yields over both intervals were 8-66, 9-62 and 8-17 t ha" ' in 1984, 1985 and 1986 respectively while under grazing the corresponding yields were 7-65, 8 • 63 and 7 • 501 ha " '. The mean annual yield of herbage defoliated at 3-week intervals was 7-50, 8 • 64 and 7 • 201 OM ha "' compared with 8-80. 9-60 and 8 461 OM ha"' for swards defoliated at 4-week intervals in the three years respectively.The nitrogen (N) content of both the available and the residual herbage was consistently higher under grazing than under cutting. Available herbage contained 31-3 and 27-7 g N kg OM "' and residual herbage 26-1 and 22-7 g N kg OM "' under grazing and cutting respectively.The mean yield of N under cutting was 284 kg ha"' compared with 304 kg ha"' under grazing. Defoliation interval had no effect on N yield, the overall mean yield being 294 kg ha "' under both 3-and 4-week defoliation intervals. The effect of the treatments on tiller population was slight and inconclusive.