ALBERT

Notes: Grazed sward surface height was controlled within the range 3·25–4·75 cm during spring and summer in two experiments. In Experiment 1, the effects of stocking two breeds of ewe of similar size but different potential levels of reproductive performance [Brecknock Cheviot (C) and Beulah Speckled Face (B)] at different annual stocking rates of twelve (SR12) and twenty (SR20) per hectare, rates of nitrogen fertilizer of 100 (N100) and 200 (N200) kg N ha−1 annum−1 and different lamb:ewe ratios (C1·2, B1·2 and B1·5) were measured in four treatments (SR20N200C1·2; SR20N200B1·2; SR20N200B1·5; SR12N100C1·2) replicated three times. In each of three years animal performance and yield of silage from areas of pasture surplus to grazing requirements were measured. In Experiment 2, breed B was compared with the Welsh Mule (W) breed, a larger with a higher potential reproductive performance, at two stocking rates, two rates of nitrogen fertilizer and two lamb:ewe ratios set on the basis of results from Experiment 1 (SR18N200B1·5; SR12N100B1·5; SR18N200W1·5; SR18N200W1·7). The treatments were replicated three times. The same terminal sire (Suffolk) was used in both experiments. A primary aim of the experiments was to test the validity of the experimental procedures used for comparing breeds of sheep where nutrition is provided predominately from grazed pastures. In Experiment 1, there was no difference between breeds C and B in the live weights of individual lambs at weaning at the same SR (20), N rate (200) and lamb:ewe ratio (1·2). Breeds C and B produced similar total yields of lamb (633 kg lamb ha−1± 10·5) and silage (193 kg DM ewe−1± 37·7), but breed B had a higher level of potential reproductive performance (1·59 vs. 1·37 lambs ewe−1: P 〈 0·001). The treatments SR20N200B1·5 and SR12N100C1·2 produced, respectively, greater and lesser yields of lamb (725 vs. 384 kg lamb ha−1, P 〈 0·001) and lesser and greater yields of silage (123 vs. 327 kg DM ewe−1, P 〈 0·001). In Experiment 2, the live weight of lambs at weaning from breed W were heavier than from breed B (29·1 vs. 26·2 kg lamb−1, P 〈 0·01) but there was no significant difference in total yield of lamb weaned between breeds W and B at the same SR (18), N rate (200) and lamb:ewe ratio (1·5) (747 kg lamb ha−1± 19·2), or in the yield of silage (66 kg DM ewe−1± 16·4), but breed W had a higher potential reproductive performance (1·85 vs. 1·58 lambs ewe−1, P 〈 0·05). The treatments SR18N200W1·7 and SR12N100B1·5 produced, respectively, greater and lesser yields of lamb (840 vs. 473 kg lamb ha−1, P 〈 0·001) and similar and greater yields of silage (60 vs. 141 kg DM ewe−1, P 〈 0·05). The experimental approach adopted and the management protocols used provided a basis for ranking the performance of the breeds of ewes examined at appropriate levels of annual stocking rate, N-fertilizer input and lamb:ewe ratio.