Abstract:
BODYWEIGHT AND OVULATION RATE IN SHEEP By T. N. EDEY* Summary The relationship between bodyweight at the time of ovulation and ovulation rate has been investigated using 715 observations on mature Merino ewes. Below 35-37.5 kg, ovulation rate was approximately lOSc/;, and varied little with bodyweight. Above this level, for each weight increase of 2.5 kg there was at least a 5% increase in ovulation rate up to 53.5 kg, and at least a 10% increase per 2.5 kg within the range 40.4-48.4 kg. I. INTRODUCTION Heape (1899) concluded that there was a close positive relationship between the weight of ewes at mating and number of lambs produced, but his concept was substantially overlooked for sixty years in studies of pre-mating nutrition (Moule 1962). The relationship between bodyweight (B.W.) and ovulation rate has not been examined in detail in the ewe, though a positive relationship was long ago established in the rabbit (Gregory 1932), bank vole (Brambell and Rowlands 1936), wild brown rat (Perry 1945), mouse (MacArthur 1949) and more recently the snowshoe hare (Newson 1964). In the ewe, Wallace (1961), Allen and Lamming (1961) and Killeen (1967) observed that, independently of a bodyweight increase before mating due to increased feed supplies (flushing), heavier ewes produced more ova than lighter ones, but the small number of ewes prevented detailed analysis of the nature of the relationship. In the present paper, ovulation data collected over four seasons have been analysed for this purpose. II. MATERIALS AND METHODS (a) Source of Data During the period 1964-67, mature Peppin Merino ewes were purchased through local saleyards for use in studies of embryo mortality. In 11 such experiments, during the months March to May, ewes of known B.W. at the time of ovulation were examined either by laparotomy or after slaughter so that corpora lutea could be counted. In Table 1, details of the origin of 715 such observations are given, 72% of them being made in May. Ewes remained in the flock for one to three years, for a mean of 1.6 observations per ewe. In all experiments, the aim was to have B.W. stable at the time of ovulation, and the small deviations from this condition are indicated in the table. In eight experiments, (618 observations), the sheep were at pasture; the remainder were fed in pens. *Department of Livestock Husbandry, University of New England. Armidale, N.S.W. 188 TABLE 1 Source of Data All observations were pooled and, using B.W. at time of ovulation, the ewes were classified into 2.5 kg B.W. intervals. From the corpora lutea counts, the ovulation rate (ovulations per 100 ewes) for each class was calculated. Using the method of orthogonal polynomials, the regression line of best fit was calculated. Gradients of various sectors of the line were then examined. III. RESULTS Ovulation rates for the various B.W. classes, and the number of ewes in each class are shown in Figure 1. The number of ewes per class falls below 30 only at the extremities of the distribution. The regression of ovulation rate (y) on bodyweight (x) was and the regression line is plotted in Figure 1. Analysis of variance showed that although the linear component accounted for a high proportion of the variance (P < O.OOl), the cubic component was also significant (P < 0.01). In Figure 1, the histogram of the original data indicates that was relatively constant below 37.5 kg B.W. and then increased apparently reaching a plateau at about 52.5 kg. The fitted curve the ovulation rate began to ,rise at about 35 kg and approached a 55 kg. . ovulation rate rapidly before indicated that plateau above Table 2 was compiled by plottin g the gradient of the curve, (gradient)x = 10.93 + 2(0.362)x - 3(0.155)x2, and it shows the B.W. intervals over which specified increases of ovulation rate might be expected. Thus, between 35.3 and 53.5 kg, each 2.5 kg increase in B.W. appeared to produce at least a 5% increase in ovulation rate. The maximum gradient of more than 10% per 2.5 kg increase occurred between 40.4 and 48.4 kg. A 3% gradient still existed at 55 kg but the numbers of ewes above this weight were considered inadequate to allow firm conclusions to be drawn about the real nature of the curve in this area. 189 IV. DISCUSSION These results indicate that, in a fairly heterogeneous sample of Merino ewes, there was a constantly low twin-ovulation rate below bodyweights of 35-37.5 kg. Coop (1962) found little twinning in New Zealand Corriedales below about 41 kg, and that between approximately 41 and 64 kg B.W. there was a linear increase of about 3.3% in twinning rate (based on lambs born) for each 2.5 kg B.W. increase at mating. In the present work, the relationship was not linear, but over the range 35.3-53.5 kg there was at least a 5% gradient. The higher gradient in our data as compared with Coop's is partly explicable in terms of the expected losses due to failure of fertilization, and embryonic and neonatal deaths, which would occur before the data would become comparable to Coop's. In parTABLE 2 Bodyweight (B. W.) ranges over which specified increases in o,vulutio+z rate might be expected 190 titular, relatively high losses could be expected amongst twin ovalators at the embryonic (Edey 1966) and neonatal stages (Thomson and Thomson 1949; Purser and Young 1964). However, lambing data are not available for a sufficient proportion of ewes in this work to allow comparison of ovulation and lambing performance. No attempt has been made to assess objectively the degree of fatness of the ewes, but the apparent plateau in ovulation rate at the highest bodyweights could be partly explained if most ewes in this class had approached maximum bodyweight. The conclusions reached in this study will not be directly applicable to ewes with a different frame size and/or different genetic potential for ova production, but it would appear safe to conclude that in many circumstances, there will be a weight range of 15-20 kg over which bodyweight increments at mating will give important increases in ova production. V. ACKNOWLEDGMENTS Thanks are expressed to Dr. V. Bofinger, Department of Mathematics, University of New England for assistance with the statistical analysis. VI. REFERENCES A LLEN, D. M., and L AMMING, G. E. ( 1961). J. agric. Sci., Camb. 56: 69. B RAMBELL , F. W. R., and R OWLANDS , I. W. ( 1936). Philos. Trans. B, 226: 71. C OOP, I. E. ( 1962). N.Z. Jl. agric. Res. 5: 249. E DEY, T. N. (1966). J. agric. Sci., Camb. 67: 287. G REGORY , P. W. ( 1932). J. exp. Zool. 62: 27 1. HEAPE, W. ( 1899). JZ. R. agric. Soc. 10: 217. K ILLEEN, I. D. ( 1967). Aust. J. exp. Agric. Anim. Hub. `7: 126. M ACA RTHUR . J. H. (1949). Genetics 34: 194. M OULE, G. R. ( 1962). Proc. Aust. Soc. Anim. Prod. 4: 195. NEWSON, J. (1964). Can. J. 2001. 42: 987. P ERRY, J. S. ( 1945). Proc. 2001. Soc. Lond., 115: 19. P URSER , A. F., and Y OUNG, G. B. (1964). Anim. Prod. 6: 321. T HOMSON, A. M., and THOMSON, W. (1949). Br. J. Nutr. 2: 290. W ALLACE , L. R. ( 196 1) . Proc. kakura Fmrs'. Conj. Week. p. 14. 191