Effect of seasonal conditions on fleece weight and its components in Merino sheep grazing Mitchell grass pasture in north west Queensland

Abstract

Animal Production in Australia 1998 Vol. 22 EFFECT OF SEASONAL CONDITIONS ON FLEECE WEIGHT AND ITS COMPONENTS IN MERINO SHEEP GRAZING MITCHELL GRASS PASTURE IN NORTH WEST QUEENSLAND P. M. PEPPERA, MARY ROSEA and G. M. McKEONB A B Department of Primary Industries, Animal Research Institute, Yeerongpilly, Qld 4105 Department of Primary Industries, PO Box 631, Indooroopilly, Qld 4068 SUMMARY Seasonal conditions in the pre to post natal period and selected periods before and during wool growth were described using climatic measures and estimates of the quality and quantity of pasture on offer derived from a validated pasture production model (GRASP). The variation in greasy and clean fleece weight, yield, staple length, fibre diameter, neck and side wrinkle score of Merinos grazing Mitchell grass in north west Queensland was explained in terms of these pasture and climatic measures and animal characteristics such as reproductive status, age and skin area. Multiple regression equations predicting clean and greasy fleece weight from the proportion of days in the wool growth period that the green pool in the pasture was less than one kg/ha, the percentage utilisation of the pasture, age, reproductive status and skin area of the ewes explained 87% and 79% of the variation respectively. Equations with similar predictors explained 58 to 85% of the variation of the other components. The inclusion of pasture conditions in the pre to post natal period did not significantly improve the predictions of the animals later performance. Keywords: Merinos, wool, Mitchell grass, seasonal conditions INTRODUCTION Rose (1974, 1982) reported the effects of age, year and lambing performance on greasy fleece weight, percentage yield, fibre diameter and staple length of an experimental flock grazing Mitchell grass pasture on Toorak Sheep Field Research Station, Julia Creek, Queensland. The year effect, which was the effect of year of measurement compounded with year of drop, was significant in this area of harsh and variable seasonal conditions (Rose 1982). The maturation of secondary follicles can be retarded if the nutrient supply is restricted in the late pregnancy of the ewe and wool growth can be affected if the young lamb is severely undernourished (Corbett 1979). This paper attempts to unravel the effects of seasonal conditions during the pre to post natal period from the period of actual wool growth. The longer term aim of the study is to predict fleece weight and its components from seasonal conditions and animal characteristics. Table 1. Description of wool and animal characteristics Variable gfw cfw yield sl fd nws sws sa age preg Description greasy fleece weights (kg) clean fleece weight (kg) yield (%) staple length (cm) fibre diameter (mm) neck winkle score side wrinkle score 2/3 area of skin estimated by 0.09 weight Years reproductive status: failed to lamb, lambed and (LR) lost (LL), lambed and reared Range 2.0-4.8 0.8-2.7 38-60 2.3-3.8 16-22 1.6-4.5 1.0-3.0 0.7-1.2 1.25-9.75 249 Animal Production in Australia 1998 Vol. 22 MATERIALS AND METHODS The management of the flocks and the environment have been described previously (Beattie 1961; Rose 1972). Some of the cohorts in this study were born in spring 1959 to 64 onto normally poor pasture in an area where the summer rainfall dominates. Other cohorts were born in autumn 1966 to 72 when more favourable conditions usually apply. There were 153 cohorts consisting of one to 120 ewes. As the distribution of rainfall rather than the total annual rainfall affects seasonal conditions, various climatic measures which a wool grower would be able to readily calculate were computed. In addition, a pasture production model, GRASP, (McKeon et al. 1990) validated for Mitchell grasses on clay at Toorak Research Station and Gilruth Plains (McKeon et al. 1990; Hall 1996) was used to predict the quality and quantity of pasture on offer. Numerous possible measures of quality and quantity of pasture on offer were computed using GRASP for both the pre to post natal period and selected periods in the previous two years before a shearing. The pre to post natal period in this study was defined as the five months prior to marking. Fleece weight and the other characters would be expected to be correlated. These were combined into meaningful groups from the correlation matrix and principal component analyses. Models relating these groups to the animal characteristics and the climatic and GRASP measures, some of which would be expected to have similar prediction value, were determined using the step-forward multiple regression method of successively including the next best predictor. In all analyses, the number of observations in each cohort was used as a weighting factor. The screening analyses, done on the scores from the principal component analyses, reduced the number of predictors and ensuring the same predictors (or a subset of them) for each set of correlated characters. A model for each character was then determined from these reduced number of predictors using step forward multiple regression. In this way, the best overall predictors could be determined for each set of correlated characters. RESULTS From the correlation matrix in Table 3 it can be seen that, as expected, cfw, gfw, and sl were highly correlated as were nws and sws. Fibre diameter showed some correlation with fleece weight (cfw and gfw) and with wrinkle score (nws and sws). Vector two of the principal component analysis of all variates (Table 4) represented a difference between cfw, gfw, sl and fd, nws, sws; vector three, yield. On the basis of these results and the correlations, the characters were grouped into three sets for further principal component analyses (Table 4) and for model fitting: Set 1 (cfw, gfw, sl), Set 2 (fd, nws, sws) and Set 3 (yield). The results of step forward multiple regression analyses of yield and the first vector of the Sets one and two on the possible predictors are given in Table 5. None of the GRASP measures for the pre to post natal period significantly improved the predictions. Table 2. Description of climatic and GRASP measures finally selected for pr ediction Variable rd i Description the rainfall in the growing season in the year of wool growth (mm) (the beginning of the growing season for Mitchell grass being defined as the date after 1 July when the rain over three days was greater than 10 mm and the average temperature o greater than 14 C and the end of the season by the last similar event before 30 June of the next year) the total rainfall in the current and previous growing season the number of days between the previous growing period to the beginning of the next the number of rain days in the growing period. average nitrogen content of pasture on offer over previous pasture growing season average nitrogen content of pasture on offer over the pasture growing season the average nitrogen of pasture in wool growth period kg/ha/day the proportion of days in wool growth period the green pool (leaf & stem) �1 kg/ha the proportion of days in wool growth period the green leaf �5 kg/ha % utilisation of pasture (total consumed/total growth) over the wool growth period Range 118-503 rd dry rddays gN1 gN2 avN sgp1 sgl5 util 419-890 187-273 10-32 6-15 6-15 7-15 0.07-0.54 0.18-0.62 10-43 250 Animal Production in Australia 1998 Vol. 22 Table 3. Corr elation matrix of fleece weight and its components Wool gfw yield sl fd nws sws Variate cfw .925 .428 .720 .418 .160 .166 All variates Vector 1 Vector 2 -0.19 -0.19 -0.10 -0.38 0.25 0.60 0.59 29.1% gfw .062 .698 .472 .111 .114 yield sl fd Nws .262 -.038 .093 .074 Set 1 variates .243 -.161 -.116 .448 .423 .867 Table 4. Principal component vectors Set 2 variates Vector 1 Vector 2 Vector 3 0.10 -0.26 0.88 -0.02 -0.36 0.12 0.11 15.1% Vector 1 -0.60 -0.59 -0.54 Vector 2 0.34 0.42 -0.84 -0.46 -0.63 -0.62 72.9% 0.89 -0.31 -0.35 22.7% cfw gfw yield sl fd nws sws % variation explained -0.54 -0.51 -0.19 -0.41 -0.38 -0.23 -0.24 43.1% 85.6% 11.8% Table 5. Additional % variation explained with successive best pr edictors included in multiple regression equations Predictors sgp1 + age + age + sa + util + preg 2 Vector 1 of Set 1 30.0% 22.7% 18.4% 6.7% 3.8% Predictors dry + gN2 + preg 2 + age + age + gN1 + sgl5 Vector 1 of Set 2 28.3% 28.9% 6.5% 13.5% 3.5% 1.1% Predictors Rd + avN + preg 2 + age + age Yield (Set 3) 46.7% 5.6% 3.3% 2.5% Prediction equations for fleece weight and its components from multiple regressions analyses using the final predictors from Table 5 were as follows:cfw gfw sl fd = 0.93�.14 = 1.85�.33 = 3.17�.17 = 20.43�.45 -1.83�.08 sgp1 -.107�.032 LL -3.09�.19 sgp1 -.185�.074 LL -1.07�.10 sgp1 -.082�.039 LL -1.55�.21 LL +.118�.029 gN2 -.0164�.00079 dry -.200�.070 LL -0.013�.0006 dry -.022�.0031 age2 +.0081�.0014 rd +.59�.29 age +1.76�.16 sa -.161�.024 LR +2.83�.37 sa -.289�.056 LR +.94�.20 sa -.109�.030 LR +.081�.024 age +.215�.055 age -.058�.030 age -.0111�.0021 age2 -.014�.0012 util (86.6% variation explained) -.0252�.0048 age2 -.013�.0027 util (79.3% variation explained) -.0028�.0026 age2 -.008�.0015 util (76.1% variation explained) -2.18�.55 sgl5 -.057�.011 age2 (59.8% variation explained) +0.354�.039 age -.030�.0036 age2 (84.6% variation explained) +.269�.033 age (78.0% variation explained) -1.39�.37 LR (58.1% variation explained) -1.99�.15 LR +.71�.12 age -.0078�.0024 dry +.248�.012 gN2 -.406�.051 LR +.148�.010 gN2 -.179�.058 LL +.368�.088 avN -.068�.027 age2 +.143�.011 gN1 -1.40�.19 sgl5 +.063�.009 gN1 -.288�.042 LR -0.74�.51 LL n w s = 3.29�.19 sws = 2.44�.15 yield = 41.77�.87 Restraining the predictors to the same ones for each set lost little in effectiveness of prediction. A similar procedure was undertaken to determine the effectiveness of the climatic measures only. cfw = -0.14�.21 +3.09�.23 sa -.056�.007 age -.00376�.00048 dry (56.5% variation explained) 251 Animal Production in Australia 1998 Vol. 22 gfw sl fd = 0.19�.44 = 2.47�.19 = 20.82�.48 +5.21�.48 sa +1.66�.24 sa -.0088�.0019 dry -.059�.012 age2 -.0119�.0011 dry -.196�.126 LL -.0086�.0007 dry -.228�.082 LL +.0122�.0010 rd +.0005�.0350 age 2 -.085�.015 age -.146�.033 age -1.51�.22 LL +.0015�.0004 rdi -.409�.090 LR +.0014�.0002 rdi -.346�.059 LR -.49�.53 LL +9.35�2.62 sa -.00593�.00102 dry (44.2% variation explained) -.00156�.00043 dry +.0045�.0029 age2 (60.7% variation explained) -1.99�.16 LR +0.72�.13 age (52.4% variation explained) n w s = 4.74�.27 sws = 2.88�.18 +.367�.071 age -.031�.007 age2 (48.4% variation explained) +.289�.047 age -.024�.004 age2 (55.8% variation explained) -.95�.40 LR -.27�.40 age (56.8% variation explained) yield = 35.40�2.14 DISCUSSION For greasy and clean fleece weight and staple length a combination of animal characteristics (age, reproductive status and skin area) and pasture measures (proportion of days in the wool growing period when the green pool was less than one kg/ha and the percent utilisation of the pasture) gave reasonable predictions. Skin area could be estimating intake or alternatively, the larger ewes could just be growing more wool. The amount of variation explained, 76 to 87%, was reduced to 44 to 61% if only climatic measures and animal characteristics were considered. For fibre diameter and wrinkle scores the percentage variation explained dropped from 60 to 85% to 48 to 56%. It was thus concluded the inclusion of GRASP measures such as the proportion of days in the wool growth period for which the green leaf or green pool is less than a threshold and the utilisation of the pasture, could improve the predictions; a result similar to that was reported by McCown et al. (1981) for cattle. Yield was poorly estimated with and without the GRASP measures (58 and 57%). Common predictors, the rainfall over the last two growing seasons, reproductive status and age, explained 53.4% variation. The average nitrogen in the pasture over the wool growth period was replaced by skin area when only climatic measures and animal characteristics were considered. The explanation of the inclusion of skin area could be that it is correlated with another unmeasured factor which does affect yield. Kelly et al. (1996) found that lambs of ewes on a submaintenance diet produced 0.1 kg less clean wool at lamb shearing, at 0.4 years of age, than those of ewes on a maintenance diet (P<0.01) while at hogget shearing, at 1.4 years of age, the difference was not significant (P=0.10). Their hogget wool was 0.1 mm broader (P<0.05) although there was no significant differences in yield, staple length and strength. The early disadvantage of the progeny of ewes on submaintenance diets was largely overcome by hogget shearing. The ewes in our study ranged in age from 1.25 to 9.75 years. Because the GRASP estimates of pasture quality and quantity over the pre to post period did not significantly improve prediction, it appears that nutritional status of the pasture during the wool growth period overshadowed any effect the pre to post natal nutrition may have had. Corbett (1979) hypothesised that, with the rate of wool growth directly related to feed intake, the undernutrition of young sheep, unless exceptionally severe, will have little effect on their subsequent ability to grow wool. Being able to predict fleece weight and its components in different seasons could be a useful management tool for wool growers especially if they could easily obtain the necessary predictors. REFERENCES BEATTIE, A.W. (1961). Qld. J. Agric. Sci. 18, 437-445. CORBETT, J.L. (1979). In Physiological and Environmental Limitations to Wool Growth. (Eds J.L. Black and P.J. Rees) pp. 79-98 (University of New England Publishing Unit: Armidale). HALL, W.B. (1996). Near real time financial assessment of the Queensland wool industry on a regional basis. PhD Thesis, University of Queensland, Brisbane, 465 pp. KELLY, R.W., MACLEOD, J., HYND, P. and GREFF, J. (1996). Aust. J. Exp. Agric. 36, 259-67. McCOWN, R.L., GILLARD, P. WINKS, L. and WILLIAMS, W.T. (1981). Agr. Systems 6, 303-17. McKEON, G.M., DAY, K.A., HOWDEN, S.M., MOTT, J.J., ORR, D.M. SCATTINI, W.J. and WESTON, E.J. (1990). J. Biogeography 17, 355-72. ROSE, Mary (1972). Proc. Aust. Soc. Anim. Prod. 9, 48-54. ROSE, Mary (1974). Proc. Aust. Soc. Anim. Prod. 10, 367-371. ROSE, Mary (1982). Proc. Aust. Soc. Anim. Prod. 14, 463-466. 252