The Effect of Time of Shearing on Wool Production and Management of a Spring-Lambing Merino Flock

The Effect of Time of Shearing on Wool Production and Management of a Spring-lambing Merino Flock

Angus John Dugald Campbell, BVSc(Hons) BAnimSc

Submitted in total fulfilment of the requirements of the degree of Doctor of Philosophy

November 2006

University of Melbourne School of Veterinary Science

ABSTRACT

Choice of shearing time is one of the major management decisions for a wool-producing Merino flock and affects many aspects of wool production and sheep health. Previous studies have investigated the effect of shearing on only a few of these factors at a time, so that there is little objective information at the flock level for making rational decisions on shearing time. This is particularly the case for flocks that lamb in spring, the preferred time in south-eastern Australia.

A trial was conducted in a self-replacing, fine wool Merino flock in western Victoria, from January 1999 to May 2004, comparing ewes shorn annually in December, March or May. Within each of these shearing times, progeny were shorn in one of two different patterns, aligning them with their adult shearing group by 15–27 months of age.

Time of shearing did not consistently improve the staple strength of wool. December-shorn ewes produced significantly lighter and finer fleeces (average 19.1 µm, 3.0 kg clean weight), whereas fleeces from March-shorn ewes were heavier and coarser (19.4 µm, 3.1 kg). Fleeces from ewes shorn in May were of similar weight to fleeces from March-shorn ewes (3.1 kg), but they were of significantly broader diameter (19.7 µm). In young sheep, beneficial changes in some wool characteristics for each shearing group were offset by undesirable changes in others.

Shearing ewes in March or May, and weaners in March, May or June, significantly increased the risk of post-shearing mortality about three- and four-fold, respectively, compared to unshorn sheep. Substantial, highly significant associations in young sheep between post-weaning mortality, bodyweight and growth rate were also quantified using various survival analysis techniques. For example, the lightest 20% of weaners at weaning contributed 31% of all deaths in the year following weaning, and increasing average growth rate over summer and autumn from 250 to 500 g/month reduced the risk of death by 74%. These results could be used to develop supplementary feeding systems that efficiently reduce weaner mortality, which is a significant animal welfare issue in many Australian Merino flocks.

Mortality effects were incorporated into estimates of the total value of wool produced by the different shearing times between birth and culling at 6¼ years of age. Using median historical (1991–2006) wool prices, shearing ewes in March and their progeny first in June, or October (weaner)-December (ewe) shearing produced the greatest total value of wool ($111/head). March (weaners)-March (ewes) shearing had a wool value of $107/head and December (weaners)-December (adults) shearing $103/head. May-shorn ewes produced the smallest value of wool, irrespective of whether their progeny were first shorn in May or July ($93–96/head).

No shearing time consistently improved all animal health measures. May-shorn ewes had significantly more fleece rot in late autumn than the other shearing groups (odds ratio 2.5) and were up to 0.4 condition score lighter during winter, although they had a lower cost of dag (average $0.64/head) and significantly less breech strike risk in spring, compared to December-shorn ewes (odds ratio 0.18). December-shorn ewes had the greatest cost of dag ($1.50/head). March-shorn ewes had an intermediate cost of dag ($1.03/head) but significantly less breech strike than May-shorn ewes (odds ratio 0.38).

Overall, December and March shearing were shown to be appropriate alternatives for a self-replacing Merino flock in south-eastern Australia, whereas May was an undesirable shearing time.

DECLARATION

This is to certify that:

(i) the thesis comprises only my original work

(ii) due acknowledgement has been made in the text to all other material used,

(iii) the thesis is less than 100,000 words in length, exclusive of tables, maps, bibliographies and appendices

Angus JD Campbell

iii

All we like sheep have gone astray; we have turned every one to his own way; and the Lord hath laid on him the iniquity of us all.—Isaiah 53:6

ACKNOWLEDGEMENTS

I wish to thank the Vizard Foundation and Mr Maurice and Mrs Jill Glover, who generously made their farms, facilities and sheep available for this trial, and the staff of ‘South Roxby’, particularly Mr Peter Lindeman, for the day-to-day management of the sheep. This work received financial support from the Vizard Foundation and the Australian Sheep Industry Cooperative Research Centre (CRC). The CRC also kindly supported me with a top-up scholarship.

I am very grateful to my supervisors, Dr John Larsen and Associate Professor Andrew Vizard, for their guidance and assistance throughout the study. Associate Professor Vizard devised and designed the trial and it was overseen by Dr Larsen, along with other members of the Mackinnon Project at the University of Melbourne, from December 1998 to January 2002, when I took over. Mr Garry Anderson happily and ably provided me with guidance for the statistical analyses that I performed, for which I am very thankful.

The amassing of such a large amount of data is due in no small part to the dedicated and meticulous work of Ms Dianne Rees. I will be eternally grateful for her technical assistance and cheerful smile, which made 5am starts on shearing days bearable.

I would also like to thank the following people, who generously provided me with information or assistance during my work: Mr Rod Agar (Australian Wool Testing Authority Ltd) for helping with processing of the dyebanded wool samples; Dr Renick Peres (Department of Primary Industries, Geelong) for providing the weather data; and Dr Roger Thompson (Pasture and Veterinary Institute, Hamilton) for kindly providing me with the results of his lamb shearing experiments.

This work is dedicated to my partner, Sarah, for her abiding love and support, and to my son, Callum, for making anything I do worthwhile.

vii

TABLE OF CONTENTS

Abstract ...... ii Declaration...... ii Acknowledgements...... ii Table of Contents...... ii List of Tables...... ii List of Figures...... ii C h a p t e r 1 General Introduction ...... 2 1.1 Background...... 2 1.2 Determinants of Wool Enterprise Profitability & Their Relationship to Shearing Time...... 2 1.2.1 Stocking Rate...... 2 1.2.2 Wool Production per Sheep ...... 2 1.2.3 Average Price per Kilogram of Wool ...... 2 1.2.4 Enterprise Costs...... 2 C h a p t e r 2 Literature Review ...... 2 2.1 The Effect of Shearing Time on Wool Quality...... 2 2.1.1 Staple Strength & Position of Break...... 2 2.1.1.1 Components of Staple Strength & Position of Break...... 2 2.1.1.1.1 Fibre Diameter Properties...... 2 2.1.1.1.2 Intrinsic Fibre Strength...... 2 2.1.1.1.3 Fibre Shedding ...... 2 2.1.1.2 Associations between Staple Strength, Position of Break & Shearing Time ...... 2 2.1.1.3 Shearing Time & Recurring Factors Affecting Staple Strength...... 2 2.1.2 Fleece Weight, Yield & Fibre Diameter ...... 2 2.1.3 Staple Length...... 2 2.1.4 Vegetable Matter Contamination...... 2 2.1.5 Time of Shearing & Wool Quality—Summary ...... 2 2.2 The Effect of Shearing Time on Flock Management & Sheep Health ...... 2 2.2.1 Sheep Nutritional Requirements & Farm Stocking Rate ...... 2 2.2.2 Reproductive Efficiency ...... 2 2.2.2.1 Shearing During Mating ...... 2 2.2.2.2 Shearing & Lamb Birthweight...... 2 2.2.2.3 Shearing & Ewe Sheltering Behaviour ...... 2 2.2.3 Fleece Rot...... 2 2.2.4 Dag ...... 2 2.2.5 Flystrike...... 2 2.2.6 Other Health Issues...... 2 2.2.7 Mortality...... 2 2.3 Mortality of Weaner Merino Sheep ...... 2 2.3.1 The Extent of Merino Weaner Mortality in Australia...... 2 2.3.2 Limitations of Weaner Mortality Investigations...... 2 2.3.3 Causes of Mortality & Illthrift...... 2 2.3.4 Factors Associated with Weaner Mortality...... 2 2.3.4.1 Bodyweight...... 2 2.3.4.2 Growth Rate...... 2 2.3.4.3 Bodyweight & Death from Causes Other Than Malnutrition ...... 2 2.3.4.4 Physiology Underlying the Association between Mortality & Bodyweight ...... 2 2.3.4.5 Season & Year of Birth...... 2 2.3.4.6 Disease...... 2 2.3.4.7 Maternal Factors ...... 2 2.3.4.8 Sex ...... 2

2.3.5 Weaner Mortality—Summary ...... 2 2.3.6 Methodologies for Analysing Survival & Mortality...... 2 2.4 Conclusion...... 2 C h a p t e r 3 Experiment Introduction and Materials & Methods...... 2 3.1 Introduction & Trial Overview...... 2 3.2 Experimental Site ...... 2 3.3 Animals...... 2 3.4 Flock Management ...... 2 3.5 Shearing...... 2 3.5.1 Periodic Wool Growth...... 2 3.6 Measurement ...... 2 3.7 Analysis ...... 2 3.7.1 Wool Production...... 2 3.7.2 Body Condition & Weight...... 2 3.7.3 Dag ...... 2 3.7.4 Fleece Rot...... 2 3.7.5 Flystrike...... 2 3.7.6 Ewe Reproduction ...... 2 3.7.7 Mortality...... 2 3.7.7.1 Multivariate Weaner Survival Analysis...... 2 3.8 Calculation of Fleece Values ...... 2 3.9 ‘Lifetime’ Wool Production ...... 2 C h a p t e r 4 Results—Weather Observations & Wool Production...... 2 4.1 Weather Observations...... 2 4.2 Overview of Sheep Numbers & Wool Production Results...... 2 4.3 Staple Strength & Position of Break...... 2 4.3.1 Ewes ...... 2 4.3.2 Weaners ...... 2 4.4 Fleece Weight & Yield ...... 2 4.4.1 Ewes ...... 2 4.4.1.1 Patterns of Annual Wool Growth in Ewe Shearing Groups...... 2 4.4.2 Weaners ...... 2 4.5 Fibre Diameter...... 2 4.5.1 Ewes ...... 2 4.5.2 Weaners ...... 2 4.6 Staple Length...... 2 4.6.1 Ewes ...... 2 4.6.2 Weaners ...... 2 4.7 Summary of Wool Characteristics...... 2 4.8 Fleece Value ...... 2 4.8.1 Ewes ...... 2 4.8.2 Weaners ...... 2 4.9 Lifetime Wool Production Results ...... 2 C h a p t e r 5 Discussion—Wool Production...... 2 5.1 Staple Strength...... 2 5.1.1 Ewes ...... 2 5.1.2 Weaners ...... 2 5.2 Fibre Diameter & Fleece Weight...... 2 5.3 Fleece Value ...... 2 5.3.1 Ewes ...... 2 5.3.2 Weaners ...... 2 5.3.3 Lifetime Fleece Value...... 2 C h a p t e r 6 Results—Sheep Health & Mortality...... 2 6.1 Body Condition & Growth Rate ...... 2 6.1.1 Ewe Condition Score ...... 2 6.1.2 Weaner Growth Rate Following Shearing...... 2 6.2 Fleece Rot...... 2 6.2.1 Ewes ...... 2 6.2.2 Weaners ...... 2

6.3 Dag & Crutching ...... 2 6.3.1 Ewes ...... 2 6.3.2 Weaners ...... 2 6.4 Flystrike...... 2 6.4.1 Ewes ...... 2 6.4.2 Weaners ...... 2 6.5 Reproduction ...... 2 6.6 Mortality...... 2 6.6.1 Ewes ...... 2 6.6.2 Weaners ...... 2 6.6.2.1 Summary Mortality Data ...... 2 6.6.2.2 Mortality & Bodyweight...... 2 6.6.2.3 Mortality & Mean Seasonal Growth Rate...... 2 6.6.2.4 Mortality and sex ...... 2 6.6.2.5 Mortality & Weaner Shearing Time ...... 2 6.6.2.6 Multivariate Survival Analyses...... 2 6.7 Value of Lifetime Wool Production, Accounting for Survivorship...... 2 C h a p t e r 7 Discussion—Sheep Health & Mortality...... 2 7.1 Body Condition & Growth Rate ...... 2 7.1.1 Ewes ...... 2 7.1.2 Weaners ...... 2 7.2 Fleece Rot...... 2 7.3 Dag & Crutching ...... 2 7.4 Flystrike...... 2 7.5 Reproduction ...... 2 7.6 Mortality...... 2 7.6.1 Ewes ...... 2 7.6.2 Weaners ...... 2 7.7 Value of Lifetime Production, Accounting for Survivorship...... 2 C h a p t e r 8 Conclusion ...... 2 Bibliography ...... 2 A p p e n d i c e s ...... 2 A p p e n d i x 1 Fertiliser Application & Pasture Management Details on Farm A...... 2 A p p e n d i x 2 Fleece Values Based on First & Ninth Decile Historical Micron Basis Prices...... 2 A2.1 Ewes ...... 2 A2.2 Weaners ...... 2 A2.3 Lifetime Fleece Values ...... 2 A p p e n d i x 3 Price Schedules Used to Calculate Fleece Values...... 2

List of Tables

C h a p t e r 1 Table 1.1: Summary of shearing times reported in different regions of Australia...... 2

C h a p t e r 2 Table 2.1: Summary of experiments comparing shearing time and staple strength in Merinos, measured by % of tender fleeces, N/ktex and/or position of break...... 2 Table 2.2: Summary of results of studies comparing greasy fleece weight (GFW), fibre diameter (FD) and clean fleece yield of Merinos shorn annually at different times of the year...... 2 Table 2.3: Summary of trials comparing the net returns per head from shearing young Merino sheep once (shearing B only) or twice (shearing A then shearing B) ...... 2 Table 2.4: Summary of the results of studies examining changes in sheep feed requirements following shearing at different times of the year...... 2 Table 2.5: Mortality of Merino and Merino-cross sheep between weaning and approximately 18 months of age reported in Australian commercial enterprises or field experiments...... 2 Table 2.6: Summary of methodologies of studies reporting risk factors for post-weaning mortality...... 2 Table 2.7: Post-weaning mortality of sheep in different nutritional groups reported by Allden (1968c).2 Table 2.8: Mortality of weaners with different weaning weights and autumn growth rates reported by Hodge (1990)...... 2 Table 2.9: Maternal factors associated with weaner mortality ...... 2

C h a p t e r 3 Table 3.1: Calendar of flock management procedures and shearing times...... 2 Table 3.2: Statistical methods used to analyse differences between shearing groups in ewe and weaner results, and corresponding Stata® (StataCorp 2005) commands...... 2

C h a p t e r 4 Table 4.1: Number of ewes in each age group, and mean age and standard deviation (years) of the ewe flock at each shearing ...... 2 Table 4.2: Duration of wool growth (days) between, and dates of, shearings in each ewe shearing group ...... 2 Table 4.3: Duration of wool growth (days) between birth and shearing 1, and shearing 1 and shearing 2 for weaner shearing groups in each birth-year cohort...... 2 Table 4.4: Least squares mean greasy fleece weight (GFW), yield, clean fleece weight (CFW), fibre diameter (FD), staple length (SL), staple strength (SS) and percentage of mid-breaks of wool from the ewe shearing groups...... 2 Table 4.5: Least squares mean greasy fleece weight (GFW), yield, clean fleece weight (CFW), fibre diameter (FD), staple length (SL), staple strength (SS) and percentage of mid-breaks of wool produced at shearing 1 and 2 by each weaner shearing group ...... 2

xii

C h a p t e r 6 Table 6.1: Mean condition score prior to joining in March (join), during winter in July (wint) and at weaning in December (wean) of each ewe shearing treatment...... 2 Table 6.2: Mean fleece-free bodyweight at weaning and 95% CI (kg) of progeny from each ewe shearing group ...... 2 Table 6.3: Mean growth rate and 95% CI (kg/month) between May and December (8–15 months old) of weaners that, in May, were shorn (MAY-MAY), or were carrying two (MAR-MAR) or five (DEC-DEC) months wool...... 2 Table 6.4: Prevalence of severe fleece rot (score ≥3) in ewe shearing groups ...... 2 Table 6.5: Odds ratios (95% CI) of weaner shearing groups having severe fleece rot, relative to unshorn weaners or weaners carrying 12 months wool, at different times between 6 and 20 months of age..2 Table 6.6: Mean cost of dags (cents/head) in each dag score category; mean cost in each shearing group, weighted by prevalence of each dag score category, in 2002–03 (only crutched prior to shearing) and 2003–04 (crutched prior to lambing and shearing); and mean total annual cost of dag ...... 2 Table 6.7: Odds (95% CI) of severe dag (score ≥3) in weaner shearing groups in March, October & March (6, 13 & 18 months old, respectively), relative to unshorn weaners or weaners carrying 12 months wool ...... 2 Table 6.8: Flystrike prevalence in ewe shearing groups during November and early December...... 2 Table 6.9: Proportion of weaners in each shearing group affected by flystrike in May (8 months old) in two years of the trial ...... 2 Table 6.10: Proportion of weaners in each shearing group affected by flystrike and odds of flystrike, relative to DEC-DEC weaners, in November (14 months old) in three years of the trial...... 2 Table 6.11: Proportion of weaners in each shearing group affected by flystrike and odds of flystrike, relative to MAR-MAR weaners, in March (18 months old) in two years of the trial ...... 2 Table 6.12: Proportion of ewes in each shearing group lactating at lamb marking and odds (95% CI) of ewes lactating at marking, relative to DEC ewes...... 2 Table 6.13: Number of lambs weaned, number of ewes and weaning percentage in each year ...... 2 Table 6.14: Mean weaning weight and 95% CI (kg) of weaners dying and surviving during the post- weaning period in each year-trial cohort...... 2 Table 6.15: Number of weaners in each weaning weight quintile (range in kg) that died and were at risk during the post-weaning period, total mortality, cumulative mortality rate, and mortality risk (95% CI), relative to the middle quintile...... 2 Table 6.16: Average weaner growth rate (kg/month; ‘GR’) at different times throughout the post- weaning period (PWP) and total mortality ...... 2 Table 6.17: Total mortality, cumulative mortality rate (deaths/1000 weaners/month) and mortality risk of males relative to females ...... 2 Table 6.18: Mortality rate (deaths/1000 weaners/month) of shorn and unshorn weaners, mortality rate difference, and mortality rate ratio between shorn and unshorn weaners at each shearing...... 2 Table 6.19: Coefficients of statistically significant terms in Weibull, Cox, cubic spline, log-logistic and interval-censored Weibull survival analysis models, in analyses of a) all data, and b) Trial 2 data only...... 2 Table 6.20: Mean annual (Ann.) and cumulative (Cum.) survival prior to each shearing of sheep in each shearing group between birth and 6 years old (y.o.)...... 2

xiii

Appendices Table A-1.1: Annual fertiliser application and pasture management details for Farm A ...... 2 Table A-3.1: First, median and ninth decile basis prices in the period July 1992–February 2006 for wool of specified fibre diameter (FD; µm), and first, median and ninth decile percentage changes from basis price for wool of specified staple length (SL; mm) and staple strength (SS; N/ktex) ...... 2

xiv

List of Figures

C h a p t e r 1 Figure 1.1: Average contribution of individual factors to variation in quarterly clean wool price April 1996–June 1998, after Hatcher (2000)...... 2

C h a p t e r 2 Figure 2.1: An example of the effect of rate of change of fibre diameter on staple strength. The two wool fibres have a similar minimum fibre diameter and peak breaking force, but the bottom fibre has a greater linear density and therefore a lower staple strength...... 2 Figure 2.2: An example of how shearing time (dotted arrows) locates position of break (POB) on a wool fibre. ABCD is the pattern of diameter change along the wool fibre. Shearing at A & C places POB in the middle of the fibre. Shearing at B & D places POB at the end of the fibre...... 2 Figure 2.3: Price discounts due to staple length of 18 µm, 35 N/ktex wool at first, median and ninth decile prices from 1992–2006...... 2 Figure 2.4: Average vegetable matter levels in Australian states (Source: AWEX sale data 2000–01 to 2002–03, inclusive, compiled by B. Swain) ...... 2 Figure 2.5: Mortalities in Merino weaner sheep, classified by weight at weaning, from: 3–12 months of age on pasture (Lloyd Davies 1983), 3–6 months of age in a drought feedlot (Lloyd Davies et al. 1988), and 4½–6 months of age on pasture (Holmes 1992) ...... 2

C h a p t e r 3 Figure 3.1: Timing of ewe and progeny shearing treatments (age in months at weaner shearings) ...... 2

C h a p t e r 4 Figure 4.1: Monthly cumulative rainfall and average daily maximum and minimum temperatures at Farm A during the Time of Shearing Trial ...... 2 Figure 4.2: Mean staple strength of fleeces from ewe shearing groups (error bars indicate 95% confidence interval (CI); LSM: least squares means) ...... 2 Figure 4.3: Mean proportion of staple mid-breaks in fleeces from ewe shearing groups (error bars indicate 95% CI)...... 2 Figure 4.4: Mean staple strength of fleeces from shearing 1 and shearing 2 of each weaner shearing group (error bars indicate 95% CI) ...... 2 Figure 4.5: Mean proportion of staple mid-breaks in fleeces from shearing 1 and shearing 2 of each weaner shearing group (error bars indicate 95% CI) ...... 2 Figure 4.6: Mean greasy fleece weight of ewe shearing groups, adjusted to 365 days growth (excluding October-shorn maiden ewes; error bars indicate 95% CI) ...... 2 Figure 4.7: Mean clean fleece yield of ewe shearing groups (error bars indicate 95% CI)...... 2 Figure 4.8: Mean clean fleece weight of ewe shearing groups, adjusted to 365 days growth (excluding October-shorn maiden ewes; error bars indicate 95% CI) ...... 2

Figure 4.9: Mean clean wool growth rates of ewes from each shearing group in each dyeband period (■ denotes shearing) ...... 2 Figure 4.10: Mean weight of clean fleece grown by ewes in each shearing group during each dyeband period ( denotes shearing) ...... 2 Figure 4.11: Mean clean fleece weight of each weaner shearing group at shearing 1 ( ) and shearing 2 ( ) (LSM: least squares means)...... 2 Figure 4.12: Mean clean wool growth rates of each weaner shearing group between birth and shearing 1 (“Shearing 1”) and shearings 1& 2 (“Shearing 2”). Error bars indicate 95% CI...... 2 Figure 4.13: Mean yields of clean fleece from shearing 1 and 2 of each weaner shearing group (error bars indicate 95% CI) ...... 2 Figure 4.14: Mean fibre diameter of ewe shearing groups (error bars indicate 95% CI) ...... 2 Figure 4.15: Mean fibre diameter of weaner shearing groups at shearing 1 and shearing 2 (error bars indicate 95% CI)...... 2 Figure 4.16: Mean staple length of ewe shearing groups, excluding October-shorn maiden ewes (error bars indicate 95% CI) ...... 2 Figure 4.17: Mean staple lengths of each weaner shearing group from shearing 1 and shearing 2 (error bars indicate 95% CI) ...... 2 Figure 4.18: Estimated fleece value of each ewe shearing group, based on median historical basis price and median, first decile (lower bar) and ninth decile (upper bar) staple length and strength premiums and discounts...... 2 Figure 4.19: Estimated fleece values from first ( ) and second ( ) shearing of each weaner shearing group, calculated using median historical micron basis prices and length & strength discounts...... 2

C h a p t e r 5 Figure 5.1: Significant seasonal events during years 1 & 2, and hypothesised fibre diameter profile, showing the effect of different shearing times on the position of break and measured staple strength ...... 2

C h a p t e r 6 Figure 6.1: Mean condition score of ewes in each shearing group at different times during the trial (bars indicate 95% CI)...... 2 Figure 6.2: Mean fleece-free bodyweight of weaner shearing groups between weaning and 15 months of age (December to December; bars indicate 95% CI) ...... 2 Figure 6.3: Proportion of each weaner shearing group with severe fleece rot (score ≥3) between weaning and 20 months of age, amongst weaners born in 2000, 2001 and 2002 ...... 2 Figure 6.4: Proportion of each ewe shearing group with no (dag score (DS) 0), mild (DS 1–2) and severe (DS 3–5) dag (arrows indicate time of crutching or shearing)...... 2 Figure 6.5: Proportion of each weaner shearing group with severe dag (score ≥3) between weaning and 20 months of age, amongst weaners born in 2001 and 2002 ...... 2 Figure 6.6: Cumulative mortality rates of ewes in each shearing group at different times of the year.....2 Figure 6.7: Turnbull survival estimates of year-trial cohorts...... 2 Figure 6.8: Number of lambs weaned and deaths in each censoring interval of the PWP, by year-trial cohort (timelines not to scale)...... 2 Figure 6.9: Turnbull survival estimates of within-cohort weaning weight quintiles ...... 2 Figure 6.10: Turnbull survival estimates of males and females ...... 2

xvi

Figure 6.11: Mortality rate ratio of shorn and unshorn weaners, with different growth rates during the early (‘GR(early)’) or middle (‘GR(mid)’) post-weaning period, estimated by the cubic spline (──) and Cox (─ ─) models (bars indicate 95% CI) ...... 2

Figure 6.12: Effect of shearing in March on death rates of female weaners with GRearly = 1 kg/month (──) or 2 kg/month (─ ─), estimated by the cubic spline model* ...... 2 Figure 6.13: Mortality rate ratio for a 2 kg increase in weaning weight or time-varying bodyweight (BWT) from the labelled value, estimated by Weibull, Cox and cubic spline models (bars indicate 95% CI)...... 2

Figure 6.14: Mortality rate ratio for a 0.25 kg/month increase in GRearly from the labelled value for Weibull, Cox and cubic spline models (bars indicate 95% CI) ...... 2 Figure 6.15: Smoothed mortality rate estimate from all data, and estimated Weibull & cubic spline hazard functions for female weaners with median covariate values (weaning weight 17 kg; GRearly, GRmid and GRlate = 0.9, 1.6 and 2.8 kg/month, respectively) ...... 2 Figure 6.16: Smoothed curve of the scaled Schoenfeld residuals for growth rate early in the post- weaning period (GRearly) vs. time since weaning ...... 2 Figure 6.17: Per head value of wool production of each shearing group to 6.25 years of age, accounting for cumulative survival, at median historical micron basis prices using first, median and ninth decile historical staple length and strength premiums and discounts...... 2

Appendices Figure A-2.1: Estimated fleece value ($/head) produced by each ewe shearing group, based on median, first decile (lower bar) and ninth decile (upper bar) micron basis prices and median staple length and staple strength discounts ...... 2 Figure A-2.2: Estimated fleece values ($/head) from shearing 1 ( ) and shearing 2 ( ) of each weaner shearing group, calculated using first decile historical basis prices and median staple length & strength premiums and discounts...... 2 Figure A-2.3: Estimated fleece values ($/head) from shearing 1 ( ) and shearing 2 ( ) of each weaner shearing group, calculated using ninth decile historical basis prices and median staple length & strength premiums and discounts...... 2 Figure A-2.4: Estimated value ($/head) of wool produced over the lifetime (from birth to 75 months of age) of ewes shorn in each weaner and ewe shearing combination, net of shearing costs, at first, median and ninth decile historical micron basis prices with median staple length and strength premiums and discounts...... 2 Figure A-2.5: Estimated value ($/head) of wool produced over the lifetime of ewes shorn in each weaner and ewe shearing combination, net of shearing costs & accounting for cumulative survival, at first, median and ninth decile historical micron basis prices with median staple length and strength discounts ...... 2

xvii

xviii

C H A P T E R 1

GENERAL INTRODUCTION

“The question of ‘When should I shear?’ is one which should be considered seriously by every grazier for such a decision will affect farm efficiency, labour, markets, and returns.”—Sheep & Wool Branch, Tasmanian Journal of Agriculture 1964

1.11.11.1 Background The two most important management decisions at the discretion of a wool producer with a self- replacing Merino flock are when to lamb and when to shear (Morley 1994). In the winter rainfall zone of Australia, there is a large body of research to demonstrate that spring is the preferred lambing time (Lloyd Davies 1987; Foot and Vizard 1993). However, there is little holistic, objective evidence of a preferred time of shearing for spring-lambing flocks.

Table 1.1: Summary of shearing times reported in different regions of Australia Location & Stock Time of shearing Reference Class Victoria central Victoria majority (>80%) August–November Foot & Vizard (1993) Gippsland spring: 57% Irving (1991) summer: 25% autumn: 7% winter: 2% state-wide spring: 52% Court and Lawless (1995) summer: 18% autumn: 13% winter: 17% Tasmania Ewes mainly later winter; also significant Statham (2004) numbers late autumn and late spring Wethers majority spring Weaners more in spring but remainder spread throughout the year Western Australia spring: 36% McFarland & Shaw (1998); summer: 25% Bell (1993) autumn: 22% winter: 17% South Australia ~50% summer or autumn Ashton (1992) Victoria, NSW & South Reeve & Thompson (2004) Australia

1 General Introduction

On Australian farms, adult Merinos are usually shorn annually. Shearing times vary widely within and between regions in Australia (Table 1.1). In surveys of Victorian and South Australian woolgrowers, availability of labour was the most common reason given for choosing a particular shearing time, although flystrike risk, reliability of weather and available daylight hours were also considered to be important (Irving 1991; Ashton 1992). Notably, wool quality issues were not cited as important determinants of shearing time, but this may be because these surveys were conducted before staple strength was commonly measured.

Studies have shown that shearing time affects many aspects of wool production and sheep health. For example, time of shearing affects wool quality characteristics such as staple strength, position of break, fleece weight, yield and fibre diameter (Arnold et al. 1984). It can also influence stocking rate, a key determinant of farm profitability, by affecting the pattern of nutrient demand of the flock and the timing of cull stock sales (Dabiri et al. 1996; Salmon et al. 2006). Time of shearing can also influence the susceptibility of sheep to diseases such as flystrike and fleece rot (Raadsma 1988). These studies have investigated the effect of changing time of shearing on each of these factors in isolation. However, no study has estimated the combined effect of shearing time on all of the factors influencing wool enterprise profitability. Therefore, the aim of this study was to examine the totality of effects of shearing time on a spring-lambing, self-replacing Merino enterprise in south-eastern Australia. The remainder of Chapter 1 provides an overview of the important determinants of profitability for a wool enterprise in south-eastern Australia. Chapter 2 reviews the literature on the relationship between time of shearing and these factors.

1.21.21.2 Determinants of Wool Enterprise Profitability & Their Relationship to Shearing Time Together, stocking rate, wool production per sheep, average price received for wool and enterprise costs determine the gross margin of a wool enterprise (Quinn et al. 2005 Equation 1.1).

GM = SR × ( W × p – E )...... Equation 1.1 where: GM = gross margin ($/ha) SR = stocking rate (DSE/ha) W = wool production per sheep (kg/DSE) p = average wool price received per kilogram of wool ($/kg) E = enterprise costs ($/DSE)

The factors influencing each of these will be examined in turn.

1.2.11.2.11.2.1 Stocking Rate Stocking rate refers to the number of sheep grazed per unit area. It is commonly expressed as the number of dry sheep equivalents (DSE) per hectare, where DSE is the amount of feed consumed by an

2 General Introduction adult Merino sheep in medium body condition (Morley 1994). The use of DSE allows a direct comparison of animals with different feed intakes per head.

There is an important interaction between stocking rate and time of lambing. In south-eastern Australia, it is well recognised that more sheep can be grazed per hectare in spring-lambing flocks than in those lambing at other times (Caple et al. 1989; Lean et al. 1997). This is because lambing in spring best matches the flock’s peak nutritional requirements during late pregnancy and lactation with the time of highest pasture growth rate. Time of shearing is considered to have considerably less impact on stocking rate than time of lambing and, for this reason, it has been recommended to choose time of lambing before time of shearing (Foot and Vizard 1993). Time of shearing affects farm stocking rate in two different ways. Firstly, the feed requirement of sheep can increase after shearing and thereby influence stocking rate (Black and Bottomley 1980). Secondly, cull and cast-for-age sheep are usually sold soon after shearing, and consequently different times of shearing will result in different seasonal grazing pressures on a farm. These issues are discussed in greater detail in Section 2.2.1.

1.2.21.2.21.2.2 Wool Production per Sheep At any given stocking rate, genetics fundamentally determines average wool production per sheep (Morley 1994). Genetic improvement is the dominant method used in Australia to increase per head wool production of Merinos and allows improvements to be made without necessarily compromising other fleece characteristics (Hatcher 2000). Although shearing at certain times of the year can result in a relatively small increase in fleece weight (Lightfoot 1967; Arnold et al. 1984), it is accompanied by an increase in fibre diameter, which lowers the wool’s value. Time of shearing has therefore not been considered an important method for increasing wool production per sheep (Arnold et al. 1984). The relationship between shearing time, fleece weight and fibre diameter is discussed further in Section 2.1.2.

1.2.31.2.31.2.3 Average Price per Kilogram of Wool Different studies report slightly varying contributions of the various wool quality factors to wool price, but all show that average fibre diameter is the dominant determinant of price received by a grower for their wool (Figure 1.1). Variation in fibre diameter typically accounts for 50–70% of the variation in price received (Couchman et al. 1993; Hatcher 2000). Fibre diameter is one of the more heritable wool characteristics, with a heritability of approximately 0.50 (Morley 1994). Selection indexes can be devised that result in fibre diameter reduction, whilst avoiding or minimising deleterious changes to other aspects of wool production, such as fleece weight, and have been used widely in Australia (Hygate et al. 2006). Shearing time has been shown to influence fibre diameter (Arnold et al. 1984), but because greater, and permanent, progress in decreasing fibre diameter has been achieved through genetic selection, manipulating time of shearing has been a secondary consideration in strategies to lower fibre diameter.

3 General Introduction

After fibre diameter, wool quality factors that have a large effect on first stage processing performance have the most significant impact on wool price. These factors are staple strength, ‘position of break’ (the point along the staple where it breaks under tension), vegetable matter content and staple length (Hatcher 2000). Wools with a fibre diameter less than 20.5 µm tend to receive price penalties when staple strength is less than about 36 N/ktex (Oldham 2000). For example, between 1992 and 2006, 19 µm wool with a staple strength of 26 N/ktex received up to a 24% price discount compared to 36 N/tex wool of the same fibre diameter (Independent Commodity Services 2005). Position of break also tends to reduce wool price if more than 55% of tested staples break in the middle, although the discounts are usually only in the order of 1–2% (AWI 2006).

The presence of vegetable matter in wool affects processing efficiency and also lowers wool price (Charlton et al. 1981). Wool containing small or linear-shaped vegetable matter particles, such as seed and shive, tended to receive higher price penalties than burr because these contaminants were more difficult to remove (Atkinson 1989).

Time of shearing can have substantial and direct effects on staple strength, position of break and vegetable matter contamination (Warr et al. 1979; Arnold et al. 1984; Adams et al. 2000), and hence on the price growers receives for their wool.

Wool with staple length less than about 65 mm has received substantial price discounts because it is unsuitable for worsted processing (Adams et al. 2000; Independent Commodity Services 2005). Adult Merinos shorn annually invariably produce wool of adequate length, irrespective of when they are shorn (Arnold et al. 1984). However, sometimes adult sheep are shorn prematurely prior to sale, which can result in wool of short staple length that is heavily discounted (Salmon et al. 2006). Similarly, short staple length wool is commonly produced from young sheep whose shearing time is being aligned

Figure 1.1: Average contribution of individual factors to variation in quarterly clean wool price April 1996–June 1998, after Hatcher (2000)

11%

2% 1%

6% Fibre diameter Staple strength Staple length 7% Vegetable matter Colour Style Wool marketing 64% 9%

4 General Introduction with the adult flock. Consequently, short shearing interval is a dominant cause of short staple length wool (Donnelly 1991a).

1.2.41.2.41.2.4 Enterprise Costs Enterprise, or variable, costs are those which are directly attributable to a particular enterprise and that vary with its scale (Morley 1994). The main variable costs in a wool enterprise include contract services associated with shearing, supplementary feeding, animal health management and pasture maintenance (Quinn et al. 2005). Shearing time may influence costs by increasing sheep’s nutritional requirements and influencing their susceptibility to diseases such as fly strike and fleece rot (Black and Bottomley 1980; Irving 1991).

In summary, time of shearing is not considered to be the dominant factor influencing stocking rate, wool production per head, wool price received or enterprise costs, although it impacts upon all to some degree. The only factor primarily influenced by shearing time is staple length. The relatively small and diverse effects of shearing time on many aspects of a wool enterprise, rather than large effects on one or two factors, may partially explain the variety of observed shearing times that occur even within a given region.

C H A P T E R 2

LITERATURE REVIEW

This literature review discusses the effects of shearing time on wool quality characteristics and animal health, and survival of sheep after weaning more generally. Wool quality, which is discussed in Section 2.1, is difficult to define, although it may be best measured via the characteristics that most influence spinning performance and yarn quality (Vizard and Hansford 1999). Staple strength, mean fibre diameter, staple length and level of vegetable matter contamination are amongst the most important of these characteristics. The ways in which time of shearing affects sheep nutrition, health and mortality are reviewed in Section 2.2, with a more detailed examination of Merino weaner mortality in Section 2.3.

2.12.12.1 The Effect of SShearinghearing Time on Wool Quality

2.1.12.1.12.1.1 Staple Strength & Position of Break Staple strength (SS) is a measurement that represents the tensile strength of a staple of wool fibres (Lamb 2004). As outlined in Chapter 1, staple strength influences the price farmers receive for their wool because it is related to the performance of wool when it is processed into top. The adverse effect of low staple strength on processing performance “…is of considerable manufacturing significance” (Bigham et al. 1983).

Staple strength is calculated from several fibre measurements. It is the quotient of the tensile force required to break a wool staple and its linear density (Schlink 2000):

peak breaking force (N) (N/ktex)SS = ...... Equation 2.1 linear density(ktex)

The linear density of the wool fibre is its average mass per unit length and is measured in kilotex (ktex):

clean fibre mass (g) linear ktexdensity )( = ...... Equation 2.2 fibre length (m)

Staple strength was originally assessed subjectively, according to how easily a staple was broken when subjected to manually applied tension and flicked with a finger. Although this ‘flick test’ is a crude

7 Literature Review estimate of staple strength, it broadly classified wool as being sound (>30 N/ktex) or tender (<30 N/ktex), a distinction that was related to the wool’s processing performance (Bigham et al. 1983; Hansford 1987; Couchman et al. 1993). The tender classification can be further divided into part- tender (25–30 N/ktex), tender (18–24 N/ktex) or rotten (<18 N/ktex). Objective measurement of staple strength was incorporated throughout the 1990s into the set of objective characteristics that now form the basis for the sale of wool in Australia, a process known as ‘sale by description’.

Staple strength is routinely measured with either an ATLAS or AgriTest Staple Breaker machine (Vizard et al. 1994), both of which produce comparable results (Schlink 2000). The ATLAS and Staple Breaker machines first measure staple length under a standard extension, then the force required to pull the staple apart. The two resultant pieces are weighed to calculate linear density, and thence staple strength (Adams et al. 2000; Lamb 2004).

The location of the staple’s breaking point along its length, or ‘position of break’ (POB), is usually assessed alongside staple strength. As discussed in Section 1.2.3, POB is also an important characteristic of wool quality. It is calculated from the relative weights of the two pieces of broken wool from strength testing, and a minimum of 40 staples are broken to give a result describing the percentage of staples breaking in the lower third (that closest to the skin, or ‘base’), middle (‘mid’) or upper third (‘tip’) of the staple (Australian Wool Testing Authority 2004).

A significant proportion of wool grown in southern Australia has been classified as tender (Couchman et al. 1993). Foot and Vizard (1993) described this as the biggest limitation to quality of wool from ewes and young sheep in south-eastern Australia. For example, when that statement was made, 25% and 35% of the wool sold in Victoria and Western Australia, respectively, was assessed as weak (Doyle et al. 1993). Adams et al. (2000) presented unpublished survey data suggesting that between 35–45% of wool from southern Australia was “[financially] penalised” (p. 61) for poor staple strength.

2.1.1.1 Components of Staple Strength & Position of Break In order to discuss the effect of shearing time on staple strength, it is necessary to consider the physical wool properties that determine staple strength and POB.

A great deal of research has investigated the wool fibre properties that determine staple strength, and this has been well summarised in reviews such as those by Bigham (1983) and (1994). Because staple strength is calculated from several different wool staple measurements, it has been described as a “summary” of several “biological” properties of the wool fibre (Schlink et al. 2000a p. 21). The properties of interest are (Schlink and Hynd 1994):

• components of fibre ‘diameter’, comprising minimum diameter, and diameter variation along and between fibres • intrinsic fibre strength • fibre shedding

8 Literature Review

Although there has been debate in the past about the different factors that influence staple strength, it is now broadly accepted that the wool fibre’s size and shape, which are related to fibre diameter, have a greater influence on staple strength than properties of the wool material itself (Yang and Lamb 1997).

2.1.1.1.1 Fibre Diameter Properties The breaking force of a wool fibre or staple is mostly determined by the size of its minimum cross- sectional area, rather than differences in the intrinsic strength or chemical composition of the wool material present in the cross section (Adams et al. 2000). Less force is required to break a wool fibre of small diameter (and hence lower cross-sectional area) than a fibre of large diameter.

Initial examinations of the relationship between fibre diameter and staple strength focused on the relationship between minimum fibre diameter and staple strength. Various authors have reported 20– 66% of the variation in staple strength is accounted for by variation in minimum fibre diameter (Bigham et al. 1983; Hunter et al. 1983; Denney 1990; Hynd et al. 1997; Thompson and Hynd 1998). However, other studies have shown that rate of change of diameter along the fibre was correlated with staple strength to a similar or greater extent than minimum diameter, with fibres having a larger diameter change in a given length being weaker (Hansford and Kennedy 1988; Hansford and Kennedy 1990b; Friend et al. 1996; Peterson et al. 1998; Brown et al. 2002).

Minimum diameter and rate of diameter change along fibres help determine the overall shape of the wool fibre, which is referred to as the fibre diameter profile (Schlink et al. 2000a). Different aspects of the fibre diameter profile affect different components from which staple strength is calculated, explaining the association between staple strength and minimum fibre diameter, as well as rate of fibre diameter change. Minimum fibre diameter determines the fibre’s peak breaking force—the numerator of Equation 2.1—because it is related to the quantity of wool material present at the fibre’s weakest point. On the other hand, diameter variation along the fibre influences its linear density, which is the denominator of Equation 2.1. A fibre whose diameter changes gradually has a lower linear density

Figure 2.1: An example of the effect of rate of change of fibre diameter on staple strength. The two wool fibres have a similar minimum fibre diameter and peak breaking force, but the bottom fibre has a greater linear density and therefore a lower staple strength.

↕ 16 µm

9 Literature Review than one whose diameter changes suddenly along its length. If the two fibres have the same minimum diameter (i.e. the same peak breaking force), the fibre with the greater rate of fibre diameter change will have be weaker because of its greater linear density (Adams et al. 2000). This is illustrated in Figure 2.1.

POB usually occurs at the wool fibre’s point of minimum fibre diameter (Bigham et al. 1983; Schlink et al. 1998; Schlink et al. 2000b), although it is possible for it to occur at other locations due to the mechanics of fibre and staple breaking(Orwin et al. 1980; de Jong et al. 1985).

Fibres with thin regions of varying length will vary in their elongation and load-bearing properties, and thus staple strength. Fibres with a short thin section elongate less, and fail under a lower load, than fibres with a long thin section. Analyses that include fibre elongation have very high correlations with staple strength (Masters et al. 1998; Lamb 2004). This may help to explain why fibres with large diameter variation and short, thin sections have a POB at the sudden diameter change, and are often weaker than ones whose diameter changes more gradually (Brown 1971).

Diameter variation between fibres is also related to staple strength, although it is not part of the fibre diameter profile (Adams et al. 1997). Between-fibre variation is thought to be both indirectly and directly associated with staple strength (Lamb 2004). The indirect association occurs because fibres of different diameters tend to also have different lengths. When tested to breaking point, fibres of different lengths come under load at different times (Masters et al. 2000). The direct association occurs because fibres with varying diameters elongate different amounts before breaking (Lamb 2004). The indirect and direct associations have the same result: fibres within the staple do not load simultaneously, so the peak force is shared by fewer fibres at one time. Thus, there is less wool material resisting the breaking force than if all fibres in the cross section were resisting the tension together. As a result, a staple containing fibres of widely differing diameters breaks at a lower peak force than one with less between-fibre diameter variation (Adams et al. 2000; Schlink et al. 2000a, Peterson, 1998 #516).

2.1.1.1.2 Intrinsic Fibre Strength Historically, great emphasis was placed on an association between intrinsic fibre strength—the strength of the wool material itself—and staple strength. However, intrinsic fibre strength only varies moderately between individual Merinos (Hunter 1980; Yang and Lamb 1997), and even when variation in intrinsic fibre strength has been observed, it is very poorly correlated with staple strength (Lamb 2004). For example, no correlation between intrinsic fibre strength and staple strength was found in a Western Australian study of fine and broad wool Merino hoggets receiving three different levels of supplementary feeding in summer and shorn in spring (Peterson et al. 1998).

Differences in the cellular or chemical composition of wool that are related to staple strength have not been observed in Merinos (Hansford and Kennedy 1990a), although they have been in Romney sheep (Orwin et al. 1980). Even the effects on staple strength of physiological processes such as pregnancy and lactation do not appear to be mediated through changes in intrinsic fibre strength. For example, no

10 Literature Review changes in intrinsic fibre strength were observed even when large changes in staple strength were observed in lambing ewes under different nutritional regimens (Hunter et al. 1990).

Any relationship between intrinsic fibre strength and staple strength may be overridden by other fibre characteristics that cause more variation in staple strength. Reis (1992) pointed out that the breaking force of wools with the same average fibre diameter vary significantly and proposed that intrinsic fibre strength played a significant role in staple strength. However, this overlooked the possibility that wools with the same average diameter may have different minimum diameters and peak breaking forces, as explained previously.

2.1.1.1.3 Fibre Shedding Fibre shedding occurs when a wool follicle stops producing a wool fibre and is expelled from the follicle before it is shorn off. Shed fibres do not span the distance between the jaws of the testing machine and so do not bear any load but still contribute to the staple’s linear density and consequently decrease staple strength (Schlink et al. 2000a). Fibre shedding has been reported to be well correlated in wool of staple strength less than 30 N/ktex, but not in stronger wool (Schlink and Hynd 1994), and its inclusion in analyses does not seem to explain staple strength variation any more than minimum fibre diameter and fibre diameter variation alone (for example, see Hynd et al. 1997).

2.1.1.2 Associations between Staple Strength, Position of Break & Shearing Time Position of break and the along-fibre components of staple strength (minimum diameter and rate of diameter change) occur at specific points in time as wool grows. Thus, these are the components that can potentially interact with shearing time, which also happens at a single point in time along the wool staple. In contrast, other components of staple strength, such as between-fibre variation or intrinsic fibre strength, affect the staple along its entire length. It is therefore unlikely that a single event could interact with them in a consistent way.

A number of surveys have shown an association between time of shearing and staple strength. For example, Curtis and Stanton (2000 cited by Oldham 2000) demonstrated a seasonal variation in staple strength of wool produced in all states of Australia, based on average figures from wool sold between 1995 and 2000. This analysis assumed that shearing occurred 1–2 months before sale. In Victoria, staple strength was greatest in wool sold in August (and therefore probably shorn in June or July), with an average staple strength of 38 N/ktex, whilst wool sold in December (shorn October or November) had the worst staple strength of 34 N/ktex. The staple strength of wool sold from other states also varied throughout the year, although the timing of the maximum and minimum strength differed between states. South Australia and New South Wales had similar patterns to Victoria. Western Australia had a similarly shaped curve, although the peak coincided with autumn shearing (36 N/ktex) and the trough with spring shearing (31 N/ktex). Queensland wool showed less variation throughout the year: wool sold in spring was strongest (38 N/ktex) and strength wool from the rest of year was only 2–3 N/ktex less. Wool from Tasmania sold between January and July had an average strength of

11 Literature Review about 34 N/ktex but then there was a sudden increase to about 40 N/ktex for wool sold between August and December.

It should be noted that whilst Curtis and Stanton’s data showed a definite pattern to staple strength throughout the year, in Victoria, New South Wales, Queensland and Western Australia, the annual range was only 3–5 N/ktex. The range in staple strength of wool from Tasmania, the state with the largest variation, was still only 8 N/ktex. Nonetheless, average staple strength was still less than the critical value of 36 N/ktex for more than half the year in all states except New South Wales and Queensland.

Other authors have reported a similar association with shearing time across Australia (Douglas 1989; Couchman et al. 1993) and in Western Australia (Arnold and Gordon 1973; J. Stanton 1987 cited by Hansford 1987), although these data may have been biased because staple strength was not measured in all wool sold at those times. Furthermore, the patterns are not consistent in every year. In the 1991/92 selling year, Baker (1993) reported that Western Australian wool sold in winter was weakest (30 N/ktex), whereas Curtis and Stanton found wool sold in winter had the second greatest average staple strength after late autumn.

Baker et al. (1994) conducted a more detailed survey of sheep, management and environmental factors affecting staple strength of Western Australian fleece wool sold between 1989/90 and 1991/92. The survey examined how much variation in staple strength was explained by average fibre diameter, sheep

Table 2.1: Summary of experiments comparing shearing time and staple strength in Merinos, measured by % of tender fleeces, N/ktex and/or position of break Location Shearing Staple strength Lambing Reference time measurement time WA March 4%, 10%* May (McGarry and Stott September 23%, 60% 1960)

WA March 2% June (Arnold et al. 1984) March 8% March October 9% June October 11% March

Tasmania July 43, 30 N/ktex† - (Butler et al. 1994) November 30, 27 N/ktex

Sth Africa February 35 N/ktex; 35% mid-breaks March (Cloete et al. 2000) September 27 N/ktex; 60% mid-breaks

WA January 37 N/ktex; tip‡ August (Hansford 1997) February 32 N/ktex; tip July April 37 N/ktex; tip July August 35 N/ktex; middle June August 21 N/ktex; middle/base April September 33 N/ktex; middle May October 36 N/ktex; tip July

Results from: * year 1, year 2; † less nutritional variation, more nutritional variation experienced by flock ‡ Position of break

12 Literature Review age and sex, shearing time, length of growing season, timing of seasonal break, and number of false seasonal breaks. Time of shearing was significantly associated with staple strength in all years but was only of mid to low importance in explaining staple strength variation. Frustratingly, the study did not report which shearing time was associated with greater staple strength.

Several field experiments, summarised in Table 2.1, have also demonstrated that shearing time affects staple strength and indicated which times produce stronger wool. In a Western Australian trial spanning two years, March shearing reduced the likelihood of autumn (May)-lambing Merino ewes producing ‘tender’ fleeces by 83%, compared to September shearing (McGarry and Stott 1960). This result was the same in both years, despite the proportions of tender fleeces varying considerably between the two years. A similar, more recent experiment involving autumn (March)-lambing Merinos in a Mediterranean-type environment in South Africa made similar findings (Cloete et al. 2000). Shearing ewes in February (before lambing) produced wool of greater staple strength than shearing in September (before joining) and reduced the proportion of ewes with mid-staple breaks.

The previous two experiments only examined ewes lambing in autumn, however a similar Western Australian trial showed that there was a statistically significant interaction between shearing and lambing time (Arnold et al. 1984). March shearing and June lambing produced the fewest tender fleeces. March shearing and March lambing produced the next lowest proportion of tender fleeces, although it was similar to October shearing with March or June lambing. Thus, March shearing generally reduced tenderness, although the largest improvement in staple strength came from the choice of shearing and lambing time. Despite the effect of lambing time on staple strength, only shearing time affected processing performance.

Hansford (1997) directly measured staple strength, rather than the proportion of tender fleeces, of Merino wool from seven farms in the Kojonup region (Western Australia) between April 1993 and February 1994. In this trial, shearing time was clearly associated with POB. The POB was at the end of the staple in flocks shorn in summer or autumn shearing (3 flocks) but was in the middle of the staple in flocks shorn in winter or spring (4 flocks). In contrast, shearing time was not clearly associated with staple strength. Thus, shearing time appeared to influence POB but not staple strength.

Complex interactions between shearing time, nutrition and staple strength have also been demonstrated in Merino wethers in Tasmania (Butler et al. 1994). In contrast to the Western Australian ewe experiments, wool shorn in July had a greater staple strength (43 N/ktex) than wool shorn in November (27 N/ktex). However, amongst wethers that experienced sudden changes in nutrition, such as changes in grazing rotation length and multiple changes from pasture to crop paddocks, July shearing produced wool only marginally stronger (30 N/ktex) than the November shearing. As observed in the Western Australian study, any influence of shearing time on staple strength can be tempered by other extraneous factors.

McGarry et al. (1960), Arnold et al. (1984), Butler (1994), Hansford (1997), Cloete et al. (2000) and Peterson et al. (2000) all speculated that shearing improved staple strength when it coincided with the

13 Literature Review

POB. This is illustrated in Figure 2.2. The timing of shearing determines where the minimum fibre diameter, and hence POB, is situated relative to the ends of the staple. What might otherwise be a mid- break in the staple can be a tip or base break if shearing occurs just before or after the minimum fibre diameter. This will affect the measured staple strength because 10–20 mm of wool at either end of the staple is held in the jaws of the ATLAS machine used to measure staple strength (Hansford 1997). If the POB is clamped, the staple must break at a different, stronger location, resulting in a higher measured peak breaking force and staple strength (Butler et al. 1994). Although the greater staple strength may seem to be an artefact of the testing methodology, such wools have improved processing performance (Arnold et al. 1984; Hansford 1997).

2.1.1.3 Shearing Time & Recurring Factors Affecting Staple Strength If the point of minimum fibre diameter corresponds to a recurring annual event, such as the start of lambing or the onset of autumn, then when shearing occurs relative to this event will determine the POB. The final part of this section briefly discusses the ‘recurring events’ that have been associated with the point of minimum fibre diameter. If such an event occurred reliably each year, staple strength may be improved by shearing at that time.

In Merinos, diameter variation along a wool fibre and POB are mainly influenced by seasonal variation in a sheep’s nutritional intake and requirements (McManus et al. 1964; Allden 1979). Variation in staple strength was shown to be mainly caused by the effects of nutritional changes on the fibre diameter profile by Thompson and Hynd (1998). They measured minimum fibre diameter, and along- and between-fibre diameter variation in Merino wethers, bred for high or low staple strength, that either maintained or lost then gained weight from consuming diets containing different amounts of metabolisable energy. Minimum diameter and along-fibre variation accounted for 75% of the variation

Figure 2.2: An example of how shearing time (dotted arrows) locates position of break (POB) on a wool fibre. ABCD is the pattern of diameter change along the wool fibre. Shearing at A & C places POB in the middle of the fibre. Shearing at B & D places POB at the end of the fibre.

POB

A B C D

POB POB

POB

14 Literature Review in staple strength, and these two factors were more strongly associated with nutrition than genotype. In contrast, between-fibre variation, which was attributed to genetics, only accounted for 8% of the variation. Mata et al. (2002) reported similar effects on the fibre profile, but not staple strength, in sheep grazing pasture.

In sheep grazing pastures in southern Australia, rapid nutritional change frequently occurs at or near the start of autumn, and many authors have observed that POB also occurs at this time (McManus et al. 1964; Allden 1979; Reis 1992; Earl et al. 1994). Prior to the autumn break, sheep have grazed dry pastures of low nutritional value. Once pasture growth resumes after the opening rains, sheep begin to graze increasing quantities of highly digestible, high-protein, green feed, which causes substantial increases in wool growth rate and fibre diameter (Mata et al. 1990). POB is particularly sensitive to such changes in protein nutrition (Schlink et al. 1998; Schlink et al. 2000b). In Mediterranean-type environments, sheep experience changing climatic conditions, as well as nutritional ones, at the autumn break. However, it has been shown that the change in diet, rather than exposure to rain and/or cooler temperatures, affects wool growth, fibre diameter and position of break (Mata et al. 1999; Woodgate et al. 2000; Woodgate et al. 2001).

Variations in pasture availability and feed intake, rather than pasture quality, can also alter the fibre profile and POB (Butler et al. 1994; Schlink et al. 1999; Mata et al. 2002). This is particularly seen in environments where there is less seasonal variation in pasture quality, such as regions of Tasmania (Butler 1994). In these environments, POB occurs during winter, when pastures tend to be at their lowest availability.

The physiological status of the sheep also influences fibre diameter properties and hence staple strength (Hansford 1987). In particular, ewe nutritional requirements vary throughout the year according to reproductive state, and wool growth and reproduction compete for nutrients (Corbett 1979). Many authors have observed that the minimum fibre diameter and the POB in ewes can coincide with pregnancy or lactation (Hunter et al. 1990; Robertson et al. 1996; Masters et al. 1997). Thus, in the one environment, the POB may occur in autumn in wethers but in winter or spring in spring-lambing ewes, depending on how well pasture availability matches ewe requirements during pregnancy or lactation. For example, Thornberry et al. (1988) examined staple strength in 14 spring-lambing flocks across NSW and found that POB in ewes occurred at widely varying times between early pregnancy and post-weaning, but consistently occurred in autumn in wethers on the same farms. This suggests that POB in ewes is quite sensitive to the match between nutritional requirements and intake, which in turn is significantly influenced by seasonal conditions, time of lambing and number of lambs born and reared. This has been confirmed experimentally. For example, nutritional restriction of pregnant ewes lasting only a fortnight reduced staple strength by up to 27 N/ktex, with the largest reductions occurring in twin-bearing ewes (Robertson et al. 2000b). The staple strength of ewes grazing about 550 kg dry matter (DM)/ha throughout pregnancy and lactation was 34 N/ktex lower than ewes grazing >1500 kg DM/ha (14 vs. 48 N/ktex Robertson et al. 2000a). Similarly, a Victorian field trial showed that changing lambing time by just 3 weeks (from late winter to spring) increased staple strength by 4

15 Literature Review

N/ktex and reduced the likelihood of producing tender wool nearly three-fold (Scrivener and Vizard 1997).

It is important to note that, in spring-lambing ewes, the two factors most commonly associated with the POB, lambing and autumn nutrition, occur about six months apart. Most work has concentrated on one factor or the other and little attention has been paid to their combined influence on the POB within and between years (Masters et al. 1997). For example, if the POB in ewes varies between spring and autumn from year to year, no annual shearing time will consistently improve staple strength.

In summary, it can be seen that staple strength is substantially influenced by the minimum fibre diameter and along-fibre variation, and that these fibre properties are largely determined by nutrition, relative to a sheep’s physiological requirements. Shearing time is thought to increase staple strength if it places the POB at the end of the wool staple. As explained previously, spring is the preferred lambing time in southern Australia because it improves the profitability of a self-replacing Merino flock, yet no experiments have compared the effects of different shearing times on staple strength in spring-lambing flocks. Furthermore, the POB may vary substantially with changes in nutrition or nutritional requirements, and no trials have been conducted for sufficient time to test if one particular shearing time consistently improves staple strength.

2.1.22.1.22.1.2 Fleece Weight, Yield & Fibre Diameter Although shearing time affects both fibre diameter and fleece weight, the main way to manipulate these characteristics is through genetic selection, as discussed previously. Several trials, all of which were conducted for two or three years, have investigated the effect of shearing time on fleece weight, fibre diameter and yield in Merinos, and their results are summarised in Table 2.2. These studies were conducted in Western Australia, Tasmania, NSW and South Africa in wether flocks, and ewe flocks with different lambing times. Five of eight studies showed that fleece weight differed with shearing time. In four of these trials, sheep shorn in autumn produced greasy fleeces approximately 10% heavier than sheep shorn at other times of the year (McGarry and Stott 1960; Lightfoot 1967; Arnold et al. 1984; Cloete et al. 2000). In the fifth trial, a similar increase in fleece weight was observed in wethers shorn in July, compared to wethers shorn in November (Butler et al. 1994). Of the three studies that did not observe an association between fleece weight and shearing time, two were conducted in ewes in Tasmania (McLeod 1967). In these trials, although ewe feed intake was not reported, the lack of an association between fleece weight and shearing time may have been due to low winter pasture availability that restricted the feed intake and consequent wool growth of autumn-shorn ewes after shearing (Dabiri et al. 1994). The final study showing no effect of shearing time was conducted in western NSW, an environment without a strong seasonal rainfall and pasture growth pattern (Kennedy et al. 1982). In summary, whilst there is evidence that autumn-shorn wethers and non-pregnant ewes produce heavier fleeces than sheep shorn at other times of the year, in the flocks lambing in spring, the preferred lambing time in southern Australia, this association may be modulated by other factors such as winter pasture availability.

Table 2.2: Summary of results of studies comparing greasy fleece weight (GFW), fibre diameter (FD) and clean fleece yield of Merinos shorn annually at different times of the year

Stock class (lambing Location Shearing GFW FD Yield Comment Reference time) times (kg) (µm) Ewes (May) WA March 5.3 - 62% (McGarry and Stott 1960) September 4.8 62% Ewes (late winter*) Tasmania January 4.4 - - Jan. > Oct. in 2 of 3 years (McLeod 1967) July 4.1 October 4.4 Ewes (late winter*) Tasmania March 4.4 - - Oct. > Mar. in 2 of 3 years (McLeod 1967) October 4.5 Literature Review Review Literature Wethers WA April 5.9 66% (Lightfoot 1967) October 5.2 71% 17 Ewes (March & June) WA March 5 23.0 67% (Arnold et al. 1984) October 4.1 22.3 71% Ewes (winter) western March 3.6 - - No effect of shearing time (Kennedy et al. 1982) NSW May 3.5 once variation in rainfall August 3.8 accounted for November 3.9 Ewes (February) South Africa February 5.7 23.5 69% FD difference not (Cloete et al. 2000) September 4.9 23.0 75% significant Wethers Tasmania July 4.3 - - (Butler et al. 1994) November 4.0

* not explicitly stated - : not reported

Literature Review

Table 2.2 also shows that autumn-shorn wools have lower yields than wools from other shearing times. Arnold et al. (1984) suggests that this occurs for two reasons. Firstly, autumn-shorn sheep have longer fleeces during summer than sheep shorn at any other time of the year, and this longer wool accumulates more dust and environmental contaminants than shorter fleeces do. Secondly, the increased wool growth stimulated by autumn shearing is accompanied by extra wax and suint production, and this further lowers the yield of autumn-shorn fleeces. However, this lower yield is not large enough to completely offset the increase in greasy fleece weight, so that autumn shearing still tends to produce heavier clean fleece weights than other shearing times.

Shearing time is thought to affect fibre diameter, fleece weight and yield through the same mechanism—by stimulating feed intake (Arnold et al. 1984)—so they will be considered together. The relationship between shearing time and feed intake is discussed in detail in Section 2.2.1 and only the underlying mechanisms are summarised here. Removing a sheep’s fleece substantially increases heat loss and this stimulates feed intake in an effort to maintain body temperature. In turn, a higher feed intake increases wool growth rate (Arnold et al. 1964; McFarlane 1965; McGuirk et al. 1966; Allden 1979). This stimulation of wool growth may persist for weeks to months after shearing. The extent of additional heat loss following shearing is determined by ambient temperatures and, consequently, shearing at different times of the year will have different effects on feed intake and hence wool growth rate. Furthermore, in southern Australia, factors that modulate feed intake, such as pasture availability and sheep body condition, vary considerably with season, and therefore will have differing influences on wool growth rate at different shearing times. For example, autumn shearing exposes sheep to cooler ambient temperatures than sheep shorn in spring. Additionally, sheep are usually in lighter body condition in autumn than in spring and so have a greater capacity to increase their feed intake following shearing (Freer et al. 1997). Arnold (1984) proposed that the combination of these factors results in a larger increase in feed intake after shearing in autumn than in spring and, thus, autumn-shorn sheep produce heavier fleeces than sheep shorn at other times. When the growth rate of wool increases, mean fibre diameter increases (McFarlane 1965; Fernandez Abella et al. 1991; Doyle et al. 1995; Masters et al. 1998), and so when it has been observed that autumn shearing is associated with increased fleece weight, the average fibre diameter of autumn-shorn wool is also greater than wool shorn at other times. Depending on the price premiums for wool of finer fibre diameter, the monetary value associated any increase in fleece weight due to autumn shearing may be reduced, or even negated, by the reduction in wool price per kilogram. The potential for shearing time to have opposite effects on wool characteristics has been previously noted (Foot and Vizard 1993) but the overall impact of shearing time on a commercially managed, spring-lambing enterprise needs to be quantified.

2.1.32.1.32.1.3 Staple Length As explained previously, top-makers have traditionally aimed to produce a top with hauteur of 65–90 mm (Oldham 2000) and wool shorter than 65 mm attracts considerable price discounts because it reduces hauteur (Independent Commodity Services 2005; Figure 2.3). In particular, the staple length of

18 Literature Review

wool from weaner sheep is often less than 65 mm and consequently attracts significant discounts (Taylor 1989).

In commercial sheep flocks, the shearing time of young sheep must eventually be aligned with the shearing time of the adult flock. Seven studies, listed in Table 2.3, have compared different weaner shearing patterns that aligned a weaner flock to a single adult shearing time, and all found that shearing weaners twice was less profitable than shearing them once. In weaners that are shorn twice, the first shearing usually occurs soon after weaning and produces wool approximately 30 mm long (Donnelly 1991a; Rogan et al. 1995). This short staple length wool attracts a substantially lower price than longer wool and, in five of the seven trials listed in Table 2.3, short length was the most or second most important determinant of the lower profitability of twice-shorn weaners. The cost of the extra shearing was the other major factor that reduced the profitability of a two-shearing pattern. Flocks with the same lambing date but differing adult shearing times will inevitably have different weaner shearing

Figure 2.3: Price discounts due to staple length of 18 µm, 35 N/ktex wool at first, median and ninth decile prices from 1992–2006

10% 0% -10% -20% -30% -40% -50%

Clean Clean price discount -60% -70% 16-25 26-35 36-45 45-54 55-64 65-74 75-84 85-94 95-104 105- 114 Staple length (mm)

Median 1st decile 9th decile basis prices

Table 2.3: Summary of trials comparing the net returns per head from shearing young Merino sheep once (shearing B only) or twice (shearing A then shearing B) Time of Shearing A–Shearing B Wool returns Difference Reference birth Once Twice winter December–May $8.06 $5.70 -29% (Rogan et al. 1995) spring December–September $20.41 $15.07 -26% (Irving 1991) spring January–July $15.41 $9.22 -40% (Donnelly 1991a) - January–December $25.10 $14.94 -40% (Sackett 1990) spring February–November $15.52* $14.12* -9% (Newman et al. 1996) autumn October–May $16.20 $15.56 -4% (Little et al. 1993) - 3–12 months old $7–8/hd (Taylor 1989)

* gross returns

19 Literature Review patterns to align the young sheep with the adult flock. No study has attempted to directly compare the net effect in both weaner and adult sheep of various adult shearing times.

2.1.42.1.42.1.4 Vegetable Matter Contamination The importance of vegetable matter (VM) contamination in wool varies between regions in Australia (Foot and Vizard 1993). The effect of shearing time on vegetable matter contamination is highly dependent on geographic region because of the great variation in pasture species responsible for it and the time of year when contamination occurs. For example, vegetable matter differed by 23% between shorn and unshorn weaners during spring at Trangie, NSW (1.9% vs. 24.9%, respectively; Warr and Thompson 1976), but there was no vegetable matter difference between shorn and unshorn weaners in a similar experiment in Gippsland, Victoria (Irving 1991). Substantial vegetable matter contamination is generally a small problem in Victoria and wool from the south-west of the state tends to be free or nearly free of vegetable matter (Court and Lawless 1995). By contrast, 60% of wool produced in the pastoral region of South Australia can carry significant contamination (Foot and Vizard 1993).

Figure 2.4 suggests a relationship between sale date, and by inference shearing time, of wool and its average vegetable matter content, at a state level. Wool sold in South Australia, New South Wales and Queensland had significant vegetable matter content that varied throughout the year. However, as outlined above, such broad surveys can overlook variation in associations between vegetable matter contamination and shearing time. In contrast to the results presented in Figure 2.4, two NSW surveys of vegetable matter contamination found that there was generally no association between shearing time and total amounts of vegetable matter contamination (Cornish and Beale 1974; Warr et al. 1979). Warr and his colleagues did show that summer shearing in one region of the state reduced vegetable matter

Figure 2.4: Average vegetable matter levels in Australian states (Source: AWEX sale data 2000– 01 to 2002–03, inclusive, compiled by B. Swain)

3.5

2.5

1.5

Vegetable Vegetable matter (%) 0.5

0 Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun Month of wool sale

SA NSW QLD WA VIC TAS

20 Literature Review compared to winter or autumn shearing but in other districts, although levels of individual seed and burr were reduced by particular shearing times, overall vegetable matter levels did not change. It was suggested that vegetable matter contamination remained relatively constant in these districts because, as one potential vegetable matter contaminant species declined, it was replaced by another. Also, the slow maturation of the seeds of some contaminant species resulted in vegetable matter being available for periods beyond the protective effect of shearing.

In contrast to the New South Wales surveys, several studies in Victoria and Western Australia have shown a more widespread association between time of shearing and vegetable matter contamination of wool (Arnold and Gordon 1973; Arnold et al. 1984; Stanton 1987 cited by Hansford 1987; Bowman 1989; Ralph 1984 cited by Bell and Ralph 1993). These studies reported vegetable matter contamination in wool sold in different months and assumed that shearing occurred one to two months before sale. They showed that the highest vegetable matter levels were in wools shorn between March and May and the lowest from shearing between October and January.

Vegetable matter contamination can be influenced by time of shearing because extraneous matter accumulates more rapidly in long than in short wool. An experiment conducted in Western Australia showed that, between late spring and early autumn, autumn-shorn Merino wethers accumulated vegetable matter up to 14 times faster than spring-shorn sheep. These sheep were grazing pastures dominated by barley (Hordeum sp.) and brome grass (Bromus sp.), both of which are significant causes of vegetable matter contamination (Ritchie et al. 1992).

The price discount incurred for wool with significant vegetable matter may also vary seasonally. For example, Bowman (1989) showed that the wool price discount for vegetable matter contamination was highest for wool sold in August (about a 12% discount, compared to wool with <1% vegetable matter) and lowest for wool sold in January (a 4% discount). This probably reflected the different levels and type of vegetable matter contamination present in wool sold at different times.

In summary, the importance of vegetable matter contamination is inconsistent from region to region, and possibly from farm to farm, and consequently its significance when choosing a shearing time will vary. In the Western District of Victoria, vegetable matter contamination is a small problem and usually does not warrant particular attention through change of shearing time or any other management strategies. In other parts of Australia, the wide range of potential contaminant pasture species can mean that no shearing time effectively controls vegetable matter contamination and pasture improvement or tactical control of problem species, such as spraying or slashing of pastures, may be a more effective option (Little et al. 1993). It should be noted that, even if vegetable matter poses little threat to wool quality, it can cause significant health problems in young sheep, which is discussed in more detail in Section 2.2.6

21 Literature Review

2.1.52.1.52.1.5 Time of Shearing & Wool QualityQuality————SummarySummarySummarySummary Choosing a time of shearing to optimise wool quality is not a straightforward decision. In a self- replacing flock, shearing and lambing time interact to influence staple strength, and the extent of the effect varies between environments. Furthermore, the effect of shearing time on staple strength cannot be examined in isolation, because changing shearing time also affects fleece weight and fibre diameter, which are, arguably, more important determinants of wool income than staple strength. The shearing time-lambing time interaction is also important because it affects the staple length of wool produced by young sheep as their shearing time is aligned with the adult flock, and this can substantially influence the value of the wool that they produce. There is a need for a study that examines the effect of shearing time by simultaneously measuring all important wool characteristics and that is conducted in spring- lambing flocks.

22 Literature Review

2.22.22.2 The Effect of Shearing Time oonn Flock Management & Sheep Health

2.2.12.2.12.2.1 Sheep Nutritional Requirements & Farm Stocking Rate This review is restricted to Australian experiments conducted in Merinos or similar breeds, such as Polwarths, or non-Merino experiments, the results of which are particularly relevant to production systems in southern Australia. A number of experiments, using a variety of methodologies, have shown that shearing increases the nutritional requirements of sheep and these are listed in Table 2.4. It has been pointed out that this may have an important financial impact (Salmon et al. 2006). If shearing increases requirements at a time of the year when there is limited pasture availability, it may constrain overall stocking rate of the farm and hence profitability.

The amount by which shearing increases feed requirements and affects grazing sheep depends on a variety of factors, including the prevailing weather conditions—determined by season—(Elvidge and Coop 1974), the sheep’s reproductive status (Hudson and Bottomley 1978) and pasture availability (Arnold and Birrell 1977). In the first few weeks after shearing, nutritional requirements have been shown to increase by up to 69% in non-breeding sheep in Australia (Hutchinson and McRae 1969; Arnold and Birrell 1977; Hudson and Bottomley 1978; Love et al. 1978). The largest increase in these studies was observed following shearing in winter in Tasmania, however cold temperatures at other times of the year, particularly if they are accompanied by wind and/or rain, can also cause significant increases in feed requirements. For example, a 62% increase was observed following shearing in December in Armidale, New South Wales, when minimum temperatures were as low as 7° C, and 25 mm of rain fell soon after shearing (Wheeler et al. 1963).

Studies have shown that, in the month after shearing, the post-shearing increase in energy requirements of pregnant ewes was less than non-reproducing sheep. For example, in three studies, the increase in pregnant ewes was only 38–85% of the increase in their non-pregnant counterparts (Elvidge and Coop 1974; Hudson and Bottomley 1978; Bottomley 1979). In these studies, the largest difference between pregnant and non-pregnant ewes occurred after August shearing, whereas the difference was smaller following shearing in June. However, environmental temperature differences between these shearing times may not have been the only cause of the different increases in feed intake. It has been hypothesised that the smaller increase in energy requirement of pregnant ewes after shearing is associated with their pre-existing greater feed intake (Bottomley and Hudson 1976). The extra ruminal heat production associated with this greater intake partially offsets the effects of shearing and, consequently, the post-shearing increase in energy requirement will be less.

Although energy requirements probably increase immediately after shearing (Farrell and Corbett 1970; Holmes et al. 1992), some studies in grazing sheep have reported a lag of one to three weeks between shearing and an increase in feed intake (Wheeler et al. 1963; Wodzick-Tomaszewska 1963; Love et al.

Table 2.4: Summary of the results of studies examining changes in sheep feed requirements following shearing at different times of the year

Sheep breed & Time of Location Change in feed intake or Observation Method used to estimate Reference class shearing requirements post-shearing period (weeks) feed requirement

Merino wethers July Canberra, ACT ~20–70% 0–4 Total faecal nitrogen output (Arnold et al. 1964)

Merino & August Hamilton, Vic. 10–44% (fat–thin sheep) 0–4 Total faecal nitrogen output (Arnold and Birrell 1977) Corriedale Bakers Hill, WA 0 (but weight gain of shorn

Merino wethers July Armidale, NSW 30%, 59% 0–1, 1–2 Cr2O3 faecal marker (Hutchinson and McRae 1969)

Merino wethers ? Hamilton, Vic. 0, 14, 0% 0–1, 2–5, 6 ? (Love et al. 1978)

Merino wethers & December Armidale, NSW 42–62% (grazing) 0–2 Cr2O3 faecal marker (Wheeler et al. 1963) Review Literature ewes (non- 20–51% (penned) pregnant) 24 Polwarth ewes: July–August Cressy, Tas. Intake required to maintain (Hudson and Bottomley pregnant 26%, 9% 0–4, 4–8 bodyweight 1978) non-pregnant 69%, 27% 0–4, 4–8

Polwarth ewes July Cressy, Tas. 22% 0–7 Intake required to maintain (Bottomley and Hudson (spring-lambing) bodyweight 1976)

Polwarth ewes ? Cressy, Tas. 33%, 33% (non-pregnant) 0–4, 4–8 ? (unpublished data cited by 28%, 23% (pregnant) Bottomley 1979)

Romney ewes: Canterbury, NZ Energy associated with (Elvidge and Coop 1974) non-pregnant January–April 24% 0–13 bodyweight loss non-pregnant April–July 73% 0–10 3rd trimester June–July 50% 0–6

pregnant ewes June Swansea, Tas. 61% (lambing in October) 0–12 computer simulation (Black and Bottomley 1980) 18% (lambing in July)

Literature Review

1978). On the other hand, Hutchinson and McRae (1969) observed feed intake to increase in the week immediately after shearing. The reason for these differing observations is not clear but could reflect either true physiological or methodological differences. Acute cold stress can depress grazing behaviour in sheep (Sykes and Slee 1969), although this usually only persists for a few days after shearing (Webster and Lynch 1966; Donnelly et al. 1974). Alternatively, the chromium oxide faecal marker technique, used in several of the above studies, does not reflect small changes in feed intake well (Holmes et al. 1992), so changes in feed consumption immediately after shearing, before heat production reached its maximum (Farrell and Corbett 1970), might not have been readily detected.

It is not clear how long the period of increased energy requirement following shearing lasts. Hudson and Bottomley’s study (1978) compared the intakes of shorn and unshorn sheep for eight weeks after shearing, the longest study published. Shorn sheep were still eating more than unshorn ones at its conclusion, suggesting that feed intake is increased by shearing for at least two months. Although no field experiment has shown how long the effect of shearing on intake persists, a computer simulation suggested that residual effects of shearing on energy requirement for at least six months (Black and Bottomley 1980). This finding is consistent with Farrell’s (1970) observation on heat production in sheep following shearing. In grazing wethers shorn in June in Armidale, heat production did not return to pre-shearing levels until 19 weeks after shearing, despite the sheep being thermoneutral in their measurement environment much earlier than this (Farrell and Corbett 1970). This led to the conclusion that the increase in metabolic rate, and hence energy requirement, following shearing is due to a metabolic adjustment that, although proportional to the degree of cold exposure following shearing, lasts much longer than the stimulus does. Thus it is likely that the energy requirement of sheep shorn in winter may remain elevated for several months following shearing.

A sheep will lose weight if it cannot increase its feed intake after shearing to meet its greater energy requirement (Elvidge and Coop 1974; Bottomley and Hudson 1976). In southern Australia, this is most likely to occur with a winter shearing time, when cold conditions stimulate the greatest increase in energy requirement but feed availability often limits intake to less than that required to maintain bodyweight. Avoiding weight loss in this situation may require supplementary feeding or a reduction in stocking rate, to increase pasture availability. For example, Hudson and Bottomley (1978) estimated that the stocking rate of pregnant and non-pregnant sheep shorn in winter would have to be about 13% and 33% lower, respectively, than unshorn sheep, to avoid bodyweight loss immediately after shearing. Similarly, the computer simulation of Black and Bottomley (1980) estimated that winter shearing of ewes with an optimised mid-October lambing time lowered stocking rate by 38%, compared to spring shearing. In summary, the substantial and prolonged increase in energy requirement that may occur after shearing in winter, combined with the generally poor pasture availability at this time, probably makes it an undesirable choice of shearing time in southern Australia.

Shearing time can also affect farm stocking rate through its interaction with the timing of the sale of cull sheep. In southern Australia, surplus sheep are commonly sold in early summer, as a risk management strategy, to reduce grazing pressure before pasture availability decreases prior to the

25 Literature Review autumn break (Vizard and Foot 1994). If this occurs at, or soon after, the main shearing time, sheep leave the farm carrying only a small amount of wool, and little wool value is foregone (McLeod 1967). A main shearing time falling between late summer and early winter, however, comes several months after the preferred time for cull sales, and has two possible outcomes. Firstly, as described in Chapter 1, cull sheep can be shorn immediately prior to sale, but this produces short staple length wool that receives substantial price discounts (Independent Commodity Services 2005). Alternatively, sheep can remain on the farm to be sold after the main shearing, but simulations of a farm enterprise suggested that this can substantially reduce gross margins by reducing stocking rate (Salmon et al. 2006). The reduction occurs because selling sheep after shearing in autumn reduces winter grazing pressure, and hence annual per-hectare wool production and income. Even if extra stock are retained to maintain winter stocking rate, the gross margin is still reduced because extra supplementary feeding is often required to carry the extra animals throughout summer and autumn until they can be shorn. In Salmon et al.’s simulation, selling culls after shearing in May reduced winter stocking rate in a wether enterprise by 17% and maintaining winter stocking rate increased summer supplementary feeding costs by 54%, compared to December shearing. This is another undesirable consequence of late autumn or winter shearing in southern Australia.

In New Zealand, the disadvantage of winter shearing to ewe nutrition and hypothermia risk has been addressed through the use of ‘cover combs’, which leave a greater depth of residual fleece on the sheep (Dabiri et al. 1995). Even when cover combs are used, sheep metabolic rate and energy requirement increase after shearing (Holmes et al. 1992), which may reduce winter stocking rate. Therefore, it is difficult to advocate winter shearing until benefits can be consistently demonstrated that outweigh its known disadvantages.

There is a common belief amongst farmers that shearing young sheep increases their bodyweight, despite a number of Australian studies showing that the growth rate of Merino or Merino-cross weaners does not increase after shearing, compared to unshorn ones (Tucker 1964; Sumner 1984; Donnelly 1991a; Irving 1991; Rogan et al. 1995; R. Thompson pers. comm.). Although these studies used different times of lambing, all except the unpublished work of Thompson compared the effect of shearing three to six month old lambs at weaning. In contrast, Thompson’s study, conducted in western Victoria, compared spring-born Merinos first shorn when they were three, six, twelve or 15 months old. He also found no bodyweight differences at 15 and 27 months of age. The one study in which shearing was associated with increased bodyweight of weaners was conducted by Drinan and Ferguson (1966), who found that winter-born, second-cross weaners shorn at weaning in December were about 2 kg heavier than unshorn sheep in February. Some, but not all, of this difference was caused by weight loss in several unshorn sheep that were affected by flystrike. However in this study, as in the others, economic returns were not increased by shearing because the cost and lower wool value associated with the extra shearing were not offset by any increases in bodyweight, carcase value or wool production. In conclusion, although early shearing is not generally associated with increased bodyweight of weaners, there may be particular animal health situations, such as flystrike, in which an increase occurs.

26 Literature Review

2.2.22.2.22.2.2 Reproductive Efficiency It has been suggested that reproductive efficiency, as measured by the number of offspring successfully weaned as a percentage of the number of ewes mated, may be influenced by the time of shearing. Depending on timing, relative to the reproductive state of the ewe, shearing has the potential to directly influence several stages of the reproductive pathway, including mating (MacKenzie et al. 1975), gestation, parturition and lamb survival (Kenyon et al. 1999). The time of shearing can also have an indirect effect on reproductive efficiency through its influence on energy requirement and body condition of the ewe (Hudson and Bottomley 1978). The time of shearing, relative to reproduction, and the ways it affects the above events are discussed in the remainder of this section.

2.2.2.1 Shearing During Mating A study of Bungaree Merino ewes in New South Wales concluded that shearing three months prior to joining might improve the success of mating, compared to shearing prior to lambing (McGuirk et al. 1966). In the study, ewes were shorn either in June or December and were mated in late February. By the end of the mating period, 72% of June-shorn ewes, compared to 88% of December-shorn ewes had been mated. No comment was made about other factors that could have directly affected fertility, which may or may not have been associated with shearing time, such as ewe condition score. Despite the statistically significant difference in proportion of mated ewes, there was no significant difference in the percentage of ewes lambing between the two shearing times (33% shorn vs. 30% unshorn).

Several studies have observed deleterious effects of shearing just prior to mating on reproductive parameters. Parr et al. (1989) found that Merino ewes in Victoria shorn less than three weeks before mating had a lower incidence of oestrus than ewes shorn more than three weeks before mating. Similar observations have been made in New Zealand in Romney Marsh ewes that were naturally mated in autumn (McMillan and Knight 1982). Mackenzie et al. (1975) suggested that the main effect of shearing was to suppress behavioural oestrus, rather than inhibit ovulation. They examined Merino × Border Leicester ewes that were either left unshorn or shorn one week prior to exposure in winter to the severe, cold weather of the New South Wales Tablelands, when they were either fed or fasted. No difference in ovarian activity or ovulation rate was observed between shorn or unshorn ewes at necropsy but fewer fasted or fully-fed shorn ewes exhibited oestrus signs, compared to unshorn ewes (three, six and ten ewes out of ten, respectively). Similarly, Fernandez et al. (1996) found that Corriedale ewes mated in summer in Uruguay (at a similar latitude to south-eastern Australia) were unaffected by shearing if it occurred between 3 months and 3 weeks before mating but that ewes shorn closer to mating had a lower incidence of oestrus, despite shearing time having no effect on ovarian activity, as observed laparoscopically.

2.2.2.2 Shearing & Lamb Birthweight Birthweight is an important factor in neonatal lamb survival (McCutcheon et al. 1981). Excessively heavy lambs, usually singletons, are more likely to experience a difficult or prolonged birth, which causes anoxia or central nervous system trauma that then predisposes the lamb to the neonatal

27 Literature Review starvation-mismothering syndrome (Dalton et al. 1980; Haughey 1981). Low birthweight lambs, commonly twins, are at risk of death from starvation, as well as hypothermia in adverse weather (Alexander et al. 1980; McCutcheon et al. 1981). Thus the manipulation of lamb birthweight may alter rates of lamb survival, especially if there are fewer lighter or heavier weight lambs.

Numerous studies, many of which have been performed in Romneys in New Zealand, have investigated the effect of shearing during pregnancy on lamb birthweight but they have not produced consistent results. For example, two studies showed that shearing during pregnancy had no effect on the birthweight of single or twin lambs (Dabiri et al. 1994; Dabiri et al. 1996); three studies showed that shearing during pregnancy increased the birthweight of singletons only (Morris et al. 2000; Kenyon et al. 2002b; Revell et al. 2002); two studies showed that shearing increased the birthweight of twins only (Black and Chestnutt 1990; Morris 1997); whilst one study showed that shearing increased birthweight of both singles and twins (Kenyon et al. 2002c). The variable results may have occurred because of differing ewe nutrition, environmental conditions at the time of shearing—as determined by lambing time—or the timing of shearing in relation to gestation. Nevertheless, the effect of shearing during pregnancy on birthweight appears to be variable, difficult to control and does not always achieve the aim of increasing the birthweight of lightweight lambs without adversely the weight of heavier ones. More consistent, and greater, beneficial effects on lamb birthweight and survival are known to be achievable by ensuring that ewes are not undernourished during pregnancy (Kenyon et al. 2002b).

Some of the studies that observed an association between pre-lamb shearing and birthweight showed no improvement in lamb survival (Alexander et al. 1980; Grosser et al. 1991; Dabiri et al. 1994; Falck et al. 2001; Kenyon et al. 2002a; Kenyon et al. 2006b). Two studies conclude that pre-lamb shearing improves lamb survival but the evidence for their conclusions is weak. In one, the conclusion was based on a difference in survival observed in one of two years that was statistically non-significant (73% vs. 64% survival, P = 0.08, in South African Merinos Cloete et al. 1994). In the other, shearing increased survival of Romney-cross lambs by 6% at one trial site in New Zealand but had no effect on survival at another (Kenyon et al. 2006a).

2.2.2.3 Shearing & Ewe Sheltering Behaviour Shearing during pregnancy has also been examined to determine whether it can improve lamb survival by encouraging ewes to seek shelter around the time of lambing. An experiment showed that substantially more shorn than unshorn Merino ewes sought shelter beside windbreaks around the time of lambing during winter at Armidale, New South Wales (Lynch and Alexander 1977). A separate experiment showed that shearing up to four weeks prior to lambing stimulated sheltering behaviour, but that fewer ewes sought shelter at lambing if they were shorn eight weeks beforehand (Lynch and Alexander 1980). Despite these behavioural responses, no studies have reliably demonstrated that this leads to improved lamb survival (Lynch and Alexander 1976; Alexander et al. 1980). In Alexander and Lynch’s study, overall neonatal mortality was relatively low (12%) and there were few twin births, which may have affected the statistical comparisons, but fewer singleton lambs born to shorn ewes died within three days of birth than lambs born to unshorn ewes (7% vs. 16%, 0.05 < P < 0.1). Lynch and

28 Literature Review

Alexander’s study reported data from five consecutive years and showed no statistically significant effect of shearing on the mortality of singletons (P = 0.16) or twins (P = 0.064). Furthermore, when bad weather occurred and ewes did not have shelter, there was a much greater increase in lamb deaths from shorn ewes than from unshorn ones. The fact that shearing encouraged ewes to seek shelter but did not improve survival may have been because shearing still did not induce a large majority of ewes to seek shelter (Lynch and Alexander 1977) or because it affected ewe nutritional requirements and lactation, and hence lamb nutrition and survival. The results of these studies suggest that the provision of abundant shelter per se, or strategically locating it where ewes are most likely to use it, is much more effective in improving lamb survival than using shearing to ‘force’ ewes to use pre-existing, limited shelter (Alexander et al. 1979; Alexander et al. 1980; Lynch et al. 1980).

2.2.32.2.32.2.3 Fleece RotFleece Rot Fleece rot is a bacterial exudative dermatitis, mainly associated with Pseudomonas spp., that primarily affects young sheep (Burrell 1988; Colditz et al. 1996). It is important to animal health and wool quality because it is a major predisposing factor for body flystrike and can permanently discolour wool (Gherardi et al. 1983; Raadsma 1988). Prolonged or severe wetting of the skin, leading to maceration of the epidermis and establishment of infection, is a prerequisite for the development of fleece rot because it leads to maceration of the epidermis, which allows the infection to become established.

Wetting of skin is influenced by fleece length, so time of shearing is important to the pathogenesis of the disease. Sheep carrying 4–6 months wool growth during wetting events that were known to induce fleece rot have had up to twice the prevalence of fleece rot of sheep carrying either less or more wool (Raadsma 1988). The greater prevalence appears to occur because the fleeces 4–6 months after shearing absorb considerable quantities of water but are slow to dry out. In contrast, shorter wool dries quickly and longer wool reduces the risk of saturation of the skin in the first place (Copland 1986).

A further two experiments showed a relationship between shearing time and fleece rot, although they differed slightly from the results above (Hemsley et al. 1984). In both experiments, shearing weaners increased the subsequent prevalence of fleece rot, compared to unshorn sheep. In the first trial, fleece rot was assessed at 13 months of age and more fleece rot occurred in weaners shorn at five months of age than in unshorn weaners. In the second trial, fleece rot was assessed at seven months of age and was more prevalent amongst weaners shorn at three months of age than amongst unshorn weaners of the same age or two months younger. Although neither weather conditions nor staple length at the onset of fleece rot were reported in either experiment, the first set of results could be consistent with Raadsma’s (1988) observation of sheep with wool 4–6 months long being most susceptible to fleece rot. In the second experiment, a shorter wool length than that reported by Raadsma was associated with more prevalent fleece rot because more fleece rot developed in weaners with 0–4 months wool (shorn group) than ones with 0–7 or 0–5 months wool (unshorn groups seven and five months old, respectively). The younger unshorn weaners had less fleece rot than the shorn ones, although their staple lengths were similar, and Hemsley et al. suggested that having the wool present at birth, or ‘lambs’ tip’, still on the end of the staple reduced the development of fleece rot. Few sheep were used

29 Literature Review in this experiment and other differences between Hemsley et al.’s and Raadsma’s trials, such as sheep genotype and environmental conditions, might also have contributed to the variation in results.

2.2.42.2.42.2.4 DagDagDag Dag, or faecal soiling of the breech area in sheep, is important because it predisposes sheep to breech flystrike (Watts et al. 1979), and affects enterprise profitability through the cost of crutching and lowered income from affected wool (Larsen et al. 1995). Dag accumulates more readily in longer wool (French et al. 1996) but few studies have investigated the relationship between dag accumulation and time of shearing. Nevertheless, it is likely to be an important relationship because shearing time determines wool length at different times of the year and, in southern Australia, the diarrhoea that causes dag is a seasonal phenomenon that tends to occur during winter and early spring (Larsen et al. 1994). In a study in western Victoria, different weaner shearing times incurred different crutching costs, although the actual differences were not reported (R. Thompson pers. comm.). In this experiment, all weaners were crutched in April when they were seven months old but the presence of dag had necessitated jetting one month previously to control the risk of breech strike.

2.2.52.2.52.2.5 FlystrikeFlystrikeFlystrike Blowfly strike, or cutaneous myiasis, is one of the major diseases of sheep in Australia and has been estimated as being second only to intestinal parasitism in economic importance (Sackett 2006). Wool is an important factor in the development of flystrike because it traps urine, faeces and exudate on the sheep’s skin, which attract adult female flies and provide the moisture necessary for survival of deposited blowfly eggs and the establishment of strikes (Raadsma 1988). The principal blowfly species that causes flystrike in Australia, Lucilia cuprina, is most active during spring and autumn in southern Australia, when there is sufficient environmental warmth and humidity for its development. Thus, time of shearing largely determines susceptibility to flystrike because it dictates the length of wool carried by a sheep at times when blowflies are present. The important types of flystrike in a self- replacing Merino flock are body, breech (around the crutch and tail region) and pizzle (around the preputial opening on rams or wethers) strike. The main predisposing factors for body strike are fleece rot and mycotic dermatitis, whilst soiling around the crutch with dag or urine greatly increases the risk of breech strike, which is the most common form of flystrike in Australia (Gherardi et al. 1983; Raadsma 1988; Wardhaugh and Morton 1990). The presence of urine and other moisture are the main predisposing factors for pizzle strike (Raadsma 1988).

Shearing immediately before the onset of the main time of blowfly activity, which is usually spring in southern Australia, is a major control strategy for flystrike (Fels 1971). It has the advantage of achieving protection without the use of insecticides, it has to be performed anyway, and is effective in preventing body, breech and pizzle strike (Raadsma 1988). Shearing usually protects against breech strike for up to three months and against body strike for up to two months. However, in the case of body strike, shearing at the start of one risk period (spring) can increase susceptibility to flystrike at a later risk period (autumn) because sheep will be carrying 4–6 months wool at this time, which is the

30 Literature Review length most predisposed to developing fleece rot. Nonetheless, shearing usually improves overall flystrike control if it precedes the more significant period of fly activity. For example, a postal survey of 200 sheep producers in the New England region of New South Wales found that shearing in November was associated with a lower incidence of flystrike (Brodbeck and Hill 1984). Similarly, in western New South Wales, Kennedy et al. (1982) found that fewer spring-shorn ewes were affected by flystrike than autumn-shorn ewes in February (3.6% vs. 11.9%, respectively). Despite this difference, Kennedy et al. (1982) suggested that the greater incidence of mortality amongst spring-shorn ewes observed in a separate experiment may have been due to a higher prevalence of flystrike in these ewes during spring and summer. This appears to contradict the findings of Brodbeck and Hill, although they did not specify whether the strikes occurred before or after shearing, whether they were of a particular type (for example, an increased prevalence of fleece rot in spring-shorn sheep in summer) or were associated with particular rainfall events that increased susceptibility to strike.

2.2.62.2.62.2.6 Other Health Issues Shearing can indirectly affect other aspects of sheep health if it affects risk factors for those conditions. For example, shearing before lambing may induce pregnancy toxaemia if ewes do not have access to feed whilst they are yarded for shearing, or if shearing significantly increases their energy requirements (Irving 1991). Shearing has been shown to significantly increase a sheep’s susceptibility to photo- active toxins, such as St John’s Wort (Hypericum), because it exposes the skin to sunlight, which converts toxin molecules in the circulation into a physiologically active state (Bourke 2003). Shearing also increases sensitivity to hot ambient temperatures (Parer 1963) because, although fleece is a thermal insulator, it also provides excellent protection against heating by solar radiation (Bottomley 1979). Thus, in various situations, shearing might be chosen to avoid times of high-risk times associated with particular conditions, such as prior to lambing or sunny times of the year, if they are important on a farm.

Choice of shearing pattern is also important if a lice eradication program is being conducted on a farm, because eradication is facilitated if all sheep on a property can be shorn and treated with insecticide at the same time (Hungerford 1990). This might mean, for example, that young sheep might be shorn at the same time as adults in the year that an eradication program is being undertaken.

As discussed in Section 2.1.4, shearing is an important factor affecting the accumulation of vegetable matter in wool. In addition to its effects on wool quality, grass seed infestation can be an important sheep health issue because seed awns caught in the wool can penetrate the skin and eyes, leading to discomfort, loss of production, infection and death, particularly in young sheep (Campbell et al. 1972; Cornish and Beale 1974). Because shearing reduces the accumulation of vegetable matter in wool (Ritchie et al. 1992), its timing might be chosen to leave sheep carrying short wool before the time of major seed production by problem pasture species, such as barley grass (Hordeum spp.) and spear grass (Stipa spp.). This is a relatively common practice amongst farmers (Cornish and Beale 1974), although a study reported no difference in liveweight gain or feed intake between unshorn and shorn Merinos grazing pastures containing problem species in Trangie, New South Wales (Warr and Thompson 1976).

31 Literature Review

The authors suggested that the lack of production differences between shearing groups was because grass seeds caused discomfort to shorn sheep, making them reluctant to move around the paddock and graze, thereby counteracting the benefit of shearing. However, several other experiments have shown that shearing is likely to reduce the risk of death from grass seed infestation and improve production. Weaner mortality has been shown to be highly correlated (r2=0.52) with the density of barley and spear grass seeds in the pasture (Campbell et al. 1972) and shearing Merino-cross weaners grazing maturing barley grass pastures has been shown to give an immediate and sustained advantage in growth rate over unshorn sheep (Holst et al. 1996).

As with vegetable matter contamination and wool quality, the importance of grass seed infestation to animal health will vary according to the prevalence of problem species in pastures and, where grass seeds are a potential problem, control measures other than shearing, such as manipulation of pasture species, may be more appropriate or cost-effective. For example, Little et al. (1993) showed that shearing weaner sheep once was more profitable than shearing them twice, despite the extra shearing being timed to assist in the control of grass seed infestation and flystrike risk. In this instance, very effective control of grass seed infestation was achieved by slashing and spraying pastures, and shearing afforded sheep little extra protection but increased costs, thus reducing net wool returns.

2.2.72.2.72.2.7 MortalityMortalityMortality Shearing has been shown to increase the mortality risk of sheep, although this occurs intermittently and is influenced by a number of other factors (Hutchinson and Bennett 1962). Nevertheless, anecdotal evidence suggests that, when it occurs, post-shearing mortality can be severe. For example, during a period of windy, rainy weather in September in south-eastern South Australia, 8% of sheep that had been shorn in the preceding fortnight died (Geytenbeek 1962) and at Darkan, Western Australia, 28% of sheep shorn in the preceding month (approximately 100,000 animals) died during storms in January (Buckman 1982). In an Australia-wide survey, Hutchinson (1968) reported that, whilst about half of respondents reported no post-shearing deaths, 6% reported that more than 2% of sheep died in the month following shearing. At that time, the average mortality rate of 0.7% within a month of shearing translated to more than one million sheep dying after shearing annually in Australia, although this figure was not compared to estimates of death rates of unshorn sheep. In direct field studies comparing shorn and unshorn sheep, shearing has been shown to increase mortality risk 17-fold in Merino wethers at Armidale, New South Wales in April (Hutchinson and McRae 1969), five-fold in Border Leicester- cross ewes at Palmerston North, New Zealand in July (Dabiri et al. 1995) and by 63% amongst Corriedale ewes in California, regardless of whether they were shorn in April, July or December (Torell et al. 1969). The difference in these risk ratios may be at least partially explained by the time over which the observations were made: a fortnight, a month and three to five months in the Armidale, Palmerston North and Californian studies, respectively. Hutchinson’s (1968) and Geytenbeek’s (1962) surveys showed that the majority of post-shearing mortalities occur within a fortnight of shearing and that risk is particularly great in the first few days after shearing.

32 Literature Review

Post-shearing deaths are generally attributed to hypothermia, associated with adrenal gland insufficiency following excessive stimulation of the pituitary-adrenal system (Panaretto 1967; Hutchinson and McRae 1969). Sheep are at greater risk of dying if shearing precedes periods of cold, wet or windy weather (Hutchinson 1968), which has also been shown to be associated with altered adrenal function (Reid 1962). Under such environmental conditions, the ambient temperature may fall below the sheep’s lower critical temperature—the temperature below which it must produce extra heat in order to maintain homeothermy (Gregory 1995). The more ambient temperature falls below lower critical temperature, the greater the requirement for extra heat production and, if it cannot produce this extra heat, the sheep will become hypothermic and die. The impact of shearing on lower critical temperature is not restricted to the removal of the fleece. Sheep can experience several days of fasting around shearing as they are mustered and held away from pasture, and because acute cold stress after shearing discourages sheep from grazing (Hutchinson and McRae 1969; Sykes and Slee 1969; Donnelly et al. 1974). Fasting reduces ruminal heat production, thus raising lower critical temperature and the risk of death (Donnelly 1991b).

The particularly severe mortality risk experienced by sheep soon after shearing occurs for two reasons. Firstly, a sheep must rely on tissue catabolism and non-metabolic thermogenesis, such as shivering, to meet its need for extra heat production while appetite is suppressed in the immediate post-shearing period (Farrell and Corbett 1970; Holmes et al. 1992). Secondly, although metabolic acclimatization, to the point where non-metabolic mechanisms are no longer required, occurs within about two weeks of shearing (Donnelly et al. 1974), feed intake may take a longer time to increase (Wodzick- Tomaszewska 1963), requiring further tissue catabolism for heat production. Thus, during the fortnight after shearing, sheep are particularly vulnerable to hypothermia if non-metabolic mechanisms or body energy reserves cannot generate sufficient heat.

The important link between body condition and risk of death after shearing has also been identified in experiments and surveys showing that more thin sheep died after shearing than ones in moderate to good condition (Hutchinson and Lynch 1966; Hutchinson 1968). In the 1966 study, bodyweight and growth rate prior to shearing accounted for 82% of the variation in mortality, with rate of weight loss being much more important than absolute weight in explaining the variation in mortality. In a subsequent experiment, only rate of weight loss, and not body condition or post-shearing pasture availability, was associated with mortality risk, suggesting that sheep in a catabolic state before shearing are at greater risk of death afterwards (Hutchinson and McRae 1969). Similarly, weight loss in the months prior to shearing has been shown to be associated with increased susceptibility to cold off shears (Slee 1966). This may occur because sheep experiencing nutritional stress have a reduced metabolic rate. Donnelly et al. (1974) showed that such animals experience prolonged weight loss after shearing and, since body fat reserves are a primary source of energy for increased heat production following shearing (Holmes et al. 1992), thin body condition will impede a sheep’s ability to maintain body temperature after shearing, putting it at greater risk of death.

33 Literature Review

In contrast to the above studies, no association was reported between body condition and post-shearing mortality of sheep exposed to severe storms in January 1982 at Darkan in south-western Western Australia (Buckman 1982), although this may have been a result of the difficulty of collecting accurate bodyweight results via a survey of farmers after the mortality events had occurred. In this report, the large numbers of deaths after shearing were attributed to sheep not being metabolically acclimated to cold weather during the summer months.

Thus, shearing sheep at a time when their body condition is decreasing, when they are under nutritional stress or when cold storms are likely carries a greater risk of post-shearing mortality. In southern Australia, this combination is most likely in autumn, when sheep have lost weight in the preceding months, environmental temperatures are cooling and rain is becoming more likely (Hutchinson and McRae 1969). Furthermore, at this time of the year sheep have experienced warm environmental conditions for several months and, thus, are not acclimated to cold. Cold-acclimated sheep have greater basal heat production and are more sensitive to the thermogenic effects of circulating catecholamines, making them more able to withstand sudden exposure to colder weather (Graham and Christopherson 1981). In summary, sheep shorn in autumn will require a greater increase in basal heat production—and a greater stimulation by endogenous catecholamines to do so—than sheep shorn at colder times of the year, and are likely to have smaller body energy stores that are more slowly mobilised to increase heat production. All of these factors increase the likelihood death from hypothermia after shearing in autumn, compared to other times of the year.

No studies have specifically examined the effect of shearing time on mortality of weaner Merino sheep but it could be expected that the underlying physiological mechanisms and risk factors would be similar, although the greater body surface area to weight ratio of the smaller animal might make it more susceptible to chilling than an adult. The factors that have been shown to be associated with deaths of young sheep are reviewed in the following section.

34 Literature Review

2.32.32.3 Mortality of Weaner Merino Sheep

This review discusses the extent of Merino weaner mortality in Australia, its underlying causes and factors that have been shown to be associated with mortality in weaner sheep flocks in Australia and overseas. In Merinos, weaning is recommended to occur at approximately 12 weeks of age (Lean et al. 1997), and the term ‘weaner’ usually refers to a sheep between weaning and about 12 months of age. No maximum acceptable mortality rates of sheep between weaning and 12–18 months of age exist in the published literature, however a maximum of approximately 4% has been suggested as a realistic target (Larsen 2002; Behrendt 2003).

2.3.12.3.12.3.1 The Extent of Merino Weaner Mortality in Australia The problem of excessive mortality of weaner sheep, and Merinos in particular, is not new to the Australian sheep industry. For example, whilst visiting the Midland and Great Southern Districts of Western Australia, Sir Ian Clunies Ross reported that farmers were experiencing “considerable difficulty” rearing young sheep (1934). It is difficult to assess the current extent of Merino weaner mortality in Australia because very few recent surveys or scientifically-based estimates of weaner mortality in pasture-based grazing enterprises exist. Despite a paucity of reports concerned primarily with weaner mortality, commercial surveys and field experiments that incidentally or indirectly report mortality data allow estimates to be made. The average and range of Merino weaner mortality obtained from such studies in Australia are listed in Table 2.5. It indicates that rates of Merino weaner mortality exceeding 10% per annum occur in many regions of Australia, and that death rates have not decreased in recent times.

Excessive post-weaning mortality can have important, adverse effects on the efficiency of a farming enterprise. The death rate in Merino weaners of 24% reported by Rose (1972) in the semi-arid environment of north-western Queensland was more than twice the maximum that could be sustained by the flock if any form of genetic selection through culling were to occur. In an analysis of reproduction over 4 years in a Merino flock in a similar environment, post-weaning mortality was identified as one of the three most serious causes of reproductive wastage (Kennedy et al. 1976). Such high rates of mortality in northern Australia have also been reported more recently; in only three months 14% of weaner sheep died in an experiment conducted in the late 1980s under grazing conditions common to the region (Cobon et al. 1990).

The mortality results from some of the studies listed in Table 2.5 warrant further discussion because of particular details or because it is difficult to summarise the reported death rates. Harris and Nowara’s (1995) survey of sheep mortality was of mixed cropping-sheep enterprises in the Victorian Mallee. Mortality patterns in these enterprises may differ from grazing-only farms because of the availability of grain for supplementary feeding and the opportunity to graze crop stubbles. In this study, average Merino weaner mortality was 4% per annum, which is low when compared to the other listed studies.

35 Literature Review

However, more than one fifth (6/29) of Merino flocks had annual weaner mortality rates higher than 5% and, in 8% of all flocks surveyed, between 10 and 25% of weaners died each year.

Studies reporting specific animal health conditions, such as trace element or vitamin deficiencies, or gastrointestinal parasitism, may include mortality data. In such situations, high mortality might not be unexpected in the absence of treatment. However in the studies listed in Table 2.5, weaner mortality was still high in the groups that received treatment. For example, an average of 59% of Merino weaners died of vitamin A deficiency in a feeding trial of drought rations (Franklin et al. 1955). However, the death rate was still 16% amongst weaners that received vitamin A. In selenium supplementation trials in Victoria, Western Australia and South Australia, 6–16% of Merinos receiving selenium still died after weaning (Gabbedy 1971; Walker et al. 1979; Holmes 1992). Mulhearn (1958) reported up to 50% post-weaning mortality amongst Merinos and Corriedales in the high-rainfall zone of South Australia but could only attribute some of these cases to trace element deficiency or specific diseases.

Table 2.5: Mortality of Merino and Merino-cross sheep between weaning and approximately 18 months of age reported in Australian commercial enterprises or field experiments Post-weaning Location Reference mortality (range) 16% feedlot (Franklin et al. 1955) up to 50% south-eastern South Australia (Mulhearn 1958) 25% north-western Queensland (Moule 1966) 7% Western District, Victoria (Russell 1968) 14% Western Slopes, NSW (Drinan 1968) 6% (up to 24%) southern Western Australia (Gabbedy 1971) 24% Julia Creek, Queensland (Rose 1972) 12–22% Western District, Victoria (Anderson et al. 1976) 18% Central Victoria (McDonald 1975) 14% Broken Hill, Queensland (Kennedy et al. 1976) 7–10% feedlot (Davis et al. 1976) 10% Kybybolite, South Australia (Brown 1977) 17% (11–29%) Kangaroo Island, South (Walker et al. 1979) Australia 5% (up to 13%) Katanning, Western Australia (Norris 1984) 6% Western Slopes, NSW (Langlands et al. 1984) up to 56% Kybybolite, South Australia (Brown et al. 1985) 39% Condobolin, New South Wales (Denney et al. 1988) 22% n/a (Mulholland 1986 cited by Allworth 1994) 14% Julia Creek, Queensland (Cobon et al. 1990) 11% East Gippsland, Victoria (Holmes 1992) 4% (up to 25%) Mallee, Victoria (Harris and Nowara 1995) 4% (1–10%) Hamilton (Vic), Struan (SA) & (Fogarty et al. 2005a) Cowra (NSW) 12% (2–27%) Victoria (J. Webb Ware pers. comm.)

36 Literature Review

Field trials assessing parasite control programs have reported high mortalities even in weaner groups receiving recommended worm control treatments. For example, 12% of Corriedale weaners died in a trial in the Western District of Victoria, despite the use of double summer drenching, a strategy which remains the mainstay of effective worm control in that environment (Anderson et al. 1976). In a similar experiment involving Haemonchus, Trichostrongylus and Ostertagia infections of sheep in East Gippsland, 13% of Merino weaners died in one year despite weekly anthelmintic treatment and 19% died in groups managed under the common district approach of the time, which involved giving between three and six annual treatments (Barton and McCausland 1987). All of these results suggest that excessive weaner mortality cannot always be attributed to a single, treatable aetiology.

It is striking that virtually none of the studies listed above comment on the rates of weaner mortality that they report. This suggests that mortality rates of 10%, or even 20%, are not considered unusual or noteworthy, and may explain how excessive mortality could exist in the Australian sheep industry and receive little attention. For example, in Barton’s (1987) trial of parasite control, ‘salvage’ drenching “to…prevent…high mortality” only occurred once deaths exceeded 10%, implying that 10% mortality was not considered to be concerning or controllable. This attitude does not appear to differ from one expressed in the 1950s, when Franklin (1955) considered that supplementary feeding during drought satisfactorily reduced mortality when only 15% of Merino weaners died. In an extensive review of progress in reproductive research in the Australian sheep industry, Scaramuzzi (1988) discussed the improvements made over many decades in our understanding of drivers of neonatal lamb mortality but made no mention of the extent, or role, of post-weaning survival in the intergenerational ‘efficiency’ of sheep reproduction. Nonetheless, excessive weaner sheep mortality has been labelled one of the “biggest issues for Merino producers” (Ripper 2003) (p. 2) and a recent economic evaluation suggested it is the fourth most costly animal health issue in the Australia sheep industry (Sackett 2006). Many farmers have difficultly successfully managing weaner survival. More than one quarter of Western District woolgrowers enrolled in an extension program for innovative farmers nominated successful weaner feeding, and achieving target growth rates and bodyweights, to be their primary concern (Larsen 1999). Despite this, the fatalistic and frustrated attitude expressed by many farmers, and the very fact that so few recent reports of mortality on commercial farms exist, implies that the problem is under-reported, under-researched and possibly one that the industry feels is beyond its control.

2.3.22.3.22.3.2 Limitations of Weaner Mortality Investigations Studies that seek to elucidate the risk factors underlying mortality must necessarily be observational because it is not ethical to observe animals dying without intervening. No study has specifically examined mortality in Merino weaners grazed in southern Australia or a similar, Mediterranean-type environment. Thus, because reports of weaner mortality are frequently incidental to a study’s principal aims, associations between proposed risk factors and mortality are frequently confounded with other variables. Plausible physiological explanations of the observed data can be offered, however ethical considerations prevent their testing in controlled experiments. Furthermore, interpretation of mortality

Table 2.6: Summary of methodologies of studies reporting risk factors for post-weaning mortality

Breed/ Location Features of Observation Experimental or Analytical methodology Study Species production system period* Observational Study; n†

Merino, Kybybolite, SA‡ autumn lambing 6–12 Exp; ~20×3 descriptive (Allden and Anderson 1957) Merino ×

Merino Adelaide, SA autumn lambing 6–12 Exp; 24×4 descriptive, χ2 test of proportions (Allden 1968c)

Merino, western Vic autumn lambing ~6–12 Obs; unspecified descriptive case reports (Engel 1958) other

Rambouillet north-western India extensive grazing + 5–12 Obs; 3285 difference of least squares means (Ganai and Pandey 1996) (temperate climate) supplementary feeding Literature Review Review Literature Merino, Mallee, Vic mixed enterprises ?–12 Obs; 79 farms × 2 descriptive (Harris and Nowara 1995) other (sheep + cropping) years 38

goat Kenya (semi-arid extensive grazing + 4–24 Obs; 270 discrete-time logistic regression (Hary 2002) savannah) supplementary feeding based on cubic spline hazard functions

Merino Werribee, Vic spring lambing; feedlot ~4–12 Exp; 10×3×3 χ2 test of mortality proportions (Hodge 1990) then pasture

goat subhumid Nigeria smallholder; foraging ± 5–12 Obs; 877 univariate analysis of cumulative (Ikwuegbu et al. 1995) access to reserved mortality using maximum fodder likelihood estimates

Merino, SA autumn lambing 3–16 Exp; 591 χ2 test of mortality proportions (Kleemann et al. 1983) Merino x

Merino WA autumn, winter or 3–12 Exp; 375 χ2 test of mortality proportions (Lloyd Davies 1983) spring lambing at average to high district stocking rates

Merino NSW drought feedlot 3–6 Obs; 209 t-test of survivor vs. non-survivor (Lloyd Davies et al. 1988) weaning weights

goat French West Indies irrigated pasture 3–9 Obs; 837 parametric & semi-parametric (Mandonnet et al. 2003) survival analysis

Merino, Hamilton, Vic autumn, winter or 3–~11 Exp; unspecified χ2 test of mortality proportions (McLaughlin 1973) Corriedale spring lambing

Merino, Kangaroo Island, SA winter lambing 6–10 Exp; 20–40/group descriptive (Mulhearn 1958) Corriedale

African Kenya (sub-humid extensive grazing 3–12 Obs; 1785 semi-parametric survival analysis, (Nguti et al. 2003) sheep tropics) including time-varying covariates

Indian north-western India extensive grazing + 3–6 Obs; 1272 difference of least squares means (Nivsarkar et al. 1982) sheep (temperate) supplementary feeding

sheep north-western India extensive grazing 3–6 Obs; 10000 descriptive (Sharma et al. 1981) (temperate)

meat sheep USA feedlot 2–12 Obs; 8642 logistic regression, semi- (Southey et al. 2001) Review Literature parametric & parametric survival analysis 39

goat Ghana extensive grazing 3–12 Obs; 351 χ2-test of mortality proportions; (Turkson et al. 2004) univariate odds ratios

African Ghana extensive grazing 5–12 Obs; 453 χ2-test of mortality proportions; (Turkson and Sualisu 2005) sheep univariate odds ratios

African Ghana extensive grazing, ~5–12 Obs; 868 χ2-test of mortality proportions; (Turkson 2003) sheep & average flock size = 10 univariate odds ratios goat

Merino x n/a pen-study 1–3 Exp; 26 correlation between weaning (Walker and Hunt 1980) weight and duration of post- weaning survival

? California extensive grazing ?–22+ Obs; ~26×3 × 3 comparison of mortality (Weir and Torell 1967) years proportions

* age in months, beginning at weaning † x × y means x number in y groups ‡ Abbreviations refer to Australian states: NSW: New South Wales; SA: South Australia; Vic: Victoria; WA: Western Australia

Literature Review data is difficult in small trials where differences in mortality proportions may be due to the death of only one or two extra animals, and no Australian studies have used statistical techniques appropriate for analysing mortality data. Therefore, despite the different production systems examined, some quantitative estimates of the association between risk factors and weaner mortality must be obtained from overseas studies of sheep and goats.

2.3.32.3.32.3.3 Causes of Mortality & Illthrift Although weaner sheep can die from numerous diseases and syndromes, no single aetiological agent is consistently associated with Merino weaner mortality in Australia. It has thus been identified as a component of the ‘weaner illthrift’ syndrome (Gordon 1981) and a manifestation of illthrift in its most extreme form (McLaughlin 1967). However, its importance is usually alluded to without being specifically identified. For example, Wilkinson (1981) defines illthrift as a “failure of weaner sheep to thrive…when all other classes of sheep appear to be of satisfactory health” but does not mention mortality, despite going on to list illthrift’s primary manifestation as “survivability”.

Weaner illthrift is a syndrome of multiple aetiologies that is frequently associated with the grazing of dry pastures (McLaughlin 1967). Inadequate nutrition more generally, ranging from insufficient protein or energy intake to specific trace element deficiencies, has been identified as the primary factor predisposing Merino weaners to illthrift and high mortality in Australia (Gordon 1981). Weaners in poor body condition, particularly whilst grazing dry summer pastures and suffering insufficient energy or protein intake, are considered to be more susceptible to other diseases (Allworth 1983). After inadequate nutrition, gastrointestinal parasitism is a common secondary contributor to weaner illthrift and death (Gordon 1981). In winter rainfall environments, death solely due to parasitism is thought to be rare during summer (Wilkinson 1981). Rather, it commonly is considered that undernutrition places weaners at risk of mortality and that parasitism then precipitates death (Beveridge et al. 1985). A variety of diseases, either following these two major predisposing conditions or independent of them, have also been implicated in outbreaks of weaner mortality in southern Australia. These include myopathies (usually associated with selenium or vitamin E deficiency), acidosis due to cereal grain overload, polioencephalomalacia, grass seed infestation, flystrike, mycotic dermatitis, scabby mouth and lupinosis (Wilkinson 1981; Hungerford 1990).

The risk factors associated with weaner mortality are discussed in the following sections, and the design and methodologies of the experiments cited therein are outlined in Table 2.6.

2.3.42.3.42.3.4 Factors Associated with Weaner Mortality

2.3.4.1 Bodyweight Poor body condition and low growth rates at various ages have been shown to be associated with reduced survival of small ruminants after weaning. A postal survey of 79 enterprises in the Victorian Mallee (Harris and Nowara 1995) characterised patterns of sheep mortality on mixed farms (i.e. cereal cropping and sheep). In flocks experiencing more than 5% weaner deaths, loss of body condition was

40 Literature Review the clinical sign most frequently reported by farmers, with 86% of flocks in moderate or thin body condition when high mortalities were most likely to occur.

A number of field studies have further characterised the relationship between mortality and bodyweight (Ganai and Pandey 1996 ; Hary 2002 ; Turkson and Sualisu 2005). In general, these have shown that the relationship is independent of a weaner’s age at death and the cause of death. For example, in studies of Rambouillet lambs in India (Ganai and Pandey 1996), East African Dwarf goat kids in Kenya (Hary 2002) and Sahelian lambs in Ghana (Turkson and Sualisu 2005), animals lighter than the average birthweight of 2–3 kg, were approximately twice as likely to die after weaning than heavier- than-average birthweight animals, even when weaning occurred up to five months after birth. In all instances, the association between birthweight and mortality persisted for as long as the progeny were observed, which was to one (lambs) or two (kids) years of age.

Weight at weaning has been shown to have a similar, persistent association with post-weaning survival. Increases in weaning weight of 5–10 kg in Merinos have been associated with approximately two- to four-fold decreases in post-weaning mortality risk, as follows:

• 9% deaths at 22 kg vs. 2% at 27 kg (Allden 1968c) • 80% at 12 kg vs. 33% at 21 kg (Lloyd Davies 1983) • 18% at 11 kg vs. 6% at 15 kg (Lloyd Davies et al. 1988) • 21% at 15 kg vs. 4% at >27 kg (Holmes 1992 )

All of these results equate approximately to an odds of death ratio of 1.3–1.5 for a 1 kg decrease in weaning weight. A similar mortality pattern across weight groups was also observed in a drought feeding study, even though the sheep were only followed until four months of age (Franklin et al. 1964).

Two trace element supplementation trials showed that a greater proportion of weaners that were below average bodyweight at weaning died after weaning than ones of above average bodyweight. Amongst sheep grazing native pastures in East Gippsland, Victoria, 21%, 8% and 4% sheep died from the lightest, middle and heaviest third of weaners respectively (Holmes 1992; Figure 2.5). On Kangaroo Island, South Australia, 75–83% of weaners grazing improved pastures that died were lighter than the average breed/strain weaning weight (Walker et al. 1979). Similar results were observed in a feedlot trial involving Merino-cross weaners receiving wheat-based rations, where survivors were significantly heavier than weaners that died (6.5–9.5% deaths between 3 and 9 months of age; Davis et al. 1976).

The similar results reported by all the studies discussed above is striking because they were conducted in widely differing environments in Western Australia, South Australia and New South Wales, and in different production systems with different lambing times. They also included studies of sheep grazing pasture or being fed in drought or production feedlots.

41 Literature Review

It is difficult to compare directly the relative importance of weight at birth, weaning and later times to survival. However, two studies have shown that bodyweight had a similar association with risk of death of weaners, regardless of when it was measured. One examined survival of Red Maasai, Dorper and crossbred sheep in Kenya between weaning (at three months old) and twelve months of age (Nguti et al. 2003), during which time 31% (443/1442) of weaners died. Decreased weaning weight was associated with increased mortality rate, with the largest effect at low bodyweights. A 1 kg decrease in weaning weight increased the mortality risk of lightweight weaners by about 60%, equating to a hazard ratio of about 1.6, and this effect became progressively smaller for heavier weaners. A similar effect was estimated for bodyweight at any time in the post-weaning period if this time-varying term was used instead of weaning weight in the statistical model. Although no techniques exist to readily estimate how comprehensively one survival model accounts for observed variation compared to another, this result suggests that it is bodyweight in general, rather than birth or weaning weight in particular, that is associated with mortality.

The second study to analyse the effects on mortality of both weaning weight and general bodyweight examined deaths caused by gastrointestinal parasites of Creole goat kids grazing irrigated pastures in the tropical French West Indies (Mandonnet et al. 2003). The hazard ratio for a 1 kg decrease in weaning weight was 1.33. The hazard ratio for a bodyweight decrease of 1 kg at any time was 1.69. The reason for the difference between the hazard ratios for weaning weight and time-varying bodyweight was not discussed. The inclusion of bodyweight as a time-varying covariate eliminated significant associations between mortality and three other explanatory factors, sex, rearing method and anthelmintic treatment. The authors noted that this did not necessarily imply that bodyweight was the

Figure 2.5: Mortalities in Merino weaner sheep, classified by weight at weaning, from: 3–12 months of age on pasture (Lloyd Davies 1983), 3–6 months of age in a drought feedlot (Lloyd Davies et al. 1988), and 4½–6 months of age on pasture (Holmes 1992)

100%

90%

80%

70%

60%

50%

40%

30% Proportion of deaths 20%

10%

0% 10-12 12-14 14-16 16-18 18-23 23-27 >27 Weaning weight range (kg)

42 Literature Review only factor to significantly influence mortality. This was because kids in the study died after severe weight loss and, although other factors may have contributed to mortality, the reduction in bodyweight prior to death was so profound that other effects were statistically diminished to the point of non- significance. Only death caused by gastrointestinal parasites, which are not typically associated with acute mortality, was examined.

All the results discussed above show that similar associations between post-weaning mortality risk and bodyweight exist, regardless of whether the sheep are weighed at birth, weaning or a later time. Generally, a 1 kg weight decrease was associated with an odds or hazard ratio ranging from 1.3 to 2. The consistency of these observations is remarkable because they were made in widely ranging environments (Mediterranean-type, semi-arid and tropical), involved very different breeds and crosses of sheep and goats and were independent of the cause of death. Weaning weight has been shown to be well correlated with weight at subsequent times (Lloyd Davies et al. 1968 in Merinos), so it is likely that the association between mortality and weight at a particular time such as weaning is actually a reflection of a more general association between bodyweight and survival.

Lamb bodyweight at premature weaning has also been shown to be associated with risk of death in several experiments. Within groups of Merino lambs weaned at different ages between 4 and 12 weeks, survivors were 1–2 kg heavier when weaned onto a cereal grain and chaff diet than those that died (Franklin et al. 1964). The heavier weaning weight of survivors was associated with a heavier birthweight, and a higher pre- and post-weaning growth rate. All the lambs that died either lost weight or experienced negligible gain after weaning. Lambs weaned at a younger age also died more quickly than later-weaned lambs, despite being of similar bodyweight (7–9 kg) at weaning. For lambs weaned at the same age, there was no association between weaning weight and survival time but the range in weaning weight was small. In contrast, when Merino-cross lambs were weaned at three weeks onto a pelleted ration, there was a strong correlation (r = 0.94) between weaning weight and duration of survival after weaning (Walker and Hunt 1980). In this experiment, the distribution of weaning weights was greater and three quarters of the lambs died within four weeks as a result of poor consumption of the ration. Walker and Hunt concluded that bodyweight was associated with post- weaning survival in these young sheep because increasing weight represents the accumulation of energy stores required to sustain the lamb during adaptation to a new nutritional situation. Thus, lightweight lambs were more likely to die because their lower body energy stores were exhausted before full adaptation to the new diet rendered the reserves unnecessary. This concept is discussed in more detail below.

2.3.4.2 Growth Rate Allden’s (1968c) study of the effects of undernutrition of growing Merinos on lifetime production presented data that suggest associations between post-weaning growth rate, as well as weaning weight, and survival. The sheep were born in May in South Australia and weaned at six months of age. Nutrition was restricted before or after weaning by varying the stocking rate of ewes and lambs grazing green pasture during winter and spring, or withholding supplementary feeding when weaners grazed

43 Literature Review dry pasture throughout summer and autumn. Thus “low” and “high” pre-weaning nutrition groups were further divided by “low” and “high” post-weaning nutrition. This created four groups of 24 sheep that had experienced different nutritional regimens throughout their first year of life. The mortality proportions are summarised in Table 2.7. No vaccination against clostridial diseases was performed and 13% of lambs died of enterotoxaemia before weaning. These deaths, from a preventable disease, obscure any relationship between bodyweight, growth rate and survival in the 0–6 months period, and the 0–12 months period. However, from 6 to 12 months, more light weaning weight (average 22 kg) sheep died than heavy weaning weight (average 27 kg) sheep. In addition, more weaners that were nutritionally restricted after weaning (average weight loss 0.8 kg/month from 6–12 months of age) died than weaners fed ad libitum after weaning (average growth rate 0.8 kg/month). Allden commented that, “It is probable that under extensive grazing conditions and with no immediate access to water more of these [low pre- and post-weaning nutrition] sheep would have died”. The trial involved low numbers of animals, so no difference between proportions was statistically significant. Furthermore, because the survival of individuals was not analysed, it is not clear whether weaner survival was associated with absolute bodyweight or rate of weight change. Nevertheless, the results point to relationships between mortality and both weaning weight and post-weaning growth rate consistent with those of the more specific mortality studies discussed previously.

There was no relationship between the level of supplementary feeding and mortality in Merino weaners kept in a drought feedlot in New South Wales (Lloyd Davies et al. 1988). However, weaner growth rate and final bodyweight in each treatment were similar, despite the diets differing in metabolisable energy by 2 MJ/kg. This lack of difference in growth rate might explain why similar numbers of sheep died in each group. Similarly, weaner bodyweight and the rate of weight loss were related to survival of Merino-cross sheep being fed wheat-based rations for production in feedlots (Davis et al. 1976). In that experiment, 43% of sheep that died lost more than 100 g/day prior to death and the weaners that died were significantly lighter than those that survived. The deaths were independent of ration formulation and feedlot management.

Mulhearn’s (1958) report of Merino weaner management in South Australia contained information

Table 2.7: Post-weaning mortality of sheep in different nutritional groups reported by Allden (1968c) Nutritional treatment Average weight at Post-weaning end of period (kg) mortality pre-weaning: low 22.0 9% high 27.0 2%

post-weaning: low n/a 9% high n/a 2%

pre- & post-weaning: low-low 17.4 13% low-high 26.0 4% high-low 22.8 5% high-high 31.2 0%

44 Literature Review about bodyweight and mortality, although it was confounded with the effect of season and property- specific management practices, such as supplementary feeding. He observed that no winter-born weaners died during summer in a year when the average peak summer bodyweight was 37 kg. However 15% died in the following year when post-weaning growth rates were lower and average peak summer bodyweight was 27 kg. In the high-mortality year, groups of weaners grazing dry pasture were supplemented with oats or linseed meal ad libitum and compared to weaners receiving no supplement. Supplemented weaners gained an average of 2.5 kg/month (linseed) and 1.6 kg/month (oats) but unsupplemented weaners lost 1 kg/month between February and May. In the oats and linseed groups, 8% and 10% died, respectively, compared to 28% of weaners in the control group. Deaths also continued for longer in the control group and nearly all the weaners that died were of low weight or emaciated. In the supplemented groups, it was noted that the weaners that died were lighter prior to the start of supplementation. This pattern of mortality was similar to those described by Allden (1968c) and Lloyd Davies (1983; 1988) and Hodge (1990). Mulhearn described a situation of fewer deaths and smaller differences between groups on another property where different supplements were used. However on this farm the controls experienced only a slight weight loss whilst grazing dry pasture. This suggests that it is the bodyweight or growth rate achieved by supplementary feeding, rather than the provision of feed itself, that is associated with mortality.

In a field experiment in the same environment, 20% (3/15) of a monitor group of unsupplemented autumn-born Merino weaners died, compared to no deaths in groups receiving various supplementary feeds (Allden and Anderson 1957). During summer it became necessary to start feeding the unsupplemented group, as “many of the young sheep were…in critical condition” (p. 75). The unsupplemented sheep were at least 1 kg lighter than their counterparts at the start of feeding in January and lost 4 kg throughout February, when the deaths occurred, compared to gains of 0.2 and 2.2 kg/month by the weaners receiving supplementary feed. In another study in Victoria, approximately four times as many deaths occurred in a group of Merino weaners grazing dry pasture over summer without supplementation (McLaughlin 1973). Although detailed data were not presented, unsupplemented and supplemented weaners were approximately the same weight at weaning but unsupplemented weaners always grew at a slower rate than the supplemented groups.

Mortality patterns also coincided with bodyweight changes rather than feeding regimens in a seven- year trial of supplementary feeding strategies for weaner wool sheep in a Mediterranean-type climate in inland California (Weir and Torell 1967). Significant mortalities occurred in only one year, which was associated with light weaning weight and weight loss over summer when the nutritional quality of pastures was particularly poor. In this year, 17% of unsupplemented (control) weaner sheep died between the start of summer and the following spring, whereas there were only “minor…losses” in the groups receiving supplementary feed. These results were in contrast to those from another year, when, although similar weight loss occurred during summer, sheep had been considerably heavier at weaning and the death rate was much lower. These results are similar to those of Allden et al. (1968c), discussed above.

45 Literature Review

In the studies described above, the associations between bodyweight, bodyweight change and mortality rate were confounded with the provision of supplementary feed. However the patterns of mortality across the different bodyweight or growth rate groups were similar, regardless of the type of supplementary feed provided. Furthermore, supplementary feeding was only associated with differences in survival if it was accompanied by a growth rate response. This suggests that bodyweight or bodyweight change, rather than supplementary feeding per se, is associated with survival of weaner sheep. However the notion that generally providing supplementary feed to Merino weaners effectively prevents excessive mortalities was challenged by Norris (1986), who found no correlation between the cost of supplementary feed and Merino weaner survival on 10 different farms in south-west Western Australia. He concluded that farmers did not understand the specific protein and energy requirements of Merino weaners and therefore failed to prevent weaner mortalities because they provided inappropriate supplementary feed, especially hay.

It is difficult to make general statements about the extent to which dry pastures match the nutritional requirements of weaner sheep because the nutritional characteristics of summer pastures vary tremendously between locations and seasons. Furthermore, the nutritional requirements of weaners also change as they grow older. One study of supplementary nutrition for Merino weaners grazing typical dry, summer pastures in southern Australia showed a linear response of liveweight gain to increasing energy supplementation but a declining response to increasing levels of supplementary protein (Allden 1959). From these results Allden concluded that such pastures primarily fail to meet weaners’ energy requirements but are only nominally deficient in protein. Similarly, Foot et al. (1983) commented that weaners do not show a response to supplementary protein if even small amounts of green feed, which is high in protein, are present in the pasture being grazed. In contrast, another experiment achieved higher growth rates in Merino weaners by increasing the protein content of isocaloric supplementary rations offered from December to May (9–15 months of age) (Lloyd Davies et al. 1968). Weaners offered a supplement containing 22% crude protein (CP) grew at 1 kg/month from December to March, whereas weaners offered a supplement containing 8% CP gained 0.6 kg/month, whilst a control group offered no supplement lost 0.8 kg/month throughout this time. However, no comparison was made between these diets and the expected protein intake from grazing pastures containing different quantities of green feed.

2.3.4.3 Bodyweight & Death from Causes Other Than Malnutrition Many of the studies reporting an association between bodyweight and survival have focussed on deaths of weaners during summer, when growing sheep frequently suffer malnutrition while grazing the dry pastures typical of the Mediterranean-type environment of southern Australia (Bellotti et al. 1993). An association between bodyweight and survival of sheep grazing dry pastures and dying principally from malnutrition is unsurprising, as weight loss inevitably precedes death. However several studies have reported a similar association between bodyweight and death at other times of the year and when weaners die from other causes. For example, a study of three strains (Merryville, Peppin and Collinsville) of Merino weaners showed that bodyweight was related to risk of death due to gastrointestinal parasitism during winter in southern Victoria (Hodge 1990, summarised in Table 2.8).

46 Literature Review

In that study, lambs born in October were weaned at between nine and 20 weeks of age, in order to create different weaning weight groups. They were then fed to maintain weight until the following April. At this time they were either grazed on abundant or restricted pasture (>1500 kg DM/ha or <700 kg green DM/ha), resulting in growth of 3.3 kg/month or approximate maintenance of weight (lost or gained < 0.3 kg/month) respectively. This resulted in bodyweights at the end of the autumn break that varied approximately twofold—from 16 kg to 30 kg (for Merryville and Peppin strains), and 21 kg to 37 kg (Collinsville). Most deaths occurred in the groups that were light at weaning and then only maintained weight during autumn. Significantly fewer weaners died in the two groups that achieved an intermediate weight by the end of autumn. No deaths occurred in the group that were heaviest at weaning and then gained weight throughout autumn. It is of significance that similar deaths occurred in the two groups of equal weight at the end of autumn, regardless of how it was achieved. This suggests that it is absolute bodyweight, rather than the timing of its acquisition, that is associated with survival. No association was observed between proportion of weaners dying and their rate of growth during winter. However results for individual animals were not reported and so more detailed relationships between growth rates and mortality may have been overlooked.

An association between bodyweight and mortality that extended beyond times of malnutrition was also observed in survival analyses conducted overseas of sheep (Nguti et al. 2003) and goats (Mandonnet et al. 2003). In these analyses, the ratio between death rates of animals from different weight group classes remained constant over time. This indicated that bodyweight differences accounted for the same difference in risk of death, regardless of the time of year. Additionally, no cause of death was specified in Nguti et al.’s (2003) study, yet bodyweight was significantly associated with survival. In Hary’s (2002) study, gastrointestinal parasites, and not malnutrition, were a leading cause of death. These results are particularly significant, since bodyweight was associated with survival even though the cause of death was not directly related to nutrition.

2.3.4.4 Physiology Underlying the Association between Mortality & Bodyweight It is appropriate to consider briefly how bodyweight might confer a survival advantage on weaner sheep and whether such mechanisms are consistent with the results described above. Following a period of weight loss, death occurs when body energy reserves have been fully consumed and the function of the vital organs is compromised as they are catabolised in an attempt to meet the animal’s

Table 2.8: Mortality of weaners with different weaning weights and autumn growth rates reported by Hodge (1990) Weaning weight Autumn growth rate (kg/month) 0 3.3 low 11–22% * 4–5%

intermediate 0–5% 0%

high 0–4% 0%

* range is for different strains

47 Literature Review daily metabolic requirements (Thornton et al. 1979). Lightweight sheep are at greater risk of death from this process than their heavier counterparts for several different reasons, which are discussed in more detail below.

An important aspect of the association between bodyweight and survival is that the total body energy stores of weaner sheep increase with liveweight. A linear relationship, independent of breed, between total body energy and liveweight was demonstrated in an experiment that measured body composition and total energy of Merino and Merino × Dorset Horn sheep between weaning at 7 weeks (average liveweight 15 kg) and 10 months of age (average liveweight 45 kg; Allden 1970). Heavy sheep had a higher energy density than lightweight ones and the proportion of energy stored as fat also increased as the animals grew. For example, at 15 kg, energy density was approximately 1 MJ/kg and 13% of total body energy was stored as fat, whereas the energy density of weaners weighing 30 kg was 2.2 MJ/kg and 70% was stored as fat. The variation of fat content with bodyweight has also been shown to be independent of the growth rate of sheep of different breeds (Kellaway 1973).

These results indicate three mechanisms by which heavy weaners can mobilise body energy with less risk of dying than light weaners. Firstly, heavy weaners have a greater total body energy reserve upon which they can draw. Secondly, heavy weaners will lose less weight in mobilising a given quantity of body energy than light weaners because of their higher energy density. Finally, more of this reserve is fat, the catabolism of which spares protein in the vital organs (Thornton et al. 1979). Therefore, light weaners forced to mobilise body energy, for any reason, will lose more weight than heavy weaners, and a greater proportion of this weight loss will be protein contained in vital tissues, to the detriment of organ function and ultimately survival (Allden 1970; Doyle and Egan 1983).

Once weaners are lightweight, their mortality risk is increased further because they replace lost body protein more slowly than heavy sheep under conditions of limiting protein and energy intake (Allden 1970). Such nutritional conditions are typically experienced by sheep grazing dry summer pastures in southern Australia. For example, in a field study of Merino weaners grazing poor quality, dry phalaris/subterranean clover pastures in South Australia, heavy (average 36 kg) sheep prior to the period of summer weight loss regained bodyweight more quickly than light (28 kg) ones during the following winter (Donald and Allden 1959). This was despite the more severe nutritional circumstances during summer experienced by the heavier group.

The age at which growth is restricted also affects how quickly young sheep regain weight, which in turn influences the length of time they remain at risk of death from low bodyweight. Merinos whose growth was restricted before 6 months of age recovered bodyweight more slowly than weaners restricted between 6 and 12 months of age (Allden 1968b). When Corriedale ewes were nutritionally restricted (by weaning onto dry pasture) at 8 weeks of age, they were 10% lighter at their first mating (38 vs. 42 kg) than if they were weaned at 12 or 16 weeks (McLaughlin 1966).

Finally, weaner sheep can consume nearly 20% more highly digestible feed than adults, relative to metabolic liveweight, but their relative feed intake decreases at lower digestibilities, and is similar to

48 Literature Review adults at feed digestibilities of 46% and 56% (Egan and Doyle 1982). Thus, nutritional deficiencies of dry summer pastures are exacerbated for weaner sheep because their intake is relatively lower when grazing feeds of lower digestibility.

The nutritional deficiencies that occur when grazing dry pasture explain the improvement in survival achieved by providing weaners with appropriate supplementary feed. Providing high-digestibility supplementary feed during times of nutritional deprivation increases, or at least reduces the decline in, the fat and total body energy content of the weaner sheep’s carcase (Weir and Torell 1967). Furthermore, the growth response to supplementary feeding over a weaner’s first summer of life has been shown to be independent of the weaner’s age (Marshall 1985). This shows that weight loss, and its associated survival outcomes, is not obligatory for even young sheep grazing summer dry pastures, provided they receive appropriate supplementary nutrition.

2.3.4.5 Season & Year of Birth Studies that have spanned more than one year show that weaner mortality can vary substantially from season to season and/or year to year, although the effects have not always been reported directly or quantified. In southern Australia, most weaner deaths occur during summer. For example, in field experiments in South Australia, 82% and 79% of all weaner deaths occurred between January and March, regardless of the age of the sheep (Lloyd Davies et al. 1968; Lloyd Davies 1983). Similarly, in western Victoria, half of all the occurrences of excessive mortality in weaner flocks occurred between February and April (Engel 1958). Such seasonal peak in mortality is consistent with the decline in bodyweight, and increase in risk of death, experienced by weaners grazing the dry summer pastures of southern Australia, as discussed in the previous section.

In contrast to these findings, Harris and Nowara (1995) reported that the deaths in nearly all weaner flocks with greater than 5% mortality occurred between May and August, or during December. Unfortunately, no breakdown between Merino and non-Merino flocks was provided. This survey was primarily of cereal grain cropping enterprises, where weaners often graze crop stubbles during summer and autumn. Summer weaner mortality may therefore have been lower than in solely pasture-grazing enterprises because of the improved nutrition afforded by stubbles and residual grain soon after harvest.

In contrast to the association between dry conditions and increased weaner mortality, some studies have reported that weaner mortality is greater during seasons of higher rainfall, when it would be expected that feed quantity and quality would satisfy the nutritional requirements of weaner sheep (for example, Beveridge et al. 1985; Mandonnet et al. 2003; Nguti et al. 2003). In these situations, gastrointestinal parasites are invariably implicated in the increased mortality risk. For example, an Australian field study that reported the seasonal distribution of deaths of Merino weaner sheep compared different parasite control strategies at Kybybolite in the high-rainfall zone of South Australia (Beveridge et al. 1985). Significant deaths only occurred in the groups that were not treated for gastrointestinal parasites, and then most occurred during the winter months (up to 15%/month). However, significant numbers of untreated weaners did die between December and April in two of five

49 Literature Review years. In these instances, the poor nutrition of weaners during summer was considered to have predisposed them to death from gastrointestinal parasitism, despite the relatively low parasite burdens at this time.

A multi-year study of weaner sheep in the sub-humid tropics of Kenya quantified the seasonal differences in mortality risk (Nguti et al. 2003). During high-rainfall years, weaners died at 1.5–5.0 times the rate of weaners in low-rainfall years. The authors considered gastrointestinal parasitism to underlie this year-to-year variation in mortality, with the higher worm burdens acquired in wetter years outweighing any benefit to survival from improved feed quantity or quality at these times.

2.3.4.6 Disease Many studies have reported associations between weaner mortality and specific diseases, although there are few surveys of causes of weaner mortality in Australia. Amongst Merino weaner flocks in the Victorian Mallee, loss of condition, blowfly strike and diarrhoea were the signs most frequently reported by farmers in flocks experiencing more than 5% mortality (40%, 30% and 20% of flocks, respectively Harris and Nowara 1995). The authors more generally considered that pyrrolizidine alkaloid poisoning from grazing heliotrope (Heliotropium europaeum) was a leading cause of mortality in the Mallee and that gastrointestinal parasitism was less common because of the low rainfall environment and the fact that many sheep grazed crop stubbles for a part of each year. However they did not discuss weaner mortality in more detail.

Oral selenium supplementation reduced mortalities (6% vs. 16%) in the first 52 days after weaning amongst Merino weaners in a field trial in the East Gippsland region of Victoria, where soil deficiencies of selenium, cobalt and copper were suspected (Holmes 1992). Selenium administration appeared to have a direct effect on mortality because it was not associated with increased growth rate or bodyweight. Similar reductions, independent of bodyweight change, were observed on some farms in south-western Western Australia during a selenium supplementation trial (Gabbedy 1971). However in another study, repeated oral selenium supplementation increased weaner bodyweight by 2 kg between 3 and 12 months of age, as well as decreasing mortalities (0% vs. 18% in the untreated groups), despite untreated lambs showing no post-mortem signs of white muscle disease (McDonald 1975).

Two overseas studies have quantified the risk of death associated with weaner faecal egg count (FEC), packed cell volume (PCV) and anthelmintic treatment (Mandonnet et al. 2003 ; Nguti et al. 2003). These studies were conducted in the tropics or sub-tropics and, where specific parasite species were identified, Haemonchus contortus was the most common. An increase of 1000 eggs per gram was associated with a hazard ratio of 1.0–2.1 (weaner sheep) and 4.1 (weaner goats), similar to the hazard ratio associated with a 1 kg decrease in bodyweight. In the study by Nguti et al., a 1% decrease in PCV was associated with a 16% increase in mortality rate. Changes in PCV in this study may have been associated with disorders other than haemonchosis because all causes of death were examined. In the weaner goat study, a one standard deviation-decrease in PCV increased mortality rate by up to 3 times,

50 Literature Review although the absolute PCV values were not reported. Haemonchosis was the primary cause of death in this study, which may explain the larger hazard ratio associated with PCV. Anthelmintic treatment also reduced the risk of death by 77% in the 3 weeks following treatment. As has been discussed previously, increased bodyweight prior to acquiring nematode infection has also been shown to reduce the risk of death from parasitism (Hodge 1990).

2.3.4.7 Maternal Factors Maternal bodyweight, body condition, parity, stocking rate and milk yield, as well as size of the litter produced, have been shown to be associated with the survival of offspring following weaning (Table 2.9). However, no studies have actually investigated the mechanism of action of these associations.

The results of one study suggested that some of these factors may indirectly influence survival through their effect on weaner bodyweight (Hary 2002). Although maternal milk yield was significantly

Table 2.9: Maternal factors associated with weaner mortality

Maternal Species/Breed Measurement Magnitude of effect Reference factor of mortality

Stocking rate/ Merino % mortality 20% at 12 ewe/ha & (Lloyd Davies liveweight 48 kg vs. 1983) 9% at 4/ha & 52 kg

Parity Rambouillet % mortality lower mortality from (Ganai and lower parity ewes Pandey 1996)

Litter size Rambouillet % mortality 77% twins vs. (Ganai and 79% singletons (P < Pandey 1996) 0.05)

Age Dorper & Maasai hazard ratio 0.7–0.4 for >2 y.o. vs. (Nguti et al. sheep 2 y.o. (P < 0.001) 2003)

meat breed sheep hazard ratio 0.63: 4 yo vs. (Southey et al. younger 2001)

Weight at Indian sheep % mortality decreased as maternal (Nivsarkar et al. lambing breeds weight at lambing 1982) increased

West African cumulative mortality lowest for 2nd (Ikwuegbu et al. Dwarf Goat mortality & 3rd parity ewes (P > 1995) 0.05)

Rearing Creole goats hazard ratio 2.5 ± 1.2 artificial vs. (Mandonnet et method maternal al. 2003)

meat breed sheep hazard ratio 5.9 ± 1.2 artificial vs. (Southey et al. maternal 2001)

Milk yield East African Goat hazard ratio ≈ 2 for total dam milk (Hary 2002) yield < 22.5 kg vs. ≥ 22.5 kg

51 Literature Review associated with mortality rate, the relationship between litter size and post-weaning survival became non-significant once the statistical model was adjusted for birthweight and total dam milk production. This suggested that litter size was not a “primary risk factor” for weaner mortality but acted via direct factors such as neonatal bodyweight and growth. An indirect relationship between litter size—or other similar factors—and post-weaning survival, mediated via weaning weight, is appealing. However other mechanisms of action—increased colostrum intake and improved offspring immune function, for example—cannot be excluded. Indeed, total milk yield continued to account for variation in survival along with bodyweight. This suggests that the effect of the former was, at least in part, independent of the latter.

Other studies have also suggested associations between maternal factors and weaner survival but do not propose or elucidate mechanisms of action for them. In a study of the effect of stocking rate and time of lambing on Merino ewe and weaner production in Western Australia, post-weaning mortality was two times greater (20% vs. 9%) amongst ewes and lambs stocked at 12 ewes/ha than those stocked at 4 ewes/ha (Lloyd Davies 1983). Lloyd Davies did not comment on the cause for this difference but, in a subset of this data, average ewe liveweight declined from 52 kg to 48 kg as stocking rate was increased over this range (Lloyd Davies 1962). This, in turn, may have been associated with decreased maternal milk production, to the detriment of offspring weaning weight, since heavier dams in better condition usually produce more milk (Foot et al. 1987).

In a less sophisticated analysis than Hary’s or Lloyd Davies’, a lower proportion of weaner deaths was observed amongst the offspring of ewes that were heavier at lambing (Nivsarkar et al. 1982), although no summary data were presented. An explanation of this relationship was not proposed, although it is reasonable to assume that ewe bodyweight was related to milk production, as in the other studies. Similarly, the slightly lower survival noted amongst Rambouillet weaner sheep in India born as twins, compared to singletons, was presumed to be mediated via birthweight, although no further analysis was performed (Ganai and Pandey 1996).

In Ganai and Pandey’s study, associations between weaner survival, litter size and ewe parity were also “largely explained” by the effect of the latter two factors on birthweight, although these relationships were not explored in detail. Over the three months following weaning, more offspring of seventh parity ewes died than offspring of younger ewes (10% vs. 6%). Conversely, in two other studies, weaners born to primiparous ewes had a greater risk of death than the offspring of multiparous ewes (Southey et al. 2001; Nguti et al. 2003). A similar mortality pattern was present amongst West African Dwarf goats, where there was a tendency for lower mortalities amongst goat kids born to second and third parity does than those born to first, fourth or higher parity (Ikwuegbu et al. 1995). All of these results might be explained by younger, multiparous ewes having higher milk yields and thus weaning heavier offspring, to the benefit of weaner survival. This was demonstrated in Ikwuegbu et al.’s study, in which kids born to first parity does had lower weaning weights and post-weaning growth rates than kids born to second- and third-parity does.

52 Literature Review

Southey et al. (2001) reported that artificially reared weaner sheep in feedlots died at 5.9 times the rate of maternally reared weaners. The death rate amongst artificially reared Creole weaner goats was similarly greater than that of maternally reared kids (Mandonnet et al. 2003). In Mandonnet et al.’s study, the association between rearing method and mortality, along with other factors, became non- significant when time-varying covariates representing FEC, PCV and bodyweight were added to the model. As with the other maternal factors discussed above, such a change in the statistical model suggests that rearing method is not a primary risk factor. Instead, it is likely that its association with post-weaning survival is mediated via another factor or factors. Although likely mechanisms were not proposed by Southey et al. or Mandonnet et al., possibilities could include an effect of rearing method on lamb bodyweight or immunocompetence.

2.3.4.8 Sex Several overseas studies have consistently demonstrated that male weaners face a greater risk of death than females, although no comparable results have been reported in Australia. Southey et al. (Southey et al. 2001) reported the odds of male weaner meat sheep dying in feedlots to be 40% higher than the odds of females dying. This is consistent with the odds (OR) and hazard ratios (HR) reported in Sahelian (Turkson and Sualisu 2005, OR 1.7), and Maasai, Dorper and crossbred (Nguti et al. 2003, HR 1.4) sheep in Africa, as well as West African dwarf goats (Turkson et al. 2004, OR 1.8). Similarly, in a study of survival of Indian sheep breeds, 10% of male weaners died between 3 and 6 months of age, compared to 5% of females. In a study of Creole goat kid deaths due to gastrointestinal parasitism, males died at 3 times the rate of females. No explanation of this widely observed phenomenon has been offered. It has been previously recognised that males are more susceptible to strongyle infection than females, although the reason remains unclear (Barger 1993). However this does not explain the higher mortality rate reported in the other studies, where deaths were not only caused by parasitism. In fact, in the analysis of the subset of Maasai and Dorper weaners that died of gastrointestinal parasitism, there was no difference in mortality rate between the sexes (Nguti et al. 2003).

Since male lambs and kids, but not females, are often castrated, increased deaths amongst males may be associated with post-operative infection or other complications following this procedure. However castration was not routinely performed in all studies discussed here. When it was performed, it was often done at birth, yet sex differences in mortality sometimes persisted to 12 months of age. This is an unlikely duration of effect for a surgical procedure that heals in several weeks. Another potential reason for the difference in mortality risk could be differences in body composition between male and female weaners. Sex hormones are known to affect the deposition of muscle and fat (Mersmann 1987) and the consequent difference may influence risk of death.

2.3.52.3.52.3.5 Weaner MortalityMortality————SummarySummarySummarySummary Few surveys, and no specific mortality studies, of Merino weaners have been conducted in southern Australia but the available evidence suggests that post-weaning mortality commonly exceeds 15%,

53 Literature Review which is no better than that achieved in goat and sheep flocks run by subsistence farmers in Africa. In overseas studies, survival analyses have demonstrated associations between mortality and a variety of factors that have been incidentally suggested or remarked upon in Australian experiments involving Merinos. The outstanding feature of this review is that remarkably similar results have been obtained over a tremendous variation in countries, climatic conditions, ruminant species and production systems.

Bodyweight is the factor most frequently reported to be associated with post-weaning mortality and this association has also been the most widely investigated. Weaners of lower bodyweight invariably have been shown to be at higher risk of death than heavy weaners, regardless of when weight is measured or the actual rate of weight gain, with a 1 kg bodyweight decrease usually associated with a 30–70% increase in risk. In several studies, specifically increasing the bodyweight of lightweight weaners has reduced mortality more than weight increases amongst heavy weaners. Providing supplementary feed to weaners appears to reduce mortality by increasing bodyweight, rather than via some direct effect of the feed itself. Similarly, weaner bodyweight has appeared, at least partly, to mediate the association between mortality and maternal factors such as parity, litter size and condition score.

Seasonal and annual variation in weaner mortality is common, with more favourable survival associated with increased availability of high quality pasture to meet the critical nutritional requirements of weaner sheep. Perversely, such environmental conditions may also favour the development of increased burdens of gastrointestinal nematodes, which can lead to the death of large numbers of weaner sheep. The association between mortality and specific measures of the severity of some parasite infections, such as faecal egg count and packed cell volume, have also been quantified.

The modest body of work concerning post-weaning survival has translated into few quantitative recommendations for the management of Merino weaners in Australia. Allden’s findings that Merino weaners can resume normal growth if they reach 40% of adult weight before experiencing summer weight loss (1968a) and that weaners can substitute a substantial amount of summer pasture for supplementary feed (1969) led him to conclude that “supplementation [is un]attractive except when feeding for survival”. The principal advice given for minimising post-weaning mortality is that lambs should reach at least 45% of adult bodyweight by weaning or drying-off of pastures, whichever is the later, and then grow at 1 kg/month until the onset of autumn rains (Foot et al. 1987). Furthermore, as has been discussed here, for the Merino weaner in southern Australia, the relationships between bodyweight and growth, let alone other potential risk factors, remain unquantified.

2.3.62.3.62.3.6 Methodologies for Analysing Survival & Mortality Survival analysis is the term commonly used to describe the statistical procedures that uniquely analyse time-to-event (also known as ‘survival time’ or ‘failure time’) data. It can be applied to the analysis of the occurrence of any well-defined event, not just death. However, the event of interest is usually non- recurring.

54 Literature Review

Data describing survival or other ‘time-to-event’ outcomes typically consist of two variables for each individual: (1) a continuous variable describing how long the individual was under observation, and (2) a binary variable indicating whether or not the individual experienced the event of interest, such as death (Hosmer and Lemeshow 1999).

Analysis of such data must take into account several unique characteristics. The ‘classical’ approach to analysing survival data involves analysis of variance or linear regression of survival proportions from specific points in the time under study. These methods, however, assume that all error terms are normally distributed and have equal variance. These assumptions are infrequently met by survival time data (Hosmer and Lemeshow 1999; Hary 2002). On the other hand, logistic regression is suitable for analysing a binomially distributed outcome such as death. It provides intuitive information on the probability of an event’s occurrence but ignores the extra information offered by survival data on the timing of events (Allison 1997). No information can be gleaned from logistic regression regarding differences between an individual that dies early in a study’s course and one that dies at the end (Southey et al. 2001). Survival analysis methodology addresses these key characteristics of time-to- event data that limit the use of other statistical techniques.

Survival analysis uses a variety of techniques to examine time-to-event data (Dohoo et al. 2003). They are used to estimate the hazard and survivorship functions, make comparisons between them or estimate the effect of particular factors (or ‘covariates’ or ‘predictors’) on survivorship outcomes. Parametric models assume that the hazard is distributed as a defined function, whereas semi-parametric models make no assumption about the underlying hazard function but both techniques model multivariate effects of categorical and continuous predictors. In contrast, non-parametric techniques graphically describe survival data or make univariate comparisons between levels of a categorical covariate. The most commonly used technique is the semi-parametric Cox proportional hazards model (Cox 1972).

A more recent parametric approach to survival analysis starts with a non-parametric graph of the hazard, which is then described mathematically using functions known as cubic splines (Royston and Parmar 2002). Covariates are then modelled in terms of their effects on this cubic spline function. This approach has the advantage of generating a more accurate mathematical representation of the observed hazard function, rather than approximating it by a simpler curve, such as the commonly used Weibull hazard (Royston and Parmar 2002).

In practical terms, survival analysis must accommodate censored data (Hosmer and Lemeshow 1999). Right-censoring occurs when individuals are present throughout the observation period but do not experience the event of interest before it ends. This may have occurred because the observation period simply ended before these individuals experienced the event, or the subjects were lost to follow-up during the observation period. Left-censored observations are ones where the event has already occurred before a subject comes under study. Finally, interval censoring occurs when an event has taken place within a certain time interval but the precise time of the event is not known (Dohoo et al. 2003). Interval-censored data can take two different forms. When all individuals are observed at

55 Literature Review common time points, the data are termed synchronous interval-censored data. However if the censoring intervals are not the same for all individuals—if there is overlap, for example—then the data are termed asynchronous (Radke 2003). Interval censoring occurs almost universally in data from veterinary and agricultural experiments but is seldom taken into account during analysis (Radke 2003).

Several commonly used statistical software packages are capable of analysing interval-censored data, including SAS/STAT (SAS Institute Inc. 2004), S-plus (Insightful Corp. 2004), Minitab (Minitab Inc. 2005) and Stata (StataCorp 2005). All of these programs are restricted to parametric or non-parametric analytical techniques and cannot perform semi-parametric interval-censored data analysis (Lindsey and Ryan 1998).

Interval-censored data can be approximated by using the beginning, midpoint or end of the censoring interval as an exact failure time and then performing the analysis with standard techniques. Approximating data using any of these approaches can produce variable results (Leung et al. 1997), depending on the extent of interval censoring in a survival dataset, as well as the methodology used. For example, Radke (2003) compared parametric analyses of interval-censored data and approximated data which used the beginning, midpoint or end of the intervals as exact failure times. In this data, many observations shared common censoring intervals and 61% of cases were left-censored. He used the same parametric hazard function a priori in all four analyses but its fit of the data was not tested. The models that ignored interval censoring substantially underestimated the categorical covariate effects. For example, several of the effects estimated from the approximated data were non-significant. Even when they were significant, they were not large enough to be considered important in a practical sense.

Alternatively, it has been suggested that interval censoring in parametric analysis can often be ignored and the midpoint of each failure interval used to approximate exact failure times instead (Lindsey 1998). This author examined “difficult” datasets, although all had relatively few left-censored observations. Although indicating that the results may not always be reliable, he nevertheless stated that conclusions drawn from such analyses are remarkably robust, irrespective of the hazard function chosen for the parametric analysis.

J. C. Lindsey and L. M. Ryan (1998) compared both commercially available and novel interval- censored data analysis methods to Cox proportional hazards analysis of approximated data. All methods estimated similar covariate effects in one reasonably large data set (n=94; 54 interval-censored observations). However in smaller (n=31) and more heavily censored data with wide intervals and many right-censored observations, the Cox analysis of the approximated failure times underestimated the magnitude of the covariate effects by 25–50%. In conclusion, it appears that approximating interval-censored data using the midpoint of the censoring intervals is generally considered to produce satisfactory results, although the magnitude of the covariate effects may be underestimated.

Other techniques do exist for the analysis of interval-censored survival data. These include the piecewise constant (or piecewise exponential) hazard model, as well as non-parametric estimators and

56 Literature Review statistical tests (Lindsey and Ryan 1998). The piecewise constant model provides a result similar to the more common techniques. It analyses the survival outcomes from each censoring interval with logistic regression, which is applied to a dataset that has been transformed by what is known as a ‘link’ function (Hosmer and Lemeshow 1999). This method was used by Hary (2002) and Southey et al. (2003) in analyses of weaner survival. In both these examples, all animals were observed at common time points. Therefore the data were synchronously interval-censored. Although a technique for applying this methodology to asynchronous interval-censored data has been suggested (Carstensen 1996; Hosmer and Lemeshow 1999), it is not implemented in any commercially available statistical software packages. It is also difficult to obtain correct standard errors from this technique when it is applied to a proportional hazards model containing time-varying covariates (Farrington 1996).

Survival analysis is also suited to the analysis of time-to-event data where the covariates to be analysed may take on different values over time, such as whether an animal has received an anthelmintic drench or not in the last month (Allison 1997). Such factors are termed time-varying covariates. However no commercial software exists that can analyse time-varying covariate effects in interval-censored data (Radke 2003).

Survival analysis of interval-censored data containing time-varying covariates therefore requires a compromise. If commercial software is to be used, then one of these aspects of the data must be approximated. Whether this occurs with the time-varying effect or the interval censoring will depend on which is considered the least important to the analysis.

Several studies illustrate the advantages of survival analysis and the difference between the results produced by different methodologies when analysing weaner mortality. For example, Mandonnet et al. (2003) analysed survival of goat kids with the Cox proportional hazard and Weibull parametric models, although only the Cox model results were reported. They also used time-varying covariates to identify several factors that had important associations with survival throughout the entire post-weaning period. They did not discuss the manner in which time-to-death data were collected or interval censoring issues. Southey et al. (2001) also used Weibull and Cox models to analyse survival of meat lambs raised in a feedlot, and compared these results to those from logistic regression. The hazard ratios from the two survival analysis models and odds ratios from the logistic regression were all very similar, although the standard errors of the odds ratios were larger and the authors noted that a methodology that properly accounts for censoring, such as survival analysis, is preferable. Nguti et al. (2003) carried out a Cox survival analysis of post-weaning mortality of lambs in Kenya. Again, the method for recording deaths was not stated. It appears that weaners were yarded approximately every four to six weeks, which would have created interval-censored data if this were the time when deaths were noted. Hary (2002) estimated odds ratios for covariate effects by performing logistic regression on biweekly survival records for goat kids. Hazard curves for different covariate levels were estimated using cubic spline functions and hazard ratios were calculated from these curves. This approach appears to favour comparisons between a small numbers of categories only. It would not be practical for analysing the effect of a continuous covariate unless it could be subdivided in this way. Whilst detailed comparisons

57 Literature Review were made between different time points on various hazard rate curves, the methodology did not generate hazard or odds ratios that summarised covariate effects well across larger time periods, or indeed the entire study.

2.42.42.4 ConclusionConclusionConclusion Previous research has not clearly identified the most appropriate time for shearing a self-replacing Merino flock in southern Australia. Even when wool quality or animal health is considered alone, no single annual shearing time is clearly preferable because, for example, the effect of autumn shearing on fibre diameter and staple strength has opposite consequences for wool price. Even if the net effect on wool quality favoured autumn shearing, for instance, such an advantage would have to be weighed against the increased susceptibility of autumn-shorn sheep to flystrike in spring and post-shearing mortality, and the costs of managing such conditions might negate any increase in wool income. Similarly, a single shearing time may not always provide benefits for individual factors. For example, any given shearing date cannot reduce both spring and autumn flystrike risk. Therefore, a study is required that measures the effect of different shearing times on wool production and animal health parameters concurrently.

C H A P T E R 3

EXPERIMENT INTRODUCTION AND MATERIALS & METHODS

3.13.13.1 Introduction & Trial Overview Time of shearing affects many aspects of wool production and sheep health but no study has examined these factors concurrently in a spring-lambing, self-replacing Merino flock in southern Australia. A long-term trial was conducted to compare wool production, sheep bodyweight, fleece rot, dag, flystrike risk and mortality in spring-lambing Merino ewes and their progeny shorn at different times and managed under commercial conditions in south-eastern Australia. Adult ewes were shorn in December (DEC), March (MAR) or May (MAY), and within each adult shearing time, two patterns of shearing the progeny were examined. The shearing times were chosen in consultation with farmers to ensure that they were practical alternatives for a spring-lambing flock, and the trial was conducted for five years in order to measure the long-term effects of shearing time on production under a range of seasonal and environmental conditions.

3.23.23.2 EEExperimentalExperimental Site The trial was conducted from January 1999 to May 2004 on two farms in Western Victoria, a major wool growing area of Australia. The trial started on Farm A, 15 km west of Geelong and, following the sale of this farm, all sheep in the trial were moved in February 2003 to Farm B, 20 km north-west of Farm A. The region has an annual average rainfall of about 535 mm.

The soils on Farm A were a heavy, basalt type, to which large amounts of fertiliser had been applied for the previous 15 years (Appendix 1). A single paddock on Farm A was allocated for grazing by the ewe flock for the duration of the trial. This paddock contained a mixture of annual grasses, including barley grass (Hordeum spp.), silver grass (Vulpia spp.) and wallaby grass (Austrodanthonia spp.), as well as smaller amounts of volunteer Phalaris aquatica and subterranean clover (Trifolium subterraneum). The paddocks grazed by the weaner and hogget sheep on Farm A contained well- established pastures of subterranean clover (cv. Leura and cv. Trikkala) and phalaris (cv. Sirosa and cv. Holdfast), which were sown in autumn 1995.

Farm B had lighter, sandy soils that had not received fertiliser for many years. The pastures were dominated by annual grasses and contained mainly onion grass (Oxalis pes-caprae), silver grass and

59 Materials & Methods barley grass. They also contained some volunteer cocksfoot (Dactylis glomerata) and phalaris. The soil concentrations of phosphorus from the sites on Farm A and B in October 2003 were 18 and 8 mg Olsen P /kg, respectively.

The winter stocking rate in the trial paddocks on Farms A and B was approximately 10 dry sheep equivalents per hectare (DSE/ha). This was more than the district average but lower than that achieved elsewhere on Farm A, and was consistent with a sustainable rate for a ‘high input’ farm under best- practice management in that environment (Vizard 1997).

Monthly rainfall was measured on Farm A manually with a rain gauge and maximum and minimum monthly temperatures were obtained from the nearest Bureau of Meteorology recording site, 15 km away.

3.33.33.3 AnimalsAnimalsAnimals Six hundred fine-wool Merino ewes of the one genotype were used to establish the trial flock. Approximately one third were born in September 1995 and the remainder were born in September 1996. With this sample size, the power of detecting a 5% difference in measured characteristics between ewe shearing groups was greater than 0.9 and, if the flock’s weaning percentage were assumed to be at least 75%, the number of progeny in the six weaner shearing groups was such that the power of detecting a 10% difference between measured characteristics in weaner shearing groups was at least 0.8.

3.43.43.4 Flock Management The sheep were managed as a spring-lambing, self-replacing fine-wool flock, with the first progeny born in the trial in September 1999. The routine management calendar of the flocks is shown in Table 3.1. Sheep were managed as three different age groups: mixed-sex weaners (weaned sheep younger than 12 months of age), mixed-sex hoggets (sheep born one year prior to the current weaner flock, up to 19 months of age) and adult ewes (19 months of age or older). Within each management group, all sheep grazed together in the same paddock and were managed identically, apart from their shearing time.

Ewes were naturally mated as a group in a single paddock for five weeks to fine-wool Merino rams. Mating started on about 5th April each year and lambing therefore started on or about 1st September. Lambs were marked as a single group three weeks after the end of lambing and at this time were identified to their dams’ shearing treatment using the marked udder method (Davis et al. 1981): coloured branding fluid corresponding to each ewe shearing treatment, mixed with vegetable oil, was applied to each ewe’s udder one day prior to lamb-marking and the udder and breech area of ewes was examined to determine lactational status (Dun 1963). Ewes were classified as either lactating (‘wet’) or

60 Materials & Methods non-lactating (‘dry’). Lambs were identified to ewe treatment according to the colour of branding fluid on their heads at marking on the following day. Each lamb had a copy of a uniquely numbered tag placed in both ears, as well as another tag identifying to which of the six progeny shearing treatments it belonged (see below). Routine marking procedures were performed by experienced farm staff or a contractor. Lambs were vaccinated against Clostridium tetani, C. perfringens type D, C. novyi, C. chauvoei and C. septicum, mulesed with a ‘radical-V’ operation (Morley 1983b) and cyromazine 0.1% W/V (Vetrazin Liquid® 500 g/L, Novartis Animal Health Australasia Pty Ltd) was applied over the breech to prevent flystrike of mulesing wounds. Male lambs were castrated by placing a rubber ring around the neck of the scrotum. Lambs and ewes were returned to their paddock immediately after marking.

Lambs were weaned, received a booster vaccination and were drenched with an effective anthelmintic at the start of December, 12–13 weeks after the start of lambing.

At 19 months of age, replacement ewe hoggets were selected to enter the adult ewe flock and wether hoggets exited the trial. Selection of replacement ewes was based on a 12% micron premium index that incorporated fibre diameter, clean fleece weight, staple strength and bodyweight in its selection criteria (Ponzoni 1986). Within each year, the same number of ewe hoggets was selected from each of the six progeny shearing treatments. The number available for selection was determined by the weaner shearing treatment with the least number of ewe hoggets in it. About 24–29 sheep from each weaner shearing group entered the ewe flock each year. At the same time, the oldest ewes, ranked lowest on index, were culled from each ewe shearing treatment in order to leave, after the addition of the replacement ewe hoggets, approximately 200 ewes in each adult shearing treatment.

In order to compare the flystrike risk of the shearing groups, no sheep received prophylactic chemical treatment for flystrike. Flocks were monitored at least twice weekly throughout times of flystrike risk and individual strikes were treated by clipping wool away from the affected and surrounding area, and applying topical insecticide. Struck sheep were not removed from the trial, although withholding periods for treated fleeces were observed. A flock was jetted with cyromazine if flystrike incidence exceeded 2% strikes/week. Neither crutching nor the removal of wool from around the preputial opening of wethers (‘ringing’) was routinely performed. Some shearing treatments were crutched to reduce an anticipated fly risk on three occasions: March- and May-shorn ewes were crutched in December 2000, December- and March-shorn in July 2001, and all three ewe treatments in July 2003. The weight of crutchings removed from each sheep was measured on each of these occasions. The 18- month old wether progeny of May-shorn ewes were ringed in March 2002.

A strain of intermediate-virulence footrot (Dichelobacter nodosus) was present in some sheep on Farm A, although no clinical signs of footrot were observed in any of the sheep involved in the trial. However, at the request of the owners of Farm B, the sheep were inspected in the yards by Farm A’s manager in February 2002 before the trial flocks were moved and up to four sheep were culled from each ewe and progeny shearing treatment, based on the appearance of their feet. No footrot lesions were detected in the trial flocks at two subsequent inspections on Farm B.

61 Materials & Methods

Table 3.1: Calendar of flock management procedures and shearing times

Month Ewe flock Lamb/weaner flock Hogget flock

Sept Lambing start

Oct Record BWT *, CS, DS, FR Lamb marking & allocation Record BWT, CS, DS, FR to weaner shearing treat- † Identify lactating & dry ewes ments (see text) Shearing: OCT-DEC

Nov ------Monitor flocks for flystrike, and record & treat individual cases (ongoing—see text) ------

Dec ------Record BWT, CS, DS, FR ------1st summer drench ------

Weaning

Shearing: DEC† Shearing: DEC-DEC Shearing: DEC-DEC

Supplementary feeding as required (ongoing throughout summer & autumn)

Jan

Feb ------2nd summer drench ------

March ------Record BWT, CS, DS, FR ------

Shearing: MAR Shearing: MAR-MAR Shearing: MAR-MAR JUN-MAR

Cull ewes exit trial after Replacement ewe hoggets shearing enter adult ewe flock

Ewe hogget culls & all wether hoggets exit trial after shearing

April ------Monitor flock worm egg counts monthly and drench if necessary ------

Mating starts (~April 5)

May Mating ends after 5 weeks ------Record BWT, CS, DS, FR ------

Shearing: MAY Shearing: MAY-MAY Shearing: MAY-MAY (culls exit trial off shears) JUL-MAY (culls exit trial off shears)

June Supplementary feeding as Record BWT, DS, FR required (ongoing, determined by farm management) Shearing: JUN-MAR

July ------Record BWT, CS, DS, FR------

Shearing: JUL-MAY

[Weaner flock now considered to be Hogget flock]

Aug

* BWT: bodyweight; CS: condition score; DS: dag score; FR: fleece rot score † see Section 3.5 for description of shearing group abbreviations

62 Materials & Methods

3.53.53.5 ShearingShearingShearing The shearing times in the trial were recommended by a workshop held with approximately 100 farmers and researchers participating in a farm extension program as (Larsen et al. 2002). The recommended shearing times were all considered to be realistic alternatives for commercial farms in the district, each offering potential benefits and disadvantages that warranted full examination.

In January 1998, the ewes in the trial flock were randomly allocated, after stratification for age, to one of three annual shearing treatments: December (DEC), March (DEC) and May (MAY). Ewes allocated to December shearing were immediately shorn. Observations were commenced in December 1999, when the first 12-month DEC shearing took place.

Progeny from each ewe shearing time were randomly allocated to one of two weaner shearing patterns that aligned them with the adult shearing time over their first two shearings (Figure 3.1). Thus, there were three ewe shearing and six progeny shearing treatments in total. Progeny of DEC ewes were either first shorn at weaning (3 months old, in December) then 15 months of age (December; abbreviated as ‘DEC-DEC’), or first shorn at 13 months of age (October) then at 27 months of age (December) in the DEC ewe flock itself (OCT-DEC). Progeny of MAR ewes were either first shorn at 6 months of age (March) then 18 months of age (March; MAR-MAR), or first shorn at 9 months of age (June) then 18 months of age (March; JUN-MAR). Progeny of MAY ewes were either first shorn at 8 months of age (May) then 20 months of age (May; MAY-MAY), or first shorn at 10 months of age (July) then 20 months of age (May; JUL-MAY).

Sheep were shorn by professional shearers in about the middle of the respective month. As each sheep was shorn, the belly wool and any dags were separated from the main fleece. The unskirted fleece,

Figure 3.1: Timing of ewe and progeny shearing treatments (age in months at weaner shearings) Ewe shearing treatment DEC MAR MAY

  

Weaner December March March May May

shearing 2 (15 mths) (18 mths) (18 mths) (20 mths) (20 mths)

Weaner December October March June May July

shearing 1 (3 mths) (13 mths) (6 mths) (9 mths) (8 mths) (10 mths)



 All lambs born in September Progeny shearing DEC-DEC OCT-DEC MAR-MAR JUN-MAR MAY-MAY JUL-MAY treatment

63 Materials & Methods without belly wool, and dags were weighed separately. Approximately 100 g of wool was removed from the mid-side of the fleece of each sheep for measurement of fibre diameter, staple length, staple strength, position of break and Schlumberger dry yield by a commercial wool-testing laboratory (Australian Wool Testing, York, Western Australia). Clean fleece weight was calculated as the product of greasy fleece weight and yield. All mid-side samples were tested individually, except for those from shearing 1 of the DEC-DEC progeny treatment, when smaller mid-side samples from each sheep were aggregated and measured as a bulked sample. This was requested by the participating farms to avoid losing an excessive proportion of the lambs’ fleece wool to the testing sample, for which the farm would receive no income.

3.5.13.5.13.5.1 Periodic Wool Growth Dyebands were applied to wool staples on some sheep to determine the amount of wool grown at different times of the year. Approximately 20 ewes from each shearing treatment were randomly selected in May 2000 and had dyebands (Chapman and Wheeler 1963) applied to a mid-side wool patch in May 2000, October 2000, December 2000, March 2001, May 2001, July 2001, December 2001, March 2002, May 2002, October 2002, December 2002 and March 2003. Additional ewes, age- matched to preserve a similar age structure in each dyeband group, were randomly selected from the adult ewe flock to replace dyebanded ewes culled in March each year. Based on previously reported daily wool growth rates (Langlands 1968), with this sample size the power of detecting a 1 g/d, or approximately 10%, difference in wool growth rate between shearing groups was greater than 0.8.

Dyebanded wool staples were removed prior to shearing. They were scoured according to the method of Williams and Chapman (1966), although petroleum ether (Petroleum Spirit BR 40–60C, Nuplex Industries (Aust) Pty Ltd) was used instead of Shell X4 solvent. The staple was cut at the base of each dyeband with fine scissors and each segment was weighed. The segment’s weight as a proportion of the total staple weight was calculated, and this was multiplied by the sheep’s clean fleece weight to calculate the total weight of fleece grown by each sheep in the period delineated by adjacent dyebands.

3.63.63.6 Measurement Measurements began with the first shearing in December 1999. The flocks were yarded before each shearing, as well as at other regular times throughout the year, and every sheep was weighed, and assessed for condition score (except weaners; Bell 1994), extent of dags (Larsen et al. 1994) and fleece rot (from March 2001 onwards, Murray and Mortimer 2003). The first weaner measurements were performed within a week of weaning each year, when sex was also recorded.

At any particular weighing, sheep in each weaner shearing group were carrying varying amounts of greasy fleece, making direct comparison of weights improper. The fleece-free bodyweight of weaners was estimated using the following methods: if weaners were weighed at the same time as shearing, the greasy fleece weight was subtracted from the observed bodyweight to determine fleece-free weight.

64 Materials & Methods

For weaners that were not shorn at weighing and had not been shorn previously, the mean greasy fleece weight from weaners shorn for the first time was used to estimate fleece-free weight. If weaners were not shorn at weighing but had been shorn previously, the following formula was used: fleece-free weight = observed weight – ( mean GFWthis shearing – individual GFWprevious shearing ). Finally, when weaners were weighed at times when no sheep were shorn, the following formula was used to calculate fleece-free bodyweight: fleece-free weight = woolly weight – ( mean GFWmost recent shearing + proportion of current wool growth period – individual GFWprevious shearing ). For example, the fleece-free x bodyweight of unshorn sheep in April = woolly weightApril − [ mean GFWMarch-shorn weaners + /y × (mean

GFWMay-shorn weaners − mean GFWMarch-shorn weaners ) ], in which x = number of days since last shearing (i.e. March) and y = number of days between last shearing (i.e. March) and next shearing (i.e. May)

In addition to routine yardings, sheep were regularly inspected in their paddocks as part of routine farm management and the ear tag numbers of dead sheep were recorded. Sheep were also considered to have died if they were missing at two consecutive musters and were not found in another flock on the farm, even if no identifiable body was found.

3.73.73.7 AnalysisAnalysisAnalysis The various statistical techniques used for data analysis are listed in Table 3.2, with details given in Sections 3.7.1–3.7.7, below. All statistical analyses were performed in the software package Stata (StataCorp 2005). Fixed effect terms representing the year of the trial and sheep age, and their interactions with shearing treatment, were included in all regression models used to analyse the ewe data. A ‘trial year’ was defined so that within-year comparisons of wool production were made over the same gestation. For example, the first round of wool production was that collected at shearings in December 1999, March 2000 and May 2000, and is described as ‘year 1999-2000’ throughout the thesis. Year of birth and its interaction with shearing treatment was included as a fixed effect in the weaner data regressions. The statistical significance of interaction terms was tested with an overall Wald test, using the testparm command, and interactions were removed from the model if not statistically significant (P > 0.05).

Generalised estimating equations for panel data (‘panel data regression’: Hanley 2003; Hanley et al. 2003) were used to analyse results from the ewe flock because most sheep remained in the flock for more than one year, which meant that observations were not independent between years and could not be tested using analysis of variance. Briefly, the outcome of interest (for example, fibre diameter) was modelled as a linear combination of dummy variables representing the ewe shearing treatments, fixed effects and interactions, assuming no specific structure to the inter-year correlation matrix. The variable identifying individual ewes was specified as the one in which observations were repeated, and year was specified as the time variable over which the repeated observations occurred. Once the significant fixed effects were determined, the statistical significance of the differences between

Table 3.2: Statistical methods used to analyse differences between shearing groups in ewe and weaner results, and corresponding Stata® (StataCorp 2005) commands

Outcome Ewes Weaners Statistical method Stata command Statistical method Stata command Wool parameters (staple panel data regression xtgee analysis of variance anova strength, % of tip-, mid- & base-breaks, fleece weight, fibre diameter, yield, staple length)

Condition score panel data regression xtgee n/a

Weaning weight n/a analysis of variance anova Methods & Materials

Growth rate n/a analysis of variance anova

66 Ewe lactational status at lamb panel data regression, specifying binomial logit n/a marking distribution

2 Weaning percentage Pearson χ test tabi n/a χ2 goodness-of-fit test csgof

Fleece rot: severe vs. not odds ratios combined across years with Mantel- mhodds odds ratios combined across years with mhodds severe Haenszel weights Mantel-Haenszel weights

Dags: severe vs. not severe panel data regression, specifying binomial xtgee logistic regression mhodds distribution

Flystrike odds ratios combined across years with Mantel- mhodds odds ratios combined across years with mhodds Haenszel weights Mantel-Haenszel weights

Mortality mortality rate (after shearing) ratios combined across ir mortality rate (after shearing) ratios combined ir years with Mantel-Haenszel weights across years with Mantel-Haenszel weights survival analysis using Cox, Weibull & cubic stcox, streg, stph, spline models intcens

Materials & Methods

shearing groups within years and between groups’ least squares means was determined with an overall Wald test, using the test command.

For measurements with a binary outcome, odds ratios between shearing groups were estimated using panel data logistic regression (Hosmer and Lemeshow 2000) or the Mantel-Haenszel method (Rothman 1986b), as detailed in Table 3.2. In the latter case, a χ2 test was used to verify that odds were homogeneous across years before they were combined (Dohoo et al. 2003).

None of the ewe or weaner data was orthogonal, so ‘overall’ shearing treatment means across years were estimated as least squares means (LSM) from the regression model used in a particular analysis. In the ewe data, the LSM were adjusted for equal numbers of ewes aged between 2 and 6 years, inclusive, because the flock structure differed between years of the trial.

3.7.13.7.13.7.1 Wool Production Each ewe shearing was not exactly 365 days apart and the duration of wool growth of each weaner shearing treatment differed, so fleece weight and staple length were analysed on per day (g/d or mm/d) basis. For the weaner shearing groups, the duration of wool growth prior to shearing 1 was calculated assuming that all weaners were born on the lambing start date. Maiden ewes from the OCT-DEC weaner group were excluded from the ewe flock’s fleece weight and staple length analyses because, unlike all other maiden ewes, their duration of wool growth was greater than 12 months.

Wool growth rates determined by dyeband measurements for the same period in different years represented repeated measurements and so were analysed with panel data regression, like the other ewe measurements. If ewes were pregnant or lactating during the dyeband period being analysed, reproductive status, based on whether they were lactating or not lactating at lamb-marking, was included as a fixed effect in the regression model.

3.7.23.7.23.7.2 Body Condition & Weight Ewe condition score was analysed as a continuous variable. The growth rate of weaners with different shearing times was compared for the periods weaning–May (i.e. 3–8 months old) and May–December (8–15 months old), using estimates of fleece-free bodyweight. These periods were chosen because they were common to all weaner cohorts.

3.7.33.7.33.7.3 DagDagDag The prevalence of severe dag (dag score ≥3) in ewes and weaners was analysed using panel data regression, specifying binomially distributed data, and logistic regression, respectively. The odds of ewes having severe dag was analysed for different times between October and March, which was when the most comprehensive observations nearest the onset of the breech strike risk period were made. For

67 Materials & Methods the weaner shearing groups, only results from sheep born in 2001 and 2002 were used because regular observations were not recorded in the earlier cohorts.

The cost of dags in the ewe shearing groups was calculated using the method and values described by Larsen et al. (1995). Briefly, dag cost had two components: the difference between the values of the wool contained in crutchings and the same weight of fleece wool, and the labour cost of crutching sheep. The wool value difference was calculated from the ratio of the price of crutchings to the price of fleece wool (0.52 if dag score was ≤1 and 0.26 if dag score was ≥2) and the clean wool yield from crutchings (36.4%, 28.5%, 22.9%, 19.8%, 13.6% and 14.4% for dag scores 0–5, sequentially). The cost of labour to crutch ewes was estimated as 40, 60, 80, 100, 120 and 140 cents/head (J. Webb Ware pers. comm.) for dag scores 0–5, sequentially. The cost of dags was calculated for all sheep in a shearing group and then expressed as an average cost per head.

3.7.43.7.43.7.4 Fleece RotFleece Rot The prevalence of severe fleece rot (fleece rot score ≥3) in different months was analysed using the Mantel-Haenszel method because the low prevalence of severe fleece rot in some shearing treatments in some years precluded the use of the χ2 test or logistic regression.

3.7.53.7.53.7.5 FFFlystrikeFlystrikelystrikelystrike Flystrike prevalence at different times of the year was estimated from regular observations of the flocks in their paddocks and when they were yarded. Prevalence was analysed as a binary outcome (flystrike present or absent), regardless of severity or location of strike on the sheep, using the Mantel-Haenszel method because of the low numbers of affected sheep present in some years.

3.7.63.7.63.7.6 Ewe Reproduction Ewe lactational status (‘wet’ or ‘dry’) at lamb marking was used as a measure of whether or not a ewe had successfully reared a lamb to that point in time and was analysed with panel data regression, specifying a binomial data distribution.

The number of ewes in each shearing group each year was very similar, so differences between the numbers of lambs weaned from each shearing group per year were examined using the Pearson χ2 test. A χ2 goodness-of-fit test was performed within the data from each year to test whether the proportion of lambs weaned in each shearing group varied significantly from the expected value of one third of the total number of lambs weaned in that year.

3.7.73.7.73.7.7 MortalityMortalityMortality Sheep survival was determined from routine flock inspections, weighing and shearing records. Time of death was not precisely known, which meant that the mortality data were interval-censored, with the beginning of the interval defined by the last time a sheep was observed, and the end defined by the first

68 Materials & Methods time it was noted missing (Radke 2003). In the univariate analyses, and the multivariate analyses involving time-varying covariates (see below), the interval-censored failure times were approximated by exact failure times equal to the midpoints of the censoring intervals (Rothman 1986b). Minitab (Minitab Inc. 2005) was used to generate Turnbull non-parametric survivorship graphs (Turnbull 1976), which account for interval censoring. In the multivariate analyses, sheep prematurely removed from the trial or still alive at the end of an observation period were classified as right-censored observations (Hosmer and Lemeshow 1999).

In the univariate analyses, cumulative mortality rate (CMR) was calculated as the number of deaths divided by the total time at risk for all sheep during the time of analysis (Rothman 1986b). The association between shearing and mortality was analysed in the ewe flock using records from July 2001 to May 2004, because this was when the ewe flock was observed more frequently (every 1–3 months) and more detailed survival records were available. The ewe data were analysed by comparing the cumulative mortality rate of the three ewe treatments in the time immediately after one group’s shearing, using the Mantel-Haenszel method for incidence rate ratios (Rothman 1986a) after checking for the homogeneity of rates, as described above. Mortality rate ratios were combined across years assuming that observations were independent from year to year. In weaners, the associations between mortality and shearing time, bodyweight, growth rate and sex were analysed in a similar way, using Mantel-Haenszel incidence rate ratios, and log-rank and Wilcoxon tests (Hosmer and Lemeshow 1999). Mortality rate differences between weaner shearing groups were also calculated and combined across years using weights equal to the inverse of the year-specific incidence rate variance (Rothman 1986b). Multivariate survival analyses were also performed for weaners, as described below.

Mortality risk attributed (attributable risk, AR) to membership of a particular sub-group (e.g. lightest weaning weight quintile or male sex), attributable fraction (AF), population attributable risk (PAR) and population attributable fraction (PAF) were calculated as follows (Rothman 1986a):

AR = CMRsub-group – CMRnon sub-group

AF = AR ÷ CMRsub-group

PAR = CMRoverall – CMRnon sub-group

PAF = PAR ÷ CMRoverall

3.7.7.1 Multivariate Weaner Survival Analysis To increase power of the survival analysis, weaner survival data from another large field trial conducted on Farm A were added to the survival data of the progeny from the shearing trial. The additional survival data were from weaners in a central progeny test (Casey et al. 1995) conducted on Farm A between 1996 and 2001. About 1600 spring-born weaners were used in this trial, with detailed survival records available from weaning to about 15 months of age. Weaners in the central progeny test were managed similarly to the weaners in the shearing trial, with the exception that they were all jetted with cyromazine at or after weaning each year and were shorn in December, at 3 months of age,

69 Materials & Methods for the 1998–2000-born cohorts or February, at 5 months of age, for the weaners born in 1996 and 1997.

Multivariate survival analysis of weaners from weaning to 12–15 months of age, or during the ‘post- weaning period’ (PWP), was performed using Cox (1972), Weibull (Hosmer and Lemeshow 1999), log-logistic (Hosmer and Lemeshow 1999) and cubic spline-based (Royston and Parmar 2002) models. The analysis was modelled assuming proportional hazards and used two internal ‘knots’ for the cubic spline polynomial function representing the baseline hazard (Royston and Parmar 2002). A Weibull survival analysis of the time-constant covariates using interval-censored data was also performed.

The models included a specification that individuals were clustered at the cohort level, where possible, and were developed using purposeful selection of covariates (Hosmer and Lemeshow 1999) as detailed below. Covariates were included in a model if they were significant at the 0.25 level in univariate log- rank or Wilcoxon tests. The scale of continuous covariates was checked for linearity in the log hazard and, if this was not the case, appropriate single-term transformations were chosen using the method of fractional polynomials (Hosmer and Lemeshow 1999 p.161–3). Preference was given to the more 1 easily interpretable loge(x) or /x transformation if this model’s deviance was only slightly greater than the model containing the optimum transformation and the assumption of linearity was satisfied by the inspection of quartile design variables (Hosmer and Lemeshow 1999 p.164). Once the main terms in the model were determined, first-order interactions were added individually and retained if they were biologically plausible and their partial likelihood test was significant at the P = 0.05 level. Finally, other potential main effects were re-checked for significance in the model and all non-significant terms were removed. The proportional hazards assumption of the Cox model was checked by examining graphs of the scaled Schoenfeld residuals for each covariate and performing partial likelihood ratio tests to check that interaction terms between each covariate and loge t were not significant, using the command stphtest (Hosmer and Lemeshow 1999 p.205–210).

For each cohort, the mean growth rates of weaners during day 0–150 (GRearly), day 150–225 (GRmid) and day 225–~332 (GRlate) after weaning were calculated and assigned to all sheep in that cohort. Where included in a model, these mean seasonal growth rates, along with fleece-free bodyweight (measured at various times during the PWP) and shearing group, were expressed as time-varying covariates. Time-varying covariates were accommodated by structuring the data according to the method of Hosmer and Lemeshow (1999), whereby an individual’s survival record was split into several ‘sub-observations’ representing the different times during its survivorship experience in which the time-varying covariate took on different values. The covariate took its observed value in the sub- observations in which it was defined and the value of zero in all other sub-observations. For example, the variable GRmid represented the mean growth rate of weaners in each cohort during time after weaning, t, 150 ≤t<225. Obviously, in a model, mortality risk for t<150 could not be associated with a growth rate measured at t≥150. Thus, each weaner’s survival record was split into two sub- observations representing 0≤t<150 and 150≥t and GRmid was set equal to zero in the sub-observation for 0≤t<150 (i.e. the time before GRmid was defined) and, for t≥150, GRmid took its measured value.

70 Materials & Methods

Weaners in the central progeny test were only shorn at one time of the year, so fleece-free bodyweights were estimated based on the mean wool growth rates of weaners in the time of shearing trial. The sex of some weaners born in 1999 and 2000 was not recorded until the second time they were weighed, meaning that the sex of weaners dying before day 168 and day 143, respectively, post-weaning was not known. These cohorts were not included in the univariate analysis of sex and mortality. To enable these animals’ observations to be included in the multivariate analyses, sex was categorised as ‘female’, ‘male’ or ‘unknown’.

3.83.83.8 Calculation of Fleece Values A series of fleece values for the wool produced at each shearing was calculated so that production from the shearing groups could be compared under a range of historical market conditions. Based on the mean fibre diameter, staple length and staple strength from each shearing, arrays were created containing wool prices at historically ‘low’, median and ‘high’ micron premiums, to which low, median and high premiums and discounts for staple length and staple strength were applied. The low, median and high values used were the first, median and ninth decile prices for the period July 1992– February 2006, obtained from the software MLS Analyser (Independent Commodity Services 2005). Fleece values were then calculated by multiplying these wool prices by the mean clean fleece weight per head. The mean proportion of mid-breaks in a clip was not included in the wool price calculation because no long-term price schedules for this characteristic have been published. The cost of shearing was assumed to be $4/head.

3.93.93.9 ‘Lifetime’ Wool Production In order to compare shearing groups over the same length of time, wool production from each group over a sheep’s recommended commercial ‘lifetime’ of birth to six years of age (Morley 1994) was estimated using two different assumptions. Firstly, lifetime performance was estimated for ewes that had survived to six years of age. Secondly, lifetime performance was estimated, taking into account the probability of survival throughout the entire period. The endpoint for the comparison period was the seventh December after birth, when ewes were 75 months of age. At this time, DEC ewes would have been just shorn but MAR and MAY ewes would have been carrying some wool. The value of each shearing pattern’s lifetime wool production, assuming a ewe survived to culling at six years of age, was calculated by summing the mean fleece values of weaner shearing 1, weaner shearing 2 and maiden shearing, plus the mature ewe fleece value based on the pattern’s LSM fibre diameter, staple length and strength for ewes 3–6 years old.

The value of lifetime production incorporating the different survivorship of each shearing pattern was calculated by multiplying each fleece value from the lifetime calculation, above, by the shearing group’s ‘cumulative survival’ from birth to that shearing and then summing over all shearings.

71 Materials & Methods

Cumulative survival of a shearing pattern to a particular age was estimated using:

• mean survival of unshorn weaners from weaning to shearing 1, taken from the Kaplan-Meier survivorship estimate for unshorn weaners in all cohorts

• mean survival between shearing 1 & shearing 2 of the weaner shearing group

• mean survival between shearing 2 and the first adult shearing of ewes selected into the adult flock from the weaner shearing group

• mean survival between shearings of the ewe shearing group

MAR and MAY ewes were only present for nine and six months, respectively, in the sixth year of the ‘lifetime’ calculations and the mean survival fractions were proportionally increased for this time.

The clean fleece weight produced over the lifetime of a sheep from each weaner-ewe shearing pattern was also estimated, using the sum of the least squares mean (LSM) clean fleece weights from the first and second weaner, maiden (2 years old) and mature (3–6 years old) ewe shearings. Based on dyeband measurements, MAR and MAY ewes produced 85% and 73%, respectively, of their annual wool growth between shearing and December, and these fractions were used to calculate the clean weight of unshorn wool on sheep in the last year of these two groups.

C H A P T E R 4

RESULTS—WEATHER OBSERVATIONS & WOOL PRODUCTION

4.14.14.1 Weather Observations

Rainfall and temperature during the trial are summarised in Figure 4.1. Annual rainfall was lower than the 100-year average for four out of the five years of the trial. In 1999, there was a dry spring and below-average rainfall continued throughout 2000. Rainfall was greater than average in autumn 2001 and average rain was received throughout the rest of that year. Autumn 2002 was dry, average rain fell during winter and spring, and then very dry conditions occurred until the end of the trial, in association with drought across much of south-eastern Australia, with the exception of greater-than-average rainfall between July and October 2003.

Figure 4.1: Monthly cumulative rainfall and average daily maximum and minimum temperatures at Farm A during the Time of Shearing Trial

160 32

140 28

120 24 Results—Weather Observations Observations Results—Weather

100 20

80 16 74

Rainfall (mm) 60 12 Temperature Temperature (°C)

40 8

20 4

0 0 Oct- Jan- Apr- Jul- Oct- Jan- Apr- Jul- Oct- Jan- Apr- Jul- Oct- Jan- Apr- Jul- Oct- Jan- Apr- Jul- Oct- Jan- 98 99 99 99 99 00 00 00 00 01 01 01 01 02 02 02 02 03 03 03 03 04 Date

Monthly rainfall Average max. temp. Average min. temp.

Results—Wool Production

4.24.24.2 Overview of Sheep Numbers & Wool Production ResultsResultsResults Ewes remained in the trial for an average of 2.8 (range 1–5) years. The average age of ewes increased as the trial progressed, from 3.3 years in 1999-2000 to 5.3 years in 2003-2004 (Table 4.1). Although the age structure of the ewe flock changed as old ewes were culled and replacement ewe hoggets were selected into the flock, the age structure of each shearing group was similar. Ewe shearing dates and the times between each shearing are shown in Table 4.2. The duration of wool growth before each weaner shearing are shown in Table 4.3.

There were statistically significant interactions between shearing group and year for all the measured ewe and weaner wool characteristics, indicating that differences between groups depended on the year of observation. For some characteristics, the shearing group rankings were consistent across years and only the magnitude of the differences varied. However, for other characteristics, the shearing groups did not follow a consistent pattern from year to year.

Only a single, bulked mid-side sample was collected from DEC-DEC weaners at shearing 1, so confidence intervals could not be estimated around the results for this group.

The progeny of DEC ewes born in 1999 were shorn incorrectly. The DEC-DEC lambs were shorn correctly at weaning in December but they were then shorn the following October, instead of in December. The OCT-DEC lambs were shorn in December when they were 15 months old, instead of in October at 13 months of age. Data from these weaners when they were older than 13 months of age, when the mistake occurred, were excluded from analysis.

Table 4.1: Number of ewes in each age group, and mean age and standard deviation (years) of the ewe flock at each shearing Trial Numbers of ewes aged: Mean year 2 3 4 5 6 7 Total age 1999-2000 348 183 531 3.3 (0.5) 2000-01 348 183 531 4.3 (0.5) 2001-02 134 273 135 542 4.6 (1.4) 2002-03 137 119 233 35 524 5.0 (1.9) 2003-04 160 128 114 173 575 5.3 (2.3)

75 Results—Wool Production

Table 4.2: Duration of wool growth (days) between, and dates of, shearings in each ewe shearing group

Shearing Duration of wool growth Mean group (shearing date) Pre-trial 1999-2000 2000-01 2001-02 2002-03 2003-04 DEC 334 364 369 373 364 361 8/1/99 (8/12/99) (6/12/00) (10/12/01) (18/12/02) (17/12/03) MAR 410 350 364 373 368 373 12/2/99 (28/3/00) (13/3/01) (12/3/02) (20/3/03) (22/3/04) MAY 369 367 362 371 356 365 13/5/99 (16/5/00) (18/5/01) (15/5/02) (21/5/03) (11/5/04)

Table 4.3: Duration of wool growth (days) between birth and shearing 1, and shearing 1 and shearing 2 for weaner shearing groups in each birth-year cohort

Shearing Shearing # Weaner birth cohort Mean group 1999 2000 2001 2002 DEC-DEC 1 98 95 100 108 100 2 n/a 370 373 363 369 MAR-MAR 1 208 193 191 199 198 2 351 364 373 368 364 MAY-MAY 1 257 262 255 261 259 2 367 358 371 356 363 JUN-MAR 1 278 276 282 294 283 2 281 281 282 273 279 JUL-MAY 1 316 314 313 309 313 2 308 306 313 308 309 OCT-DEC n/a 408 409 401 406

n/a: Incorrectly shorn

76 Results—Wool Production

4.34.34.3 StaplStaplStapleStaple Strength & Position of Break

4.3.14.3.14.3.1 EweEweEwesEwesss No shearing group consistently produced the strongest wool each year, despite statistically significant differences in staple strength between shearing groups within years (Figure 4.2). Overall, the staple strength of the three shearing groups was similar (least squares mean staple strength ≈ 37 N/ktex; P = 0.73). At least one group produced wool with staple strength less than 35 N/ktex in every year of the trial, although no particular shearing group was consistently affected. MAY fleeces had the lowest or equal lowest proportion of mid-breaks in all years, irrespective of their staple strength (P < 0.05; Figure 4.3) and in four years out of five had more than 75% base breaks.

77 Results—Wool Production

Figure 4.2: Mean staple strength of fleeces from ewe shearing groups (error bars indicate 95% confidence interval (CI); LSM: least squares means)

55 50 45 40 35 DEC 30 MAR 25 MAY 20 15

Staple strength (N/ktex) strength Staple 10 5 0 1999-2000 2000-01 2001-02 2002-03 2003-04 LSM Trial year

Figure 4.3: Mean proportion of staple mid-breaks in fleeces from ewe shearing groups (error bars indicate 95% CI)

100%

90%

80%

70%

60% DEC 50% MAR MAY 40%

30%

Percentage of midbreaks Percentage 20%

10%

0% 1999-2000 2000-01 2001-02 2002-03 2003-04 LSM Trial year

78 Results—Wool Production

4.3.24.3.24.3.2 WeanerWeanerWeanersWeanersss Staple strength and position of break could not be measured in any DEC-DEC fleeces because of their short staple length. Amongst MAR-MAR weaners, only about two thirds of fleeces from sheep born in 1999 and 2002 were long enough for staple strength testing but all available results were included in the analysis. Very few samples in the other two years were long enough for testing and they were excluded from the analysis.

The staple strength of wool from the weaner shearing groups followed a similar pattern each year, particularly at shearing 1 (Figure 4.4). At shearing 1, all groups except MAR-MAR (when tested) produced wool of staple strength less than 32 N/ktex. Mean strength decreased progressively in each group that was shorn later in the year and OCT-DEC fleeces were weakest. Within years, all shearing groups except JUN-MAR and JUL-MAY differed significantly from each other. The staple strength of MAY-MAY, JUN-MAR, JUL-MAY and OCT-DEC varied by less than 5 N/ktex from year to year. Generally all groups contained more than 70% mid-breaks in all years (Figure 4.5). Overall, OCT- DEC had fewest mid-breaks (66%) and the other shearing groups were similar (79–86%).

Figure 4.4: Mean staple strength of fleeces from shearing 1 and shearing 2 of each weaner shearing group (error bars indicate 95% CI)

55 First shearing 50 45 40 35 30

25 20 15 10 5 0 1999 2000 2001 2002 LSM

Second shearing 55 50 45 40 Staple strength (N/ktex) 35 30

25 20 15 10 5 0 1999 2000 2001 2002 LSM Birth year

DEC-DEC MAR-MAR MAY-MAY JUN-MAR JUL-MAY OCT-DEC

79 Results—Wool Production

Staple strength at shearing 2 was generally greater than at shearing 1, although only JUN-MAR and JUL-MAY produced wool of staple strength greater than 35 N/ktex more than half of the time (i.e. in two or more of the four weaner cohorts). JUN-MAR weaners produced the strongest fleeces and DEC-DEC the weakest in all years. The ranking of the other, intermediate groups differed between years. Most fleeces from shearing 2 had considerably fewer mid-breaks than shearing 1. DEC-DEC contained mainly tip-breaks (69%), MAR-MAR tip- and mid-breaks (35% and 46%, respectively), and MAY-MAY, JUN-MAR and JUL-MAY fleeces mainly contained base-breaks (58–93%).

Figure 4.5: Mean proportion of staple mid-breaks in fleeces from shearing 1 and shearing 2 of each weaner shearing group (error bars indicate 95% CI)

First shearing 100% 90% 80% 70% 60%

50% 40% 30% 20% 10% 0% 1999 2000 2001 2002 LSM

Second shearing 100% 90% Percentage of mid-breaks 80% 70% 60%

50% 40% 30% 20% 10% 0% 1999 2000 2001 2002 LSM Birth year

DEC-DEC MAR-MAR MAY-MAY JUN-MAR JUL-MAY OCT-DEC

80 Results—Wool Production

4.44.44.4 Fleece Weight & Yield

4.4.14.4.14.4.1 EweEweEwesEwesss DEC ewes produced less greasy wool than either MAR or MAY ewes in four of the five years of the trial (P < 0.05) , although the difference ranged from 0.1 to 0.5 kg between years (Figure 4.6). MAR and MAY greasy fleece weights only differed in the 2000–01 trial year (P = 0.02). The least squares mean greasy fleece weight of DEC ewes was 3.86 kg, which was 0.3 kg lower than MAR (4.18 kg) and MAY (4.17 kg), which did not vary significantly from each other.

There was less than 1% vegetable matter present in all ewe and weaner fleeces and the amount did not differ between shearing groups.

The overall clean fleece yield from DEC ewes was greater than the other groups (least squares means of 77.0%, 74.3% and 75.3% for DEC, MAR and MAY, respectively; Figure 4.7). Clean fleece weights followed a similar pattern to greasy fleece weights from year to year (Figure 4.8) and, overall, the least squares mean clean fleece weight of DEC ewes was 140 and 150 g lower than MAR and MAY (2.98, 3.12 and 3.13 kg, respectively; P < 0.05). There was no statistically significant difference between MAR and MAY clean fleece weights.

Figure 4.6: Mean greasy fleece weight of ewe shearing groups, adjusted to 365 days growth (excluding October-shorn maiden ewes; error bars indicate 95% CI)

5.0

4.5

4.0

3.5

3.0 DEC 2.5 MAR

2.0 MAY

1.5

Greasy Greasy fleece weight (kg) 1.0

0.5

0.0 1999-2000 2000-01 2001-02 2002-03 2003-04 LSM Trial year

81 Results—Wool Production

Figure 4.7: Mean clean fleece yield of ewe shearing groups (error bars indicate 95% CI)

80.0

77.5

75.0 DEC 72.5 MAR MAY Yield Yield (%) 70.0

67.5

65.0 1999-2000 2000-01 2001-02 2002-03 2003-04 LSM Trial year

Figure 4.8: Mean clean fleece weight of ewe shearing groups, adjusted to 365 days growth (excluding October-shorn maiden ewes; error bars indicate 95% CI)

5.0

4.5

4.0

3.5

3.0 DEC 2.5 MAR

2.0 MAY

1.5

Clean Clean fleece weight (kg) 1.0

0.5

0.0 1999-2000 2000-01 2001-02 2002-03 2003-04 LSM Trial year

82 Results—Wool Production

4.4.1.1 Patterns of Annual Wool Growth in Ewe Shearing Groups Dyebands were applied to mid-side wool staples of ewes in each shearing group at the following times:

• 2000: May, October and December • 2001: March, May, July (except MAY ewes) and December • 2002: March, May, October and December • 2003: March

Dyebands were applied at similar intervals throughout 2003, but the samples were accidentally destroyed at the testing laboratory.

The average wool growth rates of the shearing groups in each dyeband period are shown in Figure 4.9. There were statistically significant (P < 0.001) differences between shearing groups in most periods examined. There was no statistically significant interaction between shearing group and age or reproductive status on wool growth rate (P = 0.07–0.6, depending on dyeband period analysed). Sheep that had just been shorn grew wool faster than unshorn sheep, regardless of when shearing occurred, but DEC ewes had smaller percentage increases in wool growth rate after shearing than the other shearing groups. The time of fastest wool growth for all groups generally occurred between May and December (i.e. winter–spring), rather than immediately after shearing. The only exception to this was

Figure 4.9: Mean clean wool growth rates of ewes from each shearing group in each dyeband period (■ denotes shearing)

14.0

12.0

10.0

8.0

6.0

4.0

2.0 Clean wool growth rate (g/d)

0.0 Apr-00 Jul-00 Oct-00 Jan-01 Apr-01 Jul-01 Oct-01 Jan-02 Apr-02 Jul-02 Oct-02 Jan-03 Date

DEC MAR MAY

Note: The graphs in the figure represent average wool growth rates during dyeband periods and, although they have been drawn as smoothed, continuous curves to facilitate interpretation, it should not be inferred that wool growth rate was constant during these times. Also, the subtle inflections as the curves change from one level to the next are artefacts of the smoothing process.

Figure 4.10: Mean weight of clean fleece grown by ewes in each shearing group during each dyeband period ( denotes shearing)

2000 2001 2002 2003 May-Oct Oct-Dec Dec-Mar Mar-May May-Jul Jul-Dec Dec-Mar Mar-May May-Oct Oct-Dec Dec-Mar

MAY Results—Wool Production Production Results—Wool 84 MAR Shearing group

DEC

0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5 8.0 8.5 9.0 9.5 10.0 Clean fleece weight (kg)

Results—Wool Production for DEC ewes in 2002, when wool growth rate was greatest after shearing. The slowest time of wool growth for all shearing groups occurred in the period immediately before shearing, when sheep were carrying the most wool of the year. DEC ewes only grew wool faster than the other groups in the period December–March.

The total weight of clean wool grown in each of the dyeband periods by the shearing groups is shown in Figure 4.10. Fleece weight differences between groups accumulated over time, so that, after 34 months, dyebanded DEC, MAR and MAY ewes had grown 7.5, 8.4 and 8.6 kg of clean wool, respectively. In the two years when wool growth was measured between October and December in all groups, MAR and MAY ewes grew the same amount of wool as each other but DEC ewes grew 40– 50% less. On average, MAR ewes grew 85% of their wool between shearing and December, or in 75% of their production year, and MAY ewes grew 73% of their total clean fleece weight between shearing and December, or in 50% of their production year.

4.4.24.4.24.4.2 WeanerWeanerWeanersWeanersss The patterns of greasy and clean wool growth in the weaner shearing groups were similar and only clean wool growth is reported. The shearing groups grew wool over different lengths of time, between 3 and 13 months (Table 4.3), and the mean fleece weights of each group reflected this (Figure 4.11). Nonetheless, wool production still differed between groups when compared on a grams per day basis (P < 0.01; Figure 4.12). The mean wool growth rates of the shearing groups generally followed a similar pattern each year, although there was some variation. At shearing 1, DEC-DEC weaners had the slowest wool growth rate, followed by MAR-MAR. The next greatest wool growth rates, from MAY-MAY, JUN-MAR and JUL-MAY weaners, were similar to each other. OCT-DEC weaners had

Figure 4.11: Mean clean fleece weight of each weaner shearing group at shearing 1 ( ) and shearing 2 ( ) (LSM: least squares means)

6.0 5.5 5.0 4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0 0.5 0.0 OCT OCT OCT OCT OCT Clean Clean fleece weight (kg/head) JUL..MAY JUL..MAY JUL..MAY JUL..MAY JUL..MAY JUN..MAR JUN..MAR JUN..MAR JUN..MAR JUN..MAR DEC..DEC MAY..MAY DEC..DEC MAY..MAY DEC..DEC MAY..MAY DEC..DEC MAY..MAY DEC..DEC MAY..MAY MAR..MAR MAR..MAR MAR..MAR MAR..MAR MAR..MAR 1999 2000 2001 2002 LSM SHEARING GROUP Birth year

85 Results—Wool Production the greatest wool growth rate, producing wool about 1 g/d more quickly than JUL-MAY weaners, on average. At shearing 2, DEC-DEC weaners again had a smaller mean wool growth rate than other groups. MAR-MAR and MAY-MAY weaners grew wool about 1 g/d faster than DEC-DEC, but slower than JUL-MAY. JUN-MAR weaners had the greatest mean wool growth rate. These differences were reflected in absolute fleece weights at shearing 2. The weight of DEC-DEC fleeces, which were grown over 12 months, was similar to the weights of JUN-MAR and JUL-MAY fleeces, which were grown over 9–10 months. The least squares mean total clean fleece production (i.e. shearing 1 plus shearing 2) from each weaner shearing pattern, and the duration of wool growth, were: DEC-DEC 3.2 kg (15 months); MAR-MAR 4.1 kg (18 months); MAY-MAY 4.5 kg (20 months); JUN- MAR 4.2 kg (18 months); JUL-MAY 4.5 kg (20 months); and, OCT-DEC 2.5 kg (13 months).

There was little consistent pattern in the shearing groups’ clean wool yield at either shearing (Figure 4.13). At shearing 1, DEC-DEC weaners born in 1999 and 2000 produced wool with the smallest yield but the 2002-born cohort had the greatest yield. Differences between groups were smaller at shearing 2. Overall, even the largest differences between shearing groups were relatively small: 3.9% between DEC-DEC and MAY-MAY at shearing 1, and 1.4% between MAY-MAY and JUN-MAR at shearing 2. Differences in yield between shearing groups were smaller than the differences in wool growth rate and the patterns of clean and greasy wool production were similar, as mentioned above.

Figure 4.12: Mean clean wool growth rates of each weaner shearing group between birth and shearing 1 (“Shearing 1”) and shearings 1& 2 (“Shearing 2”). Error bars indicate 95% CI.

Shearing 1 9 8 7 6

5 4 3 2 1 0 1999 2000 2001 2002 LSM

Shearing 2 14 13 12 11 10 9 8 7 Cleanwool growth rate (g/d) 6 5 4 3 2 1 0 1999 2000 2001 2002 LSM Birth year

DEC-DEC MAR-MAR MAY-MAY JUN-MAR JUL-MAY OCT-DEC

86 Results—Wool Production

Figure 4.13: Mean yields of clean fleece from shearing 1 and 2 of each weaner shearing group (error bars indicate 95% CI)

Shearing 1 82 80 78 76 74 72

70 68 66 64 62 60 1999 2000 2001 2002 LSM

Yield (%) Yield Shearing 2 82 80 78 76 74 72 70 68 66 64 62 60 1999 2000 2001 2002 LSM Birth year

DEC-DEC MAR-MAR MAY-MAY JUN-MAR JUL-MAY OCT-DEC

87 Results—Wool Production

4.54.54.5 Fibre Diameter

4.5.14.5.14.5.1 EweEweEwesEwesss DEC ewes produced significantly finer wool than MAR or MAY in all years of the trial except 1999-2000 (P < 0.05; Figure 4.14). The size of the difference varied between years, ranging from 0.1 to 1.1 µm. Over the whole trial DEC ewes produced the finest wool (least squares mean 19.1 µm), followed by MAR (19.4 µm) then MAY (19.7 µm).

Figure 4.14: Mean fibre diameter of ewe shearing groups (error bars indicate 95% CI)

21.0

20.5

20.0

19.5 m)

µ 19.0 DEC 18.5 MAR 18.0 MAY 17.5

Fibre Fibre diameter ( 17.0

16.5

16.0

15.5 1999-2000 2000-01 2001-02 2002-03 2003-04 LSM Trial year

88 Results—Wool Production

4.5.24.5.24.5.2 WeanerWeanerWeanersWeanersss The mean fibre diameter at shearing 1 of the weaner shearing groups followed a similar pattern each year, although group means, and the difference between them, varied between years (Figure 4.15). JUN-MAR and/or JUL-MAY fleeces were generally the finest, with overall means of 15.7 µm and 16.0 µm, respectively. MAR-MAR, MAY-MAY and OCT-DEC fleeces were usually broader (P < 0.05), but there were no statistically significant differences between these three groups. DEC-DEC fleeces were more than 1 µm broader than any other group (P < 0.05) each year.

In contrast to shearing 1, fibre diameter at shearing 2 followed a less consistent pattern each year, although there were still significant differences (P < 0.05) between shearing groups in all cohorts except those born in 2002. DEC-DEC produced the finest wool in two of three measured years (least squares mean 17.1 µm). Wool from JUN-MAR or JUL-MAY weaners was broadest every year (both least squares means about 18.6 µm), whilst wool from MAR-MAR and MAY-MAY weaners was of intermediate diameter (least squares mean 18.1 and 18.4 µm, respectively).

Figure 4.15: Mean fibre diameter of weaner shearing groups at shearing 1 and shearing 2 (error bars indicate 95% CI)

19.0 Shearing 1 18.5 18.0 17.5 17.0

16.5 16.0 15.5 15.0 m)

µ 14.5 1999 2000 2001 2002 LSM

Shearing 2 20.5 20.0 19.5 Fibrediameter ( 19.0 18.5 18.0

17.5 17.0 16.5 16.0 15.5 15.0 14.5 1999 2000 2001 2002 LSM Birth year

DEC-DEC MAR-MAR MAY-MAY JUN-MAR JUL-MAY OCT-DEC

89 Results—Wool Production

4.64.64.6 Staple Length

4.6.14.6.14.6.1 EwesEwesEwes The differences between the mean staple lengths of the shearing groups, adjusted to 365 days growth, were rarely significant and followed no particular pattern throughout the trial. The largest difference was 5 mm, which occurred between DEC and MAR ewes in the 2000-01 trial year (Figure 4.16).

4.6.24.6.24.6.2 WeanersWeanersWeaners The staple length of fleeces from shearing 1 of MAR-MAR weaners born in 2001 was not measured due to an oversight at the testing laboratory.

The overall, least squares mean staple lengths at shearing 1 ranged from 29 mm (DEC-DEC lambs shorn at weaning) to 100 mm for the weaners first shorn in OCT-DEC (Figure 4.17), with the staple length ranking of groups corresponding to duration of wool growth between birth and shearing 1. Mean staple length of any group varied less than 10 mm between years. The least squares mean length growth rate of DEC-DEC weaners was greatest (0.29 mm/d) and OCT-DEC smallest (0.25 mm/d). There were few significant differences in any year between the other shearing groups, which all had mean growth rates of about 0.22 mm/d.

Figure 4.16: Mean staple length of ewe shearing groups, excluding October-shorn maiden ewes (error bars indicate 95% CI)

100

90 DEC 85 MAR MAY 80 Staple length (mm)

70 1999-2000 2000-01 2001-02 2002-03 2003-04 LSM Trial year

90 Results—Wool Production

Amongst the weaners carrying 12 months wool at shearing 2 (DEC-DEC, MAR-MAR and MAY- MAY), there was no consistent year-to-year pattern in the mean staple length (Figure 4.17), nor mean rate of length growth, which varied by less than 0.02 mm/d within years. The mean length growth rate of JUN-MAR and JUL-MAY weaners was equal to or slightly greater than the other groups. These differences in rate of length growth translated to less than 6 mm difference in staple length between shearing groups over a 12 month period of wool growth.

Figure 4.17: Mean staple lengths of each weaner shearing group from shearing 1 and shearing 2 (error bars indicate 95% CI)

110 First shearing 100 90 80 70 60

50 40 30 20 10 0 1999 2000 2001 2002 LSM

Second shearing 110 100 90 Staplelength (mm) 80 70 60

50 40 30 20 10 0 1999 2000 2001 2002 LSM Birth year

DEC-DEC MAR-MAR MAY-MAY JUN-MAR JUL-MAY OCT-DEC

91 Results—Wool Production

4.74.74.7 Summary of Wool Characteristics The least squares means of the wool characteristics from the ewe and weaner shearing groups reported in the previous sections are summarised in Tables 4.4 and 4.5.

Table 4.4: Least squares mean greasy fleece weight (GFW), yield, clean fleece weight (CFW), fibre diameter (FD), staple length (SL), staple strength (SS) and percentage of mid-breaks of wool from the ewe shearing groups

Shearing GFW Yield CFW FD SL SS Mid- group (kg) (kg) (µm) (mm) (N/ktex) breaks DEC 3.9 77% 3.0 19.1 87 36.6 65% MAR 4.2 74% 3.1 19.4 88 36.2 53% MAY 4.2 75% 3.1 19.7 88 36.8 26%

Table 4.5: Least squares mean greasy fleece weight (GFW), yield, clean fleece weight (CFW), fibre diameter (FD), staple length (SL), staple strength (SS) and percentage of mid-breaks of wool produced at shearing 1 and 2 by each weaner shearing group

Shearing GFW Yield CFW FD SL SS Mid- group (kg) (kg) (µm) (mm) (N/ktex) breaks

Shearing 1 DEC-DEC 0.54 72% 0.39 17.7 29 n/a n/a MAR-MAR 1.11 74% 0.83 16.4 42 42 79% MAY-MAY 1.59 76% 1.20 16.3 57 30 81% JUN-MAR 1.73 74% 1.29 15.7 61 24 86% JUL-MAY 2.04 75% 1.53 16.0 70 21 83% OCT-DEC 3.39 75% 2.54 16.5 100 16 66% Shearing 2 DEC-DEC 3.79 74% 2.83 17.1 100 20 28% MAR-MAR 4.32 76% 3.30 18.1 100 28 46% MAY-MAY 4.39 75% 3.32 18.4 91 29 12% JUN-MAR 3.79 77% 2.91 18.6 75 39 37% JUL-MAY 3.89 76% 2.96 18.7 86 32 6%

n/a: not available because wool too short for routine testing

92 Results—Wool Production

4.84.84.8 Fleece Value

4.8.14.8.14.8.1 EweEweEwesEwesss No shearing group consistently produced a higher fleece value than the others throughout the trial (Figure 4.18). Under median staple length and strength premiums and discounts applied to median basis prices, the average fleece values of DEC, MAR and MAY ewes were $30.13, $29.89 and $27.79 per head, respectively. If first or ninth decile historical staple length and strength discounts were used, fleece values changed by less than $4/head and the shearing group rankings within years were the same as under median premiums and discounts, with the exception of 2003-04. In this year, MAR ewes produced the greatest fleece value if median or first decile premiums and discounts were used, whereas DEC ewes produced the greatest fleece value if ninth decile premiums and discounts were used. The rankings of the mean fleece values varied depending on the staple length and strength premiums and discounts used. Under the ninth decile regimen, the mean value of DEC fleeces ($31.85) was $1.12 greater than MAR ($30.73/head), which in turn was greater than MAY ($28.42/head). If first decile premium and discounts were used, MAR produced the highest mean fleece value ($28.80), followed by DEC ($28.19) and MAY ($27.04).

If first or ninth decile micron basis prices were used, fleece values differed by $5–20 from the values calculated using median prices (Appendix 2, Figure A-2.1) but the rankings of the shearing groups

Figure 4.18: Estimated fleece value of each ewe shearing group, based on median historical basis price and median, first decile (lower bar) and ninth decile (upper bar) staple length and strength premiums and discounts

$40.00

$35.00

$30.00

$25.00 DEC $20.00 MAR MAY $15.00

$10.00

$5.00 Mean Mean fleece value ($/head, gross)

$0.00 1999-2000 2000-01 2001-02 2002-03 2003-04 Mean Trial year

93 Results—Wool Production were virtually identical to those produced under median basis and median staple length and strength premium and discount conditions.

4.8.24.8.24.8.2 WeaneWeaneWeanersWeanersrsrs The gross value of fleeces from each weaner shearing group, calculated using median historical micron basis price and median staple length and strength premiums and discounts, is shown in Figure 4.19. JUL-MAY weaners produced the highest value per head in most years of the trial and this was $2– 3/head higher than MAY-MAY weaners, even though they had the same duration of wool growth. Nonetheless, the proportional differences in fleece value between the shearing groups were similar to the different lengths of time over which the fleeces were grown. For example, the duration of wool growth of JUN-MAR weaners from birth to shearing 2 (18 months) was 10% shorter than JUL-MAY weaners (20 months) and this was approximately the same as the difference between the fleece values of these groups.

The rankings of the fleece values of the weaner shearing group did not change if first decile historical micron basis prices were used instead of median basis prices. However, if ninth decile micron basis prices were used, JUL-MAY and JUN-MAR weaners produced the most valuable fleeces, followed by MAY-MAY, MAR-MAR, DEC-DEC and OCT-DEC, in decreasing order (Appendix 2, Section A2.2). These micron basis price patterns did not differ if median, first or ninth decile premiums and discounts for staple length and staple strength were applied to them.

Figure 4.19: Estimated fleece values from first ( ) and second ( ) shearing of each weaner shearing group, calculated using median historical micron basis prices and length & strength discounts

$65 $60 $55 $50 $45 $40 $35 $30 $25 $20 $15 $10 $5 $0 OCT OCT OCT OCT OCT Mean Mean fleece value ($/head, gross) JUL-MAY JUL-MAY JUL-MAY JUL-MAY JUL-MAY JUN-MAR JUN-MAR JUN-MAR JUN-MAR JUN-MAR DEC-DEC MAY-MAY DEC-DEC MAY-MAY DEC-DEC MAY-MAY DEC-DEC MAY-MAY DEC-DEC MAY-MAY MAR-MAR MAR-MAR MAR-MAR MAR-MAR MAR-MAR 1999 2000 2001 2002 Mean SHEARING GROUP Birth year

94 Results—Wool Production

4.94.94.9 Lifetime Wool Production Results Lifetime clean wool production per head up to 6.25 years of age (i.e. in December) was estimated for all shearing groups. For DEC ewes, this was all shorn wool, but for MAR and MAY ewes this estimation included the wool still carried by the sheep. Depending on their shearing pattern as weaners, MAR and MAY ewes produced between 1.0 and 1.1 kg (6%) more clean wool over their lifetime than DEC ewes shorn as DEC-DEC weaners (Figure 4.20). DEC ewes shorn as OCT-DEC weaners produced 0.2 kg (1%) more clean wool than DEC ewes shorn as DEC-DEC weaners.

The estimated values of lifetime wool production per head, net of shearing costs, from the weaner and ewe shearing combinations, based on median basis price and median premiums/discounts were: DEC- DEC $153; OCT-DEC $157, MAR-MAR $158; JUN-MAR $161; MAY-MAY $148 and JUL-MAY $151 (Figure 4.21). Compared to the least valuable (MAY-MAY), other groups’ lifetime wool values were 2–9% higher. The rankings only differed at historically high micron basis prices, when DEC ewes shorn as OCT-DEC weaners produced a greater fleece value than MAR-MAR (Appendix 2, Figure A-2.4). The estimated value of wool production for sheep in each shearing group between birth and 6.25 years of age, accounting for each group’s different mortality rates, is described in Section 6.7.

95 Results—Wool Production

Figure 4.20: Mean clean weight of wool produced from birth to 6.25 years of age by ewes shorn in each weaner and ewe shearing pattern

20.0 18.0 16.0 14.0 12.0 10.0 8.0 6.0 4.0 2.0 Clean Clean fleece weight (kg/head) 0.0 DEC-DEC OCT MAR-MAR JUN-MAR MAY-MAY JUL-MAY

DEC MAR MAY Weaner shearing group Ewe shearing group

1st weaner 2nd weaner Maiden Adult (3–6 years old, inclusive) shearing

Figure 4.21: Net value of wool, at first, median and ninth decile historical staple length and strength premiums and discounts, produced from birth to 6.25 years of age by ewes shorn in each weaner and ewe shearing pattern

$180

$160

$140

$120

$100

$80

$60

$40

$20

$0 Wool net value of shearing ($/head) DEC-DEC OCT MAR-MAR JUN-MAR MAY-MAY JUL-MAY -$20 DEC MAR MAY Weaner shearing group Ewe shearing group

1st weaner 2nd weaner Maiden Adult shearings Total value using 1st OR 9th decile L&S discounts

C H A P T E R 5

DISCUSSION—WOOL PRODUCTION

Although some wool characteristics consistently differed between adult and weaner shearing groups, no shearing time improved all of them. In the adult flock, a beneficial effect of one characteristic on wool price tended to be offset by an undesirable effect of another. Similarly, if a weaner shearing group ranked highly on a wool attribute at shearing 1, it tended to rank low at shearing 2. Thus, when the changes in all wool characteristics were taken into account over a sheep’s lifetime, there were only small differences between shearing groups overall.

The current study compared realistic shearing options, and was conducted on a commercially managed farm under a wide variety of seasonal conditions. Thus, the results should be widely applicable to commercial wool growers in south-eastern Australia. The shearing times were chosen, in conjunction with farmers, as the most practical alternatives for a spring-lambing enterprise. A winter shearing option was excluded because it lowers stocking rate unacceptably and so is detrimental to farm profitability. Many previous experiments have compared different shearing times for two or three years but this is probably not long enough to demonstrate consistent, long-term differences between shearing times. Consequently, this trial was conducted for five years, during which time a wide range of seasonal conditions occurred, including summers with high rainfall and continuous pasture growth, and drought. This was associated with considerable inter-year variation in wool characteristics but was no greater that that observed in a ten-year study of wool quality in south-western Victoria (Gloag et al. 2002).

5.15.15.1 Staple Strength The most outstanding feature of this trial was that shearing time improved staple strength in a consistent and profitable manner on only one occasion in a sheep’s life—at the second weaner shearing, which took place between 15 and 20 months of age. In adult ewes, no shearing time consistently improved staple strength and, at the first weaner shearing, any wool value increase due to improved staple strength was negated by price reductions due to short staple length. The results for ewe and weaner shearings are discussed separately in more detail below.

5.1.15.1.15.1.1 EweEweEwesEwesss Ewes in this trial produced wool of low to moderate staple strength, similar to surveys and other studies of ewes in Victoria, Western Australia and Tasmania (Barton et al. 1994; Butler et al. 1994; Scrivener

97 Discussion—Wool Production and Vizard 1997). Wool generally attracts price discounts for strength when it is weaker than 35 N/ktex (Independent Commodity Services 2005). Wool of staple strength less than 35 N/ktex was produced by at least one shearing group in every year of the trial, although each shearing group also produced wool with relatively high staple strength on several occasions.

Shearing ewes in December, March or May did not consistently improve staple strength throughout this trial. There is a widespread belief that shearing in autumn, to coincide with the seasonal break, generally produces wool of greater staple strength (Adams et al. 2000), however this was not supported by the observations of the present study. As discussed in Section 2.1.1.3, reproducing ewes can also have a position of break (POB) in winter or spring, yet no long term experiments have examined variability of the POB or staple strength between years. In the present study, POB in ewes varied widely from year to year and, similar to other studies, no shearing time consistently improved staple strength (Hansford 1987; Thornberry et al. 1988).

The mechanisms underlying the observed staple strength results can be illustrated by examining a hypothesised fibre diameter profile from a lactating ewe during the first two years of the study (Figure 5.1). Seasonal conditions varied considerably during this time. There was an indistinct autumn break in 1999, abnormally low rainfall at the start of lambing, average seasonal conditions in 2000, a false autumn break in February 2001 followed by a pronounced break in April. In 1999, the POB was probably associated with poor nutrition at lambing (‘B’ in Figure 5.1). Thus, March and May shearing resulted in low staple strength because the POB was away from the staple end and was not clamped under the jaws of the testing machine, despite being in the lower third of the staple. In contrast, shearing in December placed the POB at the end of the staple, resulting in greater staple strength, particularly because there was little decline in pasture quality in the preceding autumn to cause a POB at this time (‘A’). In the following year, December-shorn ewes had a mid-staple break, which was probably associated with autumn (‘C’), and produced the weakest wool in that year. Shearing ewes in

Figure 5.1: Significant seasonal events during years 1 & 2, and hypothesised fibre diameter profile, showing the effect of different shearing times on the position of break and measured staple strength 1999 2000 2001 Jan Mar May Jul Sep Nov Jan Mar May Jul Sep Nov Jan Mar May +++ summer Driest Sept. April Good False 150 rainfall for 15 years break spring break mm Feb rain April

A B C

December shearing: base-break (Sept ‘99) December shearing: mid-break (April ‘00)

49 N/ktex 32 N/ktex March shearing: mid-break (Sept ‘99) March shearing: mid-/base-break (Jan ‘01)

32 N/ktex 34 N/ktex May shearing: mid-break (Sept ‘99) May shearing: base-break (April ‘01)

29 N/tex 44 N/ktex

98 Discussion—Wool Production

March was not close enough to ‘C’ to place the POB under the jaws of the testing machine, and resulted in wool having similar strength to December shearing. However, shearing in May placed ‘C’ at the end of the staple, consistent with the production of wool of the greatest staple strength in that year. In summary, spring-lambing ewes have two possible positions of break, which vary from year to year. Consequently, no single shearing time can consistently place the POB at the end of the staple in order to improve staple strength.

In contrast to the results of the present study, Curtis and Stanton (2000 cited by Oldham 2000) reported that staple strength varied with time of wool sale in Victoria. However, the reported maximum annual variation in staple strength was only 4 N/ktex and, in the absence of a statistical analysis, it is possible that there was no significant difference between the strength of wool sold at different times. Additionally, shearing time was inferred from time of wool sale and was not precisely known, making it difficult to draw direct conclusions about the effect of shearing time on wool strength from the results. The authors were also unable to examine the effect of reproductive status and lambing time, and it is likely that the observed differences were not representative of spring-lambing ewes.

The only relatively constant effect of ewe shearing time on staple strength observed in the present study was that May shearing produced wool with a majority of base-breaks in four out of five years. This distribution of breaks was quite distinct from the other shearing groups, which showed more year to year variation, but was not associated with consistently higher strength. Such an observation suggests that, despite the POB occurring before shearing each year and so being located near the base of the staple, it was not always close enough to be grasped under the staple breaker’s jaws. The relatively constant proportion of base breaks associated with May shearing is at odds with the more variable location of the POB in the hypothesised fibre diameter profile shown in Figure 5.1. To investigate this in further detail would require a study of fibre diameter profiles produced by different shearing times. This would provide more insight into the location of the POB, and its variability, than is gained by simply measuring the proportion of tip-, mid- and base-breaks in a staple. Such a study might also indicate that particular shearing times stimulate wool growth and fibre diameter changes more than others, and actually change the shape of the fibre profile, rather than just repositioning the break.

This study did not investigate the precise mechanisms by which shearing time influences staple strength, but it showed that manipulating shearing time of spring-lambing ewes will not reliably improve staple strength as it is currently measured, regardless of where the position of break occurs.

5.1.25.1.25.1.2 WeanersWeanersWeaners At shearing 1, only March-shorn weaners produced wool of staple strength greater than 35 N/ktex, although their wool was long enough for testing on just two of four occasions. Staple strength of wool from all groups at shearing 2 was generally greater than at shearing 1 but staple strength was still below 35 N/ktex on 13 of 16 occasions. The poor staple strength of weaner wool in this trial is similar to other reports (Foot and Vizard 1993; Barton et al. 1994) and confirms the difficulty of producing sound wool in young Merinos in south-eastern Australia.

99 Discussion—Wool Production

Shearing time influenced weaner staple strength, with weaners shorn in October and December (i.e. ‘spring’) producing the weakest wool at shearing 1 and 2, respectively, but the implications of this were not straightforward, as discussed below. The POB for all weaner groups at both shearings was probably caused by the same annual events in the first and second years of life, namely low pasture quality during late summer and autumn, or nutritional change at the seasonal break (Saul 1988; Morcombe and Allen 1993; Peter et al. 1993; Doyle et al. 1995).

It is likely that staple strength was influenced by an interaction between shearing time and the POB that was similar to the one in the ewe flock, which was discussed above. Thus, at shearing 1, the POB in March-shorn wool was regularly clamped under the testing machine jaws, resulting in increased staple strength. In contrast, the consistently low staple strength wool produced by the other shearing times suggests that the POB was never clamped under the testing machine’s jaws in these groups. This may have been because the POB always occurred mid-staple and never near the tip. If this were the case, such a consistent finding is interesting, given the variation in summer rainfall, nutrition and wool growth that occurred during the trial. An alternative explanation is that the weaner wool fibres might have had long thin sections, instead of a distinct POB, that could never be wholly clamped during testing to lead to improved staple strength. Future studies investigating the fibre diameter profiles of wool from different weaner shearing times could investigate this in more detail. It is likely that all the shearing groups produced fleeces with a high proportion of mid-staple breaks at shearing 1 because their short staple length meant that only the middle of the staple protruded beyond the testing machine’s jaws to be exposed to the testing force and broken.

At shearing 1, both staple strength and length were influenced by shearing time, with shorter wools being of greater staple strength. This meant that the median length and strength price discounts for all fleeces from shearing 1 were about 35–45%. For example, weaners first shorn in July or October (JUL-MAY or OCT-DEC) were the only sheep that consistently produced wool longer than 65 mm, the length below which discounts for short staple length are incurred (Independent Commodity Services 2005). However, these groups also produced wool with the lowest staple strength. Thus, manipulating shearing time in the first year of life is not a useful strategy for improving weaner staple strength because any price premium achieved by shearing at a young age will be offset by the discount received for short staple length.

The situation at shearing 2 was much clearer than at shearing 1. Weaners shorn in March whose previous shearing was in June (JUN-MAR) produced the strongest wool, often greater than 40 N/ktex. This was probably because a base POB, associated with the seasonal break at about 18 months of age, was regularly clamped under the jaws of the testing machine. In contrast, March-shorn weaners that had been previously shorn in March (MAR-MAR), or weaners shorn in December (DEC-DEC), produced wool with lower staple strength. This occurred because shearing 1 was earlier in life and placed the POB, associated with autumn in the first year of life, outside the region clamped by the testing machine. Thus, at shearing 2, December-shorn wool had mainly tip-breaks and March-shorn wool mainly tip- and mid-breaks each year. Shearing weaners for the second time in May, irrespective

100 Discussion—Wool Production of their first shearing time, produced wool of lower staple strength than the June-March shearing. This wool still had a base-break but, because of the later shearing time, it was not clamped by the testing machine. As with shearing 1, it would be worthwhile to conduct further studies to characterise the fibre diameter profile of wool from the different weaner shearing times to verify these hypotheses.

In summary, the second weaner shearing was the only shearing time that consistently improved staple strength, regularly placing the position of break within the clamping region of the staple strength testing machine, without adversely affecting other wool characteristics. Nevertheless, this tactic is still vulnerable to seasonal variation. For example, the June-March weaner shearing pattern produced wool of greatest staple strength at the second weaner shearing but wool from this group was still weaker than 30 N/ktex on one of four occasions.

5.25.25.2 Fibre Diameter & Fleece WWeighteighteighteight In contrast to the staple strength results, there were more consistent differences in fleece weight and fibre diameter between shearing groups. The effect of shearing time on fleece weight and fibre diameter in the current study was similar, although smaller in magnitude, to the results of other experiments involving ewes in winter rainfall environments (McGarry and Stott 1960; Arnold et al. 1984; Cloete et al. 2000). The clean fleece weights of both March- and May-shorn ewes were approximately 5% (150 g) heavier than December ewes. However, the benefit of the extra fleece produced by March or May shearing was offset by the wool’s broader fibre diameter (average 19.4 and 19.7 µm for March and May, respectively), which lowered its price.

A sub-group of ewes was dyebanded to investigate possible reasons for the fleece weight differences between shearing groups. Wool growth rate in each group increased after shearing in a similar manner to that described in other studies and summarised by Bottomley (1979). However, December-shorn ewes generally had a smaller increase in wool growth rate after shearing than either March- or May- shorn ewes. This is consistent with the findings of McGuirk et al. (1966), who found that June shearing caused a proportionally greater increase in wool growth rate than December shearing. The autumn-shorn ewes in the present study appeared to grow more wool in a year because they had a larger increase in wool growth rate after shearing, which persisted for a longer time, than December- shorn ewes. In the autumn-shorn groups this appeared to be superimposed on a common seasonal wool growth pattern, in which wool grew fastest during winter and/or spring. It was less clear whether a similar pattern occurred in December-shorn ewes. Extra dyebanding studies would be invaluable in exploring this relationship further and additional work investigating sheep grazing pastures of known quantity and quality, with more frequent dyebanding intervals, is indicated.

As discussed in Section 2.1.2, Arnold et al. (1984) suggested that the different body condition of sheep in autumn and spring might lie behind the varying wool growth response to different shearing times. However, in the present study, ewes were not consistently in lighter body condition in autumn (see Section 6.1.1) and average wool growth rates after shearing were not always greater in lighter sheep.

101 Discussion—Wool Production

For example, ewes in all shearing groups were approximately condition score (CS) 2.6 in December 2000, and shorn ewes increased their wool growth rate from 6.3 to 8.3 g/d at that time. In the following March, ewes were heavier (approximately CS 3.3), yet March-shorn ewes increased their wool growth rate by a much greater amount after shearing—from 5.5 to 11 g/d.

In the absence of any consistent association between condition score and wool growth rate, an alternative explanation of the greater wool growth observed in autumn-shorn sheep is that it was stimulated by cooler climatic conditions experienced by sheep after shearing. Autumn-shorn sheep had greater mean wool growth rates than December-shorn sheep for most of the year, suggesting that shearing had a persistent effect on their wool growth. Previous studies have shown that sheep respond to cold temperatures for several months after shearing. For example, shorn sheep in winter required a 27% higher feed requirement two months after shearing than woolly ones (Hudson and Bottomley 1978). Other studies have suggested that cold conditions do not stimulate wool growth rate, even in sub-zero temperatures (Bottomley 1979), but these used photoperiod-sensitive breeds with low rates of winter wool growth, and so it is possible that different effects occur in Merinos.

This study observed smaller fleece weight differences between shearing groups than other experiments (McGarry and Stott 1960; Lightfoot 1967; Arnold et al. 1984; Cloete et al. 2000). This might be explained by differences in climate, pasture availability and feed intake after shearing, although no trials have quantified these interactions in detail. In addition, sheep were grazed at greater stocking rates in the current study and so it is likely that feed availability during autumn was lower than in previous studies. This would have restricted the extent to which feed intake and wool growth could increase after autumn shearing, explaining the smaller fleece weight and fibre diameter differences than previous reports. Despite the smaller effect of autumn shearing on wool production per head in high stocking rate systems, production per hectare is still maximised because the smaller increase in fleece weight per sheep is offset by the greater number of sheep grazed per hectare (Lightfoot 1967).

Amongst weaners, fleece weights from shearing 1 varied between groups because the fleeces were grown over different lengths of time, different seasons and were harvested at different ages. At shearing 2, three groups had grown wool for 12 months (DEC-DEC, MAR-MAR and MAY-MAY) but December-shorn weaners produced less wool, about the same amount as the March- and May-shorn weaners that had only grown wool for 9–10 months between shearing 1 and 2. The difference between the December-, and March- and May-shorn weaner groups was similar to that in adult ewes and might have been due to their different first shearing times. However, the effect of shearing time on fleece weight was confounded with age, because December-shorn weaners grew their wool from 3–15 months of age, whereas March- and May-shorn weaners grew wool from 6–18 and 8–20 months of age, respectively. The cumulative fleece weights from birth to 75 months of age of autumn-shorn sheep were 6% greater than December-shorn sheep but there was a 5% difference between adult fleece weights. Thus, although weaner shearing time may have made some contribution to the fleece weight differences between shearing groups, it was small in magnitude and is not a main reason for choice of shearing time overall.

102 Discussion—Wool Production

Within the December ewe shearing groups, weaners first shorn as hoggets in October produced 3% more wool between birth and maiden ewe shearing, inclusive, than weaners first shorn as lambs in December, which represented considerable extra production over the first two years of life. This may have occurred because pasture availability immediately after shearing in October was greater than in December, or because shearing newly-weaned lambs, especially in December at the onset of declining pasture quality, caused a nutritional setback that reduced wool growth. In contrast to these results, a study conducted in the north island of New Zealand (Newman et al. 1996) found that Merinos first shorn as lambs produced 17% more clean wool to 15 months of age than those first shorn as hoggets. The difference between the two studies may have been due to differences in pasture availability, quality or sheep growth rate, which have been shown to modulate the effect of shearing on wool growth in young Romney sheep (Sumner 1984).

There was also a very strong association between weaner fibre diameter and age at which weaners were shorn at both shearing 1 and shearing 2. Mean fibre diameter decreased with age of shearing to about ten months of age, before increasing with later shearing times. As a result, weaners first shorn in June or July (9–10 months of age) produced finer fleeces than the other weaner groups at shearing 1, similar to results reported in other studies (Mullaney et al. 1969; Francis et al. 2000). The fibre diameter ranking from different ages of shearing appears to be independent of weaner nutrition, although nutrition can change the absolute fibre diameter values (Masters et al. 1998). However, in the present study, the finest groups at shearing 1 invariably produced the broadest wool at shearing 2, which offset the benefit to wool price of a particular weaner shearing pattern. For example, JUL-MAY wool had the greatest average fleece value per kilogram at shearing 1 but the smallest at shearing 2. The overall effect of fibre diameter on weaner fleece value depended on the size of the price premium for fine wool, as well as changes in other wool characteristics, discussed in Section 5.3.2.

5.35.35.3 FleeceFleeceFleece ValueValueValue Because each shearing time produced desirable and undesirable changes in different wool characteristics, a per head fleece ‘value’ from each shearing time was calculated. This placed an economic weighting on the change in each wool characteristic and summarised the effects of each shearing time, allowing the overall effect on wool production to be compared under different historical market conditions. Shearing time caused surprisingly small differences in fleece value in adults, weaners and over a sheep’s lifetime.

5.3.15.3.15.3.1 EwesEwesEwes At median historical wool prices, ewes shorn in December and March produced fleeces of similar value. December shearing produced the greatest fleece value at ninth decile (i.e. high) historical micron basis prices but March shearing was more valuable at first decile micron basis prices, although even the largest difference overall was less than $2/head. May shearing always produced the lowest fleece value.

103 Discussion—Wool Production

The fibre diameter and fleece weight differences between shearing groups made important contributions to differences in fleece value, particularly because no shearing time consistently improved staple strength, but their combined effect varied in different market conditions according to the premium paid for fine wool. At median historical wool prices, a fibre diameter increase from 19 to 20 µm decreased clean wool basis price by approximately 14% (Independent Commodity Services 2005). Under these conditions, March-shorn wool had a 4.2% lower price than December (0.3 µm broader × 14%), which was almost exactly offset by its 4.7% heavier clean fleece weight. When the premium paid for fineness was larger, the price decrease due to the broader diameter of March-shorn wool was not offset by greater fleece weight, and so the fleece value from March shearing was less than from December shearing. Similarly, at first decile prices (low micron premium), March shearing produced the most valuable fleece. May shearing always produced fleeces of lower value than March shearing because they were of broader fibre diameter but were no heavier. Although May-shorn fleeces were 150 g heavier than December-shorn ones, they were 0.6 µm broader and the extra weight only offset their lower wool price under a very limited range of market conditions, namely at first decile or lower micron premiums.

The mean proportion of mid-breaks in the wool was not included in fleece value calculations because long-term price data for this characteristic have not been published. However the median discount for a large proportion of mid-breaks between 1995 and 2005 was about 24 cents, or less than 3% of the median prices for this time (AWI 2006). Thus, proportion of mid-breaks make a relatively small contribution to wool price and the estimated fleece values from the present study should not be excessively inaccurate.

5.3.25.3.25.3.2 WeanersWeanersWeaners Because shearing group and age at shearing were confounded, as discussed above, weaner fleece values are discussed as part of the value of each shearing group’s ‘lifetime’ wool production, below. However, the two weaner shearings within each adult shearing time can be compared because they had the same durations of wool growth. Shearing the progeny of March-shorn ewes first in June, at nine months old, produced fleece values about 5% greater than shearing first in March, at six months old, over a wide range of price scenarios. The same was true for shearing progeny of May ewes first in July, at ten months of age, rather than May, at 8 months old. This probably occurred because June or July shearing produced particularly fine and valuable wool compared to March or May shearing, similar to what was observed in a previous study (Donnelly 1991a). The effect was most pronounced at historically high micron basis prices, where June-shorn weaners produced the equal greatest fleece value of all weaner shearing groups, even though they did not grow wool for the longest time (Appendix 2, Figure A-2.3). Although June- and July-shorn weaners produced the broadest-diameter wool of any groups at shearing 2, the premium paid for the fine fibre diameter at shearing 1 was disproportionately greater than the lower price received at shearing 2.

104 Discussion—Wool Production

Within the December ewe shearing groups, the effect of shearing time on progeny fleece value depended on the prevailing market conditions. However, the differences between the two shearing groups were less than the cost of the extra shearing required when weaners were first shorn as lambs, and so the pattern where weaners were first shorn as hoggets in October always had a greater fleece value net of shearing costs, even when lamb shearing produced a greater gross wool value. Very similar results have been reported in other trials involving Merinos (Taylor 1989; Sackett 1990; Donnelly 1991a; Irving 1991; Rogan et al. 1995; R. Thompson pers. comm.). In all cases, the costs of an extra shearing substantially reduced the returns from a lamb shearing pattern, although discounts for short staple length wool further reduced returns from lamb shearing in some of these studies.

There were consistent differences in fibre diameter, staple length and staple strength between the weaner shearing times. However, it seems that the favourable and unfavourable changes in wool characteristics caused by shearing time largely balanced themselves out, and thus the overall differences in fleece value between the shearing groups were small. For example, the staple length of wool from a shearing time tended to be offset by its staple strength and, similarly, the effect on price of fibre diameter from shearing 1 was offset by fibre diameter from shearing 2. Kleeman (1997) reported very similar results in a comparison of weaner shearing times in north-west Queensland. In that study, autumn-born Merinos shorn first at 5, 8 or 11 months of age produced wool of similar value between birth and 18 months of age, because a higher value from the first shearing was offset by a lower value at the second, or vice versa.

Wool growth rates of weaners prior to the first shearing were calculated assuming that all sheep were born at the start of the lambing period. Although the average birth date of sheep born during a five- week lambing period would be approximately two weeks later (Morley 1983a), this date was used in calculations because it made no assumptions about the timing or spread of lambing within the lambing period. This may have differed between ewe shearing groups due to the effects of shearing on ewe fertility. If average weaner age was reduced by a fortnight, thus decreasing the duration of wool growth prior to shearing 1, the average wool growth rate of DEC-DEC weaners increased by 0.7 g/d and all other weaner groups by less than 0.5 g/d. These results only differed by a small amount from the results presented and had no effect on estimated fleece values.

5.3.35.3.35.3.3 Lifetime Fleece Value Net of shearing costs, shearing ewes in March produced the greatest total lifetime fleece value, from birth to 75 months of age, of all the adult shearing groups, followed by December then May, regardless of when weaners within these adult times were shorn. Within each adult shearing time, the later weaner shearing time (i.e. June, October or July, respectively) resulted in the greater value of lifetime production. Under no price scenarios did May shearing produce a greater lifetime fleece value than the other groups. Nonetheless, the differences in wool value between shearing times were relatively small. The lifetime fleece value of a June (weaner)–March (adult) shearing was 3% greater than an October (weaner)–December (adult) shearing (i.e. only 0.5% per annum), and 7% greater than a July (weaner)– May (adult) shearing (1.2% p.a.). These small differences in lifetime fleece value are consistent with a

105 Discussion—Wool Production

1994 survey that found no difference in net wool returns between Victorian farms with different shearing times (Court and Lawless 1995).

It should be noted that the age structure of the ewe flock varied throughout the trial, depending on the number of maiden ewes selected into the adult flock and the number of ewes that died or were culled each year. The values used in each year of the shearing groups’ lifetime wool production calculations were the least squares mean clean fleece weight, fibre diameter, and staple length and strength, averaged over each ewe age group. These values vary slightly from the means presented for the individual wool characteristics, which were least squares means for a 4.5 year old ewe, the mean age throughout the trial.

In reality, it is unlikely that a farmer would sell March- or May-shorn ewes in December, when they are carrying less than 9 months wool. Instead, the sheep would be kept until closer to the main shearing time and sold after shearing. Keeping March-shorn ewes for longer, if feed conditions allow, would only further increase their lifetime fleece value above the other shearing times because they produced the highest lifetime value in the first place. However, at median historical prices, May-shorn ewes must be kept two and six months longer than December- and March-shorn ewes, respectively, to produce the same value of wool produced by those groups up until December. Keeping May-shorn ewes this extra time would probably require a lower stocking rate or extra supplementary feeding over summer, thus reducing the profitability of this shearing time. This was shown in a simulation of a wether flock in southern NSW, where the average gross margin per hectare of May shearing was 14– 17% lower than a December-shorn enterprise with the same winter stocking rate (Salmon et al. 2006). This was because cast-for-age wethers retained until May shearing required 63% more supplementary feeding over summer, or because a premature shearing prior to sale in December produced wool of lower value.

It is also unlikely that a self-replacing Merino flock would contain no wethers. Wethers are more likely to have a consistent POB in autumn (Thornberry et al. 1988), so their fleece value may benefit from autumn shearing. Since the present study has shown that shearing a ewe flock in March produces the most valuable wool, it is likely that March shearing of a typical flock containing ewes, wethers and weaners will also produce wool of greater value than shearing in December or May.

A self-replacing flock should not be shorn in May because it produces lower value wool than December or March shearing. However, when it comes to choosing between December and March shearing, the group rankings according to fleece value varied considerably between years, and the overall difference between these two shearing times was small. This suggests that the final decision on shearing time should not be based on fleece value alone. Furthermore, the small advantage offered by March shearing, about $0.70/head per annum, could be negated if autumn shearing reduced wool income. This could occur if, for example, autumn shearing led to greater vegetable matter contamination and lower wool value, or incurred greater variable costs in the management of blowfly strike. These issues are discussed further in Chapter 7.

106

C H A P T E R 6

RESULTS—SHEEP HEALTH & MORTALITY

6.16.16.1 Body Condition & Growth Rate

6.1.16.1.16.1.1 Ewe Condition Score The mean condition score (CS) of the ewe shearing groups followed similar patterns throughout the trial (Figure 6.1). Mean ewe condition peaked in 2000, then decreased during 2002 because of poor winter pasture growth and availability on Farm A. Although condition increased slightly prior to the trial’s move to Farm B in February 2003, ewe condition reached a nadir of about CS 1.7 in 2003 due to very difficult seasonal conditions and low pasture availability, and insufficient supplementary feeding, before increasing in the spring. Within this general pattern, there were modest differences, of

Figure 6.1: Mean condition score of ewes in each shearing group at different times during the trial (bars indicate 95% CI)

4.0

3.5

3.0

2.5

2.0 Mean Mean condition score

1.5

1.0 Jan-00 Jan-01 Jan-02 Jan-03 Jan-04 Date

DEC MAR MAY

107 Results— Health & Mortality

less than 0.4 CS, between shearing groups in winter and early summer (at weaning) in 3 of 4 years in which condition score was measured (P < 0.01; Table 6.1). In winter, MAY ewes were lightest or equal-lightest (with MAR), and were 0.2–0.4 CS lighter than DEC ewes. Conversely, DEC ewes were lightest or equal-lightest, by 0.2–0.4 CS, in December.

Table 6.1: Mean condition score prior to joining in March (join), during winter in July (wint) and at weaning in December (wean) of each ewe shearing treatment Shearing 2000 2001 2002 2003 2004 group join wint wean join wint wean join wint wean join wint wean join DEC n/a 2.8a 3.3 n/a 3.4a 3.3a 2.7 2.3a 2.1a 1.8a 2.1 2.3a 2.1a MAR n/a 2.7a 3.3 n/a 3.2b 3.6b 2.7 2.2b 2.3b 1.9b 2.0 2.5b 2.0b MAY n/a 2.6b 3.3 n/a 3.2b 3.4a 2.7 1.9c 2.2b 1.9ab 2.0 2.7c 2.1ab

Within years, means with no or common superscripts do not differ significantly (P > 0.05)

108 Results— Health & Mortality

6.1.26.1.26.1.2 Weaner Growth Rate Following Shearing Within years, the mean weaning weights of the progeny of the ewe shearing groups differed by up to 2.9 kg, or 14% (P < 0.01; Table 6.2). However, no ewe shearing group produced consistently heavier or lighter weaners than the others each year, and the differences between groups were smaller than the variation in weaning weight between years.

The average growth rates from weaning (3 months old) to May (8 months old) of the weaner shearing groups ranged from 0.3–2.0 kg/month (Figure 6.2) but the groups only differed significantly amongst the weaners born in 2000. In that cohort, sheep shorn at weaning grew 0.5 kg/month faster than sheep that were not shorn at weaning (2.0 kg/month vs. 1.5 kg/month, respectively; P < 0.01).

The growth rates between May and December (8–15 months old) of shorn and unshorn sheep differed significantly (Table 6.3) in most cohorts but there was no consistent pattern to whether shorn or unshorn weaners grew faster. For example, MAY-MAY (shorn) weaners had the fastest or equal fastest growth rate in two cohorts out of four but had similar growth rates to all unshorn groups in the other two cohorts. Where the growth rates of shorn and unshorn sheep differed, the differences were less than or equal to 0.3 kg/month.

Table 6.2: Mean fleece-free bodyweight at weaning and 95% CI (kg) of progeny from each ewe shearing group Ewe shearing Weaner birth cohort group 1999 2000 2001 2002 DEC 15.6a (15.2–16.0) 16.7a (16.2–17.2) 18.9a (18.4–19.4) 15.2 (14.7–15.7) MAR 15.4ab (15.0–15.8) 17.4b (16.8–18.0) 17.6b (17.1–18.1) 15.6 (15.0–16.2) MAY 15.1b (14.7–15.5) 18.2c (17.7–18.7) 20.5c (20.0–21.0) 15.1 (14.5–15.7)

Within years, means with no or common superscripts do not differ significantly (P > 0.05)

Table 6.3: Mean growth rate and 95% CI (kg/month) between May and December (8–15 months old) of weaners that, in May, were shorn (MAY-MAY), or were carrying two (MAR- MAR) or five (DEC-DEC) months wool Weaner shearing Birth cohort group 1999 2000 2001 2002 MAY-MAY 2.6a (2.5–2.8) 2.6ab (2.4–2.8) 2.1 (2.0–2.2) 3.4b (3.3–3.5) MAR-MAR 2.4b (2.2–2.5) 2.8b (2.6–3.0) 2.1 (2.0–2.2) 3.3ab (3.2–3.4) DEC-DEC 2.5a (2.4–2.6) 2.5a (2.4–2.7) 2.0 (1.9–2.1) 3.1a (3.0–3.2)

Within years, means with no or common superscripts do not differ significantly (P > 0.05)

109 Results— Health & Mortality

Figure 6.2: Mean fleece-free bodyweight of weaner shearing groups between weaning and 15 months of age (December to December; bars indicate 95% CI)

1999-born * 50.0 45.0 40.0 35.0 30.0 25.0 20.0 15.0 Dec- Jan- Feb- Mar- Apr- May- Jun- Jul-00 Aug- Sep- Oct- Nov- Dec- 99 00 00 00 00 00 00 00 00 00 00 00

2000-born 50.0 45.0 40.0 35.0 30.0 25.0 20.0 15.0 Dec- Jan- Feb- Mar- Apr-01 May- Jun- Jul-01 Aug- Sep- Oct- Nov- Dec- 00 01 01 01 01 01 01 01 01 01 01

2001-born 50.0 45.0 40.0 35.0 30.0 25.0

Meanfleece-free bodyweight (kg) 20.0 15.0 Dec- Jan- Feb- Mar- Apr- May- Jun- Jul-02 Aug- Sep- Oct- Nov- Dec- 01 02 02 02 02 02 02 02 02 02 02 02

2002-born 50.0 45.0 40.0 35.0 30.0 25.0 20.0 15.0 Dec- Jan- Feb- Mar- Apr-03 May- Jun- Jul-03 Aug- Sep- Oct- Nov- Dec- 02 03 03 03 03 03 03 03 03 03 03 Date DEC-DEC MAR-MAR MAY-MAY JUN-MAR JUL-MAY OCT-DEC

* For weaners born in 1999, ‘DEC-DEC’ sheep were shorn in December then October and ‘OCT-DEC’ were shorn in December 2000 only

110 Results— Health & Mortality

6.26.26.2 Fleece RotFleece Rot

6.2.16.2.16.2.1 EwesEwesEwes Fleece rot was measured during the second half of the trial (July 2001–May 2004) and the proportion of ewes in each shearing group with severe fleece rot (score ≥3) during this time is shown in Table 6.4. The large majority of fleece rot observed was yellow and none was classed as having unscourable or severe colour. Severe fleece rot was most prevalent between December 2001 and May 2002. There was no difference between shearing groups in the odds of having severe fleece rot in July, October, December or March (P ≥ 0.10). This result was consistent from year to year, despite the large variations in fleece rot prevalence between years (homogeneity of odds P ≥ 0.05). In May, the prevalence of fleece rot was greater in unshorn MAY ewes than the other groups, which were similar to each other. In 2002 and 2003, the odds of MAY ewes having severe fleece rot were 2.5 (95% CI 1.6–4.0) times greater than the odds for ewes in the other shearing groups. The odds ratios could not be combined to include the 2004 results because no DEC or MAR ewes had severe fleece rot in May of this year.

Table 6.4: Prevalence of severe fleece rot (score ≥3) in ewe shearing groups Time of Ewe shearing group measurement DEC MAR MAY 2001 July 0.0% 0.0% 0.5% December 7.8% 10.0% 6.8% 2002 March 25.1% 21.3% 14.4% May 14.3% 13.2% 27.8% July 3.5% 5.0% 8.7% October 3.6% 5.7% 5.4% December 2.1% 1.6% 1.1% 2003 March 1.1% 1.1% 1.2% May 0.5% 0.0% 1.8% July 1.5% 0.5% 0.5% October 0.5% 0.0% 0.0% December 2.0% 2.5% 1.1% 2004 March 0.0% 4.1% 8.5% May 0.0% 0.0% 17.7%

111 Results— Health & Mortality

6.2.26.2.26.2.2 WeanersWeanersWeaners Detailed fleece rot observations between weaning and 20 months of age were made amongst weaners born in 2001 and 2002, and in 2000-born weaners from 12 months of age. As in the ewes, the observed fleece rot was yellow. It was only classed as having light, unscourable colour on one occasion, at the May 2003 shearing of 2001-born weaners. There was a high prevalence of severe fleece rot between December 2001 and June 2002, when 25–40% of weaners were affected, irrespective of shearing group and age (Figure 6.3). In contrast, there was low prevalence from October 2002 onwards amongst 2001- and 2002-born weaners.

There were no differences in the odds of shearing groups having severe fleece rot between weaning and 8 months old (December–May, inclusive), irrespective of whether weaners were shorn or unshorn at the time of observation. These similarities were consistent across weaner cohorts, despite the large difference in prevalence between years (homogeneity of odds P > 0.05; Table 6.5). In June, weaners shorn within the last three months (MAR-MAR and MAY-MAY) had approximately 90% lower odds of having severe fleece rot than weaners carrying six or nine months of wool (DEC-DEC and unshorn weaners, respectively). This effect was not observed in July or October but was observed in December, at 15 months of age, when weaners with 12 months wool (DEC-DEC) or short wool (JUL- MAY and OCT-DEC) had lower odds than weaners carrying six to nine months wool (MAR-MAR, MAY-MAY and JUN-MAR). The same groups, MAR-MAR, MAY-MAY and JUN-MAR, still had odds of severe fleece rot the following March and, of the three, the one not shorn in March (MAY- MAY) still had greater odds in May.

112 Results— Health & Mortality

Figure 6.3: Proportion of each weaner shearing group with severe fleece rot (score ≥3) between weaning and 20 months of age, amongst weaners born in 2000, 2001 and 2002

2000-born

45% 40% 35% 30% 25% 20% 15% 10% 5% 0% Jul-01 Apr-02 Apr-01 Oct-01 Jan-02 Jan-01 Jun-01 Feb-02 Mar-02 Feb-01 Mar-01 Nov-01 Sep-01 Aug-01 Dec-01 Dec-00 May-02 May-01

2001-born

45% 40% 35% 30% 25% 20% 15% 10% 5% 0% Jul-02 Apr-03 Apr-02 Oct-02 Jan-03 Jan-02 Jun-02 Feb-03 Mar-03 Feb-02 Mar-02 Nov-02 Sep-02 Aug-02 Dec-02 Dec-01 May-03 May-02

2002-born Proportionof shearing group with severe fleece rot

45% 40% 35% 30% 25% 20% 15% 10% 5% 0% Jul-03 Apr-04 Apr-03 Oct-03 Jan-04 Jan-03 Jun-03 Feb-04 Mar-04 Feb-03 Mar-03 Nov-03 Sep-03 Aug-03 Dec-03 Dec-02 May-04 May-03 Date

DEC-DEC MAR-MAR MAY-MAY JUN-MAR JUL-MAY OCT-DEC

113

Table 6.5: Odds ratios (95% CI) of weaner shearing groups having severe fleece rot, relative to unshorn weaners or weaners carrying 12 months wool, at different Mortality & Health Results— times between 6 and 20 months of age

Shearing March May June July October December March May 6 months old 8 m.o. 9 m.o. 10 m.o. 13 m.o. 15 m.o. 18 m.o. 20 m.o.

114 group DEC-DEC 0.9 (0.5–1.6) 0.6 (0.3–1.4) 0.6 (0.3–1.2) 0.6 (0.2–1.5) 0.6 (0.2–1.8) ref 0.1 (0.03–0.3) 0.1 (0.02–0.3) MAR-MAR ref 0.6 (0.3–1.3) 0.1 (0.04–0.5) 0.4 (0.1–1.2) 0.6 (0.2–1.9) 5.7 (2.4–13) ref 0.2 (0.1–0.4) MAY-MAY ref ref 0.1 (0.02–0.4) 0.4 (0.1–1.2)* 0.3 (0.1–1.2) 2.4 (1.0–5.9) 0.6 (0.3–1.1)* ref JUN-MAR ref ref ref 1.3 (0.6–2.8)* 0.3 (0.1–1.3) 2.4 (1.0–6.0) 0.6 (0.3–1.1)* 0.1 (0.03–0.3) JUL-MAY ref ref ref ref 0.4 (0.1–1.4)* 1.5 (0.6–3.9)* 0.2 (0.1–0.4)* 0.3 (0.2–0.7)* OCT-DEC ref ref ref ref ref 0.6 (0.2–2.3) 0.02 (0.00–0.2) 0

ref: referent group for odds ratios within column (unshorn for comparisons March–October; carrying 12 months wool for December–May) * odds not homogeneous across years

Results— Health & Mortality

6.36.36.3 Dag & Crutching

6.3.16.3.16.3.1 EwesEwesEwes Crutching occurred at variable times throughout the trial, depending on concern about the risk of breech strike and/or reduced wool quality. All ewes were crutched in December 2000, and DEC and MAR ewes were crutched prior to lambing in July 2001 and July 2003. The proportions of each ewe shearing group with no, mild and severe dag (dag scores 0, 1–2 and 3–5, respectively) are shown in

Figure 6.4: Proportion of each ewe shearing group with no (dag score (DS) 0), mild (DS 1–2) and severe (DS 3–5) dag (arrows indicate time of crutching or shearing)

DEC ↓ ↓ ↓ ↓ ↓ ↓ ↓ 100% 80% DS 3-5 60% DS 1-2

40% DS 0 20%

0% Jun-00 Jun-01 Jun-02 Jun-03 Mar-00 Mar-01 Mar-02 Mar-03 Mar-04 Dec-99 Sep-00 Dec-00 Sep-01 Dec-01 Sep-02 Dec-02 Sep-03 Dec-03 MAR ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ 100% 80% 60% 40% 20% 0% Jun-00 Jun-01 Jun-02 Jun-03 Mar-00 Mar-01 Mar-02 Mar-03 Mar-04 Dec-99 Sep-00 Dec-00 Sep-01 Dec-01 Sep-02 Dec-02 Sep-03 Dec-03 MAY ↓ ↓ ↓ ↓ ↓ ↓ 100%

80% Proportionshearing of withmildgroup no,dag severeand 60%

40%

20%

0% Jun-03 Jun-02 Jun-01 Jun-00 Mar-04 Mar-03 Mar-02 Mar-01 Mar-00 Sep-03 Dec-03 Sep-02 Dec-02 Sep-01 Dec-01 Sep-00 Dec-00 Dec-99 Date

115 Results— Health & Mortality

Figure 6.4. Severe dag was most prevalent in spring, when 10–50% of sheep in each shearing group had a dag score ≥3. Approximately 1.5–2 times more DEC ewes were affected by severe dag in spring than MAR or MAY. Based on dag scores from 2000–2003, the odds of DEC ewes having severe dag in October, at the beginning of the risk period for breech strike, were 2.5 (95% CI 1.8–3.5) and 1.9 (1.4–2.6) times greater than MAR and MAY ewes, respectively. There was no statistically significant difference between the odds of MAR and MAY ewes having severe dag in October (P = 0.16). These odds ratios were consistent across years (Wald test of treatment × year interaction, P = 0.08), despite the varying crutching treatments in different years of the trial.

Only eleven of 997 sheep throughout the entire trial had severe dag in March: eight MAR ewes in 2003, one MAY ewe in 2002 and two MAY ewes in 2003. Very few DEC ewes had any dag at all in March. As a result, there were insufficient numbers of sheep with severe dag for a reliable statistical analysis.

The estimated costs of dags from 2002–03, when ewes were only crutched at shearing, and 2003–04, when ewes were crutched prior to both lambing and shearing, are shown in Table 6.6. The cost of dags was greater in DEC ewes than in MAR or MAY at every crutching and the average dag cost of

Table 6.6: Mean cost of dags (cents/head) in each dag score category; mean cost in each shearing group, weighted by prevalence of each dag score category, in 2002–03 (only crutched prior to shearing) and 2003–04 (crutched prior to lambing and shearing); and mean total annual cost of dag Shearing Dag 2002–03 2003–04 Annual group score Shearing Pre-lamb Shearing mean DEC 0 77 56 74 1 100 79 83 2 142 110 118 3 179 162 171 4 206 206 - 5 - - - 91 80 Weighted mean 130 171 150

MAR 0 60 46 51 1 78 79 75 2 110 100 100 3 148 143 - 4 - 186 - 5 - 239 - 74 58 Weighted mean 75 132 103

MAY 0 48 57 1 73 64 2 100 84 3 121 103 4 - - 5 - - Weighted mean 69 60 64

116 Results— Health & Mortality

DEC ewes over these two years was 46% and 134% greater than MAR and MAY ewes, respectively. The cost of dags at the pre-shearing crutch was similar in MAR and MAY ewes but the absence of a pre-lamb crutch in MAY ewes meant that their annual dag cost was lower than MAR.

6.3.26.3.26.3.2 WeanersWeanersWeaners Comprehensive dag score measurements between weaning and 20 months of age were made in the weaner cohorts born in 2001 and 2002. Each weaner shearing group in these years contained 44–75 sheep. The seasonal changes in mean dag score were similar to the changes in severe dag (dag score ≥3), which are shown in Figure 6.5. The odds ratios of sheep having severe dag were consistent across years (Wald test of group × year interaction, P > 0.5), despite the prevalence differing between years. The odds of DEC-DEC weaners having severe dag in March, at 6 months of age, were 79% less than unshorn sheep (Table 6.7). No other groups were shorn at this time and they did not differ in their odds of having severe dag. In October, at 15 months old, the odds of JUL-MAY weaners having severe dag were 46% that of unshorn sheep. Other shorn groups had reduced odds of severe dag at this time, although the odds ratios between them and unshorn sheep did not differ significantly from one. The following March, at 18 months old, there was no difference in the odds of severe dag amongst MAR-MAR, MAY-MAY, JUN-MAR or JUL-MAY weaners, irrespective of the group with which the comparison was made. Neither of the recently shorn groups, DEC-DEC or OCT-DEC, had severe dag at this time.

Table 6.7: Odds (95% CI) of severe dag (score ≥3) in weaner shearing groups in March, October & March (6, 13 & 18 months old, respectively), relative to unshorn weaners or weaners carrying 12 months wool Shearing Odds ratio in (age in months): group March (6) October (13) March (18) DEC-DEC 0.21 (0.04–1.0) 1.4 (0.82–1.3) 0 MAR-MAR ref 0.80 (0.45–1.4) ref MAY-MAY ref 0.73 (0.43–1.2) 1.0 (0.35–3.0) JUN-MAR ref 0.62 (0.36–1.1) 1.2 (0.43–3.4) JUL-MAY ref 0.54 (0.31–0.93) 1.2 (0.40–3.3) OCT-DEC ref ref 0

ref: referent (unshorn) group for odds ratios within column

117 Results— Health & Mortality

Figure 6.5: Proportion of each weaner shearing group with severe dag (score ≥3) between weaning and 20 months of age, amongst weaners born in 2001 and 2002

2001-born

60% 55% 50% 45% 40% 35% 30% 25% 20% 15% 10% 5% 0% Dec- Jan- Feb- Mar- Apr- May- Jun- Jul- Aug- Sep- Oct- Nov- Dec- Jan- Feb- Mar- Apr- May- 01 02 02 02 02 02 02 02 02 02 02 02 02 03 03 03 03 03

2002-born

40% 35%

Proportionshearing of withgroup severe dag 30% 25% 20% 15% 10% 5% 0% Dec- Jan- Feb- Mar- Apr- May- Jun- Jul- Aug- Sep- Oct- Nov- Dec- Jan- Feb- Mar- Apr- May- 02 03 03 03 03 03 03 03 03 03 03 03 03 04 04 04 04 04 Date

DEC-DEC MAR-MAR MAY-MAY JUN-MAR JUL-MAY OCT-DEC

118 Results— Health & Mortality

6.46.46.4 FlystrikeFlystrikeFlystrike

6.4.16.4.16.4.1 EweEweEwesEwesss Detailed flystrike observations amongst ewes were recorded throughout November and early December (late spring) in all years except 2001 (Table 6.8). A greater proportion of DEC ewes developed flystrike at this time than MAR or MAY ewes in all years (P < 0.05). Nine per cent of DEC ewes were affected by flystrike in 2000 and all ewes were crutched at this time to control further strikes. The greatest prevalence occurred in 2002, when the incidence of strikes, which were mainly on the breech, exceeded the trial’s threshold for chemical treatment and so all ewes were jetted with cyromazine to prevent further strikes. The odds of DEC and MAR ewes having flystrike in late spring were 5.7 (95% CI 2.9–11.2) and 2.6 (1.3–3.6) times, respectively, greater than the odds for MAY ewes, and the odds of DEC ewes were 2.2 (1.3–3.6) times greater than MAR ewes. These odds ratios were consistent across the years of the trial (homogeneity of odds, P = 0.18), despite all DEC and MAR ewes being crutched in July in 2001 and 2003.

6.4.26.4.26.4.2 WeanersWeanersWeaners Comprehensive flystrike observations were recorded in the nominated weaner cohorts at the following times: May (8 months old): 2000- and 2001-drop; November (14 months old): 1999-, 2001- and 2002- drop; and, March (18 months old): 2000- and 2002-drop. The greatest overall prevalence (5.1% or 19/372) was recorded in March 2002 amongst 18-month old weaners (2000-drop), and crutching and ringing of wethers was performed in all JUL-MAY and MAY-MAY weaners at this time. The incidence of flystrike exceeded the treatment threshold in 2001- and 2002-drop weaners in November 2002 and December 2003, respectively, and so all sheep were jetted with cyromazine, as per the trial protocol.

The prevalence of flystrike in the weaner shearing groups in May (8 months old), November (14 months old) and March (18 months old) is shown in Tables 6.9, 6.10 and 6.11, respectively. No flystrike was observed in sheep carrying ≤5, 1, and 3 months wool in May, November and March, respectively. In November, across all shearing groups, 2.0% (10/493) of males and 3.7% (23/615) of females had flystrike, mostly around the breech, although there was no difference in the odds of either

Table 6.8: Flystrike prevalence in ewe shearing groups during November and early December Shearing Flystrike in November–December: group 1999 2000 2001 2002 2003 DEC 6.4% 8.9% n/a 12.7% 3.9% MAR 4.8% 3.9% n/a 6.2% 0.5% MAY 1.0% 3.7% n/a 1.6% 0.0%

n/a: not recorded

119 Results— Health & Mortality

sex being struck (P = 0.09). In March, the odds of male weaners having flystrike were 9.7 (95% CI 3.2–29) times greater than that of females. In autumn 2004, all cases of flystrike in 2002-drop weaners occurred in wethers around the pizzle. These began as covert strikes that were not detected until they had developed into more extensive and noticeable body strikes.

Table 6.9: Proportion of weaners in each shearing group affected by flystrike in May (8 months old) in two years of the trial Shearing Flystrike in May: group 2001 2002 DEC-DEC 0.0% 0.0% MAR-MAR 0.0% 0.0% unshorn 2.6% 1.1%

Table 6.10: Proportion of weaners in each shearing group affected by flystrike and odds of flystrike, relative to DEC-DEC weaners, in November (14 months old) in three years of the trial Shearing Flystrike in November: Odds ratio group 2000 2002 2003 DEC-DEC 0.0% 11.7% 8.0% ref MAR-MAR 5.0% 5.8% 0.0% 0.5 (0.2–1.1) MAY-MAY 0.0% 1.4% 2.2% 0.1 (0.03–0.6) JUN-MAR 2.0% 0.0% 5.2% 0.3 (0.08–0.8) JUL-MAY 2.4% 0.0% 4.2% 0.2 (0.07–0.8) OCT-DEC 5.4% 0.0% 0.0% 0

Table 6.11: Proportion of weaners in each shearing group affected by flystrike and odds of flystrike, relative to MAR-MAR weaners, in March (18 months old) in two years of the trial Shearing Flystrike in March: Odds ratio group 2002 2004 DEC-DEC 0.0% 0.0% 0 MAR-MAR 6.7% 6.6% ref MAY-MAY 11.3% 1.7% 1.0 (0.4–1.7) JUN-MAR 11.3% 6.1% 1.3 (0.5–3.4) JUL-MAY 1.6% 1.8% 0.2 (0.05–1.2) OCT-DEC 0.0% 1.5% 0.1 (0.01–0.9)

120 Results— Health & Mortality

6.56.56.5 Reproduction The proportions of ewes that were assessed as lactating, or ‘wet’, at marking throughout the trial are shown in Table 6.12. Overall, MAR ewes had lower odds of being wet at marking than DEC or MAY ewes, and the odds of MAY ewes were less than DEC, although the differences between MAR and MAY, and MAY and DEC were only significant at about the 90% level (P = 0.09 and 0.11, respectively). All of these results were consistent across years of the trial and the different ewe age groups examined (likelihood ratio tests of shearing group × year and shearing group × ewe age interactions P = 0.88 and 0.16, respectively). Very similar results were obtained from panel-data logistic regression accounting for repeated measurements across years and Mantel-Haenszel odds ratios combined across years. Despite the differences in the proportions of wet ewes in each shearing groups at marking, the percentage of lambs weaned from each ewe shearing group was similar (P = 0.39; Table 6.13). Similarly, each ewe shearing group weaned about one third of the total number of lambs weaned each year (P ≥ 0.06).

Table 6.12: Proportion of ewes in each shearing group lactating at lamb marking and odds (95% CI) of ewes lactating at marking, relative to DEC ewes Shearing % lactating in: Odds Ratio group 2000 2001 2002 2003 DEC 80% 80% 75% 56% ref MAR 79% 72% 64% 45% 0.67 (0.52–0.87) MAY 81% 77% 68% 49% 0.84 (0.65–1.1)

Table 6.13: Number of lambs weaned, number of ewes and weaning percentage in each year Shearing n(lambs) n(ewes) weaning % group 1999 2000 2001 2002 2003 DEC 219 189 116% 143 184 78% 181 197 92% 147 202 73% 105 205 51% MAR 244 196 124% 142 185 77% 183 201 91% 141 201 70% 77 197 39% MAY 194 190 102% 143 190 75% 181 201 90% 132 196 67% 83 193 43% P * 0.06 0.99 0.99 0.67 0.09 * P-value of χ2 goodness-of-fit test, performed within each year

121 Results— Health & Mortality

6.66.66.6 MortalityMortalityMortality

6.6.16.6.16.6.1 EwesEwesEwes Detailed observations of ewe survival recorded throughout years 3–5 of the trial (July 2001–May 2004) were included in the analysis. During this time, 26% of ewes in the adult flock died, which represented an average mortality rate of 8.5 deaths/1000 ewes/month or 10.2% per annum. The average mortality rate across shearing groups within 2–3 month periods each year varied from 0.6 to 22 deaths/1000 ewes/month, although it was sometimes greater within a shearing group (Figure 6.6). There were not great similarities to the patterns of deaths from year to year, apart from relatively high death rates amongst MAR and/or MAY ewes during March–May in both 2002 and 2003.

Figure 6.6: Cumulative mortality rates of ewes in each shearing group at different times of the year

2001-02 45.0 40.0 35.0 30.0 25.0 20.0 15.0 10.0 5.0 0.0 not July-Oct Oct-Dec Dec-March March-May Total measured

45.0 2002-03 40.0 35.0 30.0 25.0 20.0 15.0 10.0 5.0 0.0 May-July July-Oct Oct-Dec Dec-March March-May Total

2003-04 Cumulative mortality (deaths/1000 ewes/month) rate 45.0 40.0 35.0 30.0 25.0 20.0 15.0 10.0 5.0 0.0 May-July July-Oct Oct-Dec Dec-March March-May Total Observation period DEC MAR MAY

122 Results— Health & Mortality

The mortality risks of MAY and DEC ewes off shears, relative to unshorn sheep, were homogeneous across years (P > 0.7). The risk of MAY ewes dying after shearing was 3.3 (95% CI 1.4–8.0) times greater than unshorn sheep (i.e. DEC and MAR) and the attributable fraction for shearing amongst MAY ewes was 70% (95% CI 22–89%). In contrast, DEC ewes had a statistically similar risk of death after shearing to unshorn sheep (relative risk 0.6 (0.3–1.4)). The relative mortality risk of MAR ewes was not homogeneous across years or comparison groups. MAR ewes had a significantly greater risk of death than DEC ewes in all three years (relative risk 4.9 (1.7–14.3)) but their risk only tended to be greater than MAY ewes in two of three years (relative risk 1.7 (0.8–3.6)). In contrast, MAR ewes had a smaller mortality risk than MAY ewes in 2002 (relative risk 0.16 (0.04–0.53)) but a similar risk of death as DEC ewes. Between March and May, ‘autumn’-shorn ewes (i.e. MAR or MAY) had 5.4 (2.0–21) times the risk of death of December-shorn ewes and the attributable fraction of deaths was 0.81 (0.49–0.95). No groups were shorn between July and December and there were no statistically significant differences between the mortality risks of the groups during this time.

123 Results— Health & Mortality

6.6.26.6.26.6.2 WeanersWeanersWeaners Extra data from a central progeny test involving weaners born on Farm A between 1996 and 2000, inclusive, were recruited into the weaner survival analysis. In the following results, the central progeny test is referred to as Trial 1 and the main trial is referred to as Trial 2.

6.6.2.1 Summary Mortality Data In total, 3655 lambs from nine birth cohorts over seven years were weaned and 520 (14.2%) of these died during the post-weaning period (PWP). This represented a cumulative mortality rate of 14.9 deaths per 1000 weaners per month. The median length of the PWP was 332 (range 288–371) days and the median censoring interval length was 96 (31–238) days. The Turnbull survivorship functions of each year-trial cohort are shown in Figure 6.7, and the number of lambs weaned in a cohort and the timing of deaths during the PWP are detailed in Figure 6.8. In all cohorts except Trial 2-2002, the incidence of deaths was greatest early in the PWP and decreased as the time since weaning increased. The proportion of weaners that died and the survivorship functions differed between year-trial cohorts (χ2, log-rank and Wilcoxon tests P < 0.001), with fewest deaths amongst weaners in Trial 1 born in 1996 (total mortality 4.5%, cumulative mortality rate 4.0 deaths/1000/month) and the most amongst those in Trial 2 born in 2001 (26.8%, 34.7 deaths/1000/month). There were statistically significant differences between cohort survivorship functions but there appeared to be no association between mortality and the number of weaners in a cohort.

Figure 6.7: Turnbull survival estimates of year-trial cohorts

Birth 100 year (trial) 1996 (1) 95 1997 (1) 1998 (1) 1999 (1) 90 1999 (2) 2000 (1) 2000 (2) 2001 (2) 85 2002 (2) Percent

70 0 100 200 300 400 Days post-weaning

124

Figure 6.8: Number of lambs weaned and deaths in each censoring interval of the PWP, by year-trial cohort (timelines not to scale) Birth Trial Weaning Number Total Per cent year date weaned deaths mortality

1996 1 15/11/1996 357 Deaths | 12 | 4 | 16 4.5% Days post- 0 238 332 weaning

1997 1 18/11/1997 309 | 19 | 8 | 3 | 30 9.7% 0 107 265 371

Mortality & Health Results— 1998 1 16/11/1998 330 | 24 | 2 | 0 | 4 | 30 9.1% 0 122 224 273 354

1999 1 29/11/1999 398 | 75 | 19 | 1 | 95 23.9% 125 0 171 295 366

2 1/12/1999 658 | 45 | 10 | 55 8.4% 0 168 288

2000 1 11/12/2000 214 | 16 | 0 | 1 | 17 7.9% 0 137 210 353

2 5/12/2000 438 | 31 | 4 | 0 | 35 8.0% 0 143 217 314

2001 2 18/12/2001 544 | 116 | 21 | 6 | 2 | 1 | 146 26.8% 0 79 141 174 205 301

2002 2 9/12/2002 407 | 37 | 4 | 42 | 10 | 2 | 1 | 96 23.6% 0 63 99 155 190 224 303

Results— Health & Mortality

6.6.2.2 Mortality & Bodyweight Mean weaning weight differed between years (P < 0.001; Table 6.14) but the coefficient of variation of weaning weight was about 0.2 in all years, regardless of the mean. Weaners in Trial 2 born in 2002 were lightest (15.3 kg; 23.6% mortality) and those in Trial 2 in 2001 heaviest (18.9 kg; 26.8% mortality). Across all cohorts, weaners that died during the PWP were an average of 1.9 kg lighter at weaning than survivors (P < 0.001).

The cumulative mortality rate and relative risk of the five weaning weight quintiles of each year-trial cohort are listed in Table 6.15. Sixteen weaners did not have a recorded weaning weight. The mortality risks of individual weaning weight quintiles relative to the middle quintile were homogeneous between year-trial cohorts (P = 0.3–0.8, depending on the quintile being compared). The risk of death of weaners in the lightest weaning weight quintile was 3.5 (2.7–4.6) times greater than the middle quintile and 3.4 (2.9–4.1) times greater than the mortality risk of all other weaners. The attributable fraction of mortality amongst weaners in the lightest weight quintile was 0.71 (0.65– 0.76) and the population attributable fraction of this weight group was 0.31. Weaners in the second lightest quintile experienced 1.5 (1.1–2.0) times the mortality risk of the middle quintile. The mortality risk of weaners in the two heaviest quintiles was not significantly different from that of the middle quintile.

There were statistically significant differences between levels of the Turnbull survivorship functions for weaning weight quintiles (P < 0.01; Figure 6.9), with greatest mortality amongst weaners of light weaning weight.

Table 6.14: Mean weaning weight and 95% CI (kg) of weaners dying and surviving during the post-weaning period in each year-trial cohort Birth Trial Weight year Died Survived 1996 1 13.4 (11.1–15.6) 16.0 (15.6–16.3) 1997 1 17.7 (16.4–19.1) 19.0 (18.6–19.4) 1998 1 13.1 (11.8–14.4) 16.7 (16.4–17.0) 1999 1 13.7 (13.0–14.4) 16.4 (16.0–16.7) 2 14.0 (13.0–14.9) 15.5 (15.2–15.7) 2000 1 12.8 (10.7–14.8) 15.7 (15.3–16.2) 2 14.7 (13.5–16.0) 17.7 (17.4–18.0) 2001 2 17.4 (16.8–18.1) 19.5 (19.1–19.8) 2002 2 13.6 (12.9–14.2) 15.8 (15.4–16.1)

Overall 15.0 (14.6–15.3) 16.9 (16.8–17.0)

126 Results— Health & Mortality

Figure 6.9: Turnbull survival estimates of within-cohort weaning weight quintiles

Quintile grouping 100 1 (lightest) 2 3 95 4 5 (heaviest) 90

80 Percent survival 75

65 0 100 200 300 400 Days post-weaning

127 Results— Health & Mortality

Table 6.15: Number of weaners in each weaning weight quintile (range in kg) that died and were at risk during the post-weaning period, total mortality, cumulative mortality rate, and mortality risk (95% CI), relative to the middle quintile Weaning Total Cumulative Birth Trial n weight mortality mortality rate Relative risk year died total quintile (%) (/1000 / month)

OVERALL Quintile 1 224 758 29.6 35.7 (31.3–40.7) 3.5 (2.7–4.6) 2 109 738 14.8 15.6 (12.9–18.8) 1.5 (1.1–2.0) 3 79 723 10.9 11.1 (8.9–13.8) ref 4 57 744 7.7 7.6 (5.9–9.9) 0.7 (0.5–1.0) 5 50 676 7.4 7.2 (5.5–9.6) 0.7 (0.5–1.0) TOTAL 519 3639 14.3 14.9 (13.7–16.2)

1996 1 7.0–13.1 10 71 14.1 14.1 (7.6–26.2) 10.4 (1.5–449) 13.1–15.0 1 74 1.4 1.3 (0.2–8.9) 0.9 (0.01–72.1) 15.0–16.5 1 68 1.5 1.4 (0.2–9.7) ref 16.5–18.6 1 71 1.4 1.3 (0.2–9.3) 1.0 (0.0–75.1) 18.6–28.3 2 70 2.9 2.6 (0.7–10.5) 1.9 (0.1–114) Total 15 354 4.2 4.0 (2.4–6.6)

1997 1 10.2–16.2 9 63 14.3 13.4 (7.0–25.7) 2.4 (0.7–10.8) 16.2–18.2 9 61 14.8 13.4 (7–25.8) 2.4 (0.7–10.8) 18.2–20.1 4 62 6.5 5.5 (2.1–14.7) ref 20.1–21.7 4 62 6.5 5.5 (2.1–14.7) 1.0 (0.2–5.4) 21.7–26.7 4 61 6.6 5.6 (2.1–14.9) 1.0 (0.2–5.5) Total 30 309 9.7 8.6 (6.0–12.2)

1998 1 8.5–13.7 17 66 25.8 28.2 (17.5–45.4) 7.0 (2.0–37.3) 13.7–15.6 7 69 10.1 9.8 (4.7–20.6) 2.4 (0.6–14.6) 15.6–17.1 3 69 4.3 4.0 (1.3–12.5) ref 17.1–19.0 1 61 1.6 1.3 (0.2–9.3) 0.3 (0.0–4.0) 19.0–26.3 2 65 3.1 2.6 (0.7–10.6) 0.7 (0.1–5.7) Total 30 330 9.1 8.4 (5.9–12.0)

1999 1 7.1–12.5 42 80 52.5 68.8 (50.8–93) 6.7 (3.2–15.6) 12.5–14.5 23 79 29.1 30.4 (20.2–45.8) 3.0 (1.3–7.3) 14.5–16.5 9 79 11.4 10.3 (5.4–19.8) ref 16.5–19.0 11 80 13.8 12.7 (7–22.9) 1.2 (0.5–3.4) 19.0–27.9 10 78 12.8 11.5 (6.2–21.4) 1.1 (0.4–3.1) Total 95 396 24.0 23.9 (19.5–29.2)

2 5.5–12.5 22 142 15.5 18.2 (12–27.6) 4.6 (0.9–44.4) 12.5–14.5 12 143 8.4 9.4 (5.3–16.5) 2.7 (0.4–28.3) 14.5–16.0 8 111 7.2 8.0 (4–16) ref 16.0–18.5 10 157 6.4 7.0 (3.8–13.1) 0.5 (0.01–9.6) 18.5–24.5 3 104 2.9 3.1 (1–9.6) 0.5 (0.01–9.8) Total 55 657 8.4 9.4 (7.2–12.2)

2000 1 7.5–12.5 8 45 17.8 18.9 (9.5–37.8) 2.3 (1.0–5.9) 12.5–14.5 5 44 11.4 11.1 (4.6–26.6) 1.2 (0.4–3.3) 14.5–16.1 2 41 4.9 4.1 (1–16.5) ref 16.1–18.9 1 43 2.3 2.0 (0.3–14.5) 0.9 (0.3–2.6) 18.9–24.0 1 41 2.4 2.1 (0.3–14.9) 0.4 (0.1–1.6) Total 17 214 7.9 7.3 (4.5–11.8)

2 7.8–14.4 16 86 18.6 20.9 (12.8–34.1) 4.4 (1.4–18.1) 14.4–16.5 9 86 10.5 11.0 (5.7–21.2) 2.3 (0.7–10.4) 16.5–18.2 4 85 4.7 4.7 (1.8–12.6) ref 18.2–20.2 5 89 5.6 5.7 (2.4–13.6) 1.2 (0.3–6.0) 20.2–27.1 1 82 1.2 1.2 (0.2–8.5) 0.3 (0.0–2.5) Total 35 428 8.2 8.4 (6.1–11.8)

2001 2 10.0–15.5 57 112 50.9 89.0 (68.7–115.4) 3.2 (2.0–5.1) 15.5–17.5 25 107 23.4 29.3 (19.8–43.4) 1.0 (0.6–1.8) 17.5–20.0 30 133 22.6 28.0 (19.5–40) ref 20.0–22.5 15 92 16.3 18.9 (11.4–31.4) 0.7 (0.3–1.3) 22.5–35.0 19 100 19.0 22.4 (14.3–35.1) 0.8 (0.4–1.5) Total 146 544 26.8 34.7 (29.5–40.8)

2002 2 6.5–12.5 43 95 45.3 66.5 (49.3–89.6) 2.4 (1.3–4.4) 12.5–14.5 18 79 22.8 28.0 (17.6–44.4) 1.0 (0.5–2.0) 14.5–16.0 18 76 23.7 28.1 (17.7–44.6) ref 16.0–18.0 9 84 10.7 11.7 (6.1–22.4) 0.4 (0.2–1.0) 18.0–24.5 8 73 11.0 11.8 (5.9–23.6) 0.4 (0.2–1.0) Total 96 407 23.6 28.4 (23.2–34.7)

128 Results— Health & Mortality

6.6.2.3 Mortality & Mean Seasonal Growth Rate Weaners in each cohort differed in their average growth rate during day 0–150 (i.e. summer–autumn;

‘GRearly’), 150–225 (winter; ‘GRmid’) and 225–end (early spring; ‘GRlate’) of the PWP (P < 0.01; Table 6.16). When the cohort survivorship functions (Figure 6.7) were grouped into quintiles according to growth rate, there were statistically significant differences between them (log-rank and

Wilcoxon tests P = 0.02, 0.06 and 0.00 for GRearly, GRmid and GRlate, respectively). The two cohorts with greatest overall mortality (Trial 2-2001 and Trial 2-2002) had the smallest values of GRearly but beyond this there was no obvious association between the seasonal growth rate of a cohort and its mortality.

Table 6.16: Average weaner growth rate (kg/month; ‘GR’) at different times throughout the post-weaning period (PWP) and total mortality Birth Trial Average growth rate (95% CI) Mortality year GRearly GRmid GRlate (day 0–150) (day 150–225) (day 225–end PWP) 1996 1 0.78 (0.74–0.81) 0.78 (0.74–0.81) 4.41 (4.33–4.50) 4.5% 1997 1 0.65 (0.56–0.73) 1.57 (1.49–1.64) 2.61 (2.50–2.72) 9.7% 1998 1 0.93 (0.87–0.99) 1.78 (1.72–1.83) 3.10 (3.00–3.20) 9.1% 1999 1 1.65 (1.58–1.73) 1.17 (1.08–1.27) 1.17 (1.08–1.27) 23.9% 2 1.45 (1.41–1.50) 1.56 (1.49–1.63) 1.56 (1.49–1.63) 8.4% 2000 1 2.11 (2.02–2.20) 1.56 (1.42–1.70) 1.51 (1.42–1.60) 7.9% 2 1.53 (1.45–1.61) 3.65 (3.53–3.76) 2.76 (2.63–2.89) 8.0% 2001 2 0.27 (0.22–0.32) 0.31 (0.21–0.49) 3.20 (3.13–3.28) 26.8% 2002 2 0.31 (0.27–0.36) 2.80 (2.73–2.88) 3.97 (3.89–4.05) 23.6%

Total 1.06 (1.03–1.09) 1.69 (1.64–1.74) 2.67 (2.62–2.72) 14.2%

129 Results— Health & Mortality

6.6.2.4 Mortality and sex Mortality could not be analysed in weaners in Trial 2 born in 1999 and 2000 because the sex of weaners that died was not recorded. Overall mortality amongst male weaners was greater than amongst females, with males experiencing 1.4 (1.1–1.6) times the risk of death of females (log-rank and Wilcoxon tests P < 0.01; Figure 6.10 and Table 6.17). The relative risks of males and females were homogeneous between cohorts (P = 0.7). The attributable fraction of mortality amongst males was 0.21 (0.04–0.35) and the population attributable fraction of males was 0.11.

Figure 6.10: Turnbull survival estimates of males and females

Sex 100.0 male female

97.5

95.0

92.5

Percent survival 90.0

87.5

85.0 0 100 200 300 400 Days post-weaning

130 Results— Health & Mortality

Table 6.17: Total mortality, cumulative mortality rate (deaths/1000 weaners/month) and mortality risk of males relative to females

Birth Trial Sex n Total Cumulative mortality Relative risk of year died total mortality rate (95% CI) males (95% CI)

OVERALL male 235 1262 18.6% 19.5 (17.1–22.1) 1.4 (1.1–1.6) female 194 1293 15.0% 15.3 (13.3–17.6)

1996 1 male 7 177 4.0% 3.7 (1.8–7.8) 0.8 (0.2–2.4) female 9 180 5.0% 4.7 (2.5–9.1)

1997 1 male 17 160 10.6% 9.4 (5.9–15.2) 1.3 (0.6–3.0) female 12 147 8.2% 7.1 (4.0–12.5)

1998 1 male 20 171 11.7% 10.9 (7.1–16.9) 1.9 (0.9–4.6) female 10 159 6.3% 5.7 (3.1–10.6)

1999 1 male 53 192 27.6% 28.3 (21.6–37.0) 1.4 (0.9–2.2) female 42 205 20.5% 19.9 (14.7–26.9)

2000 1 male 11 115 9.6% 8.9 (4.9–16.1) 1.6 (0.5–5.3) female 6 99 6.1% 5.5 (2.5–12.3)

2001 2 male 82 261 31.4% 42.7 (34.4–53.0) 1.5 (1.1–2.1) female 64 282 22.7% 28.2 (22.0–36.0)

2002 2 male 45 186 24.2% 29.5 (22.0–39.5) 1.1 (0.7–1.6) female 51 221 23.1% 27.5 (20.9–36.2)

131 Results— Health & Mortality

6.6.2.5 Mortality & Weaner Shearing Time There was no difference between survivorship curves of progeny from the three ewe shearing groups (P = 0.5), despite the small differences in mean weaning weights of offspring of the ewe shearing groups discussed in Section 6.1.2. In the univariate analyses, statistically significant changes in mortality risk after shearing occurred in March and May, when shorn weaners had 1.6 (1.0–2.6) and 3.2 (1.3–7.6) times, respectively, the risk of death of unshorn sheep (Table 6.18). All univariate mortality rate differences and ratios were homogeneous between cohorts (P = 0.07–0.5, depending on shearing group). The multivariate Cox and cubic spline survival models estimated that shearing in March, May or June increased mortality risk (Table 6.19). In both models, March shearing increased mortality risk up to six-fold, depending on average growth rate during summer–autumn, although there was no difference between the risk of shorn and unshorn sheep when GRearly was less than about 0.4 kg/month (Figure 6.11). Examples of the mortality rates estimated by the cubic spline model of unshorn and March-shorn female weaners, with two different values of GRearly and median values of other covariates, are shown in Figure 6.12. The Cox model estimated that shearing in May and June increased mortality risk 3.9 (1.8–8.2) and 4.3 (1.3–13.9) times, respectively, independent of growth rate. In contrast, the cubic spline model estimated that the increase in risk after May shearing depended on mid-PWP growth rate and became greater as GRmid increased (Figure 6.11) but that mortality risk did not significantly increase after shearing in June.

9.0 8.0 7.0 6.0 5.0 4.0 3.0

Mortality rate ratio 2.0 1.0 0.0 0 0.25 0.5 0.75 1 1.25 1.5 1.75 2

GR(early) or GR(mid) (kg/month)

March-shorn vs. unshorn at GR(early) May-shorn vs. unshorn at GR(mid) March-shorn vs. unshorn at GR(early)

132 Results— Health & Mortality

Table 6.18: Mortality rate (deaths/1000 weaners/month) of shorn and unshorn weaners, mortality rate difference, and mortality rate ratio between shorn and unshorn weaners at each shearing

Birth year Shearing Period observed Mortality rate (95% CI) group after shearing Shorn Unshorn Difference Ratio (days post-weaning)

OVERALL DEC - 29.9* 29.9 -4.0* (-10.7–2.8) 1.0 (0.7–1.5) MAR - 39.4 24.1 10.2 (-5.6–26.1) 1.6 (1.0–2.6) MAY - 16.6 5.8 1.5 (-4.3–7.3) 3.2 (1.3–7.6) JUN - 4.8 2.1 0.3 (-4.3–5) 2.7 (0.5–15.2) JUL - 1.4 1.3 -1.2 (-7.5–5.2) 1.1 (0.0–85)

1999 DEC 0–168 4.9 13.9 -9.0 (-16.1– -2) 0.4 (0.1–1.1) MAR 118–168 10.5 14.2 -3.8 (-20.5–12.9) 0.7 (0.1–3.3) MAY 168–288 2.8 3.1 -0.4 (-7.3–6.6) 0.9 (0–17) JUN 188–288 0.0 2.3 -2.3 (-6.7–2.2) 0.0 (0–41.8) JUL 226–288 0.0 3.0 -4.9 (-14.5–4.7) 0.0 (0–43.5)

2000 DEC 0–143 15.3 13.4 1.9 (-12.7–16.5) 1.1 (0.3–3.1) MAR 99–143 52.1 13.8 37.2 (-10–84.4) 3.5 (0.8–13.7) MAY 168–217 0.0 2.0 -2.0 (-5.9–1.9) 0.0 (0–175.4) JUN 182–217 0.0 2.0 -4.2 (-12.6–4.1) 0.0 (0–121) JUL 217–314 0.0 0.0 0.0 (0–0) -

2001 DEC 0–79 120.3 86.8 33.5 (-17.8–84.8) 1.4 (0.8–2.2) MAR 79–141 20.6 26.1 -5.5 (-32.3–21.4) 0.8 (0.1–2.8) MAY 141–174 26.4 18.3 8.1 (-32.7–48.9) 1.4 (0.1–10.1) JUN 174–206 29.8 0.0 29.8 (-11.5–71.1) - JUL 260–301 0.0 0.0 0.0 (0–0) -

2002 DEC 0–63 44.0 46.5 -2.5 (-41.4–36.3) 0.9 (0.3–2.3) MAR 99–155 107.9 51.4 57 (-10.8–124) 2.1 (0.9–4.5) MAY 155–190 125.8 10.3 116 (21.2–210) 12.2 (2.3–120) JUN 190–224 14.8 0.0 14.8 (-14.2–43.9) - JUL 244–303 7.8 0.0 7.8 (-7.5–23.1) -

* Risk differences are calculated from weighted cohort risks and so vary slightly from the ‘crude’ differences between overall absolute mortality risks, which are calculated from unweighted cohort risks.

133 Results— Health & Mortality

Figure 6.12: Effect of shearing in March on death rates of female weaners with GRearly = 1 kg/month (──) or 2 kg/month (─ ─), estimated by the cubic spline model*

Estimated weaners/month) mortality (/1000 rate 0 100 200 300 400 Days post-weaning

GRearly = 1 kg/month: unshorn March shorn GRearly = 2 kg/month: unshorn March shorn

*Assuming median values of weaning weight, GRmid and GRearly (17 kg, 2 kg/month & 3 kg/month, respectively)

134 Results— Health & Mortality

6.6.2.6 Multivariate Survival Analyses The final, statistically significant coefficients of continuous covariates developed for each multivariate model and their transformations, if applicable, are listed in Table 6.19. Generally, the same covariates were statistically significant and of similar value in each of the different survival analyses. Increasing weaning weight and GRearly were both associated with decreased mortality risk (Figures 6.13 and 6.14, respectively). Bodyweight measured at any time throughout the PWP, when included in models as a time-varying covariate in place of weaning weight, GRearly, GRmid and GRlate, had a similar association with mortality rate as weaning weight (Figure 6.13). The coefficient for GRmid was only statistically significant for the Weibull and log-logistic models, where increasing GRmid by 0.25 kg/month equated to a survival time ratio of approximately 1.2 (i.e. the average survival time was 20% longer in weaners growing at the faster rate) but in the models of Trial 2-only data, the risk ratio for a 0.25 kg/month

Table 6.19: Coefficients of statistically significant terms in Weibull, Cox, cubic spline, log- logistic and interval-censored Weibull survival analysis models, in analyses of a) all data, and b) Trial 2 data only

Covariate* Coefficient (standard error) Interval cens- Weibull Cox Cubic spline Log-logistic ored Weibull a) All data:

loge (WWT) -2.72 (0.18) -2.65 (0.18) -2.61 (0.17) -2.74 (0.19) WWT-1 -3.21 (0.23) -1 GRearly 0.65 (0.04) 0.67 (0.04) 0.66 (0.04) -0.56 (0.04) 0.61 (0.04)

GRmid -0.74 (0.11) 0.63 (0.08) n/a

GRlate 0.27 (0.17) -3.72 (1.11) 0.87 (0.18) -0.21 (0.13) n/a MAL 0.50 (0.10) 0.48 (0.10) 0.49 (0.10) -0.42 (0.09) 0.49 (0.10) -1 (GRearly )×GRlate -0.37 (0.12) -0.41 (0.10) 0.30 (0.09)

loge(WWT)×GRlate 0.87 (0.36)

b) Trial 2 (shearing time trial) data only: WWT -0.17 (0.02) -0.17 (0.02) -1 GRearly 1.29 (0.08) 1.25 (0.08)

GRmid 0.34 (0.15)

GRmid×WWT -0.03 (0.01)

GRlate -1.95 (0.44) -1 (GRearly) ×GRlate -0.21 (0.06) MAL 0.45 (0.13) 0.48 (0.13) MAR 1.68 (0.43) 2.14 (0.42) -1 MAR×GRearly -0.51 (0.15) -0.63 (0.15) MAY 1.35 (0.39)

MAY×GRmid 0.72 (0.17) JUN 1.45 (0.60)

* WWT: weaning weight (kg); GRearly, GRmid, GRlate: mean growth rate (kg/month) during early, middle and late parts, respectively, of the post-weaning period; MAL: male sex; MAR, MAY, JUN: shorn in March, May and June, respectively n/a: not applicable—techniques are not readily available for analysing time-varying covariates in interval- censored survival data

135 Results— Health & Mortality

increase in GRmid amongst unshorn sheep was not statistically significant. The risk ratio of males compared to females was approximately 1.6, irrespective of the model used.

The shape of the mortality rate functions in the Weibull and cubic spline models was similar to the smoothed average mortality rate graph derived directly from the data, indicating that they fitted the data quite well (Figure 6.15). In the Cox model, the scaled Schoenfeld residuals graphs for all

Figure 6.13: Mortality rate ratio for a 2 kg increase in weaning weight or time-varying bodyweight (BWT) from the labelled value, estimated by Weibull, Cox and cubic spline models (bars indicate 95% CI)

1.00 0.95 0.90 0.85 0.80 0.75 0.70 0.65 Mortality rate ratio rate Mortality 0.60 0.55 0.50 6 8 10 12 14 16 18 20 22 24 26 28 Weight (kg)

Weibull Cox Cubic spline Cox - BWT model

Figure 6.14: Mortality rate ratio for a 0.25 kg/month increase in GRearly from the labelled value for Weibull, Cox and cubic spline models (bars indicate 95% CI)

1.0 0.9 0.8 0.7 0.6 0.5 0.4 0.3 Mortality rate ratio rate Mortality 0.2 0.1 0.0 0 0.25 0.5 0.75 1 1.25 1.5 1.75 2

GRearly (kg/month)

Weibull Cox Cubic spline

136 Results— Health & Mortality covariates had approximately zero slope and the partial likelihood ratio tests of the interaction terms -1 with loge t were not significant, except for GRearly (Figure 6.16). This suggested that the proportional hazards assumption of the models was largely satisfied.

Figure 6.15: Smoothed mortality rate estimate from all data, and estimated Weibull & cubic spline hazard functions for female weaners with median covariate values (weaning weight 17 kg; GRearly, GRmid and GRlate = 0.9, 1.6 and 2.8 kg/month, respectively)

mortality rate (deaths/1000/month) mortality 0 6 12 18 24 24 30 18 12 0 6 0 100 200 300 400 time post-weaning (days)

Smoothed hazard function 95% CI Weibull cubic spline

Figure 6.16: Smoothed curve of the scaled Schoenfeld residuals for growth rate early in the post-weaning period (GRearly) vs. time since weaning

2 1 -

early

- 1 Scaled Schoenfeld residualsGR of 100 200 300 400 Days since weaning

137 Results— Health & Mortality

6.76.76.7 Value of Lifetime Wool Production, Accounting for Survivorship Table 6.20 shows the percentage survival from one shearing to the next and the cumulative survival prior to each shearing up until 6.25 years of age. When weaner and adult ewe survivorship was combined, the OCT-DEC shearing pattern had the most favourable cumulative survival, followed by DEC-DEC, JUN-MAR, MAR-MAR, JUL-MAY and MAY-MAY, consecutively. When this was included in the calculation of the value of wool produced over a sheep’s commercial lifetime (i.e. birth to 6.25 years old), the JUN-MAR and OCT-DEC shearing patterns had the equal greatest value ($111/head), followed by MAR-MAR ($107/head), DEC-DEC ($103/head) and JUL-MAY ($96/head), with MAY-MAY producing the smallest value ($93/head), irrespective of the size of staple length and strength discounts (Figure 6.17). At high historical micron basis prices, JUN-MAR shearing had a greater lifetime value, whereas JUN-MAR and OCT-DEC were similar at low historical micron basis prices (Appendix 2, Figure A-2.5).

Table 6.20: Mean annual (Ann.) and cumulative (Cum.) survival prior to each shearing of sheep in each shearing group between birth and 6 years old (y.o.)

Shearing group Weaner 1 Weaner 2 Maiden 3 y.o. 4 y.o. 5 y.o. 6 y.o. DEC-DEC Ann. 100% 77% 98% 92% 92% 92% 92% Cum. 100% 77% 75% 69% 64% 59% 54% OCT-DEC Ann. 85% 91% 92% 92% 92% 92% Cum. 85% 78% 72% 66% 60% 56% MAR-MAR Ann. 89% 88% 98% 90% 90% 90% 93% Cum. 89% 78% 77% 69% 62% 56% 52% JUN-MAR Ann. 85% 97% 95% 90% 90% 90% 93% Cum. 85% 82% 78% 70% 63% 57% 53% MAY-MAY Ann. 86% 88% 92% 88% 88% 88% 94% Cum. 86% 76% 70% 62% 54% 47% 45% JUL-MAY Ann. 85% 94% 87% 88% 88% 88% 94% Cum. 85% 80% 69% 61% 53% 47% 44%

138 Results— Health & Mortality

Figure 6.17: Per head value of wool production of each shearing group to 6.25 years of age, accounting for cumulative survival, at median historical micron basis prices using first, median and ninth decile historical staple length and strength premiums and discounts

$130 $120 $110 $100 $90 $80 $70 $60 $50 $40 $30 $20 $10

Net wool value * survivorship ($/head) *survivorship value wool Net $0 -$10 DEC-DEC OCT MAR-MAR JUN-MAR MAY-MAY JUL-MAY

DEC MAR MAY Weaner shearing group Ewe shearing group

1st weaner 2nd weaner Maiden Adult shearings Total value using 1st OR 9th decile L&S discounts

139

140

C H A P T E R 7

DISCUSSION—SHEEP HEALTH & MORTALITY

7.17.17.1 Body Condition & Growth Rate

7.1.17.1.17.1.1 EwesEwesEwes This study provides evidence that ewes shorn in May are of lighter body condition during winter than ewes shorn at other times. May-shorn ewes were about 0.4 of a condition score lighter than December-shorn ewes during winter in three out of the four years when condition score was measured. These differences were observed despite pasture availability during winter in the trial paddocks being generally greater than the pasture availabilities of the surrounding paddocks on the commercial farm. It is therefore possible that there was sufficient winter pasture availability for May- shorn ewes to increase their feed intake following shearing. Thus, it is difficult to compare these results to those of another study in which pregnant ewes shorn in May and offered a ration estimated as maintenance for unshorn sheep lost about 1 kg/week (Elvidge and Coop 1974). On the other hand, in the present study all ewes were approximately 13 weeks pregnant in July, and May shearing might not have increased ewe feed requirements relative to the other groups because feed intake, and concurrent heat production, was already starting to increase with pregnancy. Running ewes in lighter body condition during winter will have important economic consequences. It is likely that May-shorn ewes would either require supplementary feeding during winter more frequently or, alternatively, the ewes would have to be managed at a lower stocking rate during this period. In either case, farm profitability would be detrimentally affected.

The reason why December-shorn ewes were in lighter body condition than other sheep during spring, a time when pasture availability is abundant, is not clear. There was a greater prevalence of flystrike in December-shorn ewes during spring than the other shearing groups, and it is possible that the lighter body condition reflected a greater prevalence of sub-clinical strikes in this group.

During this study, all the shearing groups were grazed together. This design may have introduced a bias to the results. For example, shearing may have increased one group’s energy requirement and pasture intake, which may have reduced pasture availability, feed intake and body condition of the other groups. The direction of such a bias will always be towards the null hypothesis. That is, without the potential bias, larger differences between groups would be observed. Ideally, treatments would have been replicated in paddocks of different, controlled, pasture availability but such a design is prohibitive in terms of scale and cost, and consequently has never been conducted.

141 Discussion—Health & Mortality

7.1.27.1.27.1.2 WeanersWeanersWeaners Similar to other studies in Merinos and crossbred sheep, this trial found no consistent effect of shearing time on weaner growth rate (Tucker 1964; Sumner 1984; Donnelly 1991a; Irving 1991; Rogan et al. 1995; R. Thompson pers. comm.). In only one year, 2000, was a difference in growth rate observed between treatments, when sheep shorn at weaning grew faster than unshorn weaners. Although the reason for this faster growth is unclear, in this year the sheep shorn at weaning were significantly lighter than the other shearing groups. Thus, the effect of shearing is potentially confounded with compensatory growth. It is interesting to note that there was no observed difference in growth rate in weaners shorn in May. Although the cold stimulus experienced after shearing has been shown to increase the feed intake of adult sheep after shearing (Arnold et al. 1964; Arnold and Birrell 1977), if sufficient pasture is available to young sheep, their relatively high feed intake generates sufficient heat to make them relatively insensitive to cold (Egan and Doyle 1982; Freer et al. 1997). Thus, compared to adult sheep, shearing does not seem to stimulate extra feed consumption in young sheep, despite the opinion of many farmers to the contrary.

The methodology used to determine the fleece-free bodyweights of weaners warrants comment. These were estimated from the weights of weaners carrying wool because there was no shearing time common to all groups when fleece-free bodyweights could be directly compared. If these estimations were inaccurate, the growth rate comparisons between shearing groups may also be incorrect. Nevertheless, the maximum size of such an error will be no greater than several hundred grams. Thus, this method is unlikely to have compromised the study’s conclusions.

7.27.27.2 Fleece RotFleece Rot Shearing time was associated with different prevalences of fleece rot at certain times of the year in both the ewe and weaner flocks. Although fleece rot is mainly a problem in young sheep (Raadsma 1988), there were times throughout the trial when up to 25% of some adult shearing groups had severe lesions. Such prevalence would have been sufficiently great to detect differences between groups if they did exist. The only month in which a consistent difference between groups was observed was in May, when more May-shorn ewes had severe fleece rot than either December- or March-shorn ewes (Table 6.4). The prevalence of fleece rot in May-shorn ewes remained constant or increased throughout summer and autumn in all years, whereas the prevalence in the other groups remained steady or fell during this time. This may have occurred because the fleece length of May-shorn ewes prevented lesions that developed in the spring from drying out, with the result that their severity increased throughout the summer and autumn. May-shorn ewes probably had more fleece rot than March-shorn ewes in May simply because they were still carrying long wool, whereas the ewes shorn in March would have had their pre-existing lesions removed at shearing. ‘Severe’ fleece rot, as defined here, is associated with exudative lesions and is much more likely to lead to body strike than simply wool discolouration associated with lower fleece rot scores (Murray and Mortimer 2003). Thus, these results have implications for environments with significant risk of flystrike in late autumn,

142 Discussion—Health & Mortality because it appears that May shearing puts adult sheep at risk of body strike for up to two months longer than December or March shearing.

Weaners carrying 6–9 months wool in June or December had increased risk of severe fleece rot, as represented by odds ratios, at these times, whereas recently shorn sheep or sheep carrying long wool had significantly smaller risk. In south-eastern Australia, having severe fleece rot in June is unlikely to increase the risk of body strike because adult blowflies are not present at this time (Raadsma 1988). However, in December, the presence of severe fleece rot, as more frequently observed in March-, May- or June-shorn weaners, will increase the risk of body strike. March shearing has previously been associated with similar increases in fleece rot prevalence in two experiments conducted in western Victoria (R. Thompson pers. comm.). In the present study, March-, May- and June-shorn weaners still had increased risk of severe fleece rot in March and May, at 18 and 20 months of age, respectively. This may have been due to the recording of lesions already observed in December because the absolute fleece rot prevalence in these groups did not increase significantly between December and March. Shearing in March removed the fleece rot lesions of March-shorn weaners (MAR-MAR and JUN-MAR), so that their risk of severe fleece rot in May was small. Similar to May-shorn adult ewes, the prevalence of severe fleece rot in May-shorn weaners increased between March and May, possibly indicating that pre-existing lesions increased in severity throughout summer.

In summary, a main flock shearing time of May with weaners also first shorn in May, significantly increases the risk of severe fleece rot and body strike in both ewes and weaners. In addition, shearing weaners first in March or May is associated with increased fleece rot, and potentially body strike risk, in spring.

Two previous trials examining weaner shearing times reported that fleece rot significantly decreased wool value (Irving 1991; R. Thompson pers. comm.). This was not the case in the current trial as, from all shearings, wool was only classed as having unscourable colour on one occasion. The production of staining pigments in fleece rot appears to vary depending on the species or strain of bacteria present (Burrell 1988). Therefore, it is possible that shearing ewes in May, or shearing weaners in March or May, might result in lower wool value if the fleece rot causes unscourable staining of the wool.

In both ewes and weaners, there was a high prevalence of fleece rot in all groups between December 2001 and February 2002, most likely associated with the significantly above-average rainfall during this time. This indicates that prevailing environmental conditions can cause an increased prevalence of fleece rot, irrespective of shearing time. It is also interesting to note that, where differences between weaner shearing times occurred, the risk of fleece rot was greatest in sheep carrying six to nine months wool. This is a longer length of wool than previously reported to be associated with greatest risk (Raadsma 1988). Differences in staple length or genetic susceptibility to fleece rot of the sheep used in the two studies, or different environmental conditions may have been responsible for these differing observations. Nonetheless, the results of the current study confirm that sheep carrying

143 Discussion—Health & Mortality a moderate length of wool, 6–9 months, rather than very short wool or wool approaching 12 months length, are most susceptible to severe fleece rot.

7.37.37.3 Dag & Crutching Shearing ewes in March or May (i.e. in autumn) more than halved their risk of having severe dag in spring, compared to ewes shorn in December. However, December shearing conferred no corresponding advantage in autumn because there was little dag at this time. In the weaner flocks, a similar effect of autumn shearing was observed. The risk was smaller in autumn-shorn ewes in spring because their short wool during winter and early spring, when scouring is most common (Larsen et al. 1994), accumulated less dag than did the longer wool of December-shorn sheep. From summer onwards, little scouring occurred and so there was only a small amount of dag in autumn-shorn groups, despite their lengthening wool. Previously accumulated dag also dried and dropped off, as demonstrated by the declining prevalence of severe dag in the autumn shearing groups over summer. Similar results occurred in the weaner shearing groups, in which the odds of having severe dag in spring increased as the time since last shearing increased. Not all odds ratios of the weaner shearing groups at this time were statistically significant, but this was probably due to the relatively low power of the experiment to distinguish between the proportions of severe dag in group sizes of 50 sheep.

These results suggest that, if severe dag is considered to be associated with increased risk of breech strike, autumn shearing will reduce the risk of breech strike in spring compared to December shearing. Furthermore, this benefit appears to occur without the trade-off of greater risk in autumn for March- or May-shorn sheep. Nonetheless, at least half of the autumn-shorn sheep had some dag between late spring and autumn each year. Since French et al. (1996) reported a linear association between dag score and risk of breech strike in weaners in England, the mild dag prevalent in the autumn shearing groups in the current study might still pose a risk for contracting breech strike in some high-challenge environments at a time when December shearing would prevent this occurring. These results are discussed in more detail in Section 7.4, below.

Autumn shearing also had lower costs associated with dag than December shearing because of the lower overall weight of crutchings, which represented less foregone fleece wool, and a lower labour cost for removal. Shearing ewes in May was particularly attractive because the timing of shearing was close enough to lambing that a pre-lamb crutch was not required.

7.47.47.4 FlystrikeFlystrikeFlystrike Shearing ewes in May decreased their risk of flystrike in November and early December about six- fold compared to December-shorn ewes and three-fold compared to March-shorn ewes. In the trial, most flystrike in adult ewes occurred on the breech and the decreased risk of flystrike in May-shorn ewes corresponded to the smaller risk of severe dag observed. Although March- and May-shorn ewes

144 Discussion—Health & Mortality had the same risk of severe dag in spring, they had different risks of flystrike at this time. Overall, the risk of flystrike between all shearing groups in spring increased with time since last shearing, suggesting that the ‘severe/not severe dag’ criterion did not accurately predict risk of breech strike. This indicates that shorter wool provided additional protection against breech strike. In contrast, French et al. (1996) found that, for sheep with the same dag score, shorter wool was associated with a greater risk of breech strike, which they suggested was because shorter wool facilitated oviposition by blowflies. That study was conducted in young crossbred sheep with much shorter wool than the adult Merino ewes in the present study. Thus, it is possible that the observed difference in risk in the current study, between wool of six and eight months length, was caused more by reduced moisture retention, rather than oviposition closer to the sheep’s skin, in the shorter wool.

The fact that the odds ratios between shearing groups in late spring were similar from year to year, irrespective of whether or not ewes were crutched prior to lambing, is in accordance with Raadsma’s (1988) statement that crutching protects against breech strike for up to eight weeks. Given the greater risk of breech strike in December-shorn ewes in late spring, it is likely that prophylactic chemical treatment or crutching of this group would be required at lamb-marking. Alternatively, long-acting insecticide could be applied prior to lambing but pre-lambing crutching is unlikely to provide protection that persists until shearing in December.

Unfortunately, occurrence of flystrike during autumn was not recorded in detail, although there was a low prevalence in most years. Thus, no comment can be made about whether the greater risk of May- shorn ewes having severe fleece rot in late autumn was reflected in a greater prevalence of flystrike in this group, nor whether additional preventive treatments would be required in autumn-shorn sheep that might balance the cost of the extra spring treatment of December-shorn ewes. A number of May- shorn ewes died during April 2002 and, although flystrike was suspected as a prime cause of these deaths, the type of strike was not recorded.

The experimental design was such that no shearing time coincided with the expected onset of flystrike risk in spring, so Raadsma’s (1988) assertion that shearing provides up to three months protection against breech strike was not directly tested. Nevertheless, it is interesting to note that risk of strike appeared to increase with time since shearing. This is in contrast to the results of an epidemiological study in southern New South Wales that found no association between time since shearing and flystrike risk (Wardhaugh and Morton 1990), although this was probably because the authors only analysed the effect of shearing if it was performed within the previous month, despite the likelihood of it providing protection for longer than this. The association between time since shearing and risk of breech strike identified in the present study suggests that, at least in environments similar to western Victoria, shearing even one or two months closer to the onset of the spring fly wave might be useful in reducing susceptibility to breech strike.

Amongst weaners, the odds of flystrike were reduced in sheep that had been shorn up to about six months previously. Thus, December or March shearing eliminated flystrike risk in May; May, June, July or October shearing reduced the odds by about 80% in November; and shearing the previous

145 Discussion—Health & Mortality

October or December reduced the odds of flystrike by at least 90% in March, at 18 months of age. These patterns were very similar to the odds ratios of having severe fleece rot, although the latter were not always statistically significant. Overall, shearing weaners in December or October, both of which aligned weaners to a December adult shearing time, reduced flystrike risk in two of the three periods of fly activity between birth and 18 months of age. The advantage of October shearing was that it reduced risk with only one shearing in this time, compared to two for the December pattern.

In March, when weaners were about 18 months old, more cases of flystrike occurred amongst wethers than ewes in all shearing groups, with the exception of the December-shorn weaners, the most recently shorn group. Most of the flystrike cases in wethers were pizzle strike and the 10-fold greater odds of flystrike for wethers compared to ewes strongly suggests that preventative measures, such as ringing, should be undertaken during summer, unless shearing occurs at this time.

7.57.57.5 Reproduction There was no association between ewe shearing time and the number of lambs weaned in this trial. However, there were consistent differences in the proportion of ewes in each shearing group observed to be lactating at lamb marking. The lactational status of ewes at lamb marking was used as a measure of whether or not they had successfully reared a lamb or lambs up until that point in time. Using this definition, the odds of March-shorn ewes rearing a lamb to lamb marking were 33% lower than December-shorn ewes. May-shorn ewes were intermediate between the other groups in their likelihood of rearing a lamb to lamb marking.

The lactational status results are difficult to reconcile with the lack of difference in the number of lambs actually weaned from each shearing group. This result may be an artefact of the relative inaccuracy of udder examination, performed several weeks after birth, in determining the presence of a suckling lamb (Dun 1963; Plant 2006). On face value, if fewer March-shorn ewes were lactating, then this group must have successfully raised more twin lambs to wean the observed similar number of lambs as the other groups. Such a situation could have arisen if March-shorn ewes had a wider range of condition scores at mating, so that more, low-condition ewes failed to conceive but more higher-condition ewes conceived twins. However the standard errors of condition scores measured around mating were similar in all shearing groups each year, which makes this hypothesis unlikely. The only condition score observation around the time of mating that may have been pertinent to reproduction was that March- and May-shorn ewes lost condition at a slightly faster rate than December-shorn ewes during that time. Although loss of body condition can reduce ovulation, and hence conception, rate (Plant 2006), differences in body condition were not observed at weaning in the current study. March shearing usually occurred about two weeks prior to the start of mating, which may have affected oestrus at the start of the mating period, and subsequent patterns of conception and lambing (MacKenzie et al. 1975; Parr et al. 1989). However, shearing was particularly late in March in 2000 but there was no difference in percentages of lactating ewes in that

146 Discussion—Health & Mortality year. In summary, although the mechanism underlying these results is uncertain, it is clear that shearing time did not affect the number of lambs weaned, which is the reproductive outcome of direct economic importance.

7.67.67.6 MortalityMortalityMortality

7.6.17.6.17.6.1 EwesEwesEwes Overall, December-shorn ewes had less risk of dying than March- or May-shorn ewes. This was primarily driven by differences in the pattern of deaths between shorn and unshorn ewes following December shearing, compared to the pattern of deaths following shearing at other times of the year. There was no difference in the relative mortality risk of December-shorn ewes after shearing, compared to unshorn sheep, in the three months following this shearing. In contrast, the post-shearing mortality rate of May-shorn ewes was more than three times that of unshorn ewes during the same period, and the attributable fraction for the May-shorn group during this period was 0.70, indicating that about 70% of deaths within that group could potentially be removed by avoiding shearing at that time. Similarly, March-shorn ewes had five times the mortality risk of December-shorn ewes in the period immediately after their shearing and about two times the risk of May-shorn ewes in two out of three years.

These findings are consistent with previous studies that have shown that post-shearing deaths are more likely to occur later in summer or autumn, when sheep are more likely to be in light body condition or losing weight (Hutchinson and Lynch 1966; Slee 1966; Hutchinson and McRae 1969), or possibly are not acclimated to cold weather (Buckman 1982; Glass and Jacob 1992). In the current study, the autumn period with greatest mortality occurred in 2003, when body condition of ewes in all shearing groups was lowest.

The increases in risk after shearing observed in the present study are lower than the 17-fold increase reported by Hutchinson and McRae (1969) in wethers in April, although in that study weather was severe, with temperatures as low as -5° C and winds as strong as 30 km/h. In the present study, the association between shearing time and mortality risk occurred in the absence of any specific instances of severe weather immediately after shearing. Severe weather events post-shearing would be expected to increase mortality rates above those observed here.

In one year, there were a greater number of deaths amongst May-shorn ewes between March and May, which were attributed to flystrike when ewes were carrying long wool. Fly waves may produce an inconsistent, but substantial, impact on mortality in some years. May-shorn ewes are at particularly high risk of autumn fly waves and, similarly, December-shorn ewes are at greater risk of a severe fly wave in late spring.

147 Discussion—Health & Mortality

The average mortality rate of ewes in this study, 10% per annum, was surprisingly high, considering that a maximum of 1–2% annual mortality is normally expected, and considered acceptable, in adult Merinos (Morley 1994). High mortality rates were observed in years when pasture was relatively abundant, as well as during drought conditions, such as those experienced in late 2002 and throughout large parts of 2003. The age structure of the trial flock did not accurately represent that expected in a typical commercial flock, in which the distribution of ages is generally more evenly spread between two and six years old. In contrast, the trial was initiated with a disproportionately young flock that, as the trial progressed, became a disproportionately aged flock. Nonetheless, deaths were not confined to only aged or young ewes. Mortality rates of this magnitude in ewes have important economic and welfare implications and further investigation is warranted to determine whether similar mortality occurs more widely in commercially managed sheep flocks.

7.6.27.6.27.6.2 WeanersWeanersWeaners Significant mortality occurred amongst the weaner sheep in this study, which were run under commercial conditions on what was considered a well-managed, highly profitable woolgrowing farm. The average mortality in the year after weaning was 14.2%, or about one death per seven weaned lambs. Total deaths approached the industry target of less than 4% prior to one year of age (Larsen 2002) in only one cohort, but were at least twice this figure in all other years and exceeded 20% in three cohorts. On average, the death rate remained greater than 2% per month for four months after weaning. These results are similar to the only other sheep mortality survey conducted in Victoria, which found 25% of weaner flocks in the Mallee had more than 5% deaths (Harris and Nowara 1995). Other studies indirectly reporting mortality data have also shown that annual death rates of 10%–25% occur under commercial conditions in a wide variety of environments in Western Australia, South Australia, Victoria and Queensland (McDonald 1975; Anderson et al. 1976; Norris 1984; Barton and McCausland 1987; Cobon et al. 1990; Mulholland 1986 cited by Allworth 1994; Fogarty et al. 2005b; Fogarty et al. 2005a). Thus, excessive weaner mortality is a widespread problem on Australian woolgrowing farms, a significant animal welfare issue and likely to substantially affect farm productivity (Nilon 1986).

Throughout the current study, lightweight weaners consistently had a greater mortality risk than heavy ones, independent of year-to-year differences between flocks and regardless of the weaners’ absolute weight. For example, in the univariate analyses, the lightest 20% of lambs at weaning in any cohort experienced more than three times the mortality risk of weaners in the middle weaning weight quintile. The risk for weaners in the next lightest quintile was still elevated, at 1.5 times the risk. In contrast, weaners in the two heaviest quintiles were at similar risk of death to those in the median weight quintile. This consistency of risks over seven years and across a wide variety of seasonal conditions suggests that the results might be widely applicable to sheep enterprises, despite the study only being conducted on one farm. The multivariate analyses produced similar results, but also showed that weaners of low absolute bodyweight experienced a disproportionately greater risk than their heavier counterparts. For example, the mortality risk of weaners weighing 8 kg was 1.8 times

148 Discussion—Health & Mortality greater than the risk of weaners weighing 10 kg, but weaners weighing 18 kg only had 1.3 times the risk of weaners weighing 20 kg. The analysis that included bodyweight measured at any time throughout the post-weaning period produced very similar results to the models that used weight at weaning. This suggests that mortality risk is associated with weight per se and that the relationship continues to be important for many months following weaning.

Average cohort growth rate at different times after weaning was even more strongly associated with mortality than bodyweight. Relatively small increases in growth rate after weaning, during summer and autumn, were associated with substantial reductions in the rate of death and this relationship was independent of bodyweight. For example, the mortality risk of weaners when the average cohort growth rate during summer and autumn was 0.5 kg/month was 74% less than the risk of weaners in a cohort with an average growth rate of 0.25 kg/month. Similar growth rate increases occurring later in the post-weaning period were associated with a smaller reduction in mortality rate. The Cox model estimated that growth rate increases later in the post-weaning period were associated with greater reductions in risk in weaners that had been lightweight at weaning, whilst the Weibull and cubic spline models indicated greater reductions in risk if growth rate were increased in weaners that had grown more slowly earlier in the post-weaning period. For example, the Cox model estimated that increasing the growth rate of a weaner by 0.25 kg/month in early spring was associated with a 58% reduction in risk if it had weighed 10 kg at weaning but only a 43% reduction if it weighed 20 kg when it was weaned.

It is not clear why the cubic spline model estimated that increasing growth rate in early spring increased mortality risk in weaners that had also grown quickly immediately after weaning. This result was not observed in the cubic spine model of the shearing time-only data and it is possible that deaths late in the post-weaning period in some of the faster-growth rate cohorts were confounded with shearing or, alternatively, the results were spurious and caused by the wide censoring intervals in some cohorts.

The average growth rate of survivors throughout the periods of weaning–150 days, 150–225 days and day 225–end PWP were assigned to all weaners within each flock. Individual growth rates could not be used because weaners that died in a particular period perished before they were weighed a second time. Average seasonal growth rates provide a quantified estimate of the nutritional experience of a weaner flock. The average growth rate integrates the effects of a diverse range of factors, such as pasture availability and digestibility, weather and grazing behaviour, and precisely reflects the sum of those conditions, as they affect the average sheep in the flock. Average growth rate can also be estimated with relative ease, and so is a practical measure that can be adopted on commercial farms.

This study’s design did not enable it to demonstrate a causal relationship between weaner bodyweight and mortality risk. However previous research has shown that increasing bodyweight does reduce a young sheep’s risk of death (Allden 1968c; Hodge 1990) and the present study has quantified the relationship more precisely. Increasing bodyweight reduces mortality risk because heavy weaner sheep have greater body energy stores than light ones and more of this energy is stored as fat (Allden

149 Discussion—Health & Mortality

1970). This relationship is independent of breed and growth rate (Kellaway 1973). Because lightweight weaners have little or no fat, any event requiring the mobilisation of body energy, such as under- or malnutrition, or disease causes protein in the vital organs to be catabolised, which compromises organ function and leads to death (Thornton et al. 1979). A heavier bodyweight or faster rate of bodyweight gain is therefore associated with the acquisition of body energy stored as fat, which can be catabolised without threatening a weaner sheep’s survival (Allden 1970; Doyle and Egan 1983).

Very similar associations between mortality and bodyweight have been reported in survival analyses of overseas sheep and goat production systems (Mandonnet et al. 2003; Nguti et al. 2003). Mandonnet et al.’s study also reported an effect of bodyweight, measured at any time after weaning, on mortality that was similar to the results of the current study. Such similar results coming from dramatically different production systems suggests that the association between bodyweight, growth rate and mortality relates to fundamental physiological processes, and that the present study’s findings will apply to weaner sheep production in a wide variety of Australian and overseas environments.

The associations between mortality, growth rate and bodyweight quantified in this study suggest that relatively small amounts of supplementary feeding could efficiently and cost-effectively address the widespread problem of excessive weaner deaths on Australian farms. This strategy is particularly attractive because it does not require knowledge of the cause of death. Instead, protection against death is expected under the broad set of conditions generally described as the ‘weaner ill-thrift’ syndrome, which has no single, easily treatable cause (Gordon 1981; Wilkinson 1981),

The quantification of mortality risk means that the increase in growth rate or bodyweight required for a specific reduction in mortality can be calculated precisely. In turn, commercial ruminant nutrition software could be used to determine the rate of supplementary feeding required to achieve targeted weight gains to reduce mortality. Growth rate, in particular, can be directly manipulated through supplementary feeding by a farmer, whereas weaning weight must be indirectly, and relatively inefficiently, manipulated via ewe nutrition. For example, it is estimated that deaths could be reduced by 74% if the growth rate of weaners grazing dry pastures in summer were increased from 0.25 kg/month to 0.5 kg/month, which could be achieved by feeding 2.5 kg/head/month of an oat and lupin mix, costing $0.40/head/month (at $165/t). This would be highly cost-effective, particularly if feeding were directed to the lightweight weaners, which are at greater risk of death. Increasing growth rate offers a mechanism for responding to a high mortality risk situation that is anticipated because of low weaning weights, for example, or has been detected through ongoing monitoring of bodyweight or deaths.

The disproportionately high mortality risk amongst lightweight weaners suggests that the growth rates and bodyweights of this group, in particular, should be increased to improve survival. The attributable fraction calculations indicate the fraction of deaths in a particular group, or the wider population, that could be eliminated if the characteristic of that group were removed. Thus, an attributable fraction of 0.71 for the lightest weaning weight quintile suggests that 71% of weaner deaths in the lightest

150 Discussion—Health & Mortality quintile could be prevented if their bodyweight were increased. The fraction of deaths in the whole population attributed to being in the lightest weaning weight quintile was 0.31, suggesting that targeted feeding of just this fifth of the flock could address nearly one third of all mortalities. Separating the lightweight portion of the flock at weaning for subsequent preferential feeding is a clear recommendation that can be made to farmers on the basis of these findings.

No model showed a difference between mortality risk of shorn sheep and unshorn sheep following shearing when it occurred at weaning, in December, or in July, at 10 months of age. However the multivariate analyses showed that shearing weaners in March, May or June increased the risk of death approximately four-fold, compared to weaners that were not shorn at those times. These effects were similar in all years, so it is unlikely that the off-shears deaths were due to isolated instances of bad weather. In addition, no difference was observed between the survivorship of offspring of the three ewe shearing groups, despite modest differences in weaning weight of the groups’ offspring in some years of the trial. This suggests that the differences in survival of weaners following shearing was due to the actual weaner shearing events, rather than some effect of dam shearing times on the subsequent survival of their offspring.

The differences in mortality risk due to different weaner shearing times were estimated from the survival analysis, which controlled for other covariate effects. This may explain why statistically significant associations between shearing and mortality were not observed in the univariate analyses (Table 6.18). Similarly, the multivariate analyses found no increase in mortality risk following shearing in December, at weaning. Shearing at weaning may have indirectly affected survival by causing bodyweight loss or reduced growth rate, which increased mortality risk. If December shearing acted in this indirect fashion, it would not have been detected in a multivariate analysis that included bodyweight.

Although the proportional risk increases for March, May or June shearing were similar, the risk difference between shorn and unshorn weaners, representing the absolute increase in deaths, was much greater after shearing in March than in May or June. This was because the underlying death rate in March, coming soon after the late-summer peak in weaner mortality, was substantially greater than in May or June. Thus, shearing in March resulted in about an extra 62 deaths/1000/month, compared to an increase of 30 deaths/1000/month following shearing in May and 21 deaths/1000/month in June.

Shearing between March and June, when weaner sheep are likely to have already faced several months of inadequate nutrition, may have placed additional, lethal pressure on animals with few body reserves to mobilise in the face of even relatively mild cold stress. Donnelly (1991b) suggested that fasting during and immediately after shearing can increase the minimum critical temperature of adult sheep, and hence their susceptibility to hypothermia. Sheep respond to cold stress following shearing by mobilising body fat reserves (Holmes et al. 1992) and lightweight sheep losing or maintaining weight are at greater risk of dying within the fortnight following shearing than heavy ones (Hutchinson and McRae 1969). Prolonged periods of weight loss in adult sheep also have been shown to be associated with increased cold sensitivity following shearing (Slee 1966). In a study of adult

151 Discussion—Health & Mortality sheep, Panaretto (1968) concluded that whilst shorn sheep may tolerate cold stress or undernutrition, a combination of these conditions may be lethal.

Shearing at weaning in December may not have had the same effect on mortality because the lambs had grown well before weaning, were better acclimated to cold or environmental conditions were warmer. In addition, mortality rate amongst all sheep increased soon after weaning, and December shearing may have had little additional effect on an already high death rate.

The latter reason may also explain why shearing times were sometimes estimated to have minimal effect on mortality at low growth rate. At low growth rates (i.e. 0.25–0.5 kg/month), deaths already may have been so high that shearing caused little further increase. At moderate growth rates (i.e. 1–2 kg/month), unshorn weaners were less likely to die, but shorn ones were not growing fast enough to have acquired body reserves to protect them from cold stress off shears, and shearing caused extra deaths. Hutchinson (1969) observed that off-shears mortality of adult Merino wethers depended on rate of weight loss preceding shearing, rather than absolute bodyweight. He suggested this was because the rate of weight change was a more sensitive indicator of current nutritional state than absolute bodyweight. However, this is at odds with the research, albeit in growing sheep, demonstrating that bodyweight, and not growth rate, determines a sheep’s fat reserves (Kellaway 1973). Conversely, the ‘growth rates’ used in the present study were the survivors’ averages. As such, they indicated the nutritional status of the whole cohort, rather than an individual sheep, and thus fit Hutchinson’s hypothesis that off-shears mortality risk is related to a measure of current nutritional status. As discussed previously, where post-shearing mortality depended on growth rate, the absolute difference in death rates between shorn and unshorn sheep was greatest at low growth rates. Similarly, weaner flocks with lower overall mortality would be expected to have fewer extra deaths off-shears in March, May or June than high-mortality flocks.

The rate of death of males was estimated to be 1.6 times that of females. Similar odds and hazard ratios have been reported in survival analyses of weaner sheep and goats in the USA (Southey et al. 2001), Africa (Nguti et al. 2003; Turkson and Sualisu 2005) and the French West Indies (Mandonnet et al. 2003). The difference between mortality of males and females has not been explained, although Barger (1993) noted that males are more susceptible to strongyle infection than females. In the present study, however, differences in survivorship existed throughout summer, when gastrointestinal parasite challenge and infection is generally low. Post-operative complications of castration may have increased mortality amongst males, although the effect persisted throughout the post-weaning period and it is unlikely that a relatively minor surgical procedure would have such a long-lasting effect. Finally, differences in body composition, especially fat deposition, between the sexes may influence mortality. More investigation of this widely reported phenomenon is warranted.

The sex of weaners from Trial 2 born in 1999 and 2000 was not recorded before deaths started to occur and so they were excluded from the overall mortality and cumulative mortality rate calculations for sex. The overall mortality risks presented in the sex analysis were higher than in the other univariate analyses because mortality in the excluded cohorts was relatively lower than in the included

152 Discussion—Health & Mortality cohorts. To allow weaner records from these cohorts to be included in the multivariate analysis, sex was categorised as ‘male’, ‘female’ or ‘unknown’ and all cohorts were analysed. The effect of sex in the multivariate analysis will have been underestimated because no surviving weaners were of ‘unknown’ sex. The estimated association between sex and mortality derived from the multivariate survival analysis is nonetheless similar to the estimations obtained from the univariate analyses and the other studies mentioned.

The survival analyses appeared to model the observed data well, with similar risk ratios estimated by all models, and the mortality rate estimated by the cubic spline model following a similar shape to the overall mortality rate estimated from the raw data (Figure 6.15). The smoothed curve from the raw data is greater than the cubic spline-estimated rate because the former is an average based on the mortality risk of all weaners from Trial 2, whereas the cubic spline curve is only the estimated mortality risk of an unshorn female weaner of median weaning weight and growth rate. The associations between mortality, and bodyweight and growth rate covariates estimated in the progeny of the shearing trial were similar to those from the entire dataset. This suggests that the data from that trial were representative of the broader dataset and that the findings of the shearing time analysis can be interpreted more widely.

Several issues relating to the structure of the data had to be resolved prior to analysis. The intermittent observations of the weaner flocks meant that the survival times were not precisely known, but could only be located within the censoring interval between observations. Although commercial statistical software can analyse interval-censored survival data, packages are not available that simultaneously accommodate time-varying covariates (Radke 2003). A primary aim of this study was to investigate associations between mortality and events such as shearing and bodyweight change that occurred later in the post-weaning period. If they had not been included as time-varying values, an event such as shearing in May, for example, could have been associated with mortalities occurring in the previous December, which is nonsensical. The data were therefore approximated by taking the midpoints of the censoring intervals as exact failure times and time-varying covariates included. Authors have variously suggested that such an approximating censoring intervals in this way may underestimate the covariate effects (Radke 2003), particularly in the Cox model (Lindsey and Ryan 1998), or that it can be safely ignored in parametric models (Lindsey 1998). However, in this study, the covariate effects estimated by all the time-varying models, an interval-censored model and the univariate analyses were almost identical (Tables 6.19 and 6.15, respectively). Therefore approximating the interval-censored data appears to have produced reliable estimates that have not been affected by the choice of a particular survival model.

-1 The smoothed curve of the Schoenfeld residuals for the covariate GRearly initially decreased then had zero slope from approximately day 120 (Figure 6.16). This suggested that the assumption of proportional hazards was incorrect and that the covariate had a larger association with rate of death early in the post-weaning period and then decreased to a constant hazard ratio after about day 120. This effect may have been an artefact of the approximation of failure times as the midpoints of

153 Discussion—Health & Mortality censoring intervals, against which the residuals were plotted: the two cohorts observed more frequently early in the post-weaning period, and hence in which deaths were recorded earlier, were also cohorts that experienced high mortality and a low GRearly. However the failure times against which other cohorts’ residuals were plotted may have been later only because they were the midpoint of wider censoring intervals, rather than because later deaths were a feature specific to those cohorts.

Nonetheless, the fact that GRearly was significantly associated with mortality in the interval-censored survival analyses suggests that a true relationship existed.

This study has quantified, for the first time in an Australian sheep production system, relationships between mortality and bodyweight, growth rate, shearing time and sex. It has added to the evidence that weaner sheep mortality on Australian farms is often unacceptably high but it has also identified that manipulation of flock growth rates, especially soon after weaning, may very efficiently address this important problem.

7.77.77.7 Value of Lifetime Production, Accounting for Survivorship When the variations in mortality risk between the different weaner and ewe shearing groups were taken into account, two shearing patterns had the equal greatest value of lifetime wool production: shearing weaners first in October and adults annually in December, and shearing weaners first in June then adults annually in March. At the other extreme, shearing weaners first in May and adults in May produced the smallest expected value of wool over the sheep’s lifetime (Figure 6.17). This is in contrast to the lifetime value of wool production estimated without survival, which was greatest for the June-March shearing pattern. Thus, the greater mortality experienced amongst March-shorn ewes and their June- or March-shorn progeny disproportionately reduced the expected value of wool production over their lifetime, compared to the October-December shearing pattern. It is interesting to note that the October-December shearing pattern produced a favourable result. Of all the weaner shearing times examined, shearing weaners first as hoggets at around 13 months of age is probably the least common option on commercial farms because woolgrowers are aware that it tends to produce wool of poor staple strength and excessive length, without appreciating its benefit to overall wool production.

Like the other ‘lifetime’ estimates of production, the lifetime value of wool incorporating survivorship was calculated to enable the different shearing patterns to be compared but it does not represent real income derived from the shearing groups. The true commercial effect of varying survival rates in the different shearing groups would be expressed in several different ways. In flocks with greater mortality, more replacement hoggets are required and, consequently, fewer are available for sale. Reduced income from sheep sales is usually the main effect of increased mortality in commercial flocks. Thus, the greater mortality amongst May- and, to a lesser extent, March-shorn ewes would reduce income from both sales of cast-for-age ewes and surplus weaners. The greater requirement for

154 Discussion—Health & Mortality replacement stock in autumn-shorn flocks would mean that they would need a greater minimum weaning percentage to be able to remain self-replacing. If insufficient replacements were available, because of low reproductive rate for example, additional sheep would have to be purchased. Wool produced by weaner sheep is often particularly valuable, and the shearing times with high mortality after the first weaner shearing, particularly March and May, forewent potentially significant income.

155

156

C H A P T E R 8

CONCLUSION

After accounting for the multiplicity of factors affected by shearing time, no ewe and weaner shearing pattern was identified as being clearly superior for a spring-lambing Merino flock in this five-year study. However, the study did show that shearing ewes in May, regardless of when their progeny were shorn, was clearly inferior to the other shearing patterns examined. A flock with an adult shearing in May produced wool of lower value than December or March, under a wide range of historical market conditions, because it produced consistently broader fleeces that were no heavier than those from March-shorn ewes. May shearing was also associated with animal health disadvantages, including moderately lighter ewe body condition during winter in some years, a greater risk of severe fleece rot in late autumn, a three- to four-fold greater mortality risk after shearing than other ewe and weaner shearing times, and greater ewe mortality in the months before shearing in some years. This increased risk of death was not offset by lower relative risk at other times of the year. Although May shearing substantially reduced dag costs and the risk of breech strike in spring, it is highly unlikely that these benefits would outweigh the lower income and increased costs associated with shearing at this time of the year.

The distinction between the choice of December or March as an adult shearing time for a spring- lambing flock was less clear and either time might be acceptable, with a final decision being based on factors that are specific to a region or farm, or a view of the future direction of the wool market. For example, the staple strength of wethers is more reliably improved by autumn shearing in Victoria (Hansford 1987), so a particular farm’s flock structure might make March shearing preferable. On the other hand, shearing in March in some areas might result in sufficient vegetable matter contamination of wool to negate a premium paid for improved strength (Bowman 1989) and favouring December shearing instead.

In the absence of consistent staple strength differences between groups in the present study, wool income was mainly influenced by the variation in fibre diameter between the different ewe and weaner shearing times and the market premium paid for finer wool. Nevertheless, at both median and low historical micron basis prices, either shearing ewes in December and progeny first in October, or shearing ewes in March and progeny first in June produced very similar values of wool over the flocks’ lifetimes, especially when cumulative survival was taken into account. It was only at high historical prices that the June-March shearing pattern clearly produced a greater value of wool, despite the greater mortality rate associated with this shearing pattern than October-December shearing. Although it was

157 Conclusion inferred from the observed results that staple strength was not consistently improved by shearing time because of the variability of the position of break along the wool staple, further studies of the fibre diameter profile of grazing, spring-lambing ewes are warranted to investigate this hypothesis.

This study also made important findings about mortality in weaner sheep that extended well beyond the effect of shearing time. Consistently high mortality rates were observed in all cohorts of the weaner flock, which are consistent with previous observations in commercial flocks. This presents significant economic and welfare issues that could be addressed with the results from this study. Small increases in the growth rate of slow-growing weaners during summer and autumn were shown to substantially reduce the risk of death. Disproportionately high mortality risk was experienced by weaners that were lightweight at weaning, reinforcing the importance of separating lightweight sheep from the rest of the flock at weaning and improving the health and nutrition of this group. The quantification of the associations between bodyweight, growth rate and mortality risk enables the cost of the supplementary feeding required to reduce mortality to be estimated precisely. Further work is now required to transfer these individual-animal results into feeding models for the range of bodyweights and predicted mortality risks in a flock of weaner sheep, to determine feeding strategies that reduce mortality in the most efficient and cost-effective manner.

The benefits of a particular shearing time on other animal health issues, aside from mortality, were mixed. Ewes shorn in May had a smaller cost of dag and reduced risk of breech strike in spring but a greater risk of severe fleece rot, and potentially flystrike, in late autumn. December-shorn ewes had the greatest cost of dag and risk of breech strike in spring but the smallest risk of severe fleece rot. Ewes shorn in March also had a low risk of fleece rot and breech strike but had an intermediate cost of dag.

In summary, this study indicates that shearing ewes in December and their progeny first in October, or shearing ewes in March and weaners in June, are both appropriate alternatives for a self-replacing Merino flock in south-eastern Australia and that mortality amongst Merino weaners may be effectively reduced by small increases in post-weaning growth rates of these sheep and targeted management of the lightweight weaners in the flock.

158

159 Bibliography

160 Bibliography

BIBLIOGRAPHY

Adams NR (1994) Improving wool strength. In ‘Proceedings of the Australian society of animal production’. (Eds DG Masters and JG Allen) pp. 46-54. (Australian Society of Animal Production: Perth)

Adams NR, Briegel JR, Ritchie AJM (1997) Wool and liveweight responses to nutrition by Merino sheep genetically selected for high or low staple strength. Australian Journal of Agricultural Research 48, 1129-1137.

Adams NR, Oldham C, Suiter R, Wilson G (Eds) (2000) ‘Managing wool production for quality and profit.’ (Premium Quality Wool Cooperative Research Centre: Perth)

Alexander G, Lynch JJ, Mottershead BE (1979) Use of shelter and selection of lambing sites by shorn and unshorn ewes in paddocks with closely or widely spaced shelters. Applied Animal Ethology 5, 51- 69.

Alexander G, Lynch JJ, Mottershead BE, Donnelly JB (1980) Reduction in lamb mortality by means of grass wind-breaks: Results of a five-year study. Proceedings of the Australian Society of Animal Production 13, 329-332.

Allden WG, Anderson RA (1957) Unthriftiness in weaner sheep. Journal of the Department of Agriculture South Australia 61, 69-86.

Allden WG (1959) The summer nutrition of weaner sheep: The relative roles of available energy and protein when fed as supplements to sheep grazing mature pasture herbage. Australian Journal of Agricultural Research 10, 219-236.

Allden WG (1968a) Undernutrition of Merino sheep and its sequelae. I. Growth and development of lambs following prolonged periods of nutritional stress. Australian Journal of Agricultural Research 19, 621-638.

Allden WG (1968b) Undernutrition of Merino sheep and its sequelae. IV. Herbage consumption and utilization of feed for wool production following growth restrictions imposed at 2 stages of early postnatal life in a Mediterranean environment. Australian Journal of Agricultural Research 19, 997-1007.

Allden WG (1968c) Undernutrition of the Merino sheep and its sequelae. III. Effect on lifetime productivity of growth restrictions imposed at two stages of early post-natal life in a Mediterranean environment. Australian Journal of Agricultural Research 19, 981-996.

Allden WG (1969) Summer nutrition of weaner sheep - voluntary feed intake, body weight change, and wool production of sheep grazing mature herbage of sown pasture in relation to intake of dietary energy under a supplementary feeding regime. Australian Journal of Agricultural Research 20, 499- 512.

Allden WG (1970) The body composition and herbage utilization of grazing Merino and crossbred lambs during periods of growth and summer undernutrition. Australian Journal of Agricultural Research 21, 261-272.

Allden WG (1979) Feed intake, diet composition and wool growth. In ‘Physiological and environmental limitations to wool growth’. (Eds JL Black and PJ Reis) pp. 61-78. (University of New England Publishing Unit: Armidale)

161 Bibliography

Allison PD (1997) ‘Survival analysis using the SAS system. A practical guide.’ (SAS Institute Inc.: Cary, NC)

Allworth MB (1983) Weaner illthrift - treatment and prevention. In ‘Proceedings no. 67: Sheep production and preventive medicine’. (Ed. TG Hungerford) pp. 435-442. (University of Sydney Post- Graduate Foundation In Veterinary Science: Sydney)

Allworth MB (1994) Flock health. In ‘Merinos, money and management’. (Ed. FHW Morley) pp. 263- 293. (Post Graduate Committee in Veterinary Science, University of Sydney: Sydney)

Anderson N, Morris RS, McTaggart IK (1976) An economic analysis of two schemes for the anthelmintic control of helminthiasis in weaned lambs. Australian Veterinary Journal 52, 174-180.

Arnold GW, McManus WR, Bush IG (1964) Studies in the wool production of sheep. 1: Seasonal variation in feed intake, liveweight and wool production. Australian Journal of Experimental Agriculture and Animal Husbandry 4, 392-403.

Arnold GW, Gordon ID (1973) An assessment of seasonal changes in wool quality in agricultural regions in Western Australia. Journal of the Australian Institute of Agricultural Science 13, 151-155.

Arnold GW, Birrell HA (1977) Food intake and grazing behaviour of sheep varying in body condition. Animal Production 24, 343-353.

Arnold GW, Charlick AJ, Eley JR (1984) Effects of shearing time and time of lambing on wool growth and processing characteristics. Australian Journal of Experimental Agriculture and Animal Husbandry 24, 337-343.

Ashton BL (1992) Later lambing - experiences of farmers in cereal-sheep areas of South Australia. Department of Agriculture South Australia, Technical report no. 195, Adelaide.

Atkinson KR (1989) Towards a better description of vegetable matter contamination in raw wool. Wool Technology and Sheep Breeding 37, 38-43.

Australian Wool Testing Authority (2004) Staple length and strength measurement. Available at http://www.awta.com.au/Publications/Marketing/Raw_Wool_Services/staple.html (verified 26 April 2004)

AWI (2006) Woolcheque. (Australian Wool Innovation: Sydney) Available at http://www.pricemaker.info/ (verified 15 April 2006)

Baker SK, Barton J, Purser DB (1993) A profile of wool quality in the Western Australian clip. In ‘Improving sheep management to improve wool production and marketability’. (Ed. R Suiter) pp. 7-20. (Western Australia Department of Agriculture: Perth)

Baker SK, Purser DB, Barton J (1994). In ‘Improving wool strength’. (Ed. NR Adams) Proceedings of the Australian society of animal production pp. 48-49. (Australian Society of Animal Production: Perth)

Barger IA (1993) Influence of sex and reproductive status on susceptibility of ruminants to nematode parasitism. International Journal for Parasitology 23, 463-470.

Barton J, Baker SK, Purser DB (1994) Staple strength in seasonal environments in south-western Australia. Proceedings of the Australian Society of Animal Production 20, 293-296.

Barton NJ, McCausland IP (1987) Production and economic returns from Merino weaner sheep subjected to four frequencies of anthelmintic administration in East Gippsland, Victoria. Australian Journal of Experimental Agriculture 27, 759-764.

Behrendt K (2003) Wipe out weaner mortalities with better nutrition. Farming Ahead 141, 54-56.

162 Bibliography

Bell KJ, Ralph IG (1993) Current sheep management: Western Australia. In ‘Proceedings of a national workshop on management for wool quality in Mediterranean environments’. (Eds PT Doyle, JA Fortune and NR Adams) pp. 60-66. (Department of Agriculture Western Australia: Perth)

Bell KJ (1994) Preliminary farm appraisal. Appendix 2: Body condition score. In ‘Merinos, money and management’. (Ed. FHW Morley) p. 8. (Post Graduate Committee in Veterinary Science, University of Sydney: Sydney)

Bellotti W, Collins W, Moore A (1993) The Mediterranean environments. In ‘Proceedings of a national workshop on management for wool quality in Mediterranean environments’. (Eds PT Doyle, JA Fortune and NR Adams) pp. 50-59. (Department of Agriculture Western Australia: Perth)

Beveridge I, Brown TH, Fitzsimons SM et al. (1985) Mortality in weaner sheep in South Australia under different regimes of anthelmintic treatment. Australian Journal of Agricultural Research 36, 857-865.

Bigham ML, Sumner RMW, Hawker H, Fitzgerald JM (1983) Fleece tenderness - a review. Proceedings of the New Zealand Society of Animal Production 43, 73-78.

Black HJ, Chestnutt DMB (1990) Influence of shearing regime and grass silage quality on the performance of pregnant ewes. Animal Production 51, 573-582.

Black JL, Bottomley GA (1980) Effects of shearing and lambing dates on the predicted pasture requirements of sheep in two Tasmanian locations. Australian Journal of Experimental Agriculture and Animal Husbandry 20, 654-661.

Bottomley GA, Hudson JM (1976) Winter shearing does have a major disadvantage. Tasmanian Journal of Agriculture 47, 95-100.

Bottomley GA (1979) Weather conditions and wool growth. In ‘Physiological and environmental limitations to wool growth’. (Eds JL Black and PJ Reis) pp. 115-126. (University of New England publishing unit: Armidale)

Bourke CA (2003) The effect of shade, shearing and wool type in the protection of Merino sheep from Hypericum perforatum (St. John’s wort) poisoning. Australian Veterinary Journal 81, 494-499.

Bowman PJ (1989) Farm management strategies for improving the quality of fine wool. Doctor of philosophy thesis, Lincoln College.

Brodbeck PS, Hill MK (1984) Blowfly strike: A field survey of the New England region. Proceedings of the Australian Society of Animal Production 15, 656.

Brown DJ, Crook BJ, Purvis IW (2002) Differences in fibre diameter profile characteristics in wool staples from Merino sheep and their relationship with staple strength between years, environments, and bloodlines. Australian Journal of Agricultural Research 53, 481-491.

Brown TH (1971) The effect of three factors on the incidence of tender wool in deferred grazing management. Australian Journal of Experimental Agriculture and Animal Husbandry 11, 607-609.

Brown TH (1977) A comparison of continuous grazing and deferred autumn grazing of Merino ewes and lambs at 13 stocking rates. Australian Journal of Agricultural Research 28, 947-961.

Brown TH, Ford GE, Miller DW, Beveridge I (1985) Effect of anthelmintic dosing and stocking rate on the productivity of weaner sheep in a Mediterranean climate environment. Australian Journal of Agricultural Research 36, 845-855.

Buckman PG (1982). In ‘Losses of sheep after shearing due to adverse weather. Proceedings of seminar by the Australian society of animal production (Western Australian branch)’. (Eds SK Baker, HM Chapman and IH Williams) p. 11 (Australian Society of Animal Production, Western Australian Branch: Perth)

163 Bibliography

Burrell DH (1988) Bacteriology and pathogenesis of fleece rot and flystrike. In ‘Proceedings no. 110: Sheep health and production’. (Ed. TG Hungerford) pp. 231-246. (University of Sydney Post-graduate Committee in Veterinary Science: Sydney)

Butler LG (1994) Factors affecting staple strength with particular reference to Tasmania. Wool Technology and Sheep Breeding 42, 213-230.

Butler LG, Gibson WR, Head GM (1994) The effect of age, shearing date and reproduction on seasonal wool growth patterns, staple strength and position of break. Proceedings of the Australian Society of Animal Production 20, 269-272.

Campbell RJ, Robards GE, Saville DG (1972) The effect of grass seed on sheep production. Proceedings of the Australian Society of Animal Production 9, 225-229.

Caple IW, Vizard AL, Allworth MB, Morley FHW (1989) Solutions to suboptimal performance in livestock in the field. In ‘Recent advances in animal nutrition in Australia’. (University of New England: Armidale, NSW)

Carstensen B (1996) Regression models for interval censored survival data: Application to HIV infection in Danish homosexual men. Statistics in Medicine 15, 2177-2189.

Casey AE, Atkins KD, Coelli KA, Cottle DJ (1995) Merino central test sire evaluation - medium wool (1989-1993) and fine wool (1990-1993). Wool Technology and Sheep Breeding 43, 30-46.

Chapman RE, Wheeler JL (1963) Dye-banding: A technique for fleece growth studies. Australian Journal of Science 26, 53-54.

Charlton D, Rottenbury RA, Eley JR (1981) The effect of vegetable matter level on worsted processing. Journal of the Textile Institute 72, 96-99.

Cloete SWP, van Niekerk FE, van der Merwe GD (1994) The effect of shearing pregnant ewes prior to a winter-lambing season on ewe and lamb performance in the southern cape. South African Journal of Animal Science 24, 140-142.

Cloete SWP, Muller CJC, Durand A (2000) The effects of shade and shearing date on the production of Merino sheep in the Swartland region of South Africa. South African Journal of Animal Science-Suid- Afrikaanse Tydskrif vir Veekunde 30, 164-171.

Clunies Ross I (1934) Parasitological and other problems in sheep in Western Australia. Journal of the Council for Scientific and Industrial Research 7, 1-8.

Cobon DH, O’Sullivan PD, Connelly PT (1990) The effect of management strategies on the productivity of weaner sheep in north west Queensland. Proceedings of the Australian Society of Animal Production 18, 466.

Colditz IG, Watson DL, Gray GD, Eady SJ (1996) Some relationships between age, immune responsiveness and resistance to parasites in ruminants. International Journal for Parasitology 26, 869- 877.

Copland RS (1986) Prediction of susceptibility to fleece rot and body strike in young sheep. In ‘Australian advances in veterinary science’. (Ed. P Outteridge) pp. 123-124. (Australian Veterinary Association: Artarmon, NSW)

Corbett JL (1979) Variation in wool growth with physiological state. In ‘Physiological and environmental limitations to wool growth’. (Eds JL Black and PJ Reis) pp. 79-98. (University of New England publishing unit: Armidale)

Cornish PS, Beale JA (1974) Vegetable fault and grass seed infestation in sheep in New South Wales. Journal of the Australian Institute of Agricultural Science 40, 261-267.

164 Bibliography

Couchman RC, Hanson PJ, Stott KJ, Vlastuin C (1993) Wool quality: Implications for worsted processing, grower receipts and R&D. In ‘Proceedings of a national workshop on management for wool quality in Mediterranean environments’. (Eds PT Doyle, JA Fortune and NR Adams) pp. 1-23. (Department of Agriculture, Western Australia:

Court J, Lawless J (1995) Improved wool returns from clip marketing analysis 1991-1994. Agriculture Victoria, Melbourne.

Cox DR (1972) Regression models and life-tables (with discussion). Journal of the Royal Statistical Society, Series B 34, 187-220.

Dabiri N, Parker WJ, Morris ST, McCutcheon SN (1994) Effects of pre-lamb and conventional full- wool shearing on the productivity of ewes. Proceedings of the New Zealand Society of Animal Production 54, 223-226.

Dabiri N, T. MS, J. PW et al. (1995) Productivity and cold resistance in ewes pre-lamb shorn by standard or cover comb. Australian Journal of Agricultural Research 46, 721-732.

Dabiri N, Morris ST, Wallentine M et al. (1996) Effects of pre-lamb shearing on feed intake and associated productivity of May- and August-lambing ewes. New Zealand Journal of Agricultural Research 39, 53-62.

Dalton DC, Knight TW, Johnson DL (1980) Lamb survival in sheep breeds on New Zealand hill country. New Zealand Journal of Agricultural Research 23, 167-173.

Davis CH, Saville DG, McInnes P (1976) Production feeding of weaner lambs with diets based on wheat - effect of diets and management. Proceedings of the Australian Society of Animal Production 11, 433-436.

Davis GH, Wallis TR, Bray AR (1981) Marked udder method (MUM) - a technique for identifying the progeny of ewes where identification at birth is not practicable. In ‘Proceedings of the New Zealand society of animal production’. (Ed. KHC Lewis) pp. 229-232. (New Zealand society of animal production (inc.): Hamilton) de Jong S, Kavanagh WJ, Andrews MW (1985) Factors contributing to the staple strength of wool. In ‘Proceedings of the 7th International Wool Textile Conference’. Tokyo pp. 147-156

Denney GD, Thornberry KJ, Sladek MA (1988) The effect of pre and postnatal nutrient deprivation on live weight and wool production of single born Merino sheep. Proceedings of the Australian Society of Animal Production 17, 174-177.

Denney GD (1990) Phenotypic variance of fibre diameter along wool staples and its relationship with other raw wool characters. Australian Journal of Experimental Agriculture 30, 463-467.

Dohoo IR, Martin W, Stryhn H (2003) ‘Veterinary epidemiologic research.’ (AVC Inc.: Charlottetown, Canada)

Donald CM, Allden WG (1959) The summer nutrition of weaner sheep: The deficiencies of the mature herbage of sown pasture as a feed for young sheep. Australian Journal of Agricultural Research 10, 199-218.

Donnelly FB (1991a) Shearing Merino sheep with significantly less than twelve months wool growth - some field experiences. Wool Technology and Sheep Breeding 39, 40-42.

Donnelly JB, Lynch JJ, Webster MED (1974) Climatic adaptation in recently shorn Merino sheep. International Journal of Biometeorology 18, 233-247.

Donnelly JR (1991b) Cold exposure and sheep survival. In ‘Proceedings of the Australian Sheep Veterinary Society’. Sydney, Australia pp. 83-87. (Australian Veterinary Association)

165 Bibliography

Douglas SAS (1989) TEAM project - presentation of report on trials evaluating additional measurements 1981-1988. Wool Technology and Sheep Breeding 37, 56-62.

Doyle PT, Egan AR (1983) The utilization of nitrogen and sulfur by weaner and mature Merino sheep. Australian Journal of Agricultural Research 34, 433-439.

Doyle PT, Fortune JA, Adams NR (Eds) (1993) ‘Proceedings of a national workshop on management for wool quality in Mediterranean environments.’ (Department of Agriculture Western Australia: Perth)

Doyle PT, Plaisted TW, Love RA (1995) Supplementary feeding pattern and rate of liveweight gain in winter-spring affect wool production of young Merino sheep on the south coast of Western Australia. Australian Journal of Experimental Agriculture 35, 1093-1100.

Drinan JP, Ferguson BD (1966) The influence of shearing on the growth and economic value of carry- over lambs. Proceedings of the Australian Society of Animal Production 6, 190-193.

Drinan JP (1968) The relative productivity of single and twin born Merino ewes. Proceedings of the Australian Society of Animal Production 7, 208-211.

Dun RB (1963) Recording the lambing performance of ewes under field conditions. Australian Journal of Experimental Agriculture and Animal Husbandry 3, 228-231.

Earl CR, Stafford JE, Rowe JP, Ross RA (1994) The effect of stocking rate on fibre diameter, staple strength and wool weight in high and low fibre diameter wool sheep on clover based pastures. Proceedings of the Australian Society of Animal Production 20, 309-312.

Egan AR, Doyle PT (1982) The effect of stage of maturity in sheep upon intake and digestion of roughage diets. Australian Journal of Agricultural Research 33, 1099-1105.

Elvidge DG, Coop IE (1974) Effect of shearing on feed requirements of sheep. New Zealand Journal of Experimental Agriculture 2, 397-402.

Engel AE (1958) Unthriftiness of weaner sheep in south-western Victoria. Australian Veterinary Journal 34, 391-397.

Falck SJ, Carstens GE, Waldron DF (2001) Effects of prenatal shearing of ewes on birth weight and neonatal survivability of lambs. Sheep & Goat Research Journal 17, 14-20.

Farrell DJ, Corbett JL (1970) Fasting heat production of sheep at pasture before and after shearing. Proceedings of the Australian Society of Animal Production 8, 267-271.

Farrington CP (1996) Interval censored survival data: A generalised linear modelling approach. Statistics In Medicine 15, 283-292.

Fels HE (1971) Flystrike control programmes. Journal of Agriculture, Western Australia 12, 241-244.

Fernandez Abella D, Surraco L, Borsani L et al. (1991) Effect of season of shearing on growth and wool production of sheep on unimproved basalt pasture. Boletin Tecnico de Ciencias Biologicas, Universidad de la Republica Regional Norte (Salto) 1, 31-48.

Fernandez Abella D, Rodriguez Palma R, Saldanha S et al. (1996) Effects of season of shearing on the fertility of ewes mated early and on wool growth on natural basalt soil pasture. Produccion Ovina 9, 63-74.

Fogarty NM, Ingham VM, Gilmour AR et al. (2005a) Genetic evaluation of crossbred lamb production. 2. Breed and fixed effects for post-weaning growth, carcass, and wool of first-cross lambs. Australian Journal of Agricultural Research 56, 455-463.

166 Bibliography

Fogarty NM, Ingham VM, Gilmour AR et al. (2005b) Genetic evaluation of crossbred lamb production. 1. Breed and fixed effects for birth and weaning weight of first-cross lambs, gestation length, and reproduction of base ewes. Australian Journal of Agricultural Research 56, 443-453.

Foot JZ, McIntyre JS, Heazlewood PG (1983) Supplements for Merino weaner sheep grazing mature pastures in summer and autumn. Australian Journal of Experimental Agriculture 23, 374-382.

Foot JZ, Egan JK, Love KJ (1987) ‘Sheep nutrition in the Victorian environment.’ (Victorian Department of Agriculture and Rural Affairs: Melbourne)

Foot JZ, Vizard AL (1993) Current sheep management: South Australia and Victoria. In ‘Proceedings of a national workshop on management for wool quality in Mediterranean environments’. (Eds PT Doyle, JA Fortune and NR Adams) pp. 67-79. (Department of Agriculture, Western Australia: Perth)

Francis SM, Bray AR, Scales GH (2000) Wool quality characteristics of purebred Merino and Merino crossbred lambs from three to twelve months of age. Wool Technology and Sheep Breeding 48, 259- 268.

Franklin MC, McClymont JL, Briggs PK, Campbell BL (1955) Maintenance rations for Merino sheep. II. The performance of weaners fed daily and weekly on rations of wheat and wheaten chaff at maintenance levels and the effect thereon of vitamin a supplements. Australian Journal of Agricultural Research 6, 324-342.

Franklin MC, Briggs PK, McClymont JL (1964) Early weaning of Merino lambs during drought. Australian Journal of Experimental Agriculture and Animal Husbandry 4, 132-140.

Freer M, Moore AD, Donnelly JR (1997) Decision support systems for Australian grazing enterprises. 2. The animal biology model for feed intake, production and reproduction and the Grazfeed DSS. Agricultural Systems 53, 77-126.

French NP, Parkin TD, Morgan KL (1996) A case control study of blowfly strike in lambs. Veterinary Record 139, 384-388.

Friend MA, Robards GE, Kennedy JP (1996) Relationships between staple strength and fibre diameter changes in Merino sheep selectively bred for either high or low clean fleece weight. Proceedings of the Australian Society of Animal Production 21, 111-114.

Gabbedy BJ (1971) Effect of selenium on wool production, body weight and mortality of young sheep in Western Australia. Australian Veterinary Journal 47, 318-322.

Ganai TAS, Pandey RS (1996) Survivability of Rambouillets in first year of life. Indian Journal of Animal Sciences 66, 839-842.

Geytenbeek PE (1962) A survey of post-shearing losses. Proceedings of the Australian Society of Animal Production 4, 185-186.

Gherardi SG, Sutherland SS, Monzu N, Johnson KG (1983) Field observations on body strike in sheep affected with dermatophilosis and fleece-rot. Australian Veterinary Journal 60, 27-28.

Glass MH, Jacob RH (1992) Losses of sheep following adverse weather after shearing. Australian Veterinary Journal 69, 142-143.

Gloag CM, Behrendt R, Ferguson MB, Beattie L (2002) Wool quality variation in south-west Victoria. Wool Technology and Sheep Breeding 50, 222-227.

Gordon HM (1981) Differential diagnosis of unthriftiness. In ‘Proceedings no. 58: Refresher course on sheep’. (Ed. TG Hungerford) pp. 543-550. (University of Sydney Post-Graduate Foundation in Veterinary Science: Sydney)

167 Bibliography

Graham AD, Christopherson RJ (1981) Effects of adrenaline and noradrenaline on the heat production of warm- and cold-acclimated sheep. Canadian Journal of Physiology and Pharmacology 59, 985-993

Gregory NG (1995) The role of shelterbelts in protecting livestock: A review. New Zealand Journal of Agricultural Research 38, 423-450.

Grosser TI, Kleemann DO, Grimson RJ et al. (1991) Effect of shearing during late pregnancy on lamb survival in South Australian Merino ewes. Wool Technology and Sheep Breeding 39, 61-63.

Hanley JA (2003) Re: “statistical analysis of correlated data using generalized estimating equations: An orientation” - the first author replies. American Journal of Epidemiology 158, 289-290.

Hanley JA, Negassa A, Edwardes MDD, Forrester JE (2003) Statistical analysis of correlated data using generalized estimating equations: An orientation. American Journal of Epidemiology 157, 364- 375.

Hansford KA (1987) Collation and analyses of available information on additional measurement relevant to improved management of sheep. Australian Wool Corporation, Melbourne.

Hansford KA, Kennedy JP (1988) Relationship between the rate of change in fibre diameter and staple strength. Proceedings of the Australian Society of Animal Production 17, 415.

Hansford KA, Kennedy JP (1990a) The relationship between the proportions of ortho-, meso- and paracortex and the fibre diameter and staple strength of Merino wool. Proceedings of the Australian Society of Animal Production 18, 484.

Hansford KA, Kennedy JP (1990b) The relationship between variation in the fibre diameter along staples and staple strength. In ‘Proceedings of the 8th international wool textile research conference’ pp. 590-598Christchurch)

Hansford KA (1997) A study of the specification and topmaking performance of Western Australian fleeces and sale lots. International Wool Textile Organisation - Technology and Standards Committee, Report No. 15, Nice.

Harris DJ, Nowara G (1995) The characteristics and causes of sheep losses in the Victorian Mallee. Australian Veterinary Journal 72, 331-340.

Hary I (2002) Analysis of survival curves in seasonally mated pastoral goat herds in northern Kenya using logistic regression techniques. Journal of Arid Environments 50, 621-640.

Hatcher S (2000) Why produce finer wool? In ‘Managing wool production for quality and profit’. (Eds NR Adams, C Oldham, R Suiter and G Wilson) pp. 5-21. (Premium Quality Wool Cooperative Research Centre: Perth)

Haughey KG (1981) Perinatal lamb mortality. In ‘Proceedings no. 58: Refresher course on sheep’. (Ed. TG Hungerford). (University of Sydney, Post-Graduate Committee in Veterinary Science: Sydney)

Hemsley JA, Jackson N, Gordon LR, Marshall JTA (1984) Shearing increases the susceptibility of weaners to fleece-rot. Proceedings of the Australian Society of Animal Production 15, 692.

Hodge RW (1990) The strategic use of supplementary feed to increase the live weight and onset of oestrus of spring born Merino ewes at 18 months of age. Australian Wool Corporation, Research Report Series No. 111, Melbourne.

Holmes CW, Kamil KA, Parker WJ et al. (1992) Effects of shearing method on the physiology and productivity of sheep. Proceedings of the New Zealand Society of Animal Production 52, 199-202.

Holmes JHG (1992) Trace element deficiency in sheep in East Gippsland Victoria. Australian Veterinary Journal 69, 292-293.

168 Bibliography

Holst PJ, Hall DG, Stanley DF (1996) Barley grass seed and shearing effects on summer lamb growth and pelt quality. Australian Journal of Experimental Agriculture 36.

Hosmer DW, Lemeshow S (1999) ‘Applied survival analysis. Regression modelling of time to event data.’ (John Wiley & Sons: New York)

Hosmer DW, Lemeshow S (2000) ‘Applied logistic regression. 2nd edition.’ (John Wiley & Sons, Inc.)

Hudson JM, Bottomley GA (1978) The effect of shearing on feed requirements. Tasmanian Journal of Agriculture 49, 17-21.

Hungerford TG (1990) ‘Diseases of livestock.’ (McGraw-Hill: Sydney)

Hunter L (1980) The effect of wool fibre properties on processing performance and yarn and fabric properties. In ‘Proceedings of the 6th international wool textile research conference’ pp. 133-193. (IWTO: Pretoria)

Hunter L, Leeuwner W, Smuts S, Strydom MA (1983) The correlation between staple strength and single fibre strength for sound and tender wools. The South African Wool and Textile Research Institution, SAWTRI Technical Report No.514, Port Elizabeth.

Hunter L, van Wyk JB, de Wet PJ et al. (1990) The effects of nutritional and lambing stress on wool fibre and processing characteristics. Wool Technology and Sheep Breeding 38, 89-91.

Hutchinson JCD, Bennett JW (1962) The effect of cold on sheep. Wool Technology and Sheep Breeding 9, 11-16.

Hutchinson JCD (1968) Deaths of sheep after shearing. Australian Journal of Experimental Agriculture and Animal Husbandry 8, 393-400.

Hutchinson KJ, Lynch JJ (1966) Bodyweight criteria and post-shearing deaths. Proceedings of the Australian Society of Animal Production 6, 229-233.

Hutchinson KJ, McRae BH (1969) Some factors associated with the behaviour and survival of newly shorn sheep. Australian Journal of Agricultural Research 20, 513-521.

Hygate L, Swan A, Brewer H et al. (2006) Central test sire evaluation results. In ‘Merino superior sires’. (CSIRO Livestock Industries: Armidale)

Hynd PI, Hughes A, Earl CR, Penno NM (1997) Seasonal changes in the morphology of wool follicles in Finewool and Strongwool Merino strains grazing at different stocking rates in southern Australia. Australian Journal of Agricultural Research 48, 1089-1097.

Ikwuegbu OA, Tarawali G, Rege JEO (1995) Effects of fodder banks on growth and survival of West African Dwarf goats under village conditions in subhumid Nigeria. Small Ruminant Research 17, 101- 109.

Independent Commodity Services (2005) ‘MLS analyser version 4.0.’ (Independent Commodity Services Pty Ltd: Wagga Wagga, New South Wales)

Insightful Corp. (2004) ‘S-plus.’ (Insightful Corp.: Seattle, Washington)

Irving RF (1991) Factors which influence the selection of a time for shearing on a property in Gippsland. Master of Veterinary Studies thesis, University of Melbourne.

Kellaway RC (1973) Effects of plane of nutrition, genotype and sex on growth, body composition and wool production in grazing sheep. Journal of Agricultural Science 80, 17-27.

169 Bibliography

Kennedy JP, Auldist IH, Popovic PG, Reynolds JA (1976) Reproduction rate of Merino sheep in arid New South Wales. Proceedings of the Australian Society of Animal Production 11, 149-151.

Kennedy JP, Auldist IH, Gray SJ, Reynolds JA (1982) Influence of the timing of shearing and joining on productivity of sheep in the arid zone of western New South Wales. Proceedings of the Australian Society of Animal Production 14, 507-510.

Kenyon PR, Morris ST, Revell DK, McCutcheon SN (1999) Improving lamb birthweight through mid- to late-pregnancy shearing: A review of recent studies. Proceedings of the New Zealand Society of Animal Production 59, 70-72.

Kenyon PR, Morris ST, McCutcheon SN (2002a) Does an increase in lamb birth weight though mid- pregnancy shearing necessarily mean an increase in lamb survival rates to weaning? Proceedings of the New Zealand Society of Animal Production 62, 53-56.

Kenyon PR, Morris ST, Revell DK, McCutcheon SN (2002b) Maternal constraint and the birthweight response to mid-pregnancy shearing. Australian Journal of Agricultural Research 53, 511-517.

Kenyon PR, Morris ST, Revell DK, McCutcheon SN (2002c) Nutrition during mid to late pregnancy does not affect the birthweight response to mid pregnancy shearing. Australian Journal of Agricultural Research 53, 13-20.

Kenyon PR, Revell DK, Morris ST (2006a) Mid-pregnancy shearing can increase birthweight and survival to weaning of multiple-born lambs under commercial conditions. Australian Journal of Experimental Agriculture 46, 821-825.

Kenyon PR, Sherlock RG, Morris ST, Morel PCH (2006b) The effect of mid- and late-pregnancy shearing of hoggets on lamb birthweight, weaning weight, survival rate, and wool follicle and fibre characteristics. Australian Journal of Agricultural Research 57, 877-882.

Kleemann DO, South MLH, Dolling CHS, Ponzoni RW (1983) Survival, growth and wool production of South Australian strong-wool Merino and 1st-cross Merino lambs from birth to 16 months of age. Australian Journal of Experimental Agriculture 23, 271-279.

Kleemann DO (1997) Increasing efficiency of wool production. CSIRO, Work-in-progress report, Canberra.

Lamb PR (2004) Staple strength and its relevance to worsted processing. Wool Technology and Sheep Breeding 52, 171-192.

Langlands JP, Wheeler JL (1968) The dyebanding and tattooed patch procedures for estimating wool production and obtaining samples for the measurement of fibre diameter. Australian Journal of Experimental Agriculture and Animal Husbandry 8, 265-269.

Langlands JP, Donald GE, Paull DR (1984) Effects of different stocking intensities in early life on the productivity of Merino ewes grazed as adults at two stocking rates. 1. Wool production and quality, lamb growth rate, and size and liveweight of ewes. Australian Journal of Experimental Agriculture and Animal Husbandry 24, 34-46.

Larsen J, Vizard A, Counsell DJ et al. (2002) Linking Australian woolgrowers with research: The South Roxby project. Wool Technology and Sheep Breeding 50, 266-273.

Larsen JWA, Anderson N, Vizard AL et al. (1994) Diarrhoea in Merino ewes during winter - association with trichostrongylid larvae. Australian Veterinary Journal 71, 365-372.

Larsen JWA, Vizard AL, Anderson N (1995) Production losses in Merino ewes and financial penalties caused by trichostrongylid infections during winter and spring. Australian Veterinary Journal 72, 58- 63.

170 Bibliography

Larsen JWA (1999) Weaners and aspirin. In ‘Mackinnon Project Newsletter’. (Mackinnon Project, University of Melbourne: Melbourne)

Larsen JWA (2002) Small ruminants S102 - weaner management and illthrift. In ‘Veterinary Science Lecture Notes’. (University of Melbourne: Werribee, Victoria)

Lean GR, Vizard AL, Webb Ware JK (1997) Changes in productivity and profitability of woolgrowing farms that follow recommendations from agricultural and veterinary studies. Australian Veterinary Journal 75, 726-731.

Leung K-M, Elashoff RM, Abdelmonem AA (1997) Censoring issues in survival analysis. Annual Review of Public Health 18, 83-104.

Lightfoot RJ (1967) The effects of shearing time and stocking rate on wool production. Journal of Agriculture of Western Australia 8, 90-98.

Lindsey JC, Ryan LM (1998) Tutorial in biostatistics: Methods for interval-censored data. Statistics In Medicine 17, 219-238.

Lindsey JK (1998) A study of interval censoring in parametric regression models. Lifetime Data Analysis 4, 329-354.

Little DL, Carter ED, Ewers AL (1993) Liveweight change, wool production and wool quality of Merino lambs grazing barley grass pastures sprayed to control grass or unsprayed. Wool Technology and Sheep Breeding 41, 369-378.

Lloyd Davies H (1962) Studies on time of lambing in relation to stocking rate in south-western Australia. Proceedings of the Australian Society of Animal Production 4, 113-120.

Lloyd Davies H, Boundy CAP, Southey IN (1968) Summer nutrition of weaner sheep in south-western Australia. Proceedings of the Australian Society of Animal Production 7, 255-258.

Lloyd Davies H (1983) Some aspects of the production of weaner sheep in the winter rainfall regions of Australia. Livestock Production Science 10, 239-252.

Lloyd Davies H (1987) Studies of time of lambing in ewes grazing subterranean clover based pasture in relation to stocking rates in south-western Australia. Western Australian Department of Agriculture Technical Bulletin.

Lloyd Davies H, McRae IR, Petrie DJ (1988) Feeding female weaner Merino sheep during drought on the central tablelands of New South Wales. Proceedings of the Australian Society of Animal Production 17, 222-225.

Love KJ, Egan JK, McIntyre JS (1978) The effect of partial shearing on the voluntary food intake and production of Merino wethers. Proceedings of the Australian Society of Animal Production 12, 269.

Lynch JJ, Alexander G (1976) The effect of gramineous windbreaks on behaviour and lamb mortality among shorn and unshorn Merino sheep during lambing. Applied Animal Ethology 2, 305-325.

Lynch JJ, Alexander G (1977) Sheltering behaviour of lambing Merino sheep in relation to grass hedges and artificial windbreaks. Australian Journal of Agricultural Research 28, 691-701.

Lynch JJ, Alexander G (1980) The effect of time since shearing on sheltering behaviour by Merino sheep. Proceedings of the Australian Society of Animal Production 13, 325-328.

Lynch JJ, Mottershead BE, Alexander G (1980) Sheltering behaviour and lamb mortality amongst shorn Merino ewes lambing in paddocks with a restricted area of shelter or no shelter. Applied Animal Ethology 6, 163-174.

171 Bibliography

MacKenzie AJ, Thwaites CJ, Edey TN (1975) Oestrus, ovarian and adrenal response of the ewe to fasting and cold stress. Australian Journal of Agricultural Research 26, 545-551.

Mandonnet N, Ducrocq V, Arquet R, Aumont G (2003) Mortality of Creole kids during infection with gastrointestinal strongyles: A survival analysis. Journal of Animal Science 81, 2401-2408.

Marshall T (1985) Effect of month of birth and 1st summers nutrition on the productivity of Merino wethers. Australian Journal of Experimental Agriculture 25, 777-782.

Masters DG, Gherardi SG, Mata G, Greeff J (1997) Managing staple strength in winter rainfall environments. Proceedings of the Association for the Advancement of Animal Breeding and Genetics 12, 709-713.

Masters DG, Mata G, Liu SM, Peterson AD (1998) Influence of liveweight, liveweight change and diet on wool growth, staple strength and fibre diameter in young sheep. Australian Journal of Agricultural Research 49, 269-277.

Masters DG, Scrivener CJ, Mata G, Hygate L (2000) Components of staple strength in young sheep from south eastern Victoria. Asian-Australasian Journal of Animal Science 13, 177.

Mata G, Peter DW, Purser DB (1990) Dietary changes, staple strength and point of break. Proceedings of the Nutrition Society of Australia 15, 132.

Mata G, Adams NR, O’Dea T et al. (1999) Is there a specific weakness in staple strength around the break of season? Australian Journal of Experimental Agriculture 39, 401-409.

Mata G, Schroder P, Masters DG (2002) Management of intake in winter to control micron blowout and improve staple strength. Wool Technology and Sheep Breeding 50, 471-476.

McCutcheon SN, Holmes CW, McDonald MF (1981) The starvation-exposure syndrome and neonatal lamb mortality: A review. Proceedings of the New Zealand Society of Animal Production 41, 209-217.

McDonald JW (1975) Selenium-responsive unthriftiness of young Merino sheep in central Victoria. Australian Veterinary Journal 51, 433-435.

McFarland IJ, Shaw SJ (1998) Sustainable wool production - a Western Australian perspective. Proceedings of the Australian Society of Animal Production 22, 400.

McFarlane JD (1965) The influence of seasonal pasture production and grazing management on seasonal wool growth. Australian Journal of Experimental Agriculture and Animal Husbandry 5, 252- 261.

McGarry WL, Stott IK (1960) Autumn v spring shearing. Journal of Agriculture, Western Australia 1, 147-149.

McGuirk BJ, Paynter JR, Dun RB (1966) The effect of frequency and time of shearing on the reproduction and wool growth of Bungaree South Australian Merino ewes. Australian Journal of Experimental Agriculture and Animal Husbandry 6, 305-310.

McLaughlin JW (1966) The effect of timing of birth and age at weaning upon the adult production of Corriedale sheep. Proceedings of the Australian Society of Animal Production 6, 246.

McLaughlin JW (1967) Management and weaner illthrift. Victorian Veterinary Proceedings 25, 60-63.

McLaughlin JW (1973) Management of weaner sheep in western Victoria. 2. The effects of supplements of oat grain or pasture hay or the periodic grazing of a green fodder crop upon current and subsequent production. Australian Journal of Experimental Agriculture and Animal Husbandry 13, 637-642.

172 Bibliography

McLeod DD (1967) When should we shear? Tasmanian Journal of Agriculture 38, 130-137.

McManus WR, Arnold GW, Paynter JR (1964) Studies in the wool production of sheep. 2: Variation in wool characteristics with season and stocking rate. Australian Journal of Experimental Agriculture and Animal Husbandry 4, 404-411.

McMillan WH, Knight TW (1982) Shearing and time of joining effects on reproduction in two-tooth ewes. Proceedings of the New Zealand Society of Animal Production 42, 45-46.

Mersmann HJ (1987) Nutritional and endocrinologic influences on the composition of animal growth. Progress in Food and Nutrition Science 11, 175-201.

Minitab Inc. (2005) ‘Minitab release 14.2.’ (Minitab Inc.: State College, Pennsylvania)

Morcombe PW, Allen JG (1993) Improving production from September-born Merino lambs with a lupin grain supplement and phomopsis-resistant lupin stubbles. Australian Journal of Experimental Agriculture 33, 713-719.

Morley FHW (1983a) Date of joining. In ‘Proceedings no. 67: Sheep production and preventive medicine’. (Ed. TG Hungerford) pp. 83, 95. (University of Sydney Post-Graduate Foundation in Veterinary Science: Sydney)

Morley FHW (1983b) Blowfly strike. In ‘Proceedings no. 67: Sheep production and preventive medicine’. (Ed. TG Hungerford) pp. 103-114. (University of Sydney Post-Graduate Foundation in Veterinary Science: Sydney)

Morley FHW (1994) ‘Merinos, money and management.’ (Post Graduate Committee in Veterinary Science, University of Sydney: Sydney)

Morris ST (1997) The influence of pre-lamb shearing on lamb birthweight and survival. In ‘Proceedings of the 27th seminar of the society of sheep and beef cattle veterinarians NZVA’ pp. 91- 96. (Foundation For Continuing Education of the New Zealand Veterinary Association: Palmerston North)

Morris ST, McCutcheon SN, Revell DK (2000) Birth weight responses to shearing ewes in early to mid gestation. Animal Science 70, 363-369.

Moule GR (1966) Ovine production in tropical Australia. Australian Veterinary Journal 42, 13-18.

Mulhearn CJ (1958) Unthriftiness of weaner sheep in South Australia. Australian Veterinary Journal 34, 383-390.

Mullaney PD, Brown GH, Young SSY, Hyland PG (1969) Genetic and phenotypic parameters for wool characteristics in fine-wool Merino, Corriedale, and Polwarth sheep 1. Influence of various factors on production. Australian Journal of Agricultural Research 20, 1161-1176.

Murray WK, Mortimer SI (2003) ‘Agfact A3.3.41: Scoring sheep for fleece rot.’ (NSW Agriculture: Orange, NSW)

Newman SAN, Maddever DC, Sumner RMW (1996) Effect of lamb shearing on fine wool Merino hogget performance. In ‘Proceedings of the New Zealand society of animal production’. (Ed. DG Elvidge) pp. 328-331New Zealand Society of Animal Production, Hamilton, New Zealand)

Nguti R, Janssen P, Rowlands GJ et al. (2003) Survival of Red Maasai, Dorper and crossbred lambs in the sub-humid tropics. Animal Science 76, 3-17.

Nilon P (1986) Nutritional ill-thrift in weaner sheep: Causes and management options. Master of Veterinary Studies thesis, University of Melbourne.

173 Bibliography

Nivsarkar AE, Acharya RM, Kumar M (1982) Note on factors influencing preweaning and post- weaning survival of Malpura and Sonadi lambs. Indian Journal of Animal Sciences 52, 835-837.

Norris RT (1984) Management, growth and mortality in weaner sheep. In ‘Western Australian department of agriculture division of animal health 1984 scientific conference’. (Eds FC Wilkinson and JG Allen) pp. 72-89. (Western Australian Department of Agriculture: Perth)

Norris RT (1986) ‘Technical bulletin: Survey of post-weaning management, growth and mortality of Merino weaners.’ (Western Australian Department of Agriculture: Perth)

Oldham CM (2000) The importance of staple strength to wool profit. In ‘Managing wool production for quality and profit’. (Eds NR Adams, CM Oldham, RJ Suiter and GM Wilson) pp. 59-72. (Cooperative Research Centre For Premium Quality Wool: Perth)

Orwin DFG, Woods JL, Elliott KH (1980) Composition of the cortex of sound and tender wools. In ‘Proceedings of the 6th international wool textile research conference’ pp. 193-205. (IWTO: Pretoria)

Panaretto BA (1967) Some factors affecting off-shear losses in sheep. Wool Technology and Sheep Breeding 14, 27-31.

Panaretto BA (1968) Some metabolic effects of cold stress on undernourished non-pregnant ewes. Australian Journal of Agricultural Research 19, 273-282.

Parer JT (1963) The effects of wool length and nutrition on heat reactions of Merino sheep in the field. Australian Journal of Experimental Agriculture and Animal Husbandry 3, 243-248.

Parr RA, Davis IF, Tilbrook AJ (1989) Effect of shearing on oestrus and ovulation in sheep. Journal of Agricultural Science 113, 411-412.

Peter DW, Doyle PT, Curtis KM (1993) Supplements for weaners. In ‘Proceedings of a national workshop on management for wool quality in Mediterranean environments’. (Eds PT Doyle, JA Fortune and NR Adams) pp. 132-141. (Department of Agriculture, Western Australia: Perth)

Peterson AD, Gherardi SG, Doyle PT (1998) Components of staple strength in fine and broad wool Merino hoggets run together in a Mediterranean environment. Australian Journal of Agricultural Research 49, 1181-1186.

Peterson AD, Greeff JC, Oldham CM et al. (2000) Management of staple strength on farm. Asian- Australasian Journal of Animal Science 13C, 22-24.

Plant JW (2006) Infertility in the ewe. In ‘Proceedings of the Australian sheep veterinarians conferences’. (Ed. CL Trengrove) pp. 105-126. (Australian Veterinary Association: Wagga Wagga & Hobart)

Ponzoni RW (1986) A profit equation for the definition of the breeding objective of Australian Merino sheep. Journal of Animal Breeding and Genetics-Zeitschrift Fur Tierzuchtung Und Zuchtungsbiologie 103, 342-357.

Quinn H, Fitzpatrick L, Whatley J, Fry P (2005) Farm monitor project summary of results 2004/2005. Department of Primary Industries, Melbourne.

Raadsma HW (1988) Flystrike. In ‘Proceedings no. 110: Sheep health and production’. (Ed. TG Hungerford) pp. 317-337. (University of Sydney Post-graduate Committee in Veterinary Science: Sydney)

Radke BR (2003) A demonstration of interval-censored survival analysis. Preventive Veterinary Medicine 59, 241-256.

174 Bibliography

Reeve IJ, Thompson L (2004) Integrated parasite management - sheep: National survey. (Institute for Rural Futures: Available at http://www.woolinnovation.com.au/LivePage.aspx?PageId=10121 (verified 13/7/2006)

Reid RL (1962) Studies on the carbohydrate metabolism of sheep. XV. The adrenal response to the climatic stresses of cold, wind and rain. XVI. Partition of ketone bodies in blood, tissues and urine. Australian Journal of Agricultural Research 13, 296-319.

Reis PJ (1992) Variation in the strength of wool fibres - a review. Australian Journal of Agricultural Research 43, 1337-1351.

Revell DK, Morris ST, Cottam YH et al. (2002) Shearing ewes at mid-pregnancy is associated with changes in foetal growth and development. Australian Journal of Agricultural Research 53, 697-705.

Ripper L (2003) Gippsland pilots weaner management course. In ‘BestWool 2010 Information Exchange’. p. 2.

Ritchie AJM, Ralph IG, Ellis MR (1992) The accumulation of vegetable matter in spring and autumn shorn wool in a Mediterranean environment. Proceedings of the Australian Society of Animal Production 19, 142.

Robertson SM, Robards GE, Wolfe EC (1996) Factors contributing to the variation in staple strength of reproducing ewes. Proceedings of the Australian Society of Animal Production 21, 166-169.

Robertson SM, Robards GE, Wolfe EC (2000a) Grazing management of reproducing ewes affects staple strength. Australian Journal of Experimental Agriculture 40, 783-794.

Robertson SM, Robards GE, Wolfe EC (2000b) The timing of nutritional restriction during reproduction influences staple strength. Australian Journal of Agricultural Research 51, 125-132.

Rogan IM, Atkins KD, Coelli KA (1995) Shearing Merino ram weaners - is it worth it? Wool Technology and Sheep Breeding 43, 56-68.

Rose M (1972) Vital statistics for an experimental flock of Merino sheep in north west Queensland. Proceedings of the Australian Society of Animal Production 9, 48-54.

Rothman KJ (1986a) Stratified analysis. In ‘Modern epidemiology’ pp. 205-206. (Little, Brown and Company: Boston)

Rothman KJ (1986b) Stratified analysis. In ‘Modern epidemiology’ p. 183. (Little, Brown and Company: Boston)

Royston P, Parmar MKB (2002) Flexible parametric proportional-hazards and proportional-odds models for censored survival data, with application to prognostic modelling and estimation of treatment effects. Statistics In Medicine 21, 2175-2197.

Russell P (1968) The development of high stocking rates on a whole farm basis. Proceedings of the Australian Society of Animal Production 7, 103-105.

Sackett DM (1990) When to shear lambs. In ‘Holmes and Sackett on-farm’ pp. 78-79. (Holmes & Sackett: Pymble, NSW)

Sackett DM (2006) Assessing the economic cost of endemic disease on the profitability of Australian beef cattle and sheep producers. Meat and Livestock Australia Limited, North Sydney.

Salmon L, Moore A, Burbidge G, Woods A (2006) Can changing shearing date improve returns for fine wool producers? (CSIRO: Canberra) Available at https://partnerportal.csiro.au/grazplan (verified 15/6/2006)

175 Bibliography

SAS Institute Inc. (2004) ‘SAS/stat.’ (SAS Institute Inc.: Cary, North Carolina)

Saul GR (1988) The effect of season of birth, breed and pasture type on the fibre strength of wool from weaner sheep. Proceedings of the Australian Society of Animal Production 17, 467.

Scaramuzzi RJ (1988) Reproduction research in perspective. Proceedings of the Australian Society of Animal Production 17, 57-73.

Schlink AC, Hynd PI (1994). In ‘Improving wool strength’. (Ed. NR Adams) Proceedings of the Australian society of animal production pp. 50-51. (Australian Society of Animal Production: Perth)

Schlink AC, Mata G, Lewis RM (1998) Consequences of differing wool growth rates on staple strength of Merino wethers with divergent staple strengths. Wool Technology and Sheep Breeding 46, 271-285.

Schlink AC, Mata G, Lea JM, Ritchie AJM (1999) Seasonal variation in fibre diameter and length in wool of grazing Merino sheep with low or high staple strength. Australian Journal of Experimental Agriculture 39, 507-517.

Schlink AC (2000) What is staple strength? In ‘Managing wool production for quality and profit’. (Eds NR Adams, CM Oldham, RJ Suiter and GM Wilson) pp. 74-77. (Premium Quality Wool Cooperative Research Centre:

Schlink AC, Peterson AD, Huson M, Thompson AN (2000a) Components of staple strength. Asian- Australasian Journal of Animal Science 13C, 21-22.

Schlink AC, Ritchie AJM, Lea JM (2000b) Effect of break of season on short-term fibre diameter changes in low and high staple strength Merino sheep. Asian-Australasian Journal of Animal Science 23, 118.

Scrivener C, Vizard AL (1997) The relationship between reproductive performance, time of lambing and the staple strength of wool from Merino ewes. Wool Technology and Sheep Breeding 45, 120-128.

Sharma MM, Mathur PB, Kumar PN, Uppal PK (1981) Causes of mortality in lambs under semi-arid climatic condition of Rajasthan. Indian Veterinary Journal 58, 859-864.

Slee J (1966) Variation in the responses of shorn sheep to cold exposure. Animal Production 8, 425.

Southey BR, Rodriguez-Zas SL, Leymaster KA (2001) Survival analysis of lamb mortality in a terminal sire composite population. Journal of Animal Science 79, 2298-2306.

Southey BR, Rodriguez-Zas SL, Leymaster KA (2003) Discrete time survival analysis of lamb mortality in a terminal sire composite population. Journal of Animal Science 81, 1399-1405.

StataCorp (2005) ‘Intercooled Stata version 8.2 for windows.’ (StataCorp LP: College Station, Texas)

Statham M (2004) Feed gaps and fibre variations. The development of along staple diameter measurement as a diagnostic tool for improved grazing management. Tasmanian Institute of Agricultural Research, Report on the 8x5 Project, Kings Meadows, Tasmania.

Sumner RMW (1984) Effect of shearing on meat and wool production from lambs reared on hill country. Proceedings of the New Zealand Society of Animal Production 44, 45-48.

Sykes AR, Slee J (1969) Cold exposure of Southdown and Welsh Mountain sheep. 1. Effects of breed, plane of nutrition and acclimatization to cold upon resistance to body cooling. Animal Production 11, 65-75.

Taylor R (1989) Hidden costs of lamb shearing. In ‘Mackinnon Project Newsletter’. (Mackinnon Project, University of Melbourne: Melbourne)

176 Bibliography

Thompson AN, Hynd PI (1998) Wool growth and fibre diameter changes in young Merino sheep genetically different in staple strength and fed different levels of nutrition. Australian Journal of Agricultural Research 49, 889-898.

Thornberry KJ, Denney GD, Sladek MA (1988) Variation in wool staple strength, staple length, and position of break among experimental sheep on commercial properties. Proceedings of the Australian Society of Animal Production 17, 475.

Thornton RF, Hood RL, Jones PN, Re VM (1979) Compensatory growth in sheep. Australian Journal of Agricultural Research 30, 135-151.

Torell DT, Weir WC, Bradford GE, Spurlock GM (1969) Effects of time of shearing on wool and lamb production. California Agriculture 23, 16-18.

Tucker (1964) Shearing carry-over lambs: Does it pay? Agricultural Gazette of New South Wales 75, 1318-1321.

Turkson PK (2003) Lamb and kid mortality in village flocks in the coastal savannah zone of Ghana. Tropical Animal Health and Production 35, 477-490.

Turkson PK, Antiri YK, Baffuor-Awuah O (2004) Risk factors for kid mortality in West African Dwarf goats under an intensive management system in Ghana. Tropical Animal Health and Production 36, 353-364.

Turkson PK, Sualisu M (2005) Risk factors for lamb mortality in Sahelian sheep on a breeding station in Ghana. Tropical Animal Health and Production 37, 49-64.

Turnbull BW (1976) The empirical distribution function with arbitrarily grouped, censored and truncated data. Journal of the Royal Statistical Society, Series B 38, 290-295.

Vizard AL, Foot JZ (1994) Animal and pasture maintenance, a partnership. In ‘Pasture management technology for the 21st century’. (Eds DR Kemp and DL Michalk) pp. 100-113. (CSIRO Australia: East Melbourne)

Vizard AL, Scrivener C, Anderson GA (1994) The agreement and repeatability of Atlas and Staplebreaker for estimating the staple strength of midside wool samples taken from Merino sheep. Wool Technology and Sheep Breeding 42, 112-119.

Vizard AL (1997) The Roxby park experience and plans. In ‘Productive pasture systems: invited papers and posters and contributed posters from the annual conference held at the Performing Arts Centre, Hamilton’. Hamilton. (Grasslands Society of Victoria Inc.)

Vizard AL, Hansford KA (1999) A comparison of the topmaking performance of wool from sheep selected by index and visual methods. Australian Journal of Experimental Agriculture 39, 941-948.

Walker DM, Hunt SG (1980) Effects of birth-weight and energy gain preweaning on survival of lambs weaned at 21 days. Australian Journal of Agricultural Research 31, 981-989.

Walker SK, Hall GP, Smith DH et al. (1979) Effect of selenium supplementation on survival, liveweight and wool weight of young sheep on Kangaroo Island, South Australia. Australian Journal of Experimental Agriculture and Animal Husbandry 19, 689-694.

Wardhaugh KG, Morton R (1990) The incidence of flystrike in sheep in relation to weather conditions, sheep husbandry, and the abundance of the Australian sheep blowfly, Lucilia cuprina (Wiedemann) (Diptera, Calliphoridae). Australian Journal of Agricultural Research 41, 1155-1167.

Warr GJ, Thompson JM (1976) Liveweight change and intake of lambs as affected by seed infestation. Proceedings of the Australian Society of Animal Production 11, 173-176.

177 Bibliography

Warr GJ, Gilmour AR, Wilson NK (1979) Effect of shearing time and location on vegetable matter components in the New South Wales woolclip. Australian Journal of Experimental Agriculture and Animal Husbandry 19, 684-688.

Watts JE, Murray MD, Graham NPH (1979) The blowfly strike problem of sheep in New South Wales. Australian Veterinary Journal 55, 325-334.

Webster MED, Lynch JJ (1966) Some physiological and behavioural consequences of shearing. Proceedings of the Australian Society of Animal Production 6, 234-239.

Weir WC, Torell DT (1967) Supplemental feeding of sheep grazing on dry range. University of California, Division of Agricultural Science Bulletin 832.

Wheeler JL, Reardon TF, Lambourne LJ (1963) The effect of pasture availability and shearing stress on herbage intake of grazing sheep. Australian Journal of Agricultural Research 14, 364-372.

Wilkinson FC (1981) Lamb survival from marking and weaner illthrift. In ‘Proceedings no. 58: Refresher course on sheep’. (Ed. TG Hungerford) pp. 193-209. (University of Sydney Post-Graduate Foundation in Veterinary Science: Sydney)

Williams OB, Chapman RE (1966) Additional information on the dye-banding technique of wool growth measurement. Journal of the Australian Institute of Agricultural Science 32, 298-300.

Wodzick-Tomaszewska M (1963) The effect of shearing on the appetite of sheep. New Zealand Journal of Agricultural Research 6, 440-447.

Woodgate RG, Chapman HM, Robertson ID, Bell KJ (2000) Summer-autumn rainfall effects on wool staple strength and position of break. I. Small-scale field simulations of rainfall onto sheep grazing dry pasture in February and April. Australian Journal of Agricultural Research 51, 523-530.

Woodgate RG, Chapman HM, Robertson ID, Bell KJ (2001) Summer-autumn rainfall effects on wool staple strength and position of break. II. Rainfall simulations, with or without wind, on sheep on days of different ambient temperatures. Australian Journal of Agricultural Research 52, 427-432.

Yang S, Lamb PR (1997) A literature survey on wool fibre intrinsic strength. CSIRO Division of Wool Technology, Report WTC 97.03, Geelong.

178

A PPENDICES

179

180

A P P E N D I X 1

FERTILISER APPLICATION & PASTURE MANAGEMENT DETAILS ON FARM A

Table A-1.1: Annual fertiliser application and pasture management details for Farm A

Date Fertiliser type Application Pasture spray management rate (kg/ha)

Ewe paddock 1988–1993 Straight super 125 1990: Lemat for red legged earthmite (RLE) 1994 Straight super 150 1993: MCPA spot control of thistles/capeweed 1995–1996 3 in1 290 1997–1998 Pasture Gold 90 1997: " 1999 Pasture Gold 80 1999: Spray top with Round Up® (glycoside) to control barley grass 2000 Pasture Gold 110 2000: Endosulfan for RLE 2001 Goldphos 10 70 2002 MCPA spot control

Weaner paddock 1995 DAP 100 1995: Lemat for RLE 1996 Pasture Gold 137 1997–1998 Pasture Gold 110 1998: MCPA spot control of thistles/capeweed 1999 Pasture Gold 90 1999: Spray top with Round Up® to control barley grass 2000 Pasture Gold 110 2000: Endosulfan for RLE 2001 Goldphos 10 70 2002: MCPA spot spray 2002 Goldphos 10 70

181

182

A P P E N D I X 2

FLEECE VALUES BASED ON FIRST & NINTH DECILE HISTORICAL MICRON BASIS PRICES

A2.1A2.1A2.1 EwesEwesEwes

Figure A-2.1: Estimated fleece value ($/head) produced by each ewe shearing group, based on median, first decile (lower bar) and ninth decile (upper bar) micron basis prices and median staple length and staple strength discounts

$60.00 $55.00 $50.00 $45.00 $40.00 $35.00 DEC $30.00 MAR $25.00 MAY $20.00

Mean fleeceMean value $15.00 $10.00 $5.00 $0.00 1 2 3 4 5 Mean Trial year

183 Appendix 2

A2.2A2.2A2.2 WeanersWeanersWeaners

Figure A-2.2: Estimated fleece values ($/head) from shearing 1 ( ) and shearing 2 ( ) of each weaner shearing group, calculated using first decile historical basis prices and median staple length & strength premiums and discounts

$50 $45 Historic 1st decile basis price $40 $35 $30 $25 $20 $15

Mean fleeceMean value $10 $5 $0 OCT OCT OCT OCT OCT DEC-DEC DEC-DEC DEC-DEC DEC-DEC DEC-DEC JUN-MAR JUL-MAY JUN-MAR JUL-MAY JUN-MAR JUL-MAY JUN-MAR JUL-MAY JUN-MAR JUL-MAY MAR-MAR MAR-MAR MAR-MAR MAR-MAR MAR-MAR MAY-MAY MAY-MAY MAY-MAY MAY-MAY MAY-MAY 1 2 3 4 Mean SHEARING GROUP Weaner cohort

Figure A-2.3: Estimated fleece values ($/head) from shearing 1 ( ) and shearing 2 ( ) of each weaner shearing group, calculated using ninth decile historical basis prices and median staple length & strength premiums and discounts

$150 Using 9th decile historic basis price $140 $130 $120 $110 $100 $90 $80 $70 $60 $50 Mean fleeceMean value $40 $30 $20 $10 $0 OCT OCT OCT OCT OCT DEC-DEC DEC-DEC DEC-DEC DEC-DEC DEC-DEC JUN-MAR JUL-MAY JUN-MAR JUL-MAY JUN-MAR JUL-MAY JUN-MAR JUL-MAY JUN-MAR JUL-MAY MAR-MAR MAR-MAR MAR-MAR MAR-MAR MAR-MAR MAY-MAY MAY-MAY MAY-MAY MAY-MAY MAY-MAY 1 2 3 4 Mean SHEARING GROUP Weaner cohort

184 Appendix 2

A2.3A2.3A2.3 Lifetime Fleece Values

Figure A-2.4: Estimated value ($/head) of wool produced over the lifetime (from birth to 75 months of age) of ewes shorn in each weaner and ewe shearing combination, net of shearing costs, at first, median and ninth decile historical micron basis prices with median staple length and strength premiums and discounts

$260 $240 $220 $200 $180 $160 $140 $120 $100 $80 $60 $40 $20 $0 -$20

Wool value net of shearing ($/head) shearing of net value Wool DEC-DEC OCT MAR-MAR JUN-MAR MAY-MAY JUL-MAY

DEC MAR MAY Weaner shearing group Ewe shearing group

1st w eaner 2nd w eaner Maiden Adult shearings Total value using 1st OR 9th decile micron basis price

Figure A-2.5: Estimated value ($/head) of wool produced over the lifetime of ewes shorn in each weaner and ewe shearing combination, net of shearing costs & accounting for cumulative survival, at first, median and ninth decile historical micron basis prices with median staple length and strength discounts

$180 $170 $160 $150 $140 $130 $120 $110 $100 $90 $80 $70 $60 $50 $40 $30 $20 $10 $0 -$10 Net wool value * survivorship ($/head) survivorship * value wool Net DEC-DEC OCT MAR-MAR JUN-MAR MAY-MAY JUL-MAY -$20 DEC MAR MAY Weaner shearing group Ewe shearing group

1st w eaner 2nd w eaner Maiden Adult shearings Total value using 1st OR 9th decile micron basis prices

185

186

A P P E N D I X 3

PRICE SCHEDULES USED TO CALCULATE FLEECE VALUES

Table A-3.1: First, median and ninth decile basis prices in the period July 1992–February 2006 for wool of specified fibre diameter (FD; µm), and first, median and ninth decile percentage changes from basis price for wool of specified staple length (SL; mm) and staple strength (SS; N/ktex)

Deciles of micron Deciles of change (%) from basis price * FD range basis price for specified staple length and strength 1 Median 9 SL range SS range 1 Median 9 15.1 15.5 1496 2685 10704 55 64 25 29 -61% -43% -30% 15.1 15.5 1496 2685 10704 65 74 20 24 -51% -38% -22% 15.6 16.0 1397 2153 4432 36 44 40 49 -73% -54% -37% 15.6 16.0 1397 2153 4432 45 54 35 39 -51% -41% -35% 15.6 16.0 1397 2153 4432 55 64 20 24 -63% -55% -39% 15.6 16.0 1397 2153 4432 55 64 25 29 -58% -45% -34% 15.6 16.0 1397 2153 4432 55 64 30 34 -56% -38% -28% 15.6 16.0 1397 2153 4432 65 74 20 24 -54% -35% -15% 15.6 16.0 1397 2153 4432 95 104 <19 -48% -35% -22% 16.1 16.5 1353 1888 3546 36 44 40 49 -70% -52% -36% 16.1 16.5 1353 1888 3546 45 54 30 34 -68% -62% -56% 16.1 16.5 1353 1888 3546 55 64 20 24 -64% -49% -34% 16.1 16.5 1353 1888 3546 55 64 30 34 -47% -39% -23% 16.1 16.5 1353 1888 3546 65 74 <19 -53% -43% -32% 16.1 16.5 1353 1888 3546 65 74 20 24 -53% -38% -27% 16.1 16.5 1353 1888 3546 95 104 <19 -38% -30% -22% 16.1 16.5 1353 1888 3546 95 104 25 29 -36% -17% -5% 16.6 17.0 1158 1700 2743 36 44 carding -62% -48% -34% 16.6 17.0 1158 1700 2743 55 64 30 34 -44% -32% -22% 17.1 17.5 1001 1531 2330 26 35 carding -67% -56% -36% 17.1 17.5 1001 1531 2330 85 94 25 29 -30% -18% -4% 17.1 17.5 1001 1531 2330 95 104 <19 -41% -25% -15% 17.1 17.5 1001 1531 2330 95 104 20 24 -36% -24% -9% 17.1 17.5 1001 1531 2330 95 104 25 29 -30% -15% -4% 17.6 18.0 941 1284 1910 26 35 carding -62% -55% -32% 17.6 18.0 941 1284 1910 65 74 30 34 -21% -14% -7% 17.6 18.0 941 1284 1910 65 74 35 39 -11% -5% -1% 17.6 18.0 941 1284 1910 75 79 30 34 -15% -9% -2% 17.6 18.0 941 1284 1910 85 94 30 34 -14% -8% -2% 17.6 18.0 941 1284 1910 85 94 40 49 -2% 6% 15%

* For wool of staple length 85–94 mm and staple strength 35–39 N/ktex

187 Appendix 3

(continued from previous page) Deciles of micron Deciles of change (%) from basis price FD range basis price for specified staple length and strength 1 Median 9 SL range SS range 1 Median 9 17.6 18.0 941 1284 1910 95 104 20 24 -34% -22% -8% 18.1 18.5 839 1135 1638 26 35 carding -59% -46% -32% 18.1 18.5 839 1135 1638 75 84 35 39 -3% 0% 3% 18.1 18.5 839 1135 1638 85 94 25 29 -22% -11% -3% 18.1 18.5 839 1135 1638 85 94 30 34 -11% -5% -1% 18.1 18.5 839 1135 1638 85 94 35 39 0% 0% 0% 18.1 18.5 839 1135 1638 85 94 40 49 -1% 4% 8% 18.1 18.5 839 1135 1638 95 104 25 29 -21% -12% -3% 18.1 18.5 839 1135 1638 95 104 30 34 -14% -7% -1% 18.1 18.5 839 1135 1638 115 124 30 34 -23% -15% -4% 18.6 19.0 774 1019 1373 75 79 35 39 -2% 0% 2% 18.6 19.0 774 1019 1373 85 94 30 34 -7% -3% -1% 18.6 19.0 774 1019 1373 95 104 25 29 -16% -9% -2% 19.1 19.5 731 934 1201 65 74 35 39 -8% -4% -1% 19.1 19.5 731 934 1201 75 79 30 34 -6% -3% -1% 19.1 19.5 731 934 1201 75 79 35 39 -2% 0% 1% 19.1 19.5 731 934 1201 75 79 40 49 -1% 1% 3% 19.1 19.5 731 934 1201 80 84 25 29 -13% -8% -2% 19.1 19.5 731 934 1201 85 94 25 29 -12% -7% -2% 19.1 19.5 731 934 1201 85 94 30 34 -5% -2% 0% 19.1 19.5 731 934 1201 85 94 35 39 0% 0% 0% 19.1 19.5 731 934 1201 95 104 30 34 -7% -3% 0% 19.1 19.5 731 934 1201 95 104 35 39 -4% -1% 1% 19.6 20.0 668 860 1064 55 64 35 39 -23% -15% -7% 19.6 20.0 668 860 1064 65 74 35 39 -8% -4% -1% 19.6 20.0 668 860 1064 75 84 30 34 -5% -2% 0% 19.6 20.0 668 860 1064 75 84 35 39 -2% 0% 1% 19.6 20.0 668 860 1064 75 84 40 49 -1% 1% 2% 19.6 20.0 668 860 1064 85 94 25 29 -12% -6% -1% 19.6 20.0 668 860 1064 85 94 30 34 -5% -2% 0% 19.6 20.0 668 860 1064 85 94 35 39 0% 0% 0% 19.6 20.0 668 860 1064 85 94 40 49 0% 1% 3% 19.6 20.0 668 860 1064 95 104 30 34 -6% -2% 0% 20.1 20.5 617 769 1017 85 94 35 39 0% 0% 0% 20.1 20.5 617 769 1017 85 94 40 49 0% 1% 2%

188

Minerva Access is the Institutional Repository of The University of Melbourne

Author/s: CAMPBELL, ANGUS

Title: The effect of time of shearing on wool production and management of a spring-lambing merino flock

Date: 2006

Citation: Campbell, A. (2006). The effect of time of shearing on wool production and management of a spring-lambing merino flock, PhD thesis, Veterinary Science, University of Melbourne.

Publication Status: Unpublished

Persistent Link: http://hdl.handle.net/11343/39234

File Description: The effect of time of shearing on wool production and management of a spring-lambing merino flock

Terms and Conditions: Terms and Conditions: Copyright in works deposited in Minerva Access is retained by the copyright owner. The work may not be altered without permission from the copyright owner. Readers may only download, print and save electronic copies of whole works for their own personal non-commercial use. Any use that exceeds these limits requires permission from the copyright owner. Attribution is essential when quoting or paraphrasing from these works.