OTHER TITLES IN THIS SERIES
1.

Guide Dogs for the Blind: their Selection, Development and Training
by Clarence J. Pfaffenberger et al. 1976 xii + 225 pp.

2.

Ethology of Free-Ranging Domestic Animals
by G.W. Arnold and M.L. Dudzinski 1978 xii + 198 pp.

3.

Bacterial Infection and Immunity in Domestic Animals
by J.B. Woolcock 1979 xii + 254 pp.

4.

Social Structure in Farm Animals
by G.J. Syme and L.A. Syme 1979 xii + 196 pp.

5.

Physiology and Control of Parturition in Domestic Animals
edited by F. Ellendorff, M. Taverne and D. Schmidt 1979 xii + 3 4 8 pp.

6.

Trends in Veterinary Pharmacology and Toxicology
Proceedings of the First European Congress on Veterinary Pharmacology and
Toxicology, held in Zeist, 2 5 - 2 8 September, 1979

edited by A.SJ.P.A.M. van Miert 1980 χ + 363 pp.

7.

Veterinary Toxicology
edited by M. Bartik and A. Piskac 1981 346 pp.

8.

Livestock Production in Europe
edited by R. Politiek 1982 viii + 336 pp.

9.

Advances in Veterinary Immunology 1981
edited by F. Kristensen and D.F. Antczak 1982 vi + 282 pp.

10.

Genetics and Animal Breeding
Part A: Biological and Genetical Foundations of Animal Breeding
Part B: Stock Improvement Methods
by J. Maciejowski and J. Zieba 1982 viii + 284 pp; 1982 viii + 206 pp.

11.

Constitutional Disorders and Hereditary Diseases in Domestic Animals
by D. Harnori 1983 730 pp.

12.


Advances in Veterinary Immunology 1982
edited by F. Kristensen and D.F. Antczak 1983 ν + 312 pp.

13.

Prostaglandins in Animal Reproduction
edited by L.E. Edqvist and H. Kindahl 1984 viii + 304 pp.

14.

Straw and other Fibrous By-Products as Feed
edited by F. Sundst^l and E. Owen 1984 xxii + 604 pp.

15.

Impact of Diseases on Livestock Production in the Tropics
edited by H.P. Riemann and M.J. Burridge 1984 χ + 6 3 2 pp.

16.

Advances in Veterinary Immunology 1983
edited by F.J. Bourne and N.T. Gorman 1984 ν + 259 pp.

17.

Safety and Quality in Food
Proceedings of a DSA Symposium "Wholesome Food for All.
Views of the Animal Health Industries", held in Brussels, 2 9 - 3 0 March 1984
edited by DSA (Bureau Européen d'Information pour le Développement de la Santé

Animale), Avenue de l'Yser 19, Β 1040 Brussels, Belgium 1984


Developments in Animal and Veterinary Sciences, 18

Reproductive
and developmental
behaviour in sheep
An anthology from " Applied Animal Ethology "

Edited by

A.F. Fraser
Professor of Comparative Physiology
Memorial University of

Newfoundland

St. John's, Newfound/and,

Canada

ELSEVIER
Amsterdam — Oxford — New York — Tokyo

www.pdfgrip.com

1985



ELSEVIER SCIENCE PUBLISHERS B.V.
Molenwerf 1
P.O. Box 2 1 1 , 1 0 0 0 AE Amsterdam, The Netherlands
Distributors for the United States and Canada:
ELSEVIER SCIENCE PUBLISHING COMPANY INC.
5 2 , Vanderbilt Avenue
New York, NY 10017

This volume consists of a selection of articles
reprinted from the journal "Applied Animal Ethology" (Vols. 1 - 1 1 ) .

ISBN 0444-42444-x (Vol. 18)
ISBN 0-44441703-6 (Series)
© Elsevier Science Publishers B.V., 1985
All rights reserved. No part of this publication may be reproduced, stored in a retrieval system or
transmitted in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without the prior written permission of the publisher, Elsevier Science Publishers B.V./Science
& Technology Division, P.O. Box 330, 1000 AH Amsterdam, The Netherlands.
Special regulations for readers in the USA — This publication has been registered with the Copyright
Clearance Center Inc. (CCC), Salem, Massachusetts. Information can be obtained from the CCC about
conditions under which photocopies of parts of this publication may be made in the USA. All other
copyright questions, including photocopying outside of the USA, should be referred to the publisher.
Printed in The Netherlands

www.pdfgrip.com


ix

Introduction
Of all the farm animals, the sheep is the one which remains closest to the

natural condition in traditional husbandry. Although intensive husbandry
is increasingly being applied to sheep, most sheep production around the
world is still based on the extensive system. Under these conditions, therefore, attempts to improve productivity require to take particular account
of reproductive and developmental behaviour. Successful sheep production
depends heavily on optimal reproduction and development. These topics
are presented here in related sections which make up this anthology of
selected publications on sheep behaviour. They have been taken from 10
years of "Applied Animal Ethology", from its inception in December 1974
to the end of its publication under that title in 1984.
During the first decade of "Applied Animal Ethology", a great many
papers were published on sheep behaviour. Of these, 39 have been drawn
together in four sections which deal with sexual behaviour, fetal and perinatal behaviour, ewe—lamb recognition and development. In order to be productive and to perpetuate themselves successfully, sheep clearly need proficiency in mating, birth and nursing. While other aspects of sheep behaviour
will deserve due attention in the future, the topic of each section, as mentioned, has clearly represented an area of special interest in applied ethology;
these topics undoubtedly relate to the realities of sheep breeding and to
biologically critical features of sheep behaviour.
The sheep is a seasonal-breeding animal. Although some breeds have breeding seasons which can extend over much of the year, most breeds only mate
around the time when there is less light than dark per day and when daily
light is diminishing. Those breeds which have had their origins in mountainous regions or where climatic conditions are most severe in the winter have
the most restricted breeding season. Such restriction results in a very sudden
emergence of reproductive activity among the males. This activity tends to
be of limited duration and, as a result, sheep libido comes close to resembling
the rut which occurs in many boreal forms of ungulate life (Fraser, 1968).
The strength of the breeding drive is obviously critical to optimum flock
reproduction.
Since it is critical that the newborn lamb has maximum viability at birth,
it is essential for it to survive the parturient experience without difficulty.

www.pdfgrip.com



χ

This calls for fetal vitality and the ability of the fetus to adopt an appropriate
posture for expulsion at birth. Awareness of fetal activity improves our
appreciation of fetal needs and obstetrical requirements in sheep birth.
Although there may be frequent lying down and rising before birth, most
ewes remain recumbent until the fetus is partially or completely expelled.
In cases of twin and multiple births, the neonates usually follow each other
within a matter of minutes. Birth is a short process, averaging 80 minutes in
duration. Most newborn lambs are able to stand within the first half-hour
following birth and nearly all are able to stand within the first two hours.
The lamb's first attempts to suck are usually unsuccessful; it often seeks out
the teat by nosing between the forelegs of the dam or any nearby object
which the lamb may feel has maternal properties. If, at this point, the newborn fails to find the teat or is prevented from doing so by the behaviour of
the mother, its drive to suck may become inhibited.
One of the most noticeable behavioural aspects of sheep in relation to a
newborn lamb is the strong maternal relationship which quickly develops.
A ewe nearly always vigorously rejects any attempts by other lambs to engage in sucking and will look after her own young exclusively. The nature of
the vital bond between ewe and lamb has been given particularly keen study.
Such scientific attention to the phenomenon is justified since this is the
fulcrum on which sheep production is balanced.
Once the newborn lamb is able to stand, it sucks and nibbles at any object
which is at hand; this may be the fleece of the ewe. While the ewe is removing
the placenta from the lamb, the latter finds its way to the region of the teats
and udder. Sometimes the lamb is prevented from undertaking sucking by
diligent efforts of the mother to remove the placenta. If the udder is too
large, the newborn may find the teat difficult to grasp. However, once the
newborn has sucked successfully its survival is largely assured. Within about
one hour of birth, approximately 60% of newborn lambs have begun to suck
and, in normal cases, nearly all lambs have sought out the udder within the

first two hours.
For the first few weeks of life, lambs stay in close proximity to their ewes.
By one month of age, young lambs spend two-thirds of their time in the
company of other lambs and play within such groups is a notable feature
of behaviour. Play becomes reduced as lambs grow and is of only occasional
occurrence as nursing terminates with weaning.
The behavioural life-cycle of sheep will have to be continually studied and
documented throughout future research work so that appropriate husbandry
can be maintained in the interests of productivity and welfare in this species
which is woven into the fabric of world-wide animal farming.
REFERENCE
Fraser, A.F., 1 9 6 8 . Reproductive Behaviour in Ungulates. Academic Press, London and
N e w York, 2 0 2 pp.

www.pdfgrip.com


SECTION I

ι

Sexual Behaviour
MATING ACTIVITY AND ITS RELATIONSHIP TO REPRODUCTIVE
PERFORMANCE IN MERINO SHEEP

D.G. FOWLER
New South Wales Department

of Agriculture,


CSIRO, Blacktown,

N.S.W.

(Australia)

ABSTRACT
Fowler, D.G., 1 9 7 5 . Mating activity and its relationship t o reproductive performance in
Merino sheep. Appl. Anim. EthoL, 1: 357—368.
Mating activity was observed directly o n 2 2 flocks of Merino sheep at five locations
spread over several hundred kilometres. The flocks were subjected t o differences in
nutritional, seasonal, climatic and topographic influences and varied in size, and in the
age and weight of ewes and rams. Each flock was joined with rams for mating under
paddock conditions.
In the first series of observations there were 1 6 flocks at five locations and it was found
that the sum of interservice intervals for all ewes, T, during an observation period was
highly correlated with the total number of services Ν (r> 0.89, Ρ < 0 . 0 0 5 ) .
This report is confined t o an analysis and interpretation of this relationship and the
second series of observations form part of the analysis.
In the second series of observations, six different flocks of Merino ewes in t w o different environments were used to study the relationship between mating activity and reproductive performance. Each flock was joined under paddock conditions and studied by
direct observation. It was found that the sum of interservice and intermount intervals for
ewes were highly correlated with the number of pregnancies (r > 0 . 8 7 , Ρ < 0 . 0 5 ) . The
ratio T/N was found t o be highly correlated with the number of foetuses for both mounts
(r = 0.88, Ρ < 0 . 0 5 ) and services (r = 0.88, Ρ < 0 . 0 5 ) .
The variables Τ and Ν are thus closely related to each other over a wide range of environments and their ratio, T/N accounts for the major proportion of the variation in the
fertility of six different flocks of Merino sheep in t w o different environments. These
results offer the possibility of the prediction of fertility in Merino sheep.

INTRODUCTION


The mating behaviour of domestic sheep under field conditions with minimal interference has not been examined although many millions of sheep are
joined with rams for mating under these conditions each year.
Studies of mating behaviour in which oestrus has been induced with exogenous hormones (Lindsay and Robinson, 1961 a and b; Lindsay, 1966 b;
Lindsay and Ellsmore, 1968; Fletcher and Lindsay, 1968; Blockey and
Cumming, 1970) or where rams and ewes are in pens (Hulet et al., 1962 a

www.pdfgrip.com


2

and b; banks, 1964; Pepelko and Clegg, 1965; Lindsay, 1 9 6 5 , 1 9 6 6 a;
Beamer et al., 1969; Bermant et al., 1969) or where rams are removed from
the flock each night (Mattner et al., 1967; Mattner and Braden, 1967;
Braden, 1971) may bear little relationship to natural mating under paddock
conditions. Removing rams from the flock influences the length of oestrus in
ewes (Hulet et al., 1962 c; Parsons and Hunter, 1967; Parsons et al., 1967)
and intensifies the initial mating activity that occurs when rams are reintroduced to the flock (Hulet et al., 1962 a; Blockey and Cumming, 1970).
In this study, particular emphasis has been placed on the average interservice and intermount intervals of time for ewes because they can be measured
under field conditions without interfering with the flock. Also, during preliminary examination of the Series 1 data, it was found that the sum of interservice intervals was very closely related to the number of services obtained
from many different flocks over a wide range of environments.
Analysis and interpretation of this relationship then commences and it is
postulated that the average interservice and intermount intervals may be
directly related to fertility. The average interservice interval for ewes is known
to increase as the number of services increases (Pepelko and Clegg, 1965) and
a relationship between number of services and fertility has been suggested
(Gerasimov, 1957; Mattner and Braden, 1967). The second series of observations were designed to examine the relationship between the interservice and
intermount intervals and fertility. The results offer the possibility of explaining
the variation in reproductive performance as it is determined at mating which
would enable the prediction of fertility.

SERIES 1

Methods
From March, 1971 to May, 1973, the mating activity of flocks of Merino
sheep was observed at four locations throughout New South Wales (Fig.l).
At a fifth location in March, 1964 (Mattner et al., 1967), data were collected
in an identical way to that collected here and is included in the present study.
Sheep
Flocks were joined with one, two or three Merino rams for 5 weeks. Each
flock was mated with a different ram or different groups of rams whose age
varied from 2 to 6 years. Flock size varied from 50 to 400 ewes and the
average age of ewes varied from 2 years in some flocks to 8 years in others.
The average weight of ewes varied from 32 kg in some flocks to 45 kg in
other flocks. Pastures on which the flocks were grazing varied from very poor,
where there was almost zero feed, to rank where there was a superabundant
growth of green feed standing 60 to 80 cm high. Paddock size varied from 1
to 30 ha and meteorological conditions varied markedly.
Each ram was fitted with a "sire-sine" harness and crayon (Radford et al.,

www.pdfgrip.com


3

Fig.l. Map of N e w South Wales showing the location of observation sites.

1960) and was individually identified with painted horns and a side brand.
Ewes were identified by means of an ear tag and by large numbers branded
on the side of the fleece.
Observations

Observations commenced at least 1 day after ewes and rams were joined.
At all times during observation, the sheep were in view of two or three observers in a mobile observation tower. Observations always commenced soon
after sunrise and continued uninterruptedly. The average duration of observation was 10.8 ± 2.7 (SE) h.
The ram and ewe involved in each mount and service were identified and
the time recorded. A mount was defined as having occurred when a ram
straddled the rear quarters of a ewe. A service was considered to have occurred
if mounting was accompanied by a distinct pelvic thrust followed immediately
by dismounting and a short period during which the ram displayed little
interest in the ewe but stood quietly, usually nearby the ewe, with his head
and neck extended and slightly lowered. When a ram mounted twice and
achieved intromission and ejaculation during the second mount, this was
defined as two mounts and a service. Mating activity was recorded directly
onto an event chart (Fig.2).

η
Ν
t
t'

Statistical analysis
The following statistics were obtained from each event chart:
= the number of services (or mounts) observed for any one ewe.
= the sum of all services (or mounts) during the observation period.
= the time interval between two successive services (or mounts) on any one
ewe.
= the sum of interservice (or intermount) intervals of time observed for
any one ewe.

www.pdfgrip.com



4

EWE
35

20
19
Ν =10 SERVICES
Τ =975 MINUTES

L
06.00

12.00

18.00

TIME—*
PERIOD OF OBSERVATION

Fig.2. The pattern of mating activity that was observed on March 10th, 1 9 7 1 at
Condobolin where one ram had been joined with 5 0 old ewes.
2000i

ONE

0

500

N-SERVICES

RAM

100
Ν-MOUNTS

Fig. 3. Showing the sample regressions of Τ on N. Within each ram number, the different
symbols refer t o the different flocks studied in Series 1. Each flock has a pair of symbols,
one for services and the other for mounts.

www.pdfgrip.com


5

Τ = the sum of all interservice (intermount) intervals for all ewes during an
observation period.
k = the number of ewes seen to be served or mounted during the observation
period.
Regression analysis and comparison of regressions (Williams, 1959) were
used to examine the data.
Results
There was a very close relationship between Τ and Ν for services ^nd
mounts, where flocks were joined with the same number of rams (Fig.3). The
relationship can be described by the general equation:
Τ = b + Nc
where Τ and Ν are defined and c and b are constants whose value depends on
the number of rams and whether the relationship is calculated from services
or mounts.

Regression coefficients and their standard errors, and the coefficients of
correlation (Table I) and the comparison of regression analysis (Table II)
show a general trend for the slopes of the regression of Τ on Ν to decrease as
the number of rams per flock increased.
Flocks having the same number of rams are markedly similar, despite the
very wide differences in the test situations from which the data were obtained.
Correlation coefficients were in excess of 0.73 (P < 0.05) for mounts and
0.89 (P < 0.005) for services.
DISCUSSION

A direct relationship between Τ and Ν is obvious. The closeness of the
TABLE I
Results on the regression analysis of Γ o n Ν for both mounts and services
Mating
activity

Slope
± SE

Intercept
± SE

Correlation

Services
1 ram
2 rams
3 rams

137.1 ± 17.4*

87.3 ± 13.8*
51.1 ± 1 0 . 5 *

- 4 8 1 ± 175*
—453 ± 2 5 6
199 ± 202

0.93**
0.89**
0.91**

157 ± 2 3 1
383 ± 332
—215 ± 3 7 3

0.78*
0.73*
0.92**

Mounts
1 ram
2 rams
3 rams

34.1 ±
15.4 ±
30.4 ±

9.2*
4.5*

5.9*

* P < 0.05; * * P < 0 . 0 1 .

www.pdfgrip.com


6

TABLE II
Comparison of regressions in Series 1
Regressions
being
compared*

Probability** values for
the comparison of
Slopes

Intercepts

IRMvs
2RM
1RM vs 3RM
2RM vs 3RM

0.348
0.799
0.085


0.472
0.103
0.351

1RS vs 2RS
1RS vs SRS
2RS vs 3RS

0.132
0.004
0.074

0.006
0.143
0.857

*1RM = Regression for one ram mounts
1RS = Regression for one ram services etc.
**A value of Ρ < 0.05 indicates a significant difference.

relationship is not, because it is known that variation between individuals
(Hulet et al., 1962 a; Lindsay, 1966 a) and environments (Braden and Baker,
1973) are major factors influencing variation in mating activity.
The variables Τ and Ν enable the calculation of the weighted average interservice or intermount interval for ewes:
T/(N-k)
The average interservice interval for ewes is known to increase as the number of services increases because the time interval between successive services
on a ewe tends to grow progressively longer. This trend has been observed
where there is one (Pepelko and Clegg, 1965; Bermant et al., 1969; Beamer
et al., 1969) or several (Hulet et al., 1962 a and c) ewes in oestrus with one
or more rams in pens, and under field conditions (Lambourne, 1956).

The average interservice interval for ewes may be directly related to fertility.
Alarge value will tend to be associated with increased numbers of services
and prolonged periods of time from first to last service or length of oestrus
(Heape, 1900). Fertility has been shown to be directly related to number of
services (Gerasimov, 1957; Mattner and Braden, 1967) and length of oestrus
(McKenzie and Terrill, 1937; Lambourne, 1956; Dunlop and Tallis, 1964;
Blockey and Cumming, 1970).
The second series of observations were conducted to examine the relationship between the average interservice and intermount intervals and fertility.
SERIES 2

Methods
Sheep
Six different flocks of Merino ewes were each joined with three Merino

www.pdfgrip.com


7

rams for 5 weeks during the Autumn of 1971 or 1973. There were three
flocks at Glen Innes and three at Richmond and individual details of these
flocks are shown in Table III. These flocks had not been involved in Series 1
observations. All sheep were individually identified as in Series 1.
Mating activity
Observations commenced at least 1 day after rams and ewes were joined.
The daily procedure was similar to that for Series 1, and each flock was observed on four separate occasions during the first 17 days of joining.
Event charts which described the mating activity for each day were constructed. From these, weighted and unweighted, interservice and intermount
intervals were calculated.
Weighted intervals were calculated by summing all the interservice or intermount intervals, then dividing by the total number of intervals:
T/(iV-fe)

Unweighted intervals were obtained by calculating the average interservice
or intermount interval for each ewe with η > 1:
t'Kn - 1)
then obtaining the average of these values for each observation period. All
values were corrected to a standard duration of observation.
Reproductive
performance
The flocks at Richmond were slaughtered in a local abattoir, 21 days after
rams were removed. Reproductive tracts were collected and examined to
determine the number of foetuses present.
At Glen Innes, ewes were examined with a sonicaid (Fraser et al., 1971)
and/or rectal palpation (Hulet, 1973) and were under close supervision.
Statistical analysis
Regression analysis (Williams, 1959) was used to compare mating activity
with reproductive performance. Comparisons were based on data collected in
the first 17 days after the introduction of rams. Mating activity was always
treated as the independent variable.
Results
Weighted and unweighted intervals, when calculated for both services and
mounts were found to be poorly correlated with the number of pregnancies
and the number of foetuses (Table IV).
Hourly service rate and the interval sums per hour, when calculated for
both services and mounts were highly correlated with the number of pregnancies and their regression terms and slopes of regression were significant (Table IV)
The ratio T/N bore a very close relationship to the number of foetuses
(Table IV) as did interval sums per hour when calculated from services.
Thus the strongest relationships to emerge from these data are between the
number of pregnancies and interval sums per hour and that between the num-

www.pdfgrip.com



www.pdfgrip.com

•

Ο

Δ

••

A

Symbol

250
200
100

200
200
150

Flock
size*

NP
NP
NP


Ρ
Ρ
Ρ

Strain,
Ρ or
NP**

Mature
Mature
Mature

Immature
Immature
Mature

Rams'
age

Mature
Immature
Mixed

Mature
Immature
Mixed

Ewes'
age


43
38
40

45
32
42

Ewes'
weight
(kg)

*These flocks had not been used in Series 1 observations.
**P = Peppin or NP = N o n Peppin.
***R — Rank; a superabundant growth of green feed standing 60—80 cm high.
Ε — Excellent; abundant green growth of short to moderate length 5 t o 15 cm.
G — Good; pasture of lowered quality as a result of drying off.

Glen Innes
1973
1973
1971

Richmond
1973
1973
1971

Location


1.2
1.5
3.0

1.5
1.5
2.0

Ram
(%)

5
5
5

30
28
10

Paddock
size
(ha)

G
G
Ε

R
R
Ε


Pasture***

Features of the flocks and the environments in which ewes and rams were joined for mating in Series 2 observations

TABLE III

00


www.pdfgrip.com

0.36
0.77
0.67
0.89*
0.88*

0.34
0.66
0.20
0.81*
0.88*

Service activity
Weighted intervals (Tf(N — ft))
Unweighted intervals (t'/(n — 1 ) )
Services/h (N)
Interval s u m s / h (T)
Ratio (T/N)


aunting activity
Weighted intervals (T/(N - ft))
U n w e i g h t e d intervals (t'/(n — 1 ) )
Mounts/h (N)
Interval s u m s / h ( T )
Ratio (T/N)

*P < 0 . 0 5 ; N S — N o t significant.

Intercept

•

NS
NS
NS
NS

*

•

NS
NS
NS

1.0
5.2
3.2

0.6
7.2

0.8
3.1
39.9
0.7
3.4
±
±
±
±
±
1.4
2.9
7.8
0.2
1.9*

± 1.0
± 1.2
± 22.0
± 0.2*
± 0.9*
87.1
22.7
132.3
57.1
-14.2


81.3
-26.4
68.3
63.8
-29.7
±
±
±
±
±

99.9
77.7
65.4
39.4
47.6

± 100.7
± 76.5
± 52.2
± 26.3
± 51.2
0.21
0.54
0.31
0.87*
0.76

0.30
0.77

0.81*
0.92*
0.60

Correlation

Slope

Correlation

Regression

Pregnancies

Foetuses

Reproductive performance

The relationship b e t w e e n various estimates o f mating activity and reproductive performance

TABLE IV

NS

•

NS
NS
NS


NS

*

NS
NS

Regression

0.5
3.4
4.0
0.5
5.0

0.5
2.4
38.6
0.6
1.8

±
±
ι
±
±

1.1
2.6
6.0

0.1*
2.1

± 0.8
± 1.0
± 13.8*
± 0.1*
± 1.2

Slope

89.6
37.5
93.5
37.9
6.6

73.7
-21.9
38.5
47.7
22.7

±
±
±
±
±

±

±
±
±
±

82.9
70.0
50.6
25.9
52.2

82.2
61.2
32.7
18.3
69.3

Intercept


10

ber of foetuses and the ratio T/N because in both instances the relationships
hold for both services and mounts. These relationships are shown in Fig.4.
The regressions of Τ on Ν obtained from three rams in Series 1 were compared with those from Series 2 and all flocks in Series 2 were compared with
each other. Error variances were homogeneous and there were no significant
differences between regression coefficients. Positions of regressions were
significant for mounts (P < 0.05) in comparisons of Series 2 data only.
Δ
SERVICES

MOUNTS
200 r

ζ
ο

ω 100

AO

80
0
AO
Τ VALUES χ 10"3 (MIN)

80

120

80

120

to

Lu
if)

ZD
hLU

Ο

till.
Ο
d
Ζ

0

AO

80

0

40

Fig.4. The regressions of number of pregnancies on the sum of intervals and the number
of foetuses o n the ratio T/N for b o t h mounts and services. The different symbols refer t o
the different flocks studied in Series 2 (see Table III).
DISCUSSION

The number of pregnancies was closely related to the sum of intervals T/h
when calculated for both services and mounts and to the number of services
per h. That the number of pregnancies should rise as the number of services
per h rises is not surprising. Nor is it surprising that T/h is directly related to

www.pdfgrip.com



11

the number of pregnancies because as services increase the average interservice
interval for ewes increases (Hulet et al., 1962 a and c; Pepelko and Clegg,
1965) and hence the sum of intervals would increase. What is surprising is the
closeness of the relationship, particularly in view of the differences between
the flocks.
The ratio T/N is more closely related with the number of foetuses than are
the weighted or unweighted interservice or intermount intervals and a possible
explanation follows.
The weighted and unweighted average intervals are calculated using a value
of Ν which ignors those ewes where η = 1, whereas the ratio T/N does not.
This may be very important because ewes who receive one service or one
mount contribute to the value of T, not through intervals of their own but by
prolonging an interservice or intermount interval t on some other ewe.
Variation in reproductive performance in sheep is known to be influenced
by genetic, nutritional, seasonal, climatic, topographic and managerial factors.
Age, live weight, paddock size and many other factors are known to operate.
The ratio T/N accounted for a major portion of the variation in the number
of foetuses in six flocks, selected so as to differ from each other in respect of
many of the above factors. The ratio must therefore be influenced by the factors which influence reproductive performance. In order that the ratio explain
variation in fertility, it must itself be able to vary.
Variation in the ratio T/N is possible because the regression of Τ on Ν
does not contain the origin so that T/N takes differing values as the number
of services or mounts increases. Variation is also possible because the slope of
regression can vary as indicated by significant differences between slopes as
the number of rams increases from one to three per flock for services.
These results must be interpreted with caution. Many differing factors
interact to produce the level of fertility and fecundity realised in any situation. Accurate prediction of fertility is possible only if the total effect of the
interacting influences can be described objectively and the ratio T/N offers

that possibility.
ACKNOWLEDGEMENTS

The assistance provided by Messrs G. Young, S. Lane, I. Brook and D.L.
Jenkins and by Mr J. Ross of "Bon View" at Wellington and by past and
present managers of the New South Wales Department of Agriculture Research Stations at Condobolin and Glen Innes is gratefully acknowledged.
Biometricians of CSIRO, Division of Mathematical Statistics and of the
New South Wales Department of Agriculture have provided useful advice,
and the Australian Wool Corporation provided partial financial support for
this work.
REFERENCES
Banks, E.M., 1 9 6 4 . S o m e aspects of sexual behaviour in domestic sheep, Ovis aries.
Behaviour, 2 3 : 249—278.

www.pdfgrip.com


12

Beamer, W., Bermant, G. and Clegg, M.T., 1 9 6 9 . Copulatory behaviour of the ram, Ovis
aries II factors affecting copulatory satation. Anim. Behav., 1 7 : 7 0 6 — 7 1 1 .
Bermant, G., Clegg, M.T. and Beamer, W., 1 9 6 9 . Copulatory behaviour o f the ram, Ovis
aries I a normative study. Anim. Behav., 1 7 : 700—705.
Blockey, M.A. de Β. and Cumming, LA., 1 9 7 0 . Mating behaviour o f Merino ewes. Proc.
Aust. Soc. Anim. Prod., 8: 344—347.
Braden, A.W.H., 1 9 7 1 . Studies in flock mating of sheep 3. Effect of undernutrition of
ewes during joining. Aust. J. Exp. Agric. Anim. Husb., 1 1 : 375—378.
Braden, A.W.H. and Baker, Α. Α., 1 9 7 3 . Reproduction in sheep and cattle. In: G. Alexander
and O.B. Williams (Editors), The Pastoral Industries of Australia. Sydney University
Press, Sydney, N.S.W., pp. 269—302.

Dunlop, A.A. and Tallis, G.M., 1 9 6 4 . The effects of length of oestrus and number of
inseminations o n the fertility and twinning rate of the Merino ewe. Aust. J. Agric. Res.,
15: 2 8 2 - 2 8 8 .
Fletcher, I.G. and Lindsay, D.R., 1 9 6 8 . Sensory involvement in the mating behaviour of
domestic sheep. Anim. Behav., 1 6 : 410—414.
Fraser, A.F., Nagaratman, V. and Callicot, R.B., 1 9 7 1 . The comprehensive use of Doppler
ultrasound in farm animal reproduction. Vet. R e c , 8 8 : 202—205.
Gerasimov, L.M., 1 9 5 7 . S o m e characters relating t o reproduction in Romanov sheep.
Anim. Breed. Abstr., 2 6 : 1 7 6 .
Heape, W., 1 9 0 0 . The sexual season of mammals and the relation of pro-oestrum t o menstruation. Q. J. Microsc. Sci., 4 4 : 1—70.
Hulet, C.V., 1 9 7 3 . Determining fetal numbers in pregnant ewes. J. Anim. Sci., 3 6 : 325—330.
Hulet, C.V., Ercanbrack, S.K., Price, D.A., Blackwell, R.L. and Wilson, L.O., 1 9 6 2 a.
Mating behaviour of the ram in the o n e sire pen. J. Anim. Sci., 2 1 : 8 5 7 — 8 6 4 .
Hulet, C.V., Ercanbrack, S.K., Blackwell, R.L., Price, D.A. and Wilson, L.O., 1 9 6 2 b.
Mating behaviour of the ram in the multi sire pen. J. Anim. Sci., 2 1 : 865—869.
Hulet, C.V., Blackwell, R.L., Ercanbrack, S.K., Price, D.A. and Wilson, L.O., 1 9 6 2 c.
Mating behaviour of the ewe. J. Anim. Sci., 2 1 : 870—874.
Lambourne, L.J., 1 9 5 6 . Mating behaviour. Proc. Rurakura farmers Conf. Dept A g r i c ,
Wellington, pp. 16—20.
Lindsay, D.R., 1 9 6 5 . The importance of olfactory stimuli in the mating behaviour of the
ram. Anim. Behav., 1 3 : 75—78.
Lindsay, D.R., 1 9 6 6 a. Modification of behavioural oestrus in the ewe by social and
hormonal factors. Anim. Behav., 1 4 : 73—83.
Lindsay, D.R., 1 9 6 6 b. Mating behaviour of ewes and its effect o n mating efficiency.
Anim. Behav., 1 4 : 419—424.
Lindsay, D.R. and Ellsmore, J., 1 9 6 8 . The effect of breed, season and competition o n
mating behaviour of rams. Aust. J. Exp. Agric. Anim. Husb., 8: 649—652.
Lindsay, D.R. and Robinson, T.J., 1 9 6 1 a. Studies o n the efficiency of mating in sheep
I — The effect o f paddock size and number of rams. J. Agric. Sci., 5 7 : 137—140.
Lindsay, D.R. and Robinson, T.J., 1 9 6 1 b. Studies on the efficiency of mating in sheep

II — The effect of freedom of rams, paddock size and age of ewes. J. Agric. Sci.,
5 7 : 141—145.
McKenzie, F.F. and Terrill, C.E., 1 9 3 7 . Estrus, Ovulation and Related Phenomena in the
Ewe. Univ. Mo., Coll. A g r i c , Agric. Exp. Stn Res. Bull. No. 2 6 4 , 8 8 pp.
Mattner, P.E. and Braden, A.W.H., 1 9 6 7 . Studies in flock mating of sheep 2 — Fertilization and prenatal mortality. Aust. J. Exp. Agric. Anim. Husb., 7: 110—116.
Mattner, P.E., Braden, A.W.H. and Turnbull, K.E., 1 9 6 7 . Studies in flock mating of sheep
1 — Mating behaviour. Aust. J. Exp. Agric. Anim. Husb., 7: 103—109.
Parsons, S.D. and Hunter, G.L., 1 9 6 7 . Effect of the ram o n duration of oestrus in the ewe.
J. Reprod. Fertil., 1 4 : 61—70.
Parsons, S.D., Hunter, G.L. and Rayner, A.A., 1 9 6 7 . Use of probit analysis in a study of
the effect of the ram o n the time of ovulation in the ewe. J. Reprod. Fertil., 1 4 : 71—80.
Pepelko, W.E. and Clegg, M.T., 1 9 6 5 . Studies of mating behaviour and s o m e factors
influencing the sexual response in the male sheep Ovis aries. Anim. Behav., 1 3 : 249—258.
Radford, H.M., Watson, R.H. and Wood, G.F., 1 9 6 0 . A crayon and associated harness for
the detection of mating under field conditions. Aust. Vet. J., 3 6 : 57—66.
Williams, E.J., 1 9 5 9 . Regression analysis. Wiley, N e w York, N.Y., 2 1 4 pp.

www.pdfgrip.com


13

THE PREDICTION OF FERTILITY AND FECUNDITY FROM THE
MATING ACTIVITY OF EWES

D.G. FOWLER* and C M . L A N G F O R D * *
*New South Wales Department
of Agriculture, CSIRO, Blacktown, N.S.W,
Hawkesbury Agricultural College, Richmond, N.S.W.
(Australia)


(Australia)

ABSTRACT
Fowler, D.G. and Langford, C M . , 1 9 7 6 . The prediction of fertility and fecundity from
the mating activity of ewes. Appl. Anim. Ethol, 2 : 2 7 7 — 2 8 1 .
The number of foetuses and corpora lutea in four different groups of Merino ewes has
been estimated from observations of their mating activity. Observed and expected values
could not be shown t o be different.

Mating activity was found to be closely related to the number of foetuses
present or lambs born in six flocks selected to differ in age of ewes and rams,
live weight of ewes, nutrition, genotype, flock size, year and location of
joining (Fowler, 1975). These factors are among the major components known
to induce variation in the reproductive process at the time of joining (Watson,
1966) and it is possible that the relationship may be applicable on a much
broader basis. This suggestion would receive support if the relationship could
be used to predcit the performance of other groups of ewes. Accurate prediction is tangible evidence that the important sources of variation and the
outcome of their interactions from group to group, place to place and time
to time may be embodied within the predicting model. This report deals with
the prediction of fertility and fecundity from the mating activity of four groups
of ewes.
Merino ewes were used and the groups differed from each other and from
groups used previously (Fowler, 1975) in a number of important ways. They
were different sheep and were joined with different rams. Some ewes were
purchased from a commercial producer and were randomized into two groups
(A and B) of equal size (93). These ewes weighed 35.0 ± 0.41 kg at the start
of joining. Another 306 ewes were divided into two groups of equal size and
the heavier ewes were placed in one group (C), 48.6 ± 0.36 kg, and the lighter
ewes were placed in the other group (D), 39.8 ± 0.29 kg.

Each group was joined with three Merino rams for 34 days from March 8th
in 1974 (A and B) and March 24th in 1975 (C and D). All sheep were individ-

www.pdfgrip.com


14

ually identified and each group was observed on six separate occasions during
the first 17 days of joining. At all times during observation the sheep were in
view of two observers who were located in a mobile observation tower. Observations commenced soon after sunrise and continued uninterruptedly for
9 h. The ram and ewe involved in each mount were identified and the time
was recorded.
The following measurements were then obtained:
Τ = The time from first to last observed mount for each ewe, summed over
all ewes mounted more than once.
Ν = The total number of mounts observed in 9 h.
The ratio of T/N was calculated as every mount occurred during the ninth
hour and the average of these provided a value for the group for that day.
The six daily values that were obtained for each group were averaged.
The equations that were used to determine the number of foetuses and
corpora lutea were:
Number of foetuses = 9.2 T/JV-33.2; r = 0.94 (P < 0.05)
Number of corpora lutea = 11.9 T/N-46.7;
r = 0.80 (NS)

(1)
(2)

These equations were derived from the data of Fowler (1975 — (1); 1976 —

(2)). They differ, slightly, from similar equations presented in these papers, by
virtue of the different method by which T/N was calculated, i.e. 9 h of observation and the calculation of T/N as described above.
All groups were slaughtered in a local abattoir 28 days after the entire rams
were removed and the number of foetuses and corpora lutea that were present
was counted.
The values of Τ, Ν and T/N, the observed values for number of foetuses
and corpora lutea, and the derivation of expected values are shown in Table I.
It shows that the value of T/N tends to increase as both Τ and Ν increase. This
is because the quantity of Τ per unit Ν tends to increase and this agrees with
the findings of others who have shown that the interval of time which separates
successive services and mounts on a ewe tends to grow progressively longer
(Hulet et al., 1962 a and b; Pepelko and Clegg, 1965).
The regression of T/N on the number of foetuses that was obtained from
the observed values of the four groups (o) was compared with regression (1)
(·). Also, the regression of T/N on the number of corpora lutea that was obtained
from the observed values of the four groups ( o ) was compared with regression (2)
( · ) . The results of the comparisons are shown in Fig.l and the tabular material
which is below it. For both comparisons, most of the variance has been accounted for by regression. Slopes and positions account for a very small proportion of the total variance. From this it can be concluded for both number of
foetuses and number of corpora lutea, that the observed and expected values
of Table I cannot be shown to be different.
There are now ten groups of sheep, six ( · ) and four (o) for which concerted
efforts have been exercised to make them as different from each other as possible. The variation produced as a result of this effort is not beyond objective
definition by means of linear models.

www.pdfgrip.com


547
999
2 180

1 609

38.3
59.7
87.9
68.5

14.3
16.7
25.0
22.1

91
93
153
153

Joined

Τ/Ν*

Τ

Ν

Ewes

Mean value of

91

92
145
149

Killed**
98.4
120.4
196.8
170.1

Foetuses
from ( 1 )

123.5
152.0
250.8
216.3

Corpora
lutea
from ( 2 )
1.08
1.29
1.29
1.11

Foetuses
per e w e

•Calculated during the ninth hour and meaned as described. N o t a straight ratio o f means.

* * E w e s w i t h o u t c o m p l e t e mating and fertility data were deleted.
* * * E w e s killed ( f o e t u s e s / e w e ) or ewes killed (corpora l u t e a / e w e ) .

A
Β
C
D

Group

Derivation of e x p e c t e d and observed values for numbers of foetuses and corpora lutea

TABLE I

1.36
1.63
1.64
1.41

Corpora
lutea
per e w e
105
107
191
175

Observed

Foetuses


98
119
187
165

Expected***

116
123
226
199

Observed

124
150
238
210

Expected***

Corpora lutea

15

www.pdfgrip.com


16


FOETUSES

CORPORA LUTEA

300r

Ο 200h

•o
op

CD

5 iooh

88

20

30

10

VARIANCE
REGRESSION

9484

20


30

VARIANCE
6430

SLOPES

2

14

POSITIONS

5

659

RESIDUAL

309

814
2-76 P-S
DEP. A
DF

Fig.l. The relationship between mating activity (T/N) and number of foetuses and corpora
lutea. The relationship derived from earlier work (Fowler, 1 9 7 5 , 1 9 7 6 ; · ) is compared with
that found in the present study ( o )


The models are:
Number of foetuses = (9.1 ± 1.0*) T/N - ( 3 2 . 9 ± 20.9); r = 0.96*
Number of corpora lutea = (11.9 ±/2A*) T/N - ( 5 7 . 6 ± 50.1); r = 0.91*
Where Τ and Ν are derived from mounting activity and * = Ρ < 0.05.
The results focus attention on the ratio T/N It takes its value from the
number and pattern of the contacts that rams and ewes make and has been
called ram-ewe contact. For reasons discussed elsewhere (Fowler, 1975) it
is known that T/N is directly related to the number of mounts and services
which occur and to the length of oestrus, both of which are directly related
to reproductive performance in sheep (Lambourne, 1956; Mattner and Braden,
1967; Land, 1970).
The groups used in this study differ from each other and differ from those
used to derive the prediction equations, in respect of a number of the factors
which are known to induce variation in the reproductive performance of
Merinos at the time of mating. It is the interaction of these factors in any
flock which determines that flock's reproductive performance. The results
suggest that ram-ewe contact as defined objectively by the ratio T/N has measured the sum total of these interactions, for in all flocks it has estimated the
number of foetuses and corpora lutea that were present to the extent that it
is not possible to show differences between the observed and expected values.
There is no evidence from the literature to suggest that any other variable or
set of variables measured at joining can do this. It is therefore reasonable to

www.pdfgrip.com


17

conclude that ram-ewe contact may be the most important variable influencing
the joining performance of Merino sheep and that other variables would be important to the extent that they influenced contact.

Attention will now be focussed upon the factors that are within individual
rams and ewes that are responsible for maximizing the value of T/N for a
group, because these factors may enable selection of individuals for high fertility and fecundity.
ACKNOWLEDGEMENTS

Biometricians of CSIRO, Division of Mathematical Statistics and of the
New South Wales Department of Agriculture have provided useful advice, and
the Australian Wool Corporation provided partial financial support for this
work.
REFERENCES
Fowler, D.G., 1 9 7 5 . Mating activity and its relationship t o reproductive performance in
Merino sheep. Appl. Anim. Ethol., 1: 357—368.
Fowler, D.G., 1 9 7 6 . Predicting the number of corpora lutea in Merino ewes. J. Reprod.
Fertil., 4 6 : 5 2 5 .
Hulet, C.V., Ercanbrack, S.K., Price, D.Α., Blackwell, R.L. and Wilson, L.O., 1 9 6 2 a.
Mating behaviour of the ram in the one sire pen. J. Anim. Sci., 2 1 : 857—864.
Hulet, C.V., Blackwell, R.L., Ercanbrack, S.K., Price, D.A. and Wilson, L.O., 1 9 6 2 b.
Mating behaviour of the ewe. J. Anim. Sci., 2 1 : 870—874.
Lambourne, L.J., 1 9 5 6 . Mating behaviour. Proc. Rurakara Farm. Conf., Dept. A g r i c ,
Wellington, pp. 16—20.
Land, R.B., 1 9 7 0 . A relationship between the duration of oestrus, ovulation rate and litter
size of sheep. J. Reprod. Fertil., 2 3 : 49—53.
Mattner, P.E. and Braden, A.W.H., 1 9 6 7 . Studies in flock mating of sheep 2 — Fertilization and prenatal mortality. Aust. J. Exp. Agric. Anim. Husb., 7: 110—116.
Pepelko, W.E. and Clegg, M.T., 1 9 6 5 . Studies of mating behaviour and s o m e factors influencing the sexual response in the male sheep Ovis aries. Anim. Behav., 1 3 : 249—258.
Watson, R.H., 1 9 6 6 . Reproductive performance of sheep in Australia. Proc. Aust. Soc.
Anim. Prod., 6: 1—9.

www.pdfgrip.com



19

THE EFFECTS OF DOMINANCE AND INFERTILITY OF RAMS ON
REPRODUCTIVE PERFORMANCE

D.G. FOWLER* and L.D. JENKINS**
*New South Wales Department
of Agriculture, CSIRO, P.O. Box 239, Blacktown,
2148
(Australia)
**Hawkesbury
College, Richmond, N.S.W., 2753
(Australia)

N.S.W.,

ABSTRACT
Fowler, D.G. and Jenkins, L.D., 1 9 7 6 . The effects of dominance and infertility of rams
on reproductive performance. Appl. Anim. EthoL, 2: 327—337.
The extent to which infertility of dominant rams can influence reproductive performance has been studied in Merinos where there are three rams with every 1 0 0 ewes. The
percentage of infertile rams in the ram group (25—66%) and their position in the dominance rank was varied. Infertility was produced by vasectomy and when such rams were
dominant, the percentage of pregnant ewes was lower (72.3%) than in control groups
where all rams were fertile (90.4%). When high percentages of the rams in any group
(33—66%) were infertile and subordinate the percentage of pregnant ewes (90.3%) was
not different to that of controls. The percentage of ewes that had multiple pregnancies
showed no consistent trends that could be ascribed to the percentage of infertile rams or
their position in the dominance rank.
The reduction in fertility that occurred cannot have been due t o ewes mating exclusively with infertile rams because most ewes carried marks which indicated their contact with
two or more rams. Also, a high percentage of non pregnant ewes carried the marks of
fertile rams. It is suggested that non pregnant ewes may have mated exclusively with

dominant infertile rams during that period of oestrus which is known to be o p t i m u m for
conception.
These studies show that the ram or rams w h o occupy the t o p positions in the dominance rank can influence flock productivity because, when they are totally infertile, flock
fertility is depressed. The reductions occurred regardless of the fact that there were adequate numbers of fertile rams to ensure high flock fertility. It can easily be argued that
the effects of dominance may be greater where ram—ewe ratios are lower than those used
here because there would be fewer subordinate rams to modify the effects of the dominant ones.
INTRODUCTION

The dominance of one individual over others and the ranking according
to degree of dominance are features of group structure in a number of different species. Where rams are in close contact, the activities of the dominant
ram can greatly influence those of the subordinate (Hulet et al., 1962;

www.pdfgrip.com


20

Marincowitz et al., 1966). However, under field conditions where the interactions between rams may be less, Mattner et al. (1967, 1973) could find no
differences between the mounting and service activity of dominant and subordinate rams.
The effects of interactions between dominant and subordinate rams on
the productivity of commercial sheep and wool enterprises is unknown. It is
known that ewes who fail to conceive at the first oestrus after rams are introduced usually return and may conceive during their next oestrus. Rams and
ewes are often left together long enough to facilitate this occurrence. It is
known that ewes are usually served by more than one ram during oestrus
(Croker and Lindsay, 1972; Knight and Lindsay, 1973), particularly where
the ratio of rams to ewes is in excess of 1 : 100. It has been suggested
(Haughey, 1959; Edgar, 1965; Lightfoot and Smith, 1968; Dawe et al., 1970)
that with mature rams and ewes, a ratio of one ram per 100 ewes is sufficient
to ensure adequate fertility.
In the light of these findings it is tempting to suggest that the occurrence

of infertility among rams at the top of the dominance rank might be insignificant while there is at least one fully fertile ram for every 100 ewes. This
report deals with a series of studies aimed at determining the relative importance of degree of infertility of the ram group and the position of fertile and
infertile rams in the dominance rank on fertility and fecundity in Merino
ewes.
METHODS

Thirteen hundred mature Merino ewes and 133 Border Leicester X Merino
ewes were used in a series of experiments conducted during the period 1969
to 1973. The experiments were conducted at Richmond which is situated
60 km west of Sydney.
Ewes were joined to groups of 2—4 mature Merino rams at the rate of
three rams per 100 ewes so that the numbers of ewes joined to each group
varied from 66 to 133. The ram groups contained varying proportions of entire (F) or vasectomised (I) rams, occupying varying positions in the dominance rank. Ram combinations used are shown in Table I. Each year there
was a control group consisting of three entire rams (FFF). In 1969 there were
two treatment groups containing two vasectomised rams (II) in positions 1
and 2 of the dominance rank and two F rams in positions 3 and 4 of the rank.
In 1970, the treatment group contained one I ram in position 1 of the dominance rank and three F rams, in positions 2, 3 and 4. Ewes were joined with
rams in February or March and joining continued for 34 days in paddocks
which varied in size from 5 to 25 ha and contained abundant green feed.
Dominance

tests

Before each series of experiments began, it was necessary to conduct domi-

www.pdfgrip.com


21


nance tests in order to form ram groups of the desired composition. These
were conducted by starving the rams until they were prepared to fight over
a limited quantity of feed in a small bucket in the corner of an exercise yard.
Two or 3 days was usually sufficient to induce aggressive behaviour, and by
matching rams in a logical sequence it was possible to rank them in a
"nutritional" dominance order.
Each group was formed by drawing rams from two or three steps apart
in the dominance rank. Vasectomy was then conducted on rams that were to
be infertile. Vasectomy was given 4 months before joining was due to commence and rams were tested repeatedly with an electro ejaculator (Blackshaw,
1954) until their ejaculate was free of spermatozoa. The ejaculates of entire
rams were tested several times prior to and once during joining using techniques outlined elsewhere (Fowler, 1965).
The activity of individual rams
Rams were fitted with a "Sire-Sine" harness and crayon (Radford et al.,
1960). The rams in groups 2, 3, 7, 8, 10 and 13 (Table I) were each fitted
with a crayon of a different colour. Ewes were examined every third day
during joining and from the colours appearing on their rumps, it was possible to estimate the number of rams that had mounted each ewe. In 1969,
1971, 1972 and 1973 the rumps of all ewes were clipped to remove coloured
wool on day 18 of joining.
In other years, all rams had crayons of the same colour and in all years the
number of ewes marked in the first, second or both periods of joining was
recorded.
Direct

observations

Ewes were individually identified with ear tags and a side brand and the
rams had painted horns. Each flock was observed directly on at least five
separate occasions during the first 17 days of joining. Observations were conducted by several observers who were located in a mobile observation tower
from which a clear and uninterrupted view of the flock was maintained. The
durations of the observation periods were either short (2 h) or long (10 h).

During all periods, it was possible to check the dominance relationships
within the ram groups. During the long periods, more detailed observations
were conducted as described elsewhere (Fowler, 1975).
During all observation periods the dispersion of the sheep was estimated at
hourly intervals by means of the following subjective scoring system.
0
All sheep in a single group.
1
Two or three sub groups; average distance separating groups estimated
as less than 100 m. Movement of sub groups not independent of each
other.
2
Two to five sub groups; average distance separating groups estimated as
over 100 m. Movement of sub groups independent of each other.

www.pdfgrip.com