Physiological Correlates of Maternal–Offspring Behaviour in Sheep

Physiology & Behavior, Vol. 67, No. 3, pp. 443–454, 1999 © 1999 Elsevier Science Inc. Printed in the USA. All rights reserved 0031-9384/99/$–see front matter

PII S0031-9384(99)00089-X

Physiological Correlates of Maternal–Offspring Behaviour in Sheep: A Factor Analysis CATHERINE M. DWYER,1 WILLIAM S. DINGWALL AND ALISTAIR B. LAWRENCE Animal Biology Division, SAC, West Mains Road, Edinburgh, EH9 3JG, UK Received 18 February 1999; Accepted 26 April 1999 DWYER, C. M., W. S. DINGWALL AND A. B. LAWRENCE. Physiological correlates of maternal–offspring behaviour in sheep: A factor analysis. PHYSIOL BEHAV 67(3) 443–454, 1999.—The onset of maternal behaviour in sheep is controlled by levels of oestrogen and progesterone in gestation and the release of oxytocin during delivery. Factor analysis was used to investigate the relationships between maternal behaviour, offspring behaviour, and maternal ovarian hormone levels during gestation in sheep. Ewes gave birth to lambs following embryo transfer between two breeds (Suffolk and Scottish Blackface), which differ in their expression of maternal behaviour. Plasma oestradiol-17b concentration was significantly higher in Blackface ewes in the last 6 weeks of pregnancy, whereas plasma progesterone was higher in Suffolk ewes in early and mid gestation. Factor analysis revealed three factors that accounted for 50% of the total variation between variables. Factor 1 described lamb activity and had positive loadings for lamb behavioural latencies immediately postpartum, and negative loadings for sucking frequency. Factor 2 described some aspects of ewe maternal behaviour, and had positive loadings for ewe–lamb separation and lamb vocalisation, and negative loadings for ewe grooming behaviour and plasma oestradiol concentration. Factor 3 described ewe and lamb-sucking interactions, and had positive loadings for ewe moving as the lamb made sucking attempts, and negative loadings for ewe and lamb vocalisation. Factor 1 scores were significantly affected by lamb breed: Suffolk lambs received positive scores, indicating a longer latency to right and stand, whereas Blackface lambs had negative scores. Maternal progesterone concentration was significantly correlated with Factor 1 scores. Factor 2 and 3 scores were significantly affected by ewe breed. This study has shown that there are two behavioural axes to maternal behaviour in sheep, and that maternal oestradiol concentration is related to affiliative behaviours (e.g., grooming) but only weakly to ewe responsiveness to lamb-sucking attempts. © 1999 Elsevier Science Inc. Maternal behaviour

Oestradiol

Sheep

cent years. The ovarian hormones, oestrogen and progesterone, play a key role in pregnancy acting as physiological primers for maternal care (29,32). They function centrally to promote synthesis of oxytocin and b-endorphin messenger ribonucleic acid (mRNA) in several brain regions associated with the expression of maternal behaviour (8,9) and increase synthesis of oxtocinergic receptors in limbic areas of the brain (23). At high pharmacological doses, and following prolonged exposure to newborn lambs, the steroid hormones are sufficient to produce maternal behaviour in some ewes (35,44,54). However, their main behavioural effect appears to be a reduction in rejection behaviours (27,54) and the immediate appearance of affiliative maternal behaviours (e.g., grooming and udder acceptance) cannot be induced by steroid hormone treatment alone (15,25). By contrast, the birth process induces immediate maternal behaviour, even in primiparous ewes, and this response can be blocked by peridural anaesthesia (34,38). The induction of short latency maternal behaviour

THE coordinated development of the mother–offspring relationship is essential for the survival of the neonate in many species. The sheep belongs to the group of ungulates classified as “followers” (36), where the neonatal lamb is precocious and accompanies its mother immediately after birth. The onset of maternal behaviour in the ewe occurs at parturition and is characterised by a strong attraction to amniotic fluids (39), resulting in licking (grooming) of the lamb, low-pitched bleating, and acceptance of the lamb at the udder. These behaviours are directed towards the formation of an exclusive olfactory attachment between the ewe and lamb. Variation in the expression of maternal behaviours and the selectivity of ewe– lamb attachment can be affected by ewe breed (2,3,16), previous maternal experience (37,52), and maternal nutritional status (58,61). The physiological mechanisms underlying the onset of maternal behaviour and the selectivity of the bond between ewe and lamb have been subjected to intensive investigation in re1To

whom requests for reprints should be addressed at SAC, Bush Estate, Penicuik, Midlothian, EH26 0PH, UK. E-mail: [email protected]

443

444

DWYER, DINGWALL AND LAWRENCE

at birth is dependent on stimulation of the vagina and cervix (33,55). This results in the release of oxytocin into the periphery and centrally (31). Intracerebroventricular infusion of oxytocin into nonpregnant ewes stimulates maternal behaviour (28) and causes an attraction for amniotic fluids in ewes delivered under peridural anaesthesia (37). These studies demonstrate that central oxytocin plays a major role in the facilitation of maternal behaviour in physiologically primed ewes. In addition, there is increasing evidence that the opioid system plays an important modifying role in the expression of maternal behaviour with the opioid antagonist, naltrexone, reducing, and morphine potentiating, the expression of maternal behaviour in the sheep (10,26,30). Previously, we have shown that a lowland breed of sheep (Suffolk) consistently displays poorer maternal behaviour than a hill breed (Scottish Blackface): Suffolk ewes show more rejection behaviours towards their newborn lambs, spent less time grooming the lamb, are less cooperative to lamb sucking attempts, and make fewer low-pitched vocalisations towards the lamb as both primiparous and multiparous ewes (16,18,20). It is possible that differences in the expression of maternal behaviour may reflect an underlying variation in maternal physiology. Preliminary studies in sheep (59) and primates (57) have suggested that differences in maternal oestrogen are related to the quality of maternal behaviour expressed, although these are based on small numbers of animals. However, an important role for gestational levels of progesterone has also been proposed. In mice, a reduction in progesterone action by specific antibodies or antagonists in late gestation is associated with impaired maternal behaviour (63,64). Progesterone, rather than oestrogen, produces the greatest synthesis of oxytocin mRNA in the brain (32), and maternal behaviour could not be induced in nonpregnant ewes receiving oestrogens alone (27). The purpose of the present study, therefore, was to examine the role of gestational levels of oestrogen and progesterone in sheep in mediating the observed behavioural differences in maternal behaviour seen in the different breeds. Embryo transfer was also used, between breeds, to investigate any effects of the breed of foetus, or behaviour of the neonatal lamb, on ewe physiology and maternal behaviour. A multivariate analysis approach was adopted as a large number of behaviour patterns were recorded of both the ewe and lamb, in addition to physiological variables. Principal components factor analysis was chosen, as this procedure can condense and summarise the relationship between variables by assigning similar loadings to behaviours that are high correlated [see (22)]. This type of analysis is frequently used in behavioural studies to investigate individual behavioural consistency or temperament across tests [e.g., (40)]. The factor scores generated by principal components analysis, therefore, represent a score on a cluster of correlated behaviours, rather than any single behavioural measurement, and may be more related to physiological factors than are single behavioural measures. MATERIALS AND METHODS

Animals This study used 90 ewes of two breeds: 51 Scottish Blackface (weighing 57.34 6 0.83 kg at conception) and 39 Suffolk (81.72 6 1.63 kg). All ewes were multiparous and had had one (n 5 22; 14 Blackface, 8 Suffolk) or two (n 5 68; 37 Blackface, 31 Suffolk) previous litters. Ewes were brought indoors in week 7 of gestation (January 1996) and fed hay and fresh drinking water ad lib. All ewes were housed in the same

building throughout the study. Concentrate feed (218 g crude protein per kg dry matter) was provided from week 14 of gestation, at a ration of 300 g ewe21 day21 (Suffolk) or 200 g ewe21 day21 (Blackface). Rations were doubled every 2 weeks until week 18 of gestation. Ewes were vaccinated with a clostridial vaccine (Covexin-8, Mallinckodt, Middlesex, UK) in week 17. Ewes lambed in large, straw-bedded pens (7 3 7 m) in groups of approximately 10 ewes of the same breed. Embryo Transfer Ewes were given single, pure-bred embryos by embryo transfer. Donor ewes (n 5 20: 10 Blackface, 10 Suffolk) were superovulated, by administration of ovine follicle stimulating hormone (FSH, Ovagen, Immuno-Chemical Products Ltd., Auckland, NZ), and inseminated by laparoscopic artificial insemination with fresh semen from one of nine sires as previously described (42). Embryos were recovered on Day 6 following insemination and transferred laparoscopically to synchronised recipient ewes (41). Embryos were transferred across breeds in a balanced design such that the following breed combinations were obtained: Blackface ewes with Blackface lamb (BB; total n 5 24), Blackface ewe with Suffolk lamb (BS; n 5 27), Suffolk ewe with Blackface lamb (SB, n 5 22), Suffolk ewe with Suffolk lamb (SS, n 5 17). Prelambing Blood-Sampling Procedure Blood samples (7 mL) were collected by jugular venipuncture between 1000 and 1200 h at two weekly intervals from week 4 after conception until week 20. Gestation length was 145.5 6 0.21 days for both breeds of ewe. In addition, similar plasma samples were collected from 10 nonpregnant ewes (five of each breed). A final blood sample was taken from all lambed ewes at 3 days after parturition. Plasma was separated by centrifugation and stored at 2708C until assayed for oestradiol-17b and progesterone concentration. Progesterone concentration was measured using a specific radioimmunoassay and nonextracted plasma as previously described (43). Samples were incubated with a normal rabbit serum second antibody (NRS) and precipitated with donkey antirabbit serum (DARS). The minimum detection level was 0.05 ng ml21. Inter- and intraassay coefficients of variation were 12.4 and 7.23%, respectively. Oestradiol-17b was assayed in extracted samples as previously described (65). The minimum detection limit was 0.9 pg mL21. The inter- and intraassay coefficients of variation were 15.6 and 10.8%, respectively. Lambing and Behavioural Observation Ewes lambed naturally over a 2-week period while under 24-h surveillance. Additionally, a continuous video record using eight cameras and a Panasonic eight-channel digital field switcher (WJ-FS20/B, Matsushita Communication Industrial, Japan) was made. Ewes were accustomed to the presence of observers in the walkways between pens in the weeks before parturition. Each animal was clearly identified by a paintbrand to facilitate recognition on the videotapes. As far as possible, ewes were allowed to give birth to and care for their lamb unaided. However, lambing assistance was provided if the ewe failed to progress through a time schedule of events—that is, 1 h after the appearance of fluids but on appearance of parts of the lamb, and/or 2 h after parts of the lamb are seen at the vulva with no other obvious progress being made. In all cases intervention was kept to a minimum necessary to ensure the welfare of the ewe and lamb, and

MOTHER–OFFSPRING BEHAVIOUR OF SHEEP

445

mainly involved correcting lamb presentation before the ewe continued the birth process unaided. Birth difficulty was assessed on a score of 1 (no assistance) to 4 (major assistance and/or Caesarean) as previously described (16). In addition, our protocol allowed for intervention after 30 min if the ewe had rejected or was continually aggressive towards the lamb, and after 2 h if the lamb had not stood or sucked. Maternal behaviour data was collected, from videotape, during four observation periods in the 2 h after the birth of a lamb using Keytime data recording software (14). Data were collected for the first 30 min after the birth of each lamb, followed by three 10-min periods, every 30 min, over the following 90 min. In addition, live data recording of ewe and lamb vocalisation, during the same observation periods, was obtained for 73 ewes, using hand-held, Atari Portfolio computers (Atari, Slough, UK), as previously described (20). Behaviours directed towards the lamb were: grooming the lamb, low-pitched bleating, allowing and facilitating sucking attempts, sniffing or nosing the lamb, withdrawing from and/or abandoning the lamb, butting the lamb, high-pitched bleating, moving away as the lamb attempts to suck. Lamb behaviours were recorded as the latency to right and stand, latency to seek the udder, and sucking behaviours. The definitions of ewe and lamb behaviours are as previously described [(16,19); Table 1a]. Lamb weight and sex were recorded at 24 h. The ewe’s response to the handling of her lamb was recorded by a modified Maternal Behaviour Score (51) as follows: 1) ewe leaves lamb, shows no interest, and does not return when handler leaves; 2) ewe leaves lamb, returns when handler leaves; 3) ewe retreats to a distance of 5 or more meters, returns at once; 4) ewe retreats to a distance of 1–5 m, returns at once; 5) ewe

remains within 1 m of lamb and handler; and 6) ewe makes contact with lambs during handling. Statistical Analysis The effects of ewe and lamb breed, lamb sex, and ewe parity on the physiological profiles of ewes throughout gestation, and ewe and lamb behaviours, were investigated by using the Restricted Maximum Likelihood procedure [REML; (53)] given the relatively unbalanced nature of the data. Ewe and lamb breed, ewe parity, and lamb sex were fitted as fixed (block) effects in the model, and sire and dam identity of the lamb were fitted as random variables. For behavioural variables that were not normally distributed (ewe high-pitched vocalisation, lamb behavioural latencies, and lamb vocalisation) the effects of ewe and lamb breed were investigated by Kruskal–Wallis statistics. Principal components analysis was performed using the following variables: proportion of time spent grooming by the ewe; the latencies of lamb behaviours; the frequency of lamb sucking attempts, ewe and lamb vocalisations, and leaving behaviours of ewe and lambs, the response of the ewe to lamb sucking attempts; the length and difficulty of parturition; lamb weight, and plasma concentrations of oestradiol-17b and progesterone. The codes for these variables are given in Table 1b. Some aspects of maternal behaviour (specifically: lamb abandonment and aggression) occurred too infrequently to be used in the analysis. As the behavioural data consisted of repeated measures performed on the same animals separate principal component analyses were carried out on data from the first 30 min after birth, and for the subsequent 90 min. Twenty-six variables were used in the first 30-min analysis and

TABLE 1a DEFINITIONS OF EWE AND LAMB BEHAVIOURS RECORDED IN THE FIRST 2 HOURS AFTER BIRTH Behaviour

Ewe Behaviours Grooming Leaves lamb Ewe vocalisations

Withdrawing Butting Prevents sucking

Abandonment/rejection Lamb Behaviours Shakes head To knees Attempts to stand Stands To udder Suck attempt Successful suck Lamb bleats Lamb leaves

Description

Licking and nibbling movements directed towards the lamb Ewe moves more than two body lengths from lamb Ewe bleats divided into: 1) Low-pitched, mouth closed 2) High-pitched, mouth open Ewe moves back directly away from the lamb at her head (21 steps) Ewe pushes lamb down or away with downwards, sideways, or forwards movements of her head Ewe movements as lamb moves to udder and attempts to suck: 1) Circling: ewe moves hindquarters away from the lamb 2) Backing: ewe moves backwards away from lamb 3) Forwards: ewe steps forwards over the top of the lamb Ewe does not lick lamb, leaves the lamb soon after birth, butts lamb if it approaches, frequently accompanied by high-pitched bleating Lamb raises and shakes head Lamb on chest, pushes up on knees, supporting part of body off the ground Lamb on knees, supports part of its weight on at least 1 foot Lamb supports itself on all 4 feet for at least 5 s Lamb in parallel inverse position with head nudging ewe in udder region Lamb in parallel inverse position, head beneath ewe in udder region, prevented from sucking by ewe movement or leaves udder region within 5 s Lamb has teat in its mouth, in correct position, appears to be sucking for at least 5 s Any vocalisation by the lamb Lamb moves more than two ewe body lengths from dam

446

DWYER, DINGWALL AND LAWRENCE TABLE 1b Codes

Bdiff D_Lsuck Lambwt L_Latstd L_Lknee L_Lshhd L_Lstd L_suck Lthlabr L_udder L_Lunsuc MBT_suck Oest_L P_Egrml Prog_L R_Eback R_Ecirc R_Eend R_Eforw R_Eleave R_Elow R_Lblt R_Lend R_Lleave R_Lsuck

Variable measured

Birth difficulty assessed on a scale from 1 to 4 Duration of sucking bouts in 2 h after birth Lamb birth weight (kg) Latency to attempt to stand (by lamb) Latency to rise to knees (by lamb) Latency to shake head (by lamb) Latency to stand (by lamb) Latency to succesful suck (by lamb) Length of labour (first appearance of fluids to lamb expulsion) Latency to reach udder (by lamb) Latency to attempt to suck (by lamb) Mean sucking bout duration Maternal oestradiol concentration in late gestation (weeks 16 to 20) Proportion of recording period that the ewe licks the lamb Maternal progesterone concentration in late gestation (weeks 16–20) Rate of ewe backing at lamb sucking attempts Rate of ewe circling at lamb sucking attempts Rate of ewe terminating sucking bouts Rate of ewe moving forward at lamb sucking attempts Rate of ewe leaving lamb Rate of ewe low-pitched vocalisation Rate of lamb bleating Rate of lamb terminating sucking bouts Rate of lamb leaving ewe Rate of lamb sucking

24 in the subsequent 90-min analysis. Principal components analysis, followed by orthogonal rotation (VMAX), was carried out using Minitab Statistical Software (Minitab Inc.). Factor scores were then calculated for each ewe and used in subsequent analyses. The effects of ewe and lamb genotype on maternal physiology and factor 1 and 2 scores were analysed using the REML procedure as described above. Factor 3 scores were not normally distributed and were analysed by Kruskal–Wallis statistics. The variables predicting ewe grooming responses in the 30- and 90-min analyses were further investigated by stepwise regression. RESULTS

Gestational Changes in Maternal Physiology The change in plasma oestradiol concentration with increase in gestational age for each class of ewe is shown in Fig. 1. There were no significant differences between Blackface and Suffolk ewes in oestradiol concentration before conception, or between ewes who subsequently became pregnant and those who did not. Plasma oestradiol concentration did not differ from nonpregnant ewes until midgestation (week 10; p , 0.05), thereafter, oestradiol increased rapidly to term, being 15-fold higher than nonpregnant ewes at week 20 (p , 0.001). Oestradiol concentration was higher in Blackface ewes than Suffolk ewes from week 14 (mean values: Blackface 5 8.20 pg mL21, Suffolk 5 5.91 pg mL21, SED 5 0.79, p , 0.05) until week 20 (Blackface 5 18.79 pg mL21, Suffolk 5 13.83 pg mL21, SED 5 1.48, p , 0.001). Lamb breed also had a significant effect on maternal oestradiol concentration in the later stages of gestation (weeks 16–20) with ewes carrying Blackface lambs having elevated oestradiol concentrations in comparison to ewes carrying Suffolk lambs (mean oestradiol con-

centration (pg mL21): BB 5 17.24, BS 5 15.55, SB 5 14.58, SS 5 9.19, SED 5 1.16, p , 0.01). Plasma progesterone profiles for each class of ewe are shown in Fig. 2. As with oestradiol, there were no significant effects of ewe breed or subsequent reproductive state on plasma progesterone concentration before conception. Progesterone concentration was rapidly elevated in the early stage of pregnancy, and was significantly greater in pregnant ewes than nonpregnant by week 6 of pregnancy (Fig. 2, p , 0.001). The number of corpora lutea present at embryo transfer had a significant effect on progesterone concentration in early gestation (weeks 4–8: progesterone concentrations (ng mL21): 1 corpus luteum 5 5.73, 2 5 7.00, 3 5 8.17, SED 5 0.732; p , 0.01), but not thereafter. There were no significant differences between ewe breeds in the number of corpora lutea present (Blackface 5 1.88 6 0.09; Suffolk 5 1.97 6 0.1, F(1, 89) 5 0.45. Progesterone concentration was significantly higher in Suffolk ewes compared to Blackface ewes in early and midgestation (weeks 4 to 14; p , 0.05). In late gestation (weeks 16–20), however, there was a significant interaction between ewe and lamb breed (progesterone concentrations (ng mL21): BB 5 12.68, BS 5 16.51, SB 5 13.41, SS 5 11.65, SED 5 1.30, p , 0.01). Both oestradiol and progesterone concentrations had declined markedly by 3 days postpartum. There were no significant effects of ewe or lamb breed on oestradiol or progesterone concentration at 3 days postpartum. Ewe and Lamb Behaviour at Parturition The main effects of ewe and lamb breed on behaviour at parturition are summarised in Table 2. Ewe behaviour around parturition was mainly influenced by ewe breed, there were

MOTHER–OFFSPRING BEHAVIOUR OF SHEEP

447

FIG. 1. Change in mean plasma oestradiol-17b (pg mL21) with gestation (weeks) for (a) Blackface ewes and (b) Suffolk ewes carrying Blackface lambs (solid symbols and lines) or Suffolk lambs (open symbols and broken lines). The levels recorded at 72 h postpartum are shown as solid (Blackface lamb) or hatched (Suffolk lamb) bars. The concentration of oestradiol in nonpregnant ewes of the equivalent breed is shown by triangle symbols. Values are means (6SE).

FIG. 2. Change in mean plasma progesterone (ng mL21) with gestation (weeks) for (a) Blackface ewes and (b) Suffolk ewes carrying Blackface lambs (solid symbols and lines) or Suffolk lambs (open symbols and broken lines). The levels recorded at 72 h postpartum are shown as solid (Blackface lamb) or hatched (Suffolk lamb) bars. The concentration of oestradiol in nonpregnant ewes of the equivalent breed is shown by triangle symbols. Values are means (6SE).

few effects of lamb breed. Blackface ewes spent more time grooming their lambs than Suffolk ewes and made more lowpitched vocalisation towards their lamb than Suffolk ewes regardless of lamb breed. Blackface ewes were also less likely to withdraw from, leave, be aggressive towards their lamb, or move as the lamb attempted to suck than Suffolk ewes (Table 2). Lamb breed had a significant effect on the proportion of time the ewe spent grooming in the second hour after delivery, with Suffolk lambs receiving more grooming attention than Blackface lambs (p , 0.05). There was also a significant interaction between ewe and lamb breed in the rate of withdrawing from the lamb, with both breeds of ewe being more likely to withdraw from a lamb that was not the same breed as themselves (Table 2; p , 0.05). Blackface lambs were more active than Suffolk lambs, with shorter latencies to stand, seek

the udder and suck successfully (p , 0.001). However, ewe breed also had a significant effect on the latency to stand and to suck, and on the rate of sucking bouts (Table 2). Suffolk lambs were more vocal than Blackface lambs, regardless of the breed of ewe. Factor Analysis Principal components analysis of the data revealed three factors that accounted for 49.5 and 48.7% of the total variation in the variables in the first 30 min and the subsequent 90min analyses respectively (hereafter 30- and 90-min analyses). Orthogonal rotation resulted in the factor loadings shown in Figs. 3 and 4. Factor 1 (23.6 and 23.5% of the total variance for the 30- and 90-min analyses, respectively) was labelled

448

DWYER, DINGWALL AND LAWRENCE TABLE 2 EFFECT OF EWE AND LAMB BREED ON MATERNAL AND NEONATE BEHAVIOUR AT PARTURITION BB

Ewe behaviours Proportion of time grooming lamb (first 30 min) Proportion to time grooming lamb (next 90 min) Rate of withdrawing from lamb

Rate ewe leaves lamb (per 10 min) Percent of ewes being aggressive to the lamb Proportion of suck attempts where ewe backs Proportion of suck attempts where ewe circles Proportion of suck attempts where ewe moves forward Rate ewe low-pitched vocalisation Rate ewe high-pitched vocalisation†

BS

SB

SS

SEM*

Effect

0.79

0.78

0.58

0.60

0.04

Ewe breed: p , 0.001

0.31

0.45

0.18

0.20

0.012

0.00

0.001

0.029

0.005

0.012

0.84

0.53

1.21

1.14

0.07

Ewe breed: p , 0.001 Lamb breed: p , 0.05 Ewe breed: p , 0.001 Ewe breed* lamb breed: p ,0.05 Ewe breed: p , 0.01 Ewe breed: x2 5 4.39, df 5 1, p , 0.05

0.00

3.85

13.64

11.76

0.002

0.008

0.029

0.013

0.007

Ewe breed: p , 0.01

0.283

0.313

0.230

0.238

0.065

NS

0.022

0.024

0.118

0.203

0.01

Ewe breed: p , 0.001

5.31 0.00 (0.0–0.06)

5.08 0.017 (0.0–0.23)

2.62 0.017 (0.0–0.30)

3.98 0.10 (0.01–0.90)

0.43

Ewe breed: p , 0.05 Ewe breed: ‡H 5 3.29, df 5 1, p 5 0.07

Latency to stand (mins)†

13.38 (8.0–20.4)

21.90 (14.1–27.3)

9.83 (7.7–13.7)

15.57 (11.0–20.3)

Latency to suck attempt (min)† Latency to successful suck (min)†

16.87 (11.6–24.9) 42.27 (28.7–60.3)

31.57 (21.5–67.1) 96.43 (60.0–150)

13.82 (8.4–22.7) 40.05 (24.9–52.4)

24.00 (21.9–29.5) 53.73 (44.6–100.4)

Lamb behaviours

Rate of lamb sucking attempts

0.842

0.564

1.267

0.878

Rate of lamb bleating†

0.13 (0.07–0.37)

0.55 (0.14–1.48)

0.25 (0.12–0.43)

0.80 (0.34–1.32)

0.136

Ewe breed: ‡H 5 6.77, df 5 1, p , 0.001 Lamb breed: ‡H 5 14.2, df 5 1, p , 0.001 Lamb breed: ‡H 5 20.94, df 5 1, p , 0.001 Ewe breed: ‡H 5 4.6, df 5 1, p , 0.05 Lamb breed: ‡H 5 23.84, df 5 1, p , 0.001 Ewe breed: p , 0.001 Lamb breed: p , 0.001 Lamb breed: ‡H 5 13.86, df 5 1, p , 0.001

*Variance is pooled standard error of means. Values are means for each group unless otherwise indicated, significance is determined by REML analysis. †Median values (with interquartile ranges). ‡Kruskal–Wallis one-way analysis of variance.

“Lamb Activity,” and had positive loadings for the latency of the lamb to perform righting movements, stand, and attempt to suck (Fig. 3a), and negative loadings for sucking frequency and duration in the first 30-min analysis. In the 90-min analysis there was also a positive loading for lamb birth weight and the length of parturition. Factor 2 (Fig. 3; 13.2 and 12.9%) was labelled “Grooming,” and had different loadings for the 30- and 90-min analyses. During the first 30 min after the birth of the lamb factor 2 had a negative loading for ewe grooming and positive loadings for the length and difficulty of parturition, lamb birth weight, and lamb bleat rate (Fig. 3a). In the 90-min analysis, factor 2 had negative loadings for ewe grooming, oestradiol concentration in late gestation, and the mean length of sucking bouts (Fig. 3b), and positive loadings for high negative responsiveness to sucking attempts (walking forward over the sucking lamb), and ewe–lamb separation ini-

tiated by the ewe. Factor 3 (Fig. 4; 12.6 and 12.2%) was broadly similar for the 30- and 90-min analyses. At 30 min (Fig. 4a) this factor had positive loadings for ewe–lamb separation and lamb bleat rate, and negative loadings for oestradiol concentration, ewe low vocalisation rate and, weakly, the amount of time spent grooming. In the 90-min analysis factor 3 had positive loadings for the ewe backing away or circling as the lamb attempted to suck, and the lamb leaving the ewe, and weakly negative loadings for ewe and lamb bleat rate and oestradiol concentration (Fig. 4b). Stepwise regressions of correlated predictor variables on ewe grooming attention for the 30- and 90-min analysis are shown in Table 3. At 30 min after lamb delivery the main effect on proportion of time spent grooming by the ewe was the latency to groom (this variables was not included in the factor analysis as it was not independent of other variables), fol-

MOTHER–OFFSPRING BEHAVIOUR OF SHEEP

FIG. 3. Factor analysis loadings for factors 1 and 2 in (a) the analysis of ewe and lamb behaviour recorded in the first 30 min after delivery; and (b) the analysis of behaviours recorded in the subsequent 90-min period. The key to the behaviours shown is given in Table 1b. See Materials and Methods for explanation of scoring system for the Maternal Behaviour Score and birth difficulty.

lowed by plasma oestradiol concentration. Plasma progesterone concentration was also included in the model, although it was not significant, as it improved the goodness of fit of the model (56.5%). In the analysis at 90 min the latency to groom was no longer correlated with proportion of time spent grooming and the main effect was that of oestradiol concentration. Although this model accounted for only 22.1% of the variability in ewe grooming behaviour, addition of other variables did not significantly improve the goodness of fit. Effects of Ewe and Lamb Breed on Factor Scores The mean factor scores for each group of ewe–lamb dyads for the 30- and 90-min analyses are shown in Fig. 5. The main effect on factor 1 (Lamb Activity) scores was lamb breed (Fig. 5a; p , 0.001). Suffolk lambs had positive scores (indicating a long labour and prolonged latencies to right and stand),

449

FIG. 4. Factor analysis loadings for factors 2 and 3 in (a) the analysis of ewe and lamb behaviour recorded in the first 30 min after delivery; and (b) the analysis of behaviours recorded in the subsequent 90-min period. The key to the behaviours shown is given in Table 1b.

whereas Blackface lambs had negative scores. There was no influence of lamb sex. There was, however, also a significant influence of ewe breed at both 30 and 90 min (p , 0.01) with lambs born to Blackface ewes having more positive scores than lambs born to Suffolk ewes. Factor 2 (Grooming) scores (Fig. 5b) were significantly affected by ewe breed (p , 0.001) in both 30- and 90-min analyses. Blackface ewes received negative scores, indicative of good maternal behaviour (i.e., high levels of grooming, long mean sucking bouts, and high oestradiol concentration in late gestation), whereas Suffolk ewes received positive scores (e.g., high levels of separation between ewe and lamb). There was also a significant interaction between ewe and lamb breed at 30 min (p , 0.05). Factor 3 scores are shown in Fig. 5c. Scores were significantly affected by ewe breed only in both analyses (30 min: Kruskal–Wallis H (1) 5 19.57, p , 0.001; 90 min: H (1) 4.03, p , 0.05).

450

DWYER, DINGWALL AND LAWRENCE min, respectively). The ratio of maternal oestradiol-to-progesterone concentration in late gestation was significantly negatively correlated with factor 3 at 30 min (Spearmans rs 5 20.372, n 5 69 p , 0.01) and with factor 1 and 2 scores in the 90-min analyses (r2 5 10.7%, p , 0.05 and r2 5 15.9%, p , 0.05 for factors 1 and 2, respectively), but not with factor 3 at this time period. Maternal Behaviour Score was not correlated with any Factor scores. DISCUSSION

FIG. 5. Mean factor scores for (a) factor 1; (b) factor 2; and (c) factor 3 for Blackface (BF) and Suffolk (S) ewes with Blackface lambs (solid bars) or Suffolk lambs (hatched bars) for analyses at 30 min (left panel) or 90 min (right panel). Values are means (6SE).

Relationships Between Factors Maternal progesterone concentration in late gestation was positively correlated with “Lamb Activity” in both analyses (r2 5 10.0%, p , 0.05 and r2 5 16.8%, p , 0.001 for 30 and 90

This study has shown that maternal oestradiol-17b and progesterone concentrations during gestation are affected by both maternal and foetal breed in sheep. In particular, oestradiol concentration is elevated in late gestation in a breed of sheep (Scottish Blackface) previously shown to express more affiliative maternal attention and less rejection behaviour (16) than a breed of sheep (Suffolk) with a lower oestradiol concentration. Factor analysis of mother–offspring behaviour in the first 2 h after birth produced three factors explaining approximately 50% of the variation between variables. The greatest variation was in lamb activity (factor 1), and this was related to maternal progesterone concentration. The variation in maternal behaviours was explained by two factors. Maternal oestradiol, and oestradiol-to-progesterone ratio, in late gestation were related to the maternal behaviours (predominantly grooming) that loaded on factor 2 but only weakly other behaviours (such as sucking cooperation) that loaded on factor 3. A composite measure of maternal behaviour, the Maternal Behaviour Score, which is related to lamb survival (51), was not related to measures of maternal behaviour taken in the immediate postpartum period on either factor. As previously discussed (16), the Maternal Behaviour Score, when measured in a pen environment, may not accurately reflect differences between the two breeds of ewe. The motivation to avoid the handler may be much greater in the less-handled Blackface breed than the intensively raised Suffolk. Additionally, the behaviour of the ewes receiving a low score differed between breeds [i.e., vocal and agitated versus indifference for Blackface and Suffolk respectively; (16)], and this difference was not reflected in the scores received. The factors produced by Principal Component Factor analysis can be regarded as causal or organisational states underlying the observed behaviour patterns (22). Lamb activity in the first 2 h after birth was explained by a single causal factor, and was apparently unrelated to variation in any of the measured maternal behaviours. This factor was significantly affected by lamb breed, and supports previous studies suggesting that variation in maternal behaviour in the early postnatal period does not influence lamb behaviour (18). There was also a significant effect of the maternal breed with Blackface ewes having a more positive loading on factor 1 scores than Suffolk ewes, i.e., lambs born to Blackface ewes had longer behavioural latencies than lambs born to Suffolk ewes. This may be due to some intrauterine growth retardation in lambs born to the smaller Blackface ewes as the ratio of lamb weight to ewe weight is greater in Blackface ewes than Suffolk ewes (19), or to the tendency for birth to be more difficult in Blackface ewes with Suffolk lambs (18). Both these factors tend to increase the latency of lamb behaviours after birth (19,60). There was a weak positive correlation between factor 1 and maternal plasma progesterone concentration in late gestation. Maternal progesterone concentration in late gestation is known to be positively associated with both placental mass

MOTHER–OFFSPRING BEHAVIOUR OF SHEEP

451 TABLE 3

STEPWISE REGRESSION MODEL OF EWE GROOMING ATTENTION Variable

Regression Coefficient

t-ratio

p-Value

20.0004 0.0069 20.071 0.006 0.003

28.96 3.11 22.92 1.82 1.13

,0.001 ,0.01 ,0.01 ,0.1 NS

Subsequent 90 min 0.0074 0.013 20.052 0.359

2.96 2.67 22.87 1.35

,0.01 ,0.01 ,0.01 ,0.1

First 30 min Latency to groom Oestradiol concentration in late gestation Ewe high pitched bleat rate Ewe low-pitched bleat rate Progesterone concentration in late gestation Oestradiol concentration in late gestation Ewe low-pitched bleat rate Frequency of ewe leaving the lamb Frequency of lamb leaving the ewe

at term and lamb birth weight (62). This correlation of maternal progesterone with lamb activity may reflect the longer labour and slower behavioural development of heavier birth weight lambs. Furthermore, there is evidence that maternal progesterone concentration can influence foetal behaviour and arousal (13,46). Inhibition of maternal progesterone synthesis increased foetal breathing movements and behavioural arousal (13), whereas progesterone administration decreased neonatal lamb activity (12). It is possible, therefore, that longterm exposure to high progesterone in utero leads to reduced activity of both the foetal and neonatal lamb. Conversely, oestrogen administration has been found to stimulate respiratory activity in new-born lambs (45), and to increase activity in neonatal piglets (21). Although there was no relationship between maternal oestradiol concentration in late gestation and lamb behaviour in the present study, maternal oestradiol increases rapidly over the last 48 h before delivery in the ewe (11,59); thus, levels of oestradiol in late gestation are very different from those at term. Unlike progesterone, oestrogens do not appear to influence foetal breathing movements, as administration of oestrogen to preterm lambs did not stimulate breathing activity (45). Maternal oestrogen during gestation, therefore, may not play a big role in determining foetal lamb behaviour. Previously, we have shown that the inactive Suffolk lamb is more vocal than the active Blackface lambs in the first 2 h after birth (20). In the present study, however, it seems that the vocal behaviour of the lamb is not strongly liked to the underlying activity level of the lamb, as it is not loaded on factor 1, but is related to environmental factors, such as the grooming behaviour of its mother (factor 2), and her response to its sucking attempts (factor 3). High lamb vocalisation was associated with reduced ewe grooming in both 30- and 90-min analyses, and with both a high level of ewe–lamb separation and a negative ewe response to lamb sucking attempts (moving forward) at 90 min. This is consistent with previous observations (20) and with the observation that isolated lambs bleat more than mothered lambs (47). Vocalisations of young animals may be indicative of need (66) and, in the sheep, may elicit increased maternal attention, particularly grooming in the perinatal period, leading to a decline in bleating activity (47). The maternal behaviours of the ewe were divided between two dimensions: one (factor 2) was predominantly related to ewe grooming attention, and was loaded primarily with variables associated with the birth process at 30 min, and with maternal physiology at 90 min. The other (factor 3), was predic-

tive of ewe–lamb separation at 30 min and of ewe responses to lamb sucking behaviour at 90 min. Both maternal oestradiol and the ratio of oestradiol-to-progesterone were associated predominantly with ewe grooming behaviour and low-pitched vocalisation rate. Oestradiol was only weakly correlated with other maternal behaviours, such as cooperation with sucking attempts of the lamb. A number of authors have suggested that maternal oestradiol, and the changes in endocrine status in late gestation, have the function of reducing aversive responses to neonatal stimuli (24,44,50). It is likely that oestradiol also acts to enhance the attractiveness of the young through its actions, at least in the rat, on the medial preoptic area and the ventral bed nucleus of the stria terminalis (50). The combined influence of an increasing attraction to, and a reduction in aversiveness of, neonatal stimuli, in particular amniotic fluids (5), with elevated oestradiol in the sheep appears to lead to enhanced grooming behaviour. In some multiparous ewes an attraction to amniotic fluids has been seen before parturition (1,4), and this is thought to be associated with high oestrogen concentrations in the maternal circulation (54). The ovarian hormones, however, have a global action and do not code for specific behaviours (29) but are given specificity through their recruitment of neural systems at parturition, in particular the oxytocinergic, noradrenergic, and endorphin systems (24,29). Oxytocin, administered after delivery under peridural anaesthesia, specifically restored an attraction to amniotic fluids (38), and naltrexone administration to primiparous ewes reduced the duration of maternal licking of the lamb (10). Thus, oxytocin and b-endorphin have been implicated in the onset of maternal grooming after birth. It is likely, therefore, that the correlation of maternal oestradiol and ewe grooming behaviour seen in the present study reflects an enhanced ewe responsiveness to the specific action of another neural system, activated at parturition, that leads to grooming behaviour. The second axis of maternal behaviour in the sheep, the response of the ewe to sucking attempts made by her lamb (factor 3 at 90 min), was weakly related to plasma levels of oestradiol in late gestation, and was not related to either progesterone concentration or the ratio of oestradiol to progesterone. In primates, a dual mechanism for maternal responses incorporating both endocrine factors and previous experience has been proposed (56). Prior maternal experience in the ewe has been shown to play only a minor role in affecting ewe grooming behaviour, but has a strong effect on the ewe’s responses to the lamb’s sucking behaviour (52). It is possible that the co-

452

DWYER, DINGWALL AND LAWRENCE

operation of ewes with the sucking attempts of their lambs contains a large learned component, and is less dependent on the physiological status of the ewe than is grooming behaviour. However, there are significant breed differences in the propensity to show inappropriate movements as lambs approach the udder in both inexperienced (16) and experienced ewes (18). These data suggest that there may be some underlying physiological difference between breeds, although this may not be related to oestradiol concentration, or that their early life experiences may differ. For example, Blackface ewes received higher scores than Suffolk ewes for this factor, and they will themselves have received more maternal grooming attention than Suffolk ewes as lambs. In addition, the Blackface is an extensively raised animal living in an outdoor and complex environment, whereas the Suffolk is usually managed intensively. These differences in early experience may affect the way in which the adult ewe responds to the novel situation of suckling a new-born lamb. In factor 3, a high rate of lamb sucking bouts ended by the ewe backing or circling was closely associated with a high frequency of the lamb leaving the ewe. Analysis of the factor scores suggested that these behaviours were also seen more frequently in Suffolk ewes, but were not affected by lamb breed. There is now considerable evidence that the sucking behaviour of the young serves an important function of mediating offspring attachment to their mother alongside nourishment and passive transfer of immunity. Lamb attachment to the ewe is facilitated by sucking activity (49). Sucking and milk ingestion in rats and humans is associated with a number of physiological changes in the infant, including a decreased heart rate and elevated pain threshold (6), and conditioned learning has been demonstrated in human infants only a few hours old (7). A rise in plasma cholecystokinin (CCK) in the lamb with successful sucking behaviour has been implicated in the establishment of a preferential relationship between the lamb and ewe (48). The high level of ewe terminating lamb-sucking bouts seen in the present study, associated with an increased frequency of the lamb leaving the ewe, may reflect disruption to the behavioural role of CCK, and consequently, to impaired learning processes in the lamb and a

weaker attachment between ewe and lamb in Suffolk ewe dyads. This may explain the larger spatial distance maintained by Suffolk ewes to their lambs until weaning, when compared to Blackface ewes (17). In conclusion, this study has shown that the maternal ovarian hormones during gestation are affected by ewe breed, and that changes in maternal oestradiol concentration are related to some aspects of maternal behaviour. Maternal behaviour in the sheep appears to be composed of two motivational dimensions. Ewe responses to lamb-sucking behaviour and ewe grooming attention may be controlled by different physiological processes, at least in multiparous ewes. Ewe grooming behaviour appeared relatively hard-wired and was associated with elevated maternal oestradiol concentration in late gestation, suggesting that this behaviour has a large physiological component. Ewe responses to the lamb’s attempts to such, however, did not appear to be greatly influenced by ewe physiology in late gestation, but may be more affected by the previous experience of the ewe. As seen in previous studies (16,18), Blackface ewes performed better than Suffolk in ewes in both aspects of maternal behaviour, regardless of the breed of lamb, and also had higher oestradiol and oestradiolto-progesterone concentrations than Suffolk ewes in late gestation. Lamb behaviour was explained by a single motivational dimension, and was strongly related to lamb breed, as seen in previous studies (19). There was no relationship between lamb behaviours and specific maternal behaviours, although ewe breed did have a significant effect on lamb activity, possibly mediated via prenatal factors.

ACKNOWLEDGEMENTS

This work was supported by the Scottish Office Agriculture, Environment and Fisheries Department. The authors would like to thank the following: K. McLean, L. Deans, J. Chirnside, and S. Calvert for assistance with data collection at lambing time; J. FitzSimons, G. Callaghan, and M. Ramsay for animal husbandry; Edinburgh Genetics for carrying out the embryo transfer procedures; I. Swanson and F. Pitt at MRC for hormone assays; and E. Austin at BioSS for statistical advice.

REFERENCES 1. Alexander, G.: Maternal behaviour in the Merino ewe. Proc. Aust. Soc. Anim. Prod. 3:105–114; 1960. 2. Alexander, G.; Stevens, D.; Kilgour, R.; de Langen, H.; Mottershead, B. E.; Lynch, J. J.: Separation of ewes from twin lambs: Incidence in several sheep breeds. Appl. Anim. Ethol. 10:301– 317; 1983. 3. Alexander, G.; Stevens, D.; Bradley, B.; Barwick, S. A.: Maternal behaviour in Border Leicester, Glen Vale & Merino sheep. Aust. J. Exp. Agric. 30:27–38; 1990. 4. Arnold, G. W.; Morgan, P. D.: Behaviour of the ewe and lamb at lambing and its relationship to lamb mortality. Appl. Anim. Ethol. 2:25–46; 1975. 5. Arnould, C.; Piketty, V.; Lévy, F.: Behaviour of ewes at parturition towards amniotic fluids from sheep, cows and goats. Appl. Anim. Behav. Sci. 32:191–196; 1991. 6. Blass, E. M.: Behavioral and physiological consequences of suckling in rat and human newborns. Acta Paediatr. (Suppl.) 397:71– 76; 1994. 7. Blass, E. M.; Ganchrow, J. R.; Steiner, J. E.: Classical conditioning in newborn humans 2–48 hours of age. Infant Behav. Dev. 7:223–235; 1984. 8. Broad, K. D.; Kendrick, K. M.; Sirinathsinghji, D. J. S.; Keverne,

9.

10. 11.

12. 13.

E. B.: Changes in oxytocin immunoreactivity and mRNA expression in the sheep brain during pregnancy, parturition and lactation and in response to oestrogen and progesterone. J. Neuroendocrinol. 5:435–444; 1993. Broad, K. D.; Kendrick, K. M.; Sirinathsinghji, D. J. S.; Keverne, E. B.: Changes in pro-opiomelanocortin and pre-proenkephalin mRNA levels in the ovine brain during pregnancy, parturition and lactation and in response to oestrogen and progesterone. J. Neuroendocrinol. 5:711–719; 1993. Caba, M.; Poindron, P.; Krehbiel, D.; Lévy, F.; Romeyer, A.; Venier, G.: Naltrexone delays the onset of maternal behaviour in primiparous parturient ewes. Pharmacol. Biochem. Behav. 52:743–748; 1995. Chamley, W. A.; Buckmaster, J. M.; Cerini, M. E.; Cumming, I. A.; Goding, J. R.; Obst, J. M.; Williams, A.; Winfield, C.: Changes in the levels of progesterone, corticosteroids, estrone, estradiol-17b, luteinizing hormone, and prolactin in the peripheral plasma of the ewe during pregnancy and at parturition. Biol. Reprod. 9:30–35; 1973. Crenshaw, M. C.; Meschia, G.; Barron, D. H.: Role of progesterone in inhibition of muscle tone and respiratory rhythm in fetal lambs. Nature 212:842; 1966. Crossley, K. J.; Nicol, M. B.; Hirst, J. J.; Walker, D. W.; Thor-

MOTHER–OFFSPRING BEHAVIOUR OF SHEEP

14. 15. 16. 17. 18. 19.

20.

21. 22. 23. 24.

25.

26.

27.

28. 29. 30. 31. 32. 33. 34.

35.

burn, G. D.: Suppression of arousal by progesterone in foetal sheep. Reprod. Fertil. Dev. 9:767–773; 1997. Deag, J. M.: Keytime: A program system for recording and analysing behavioural data. Copyright © John M. Deag 1987–1993. Dwyer, C. M.; Lawrence, A. B.: Induction of maternal behaviour in non-pregnant, hormone-primed ewes. Anim. Sic. 65:403–408; 1997. Dwyer, C. M.; Lawrence, A. B.: Variability in the expression of maternal behaviour in primiparous sheep: Effects of genotype and litter size. Appl. Anim. Behav. Sci. 58:311–330; 1998. Dwyer, C. M.; Lawrence, A. B.: Ewe–ewe and ewe–lamb behaviour in a hill and a lowland breed of sheep: A study using embryo transfer. Appl. Anim. Behav. Sci. 61:319–334; 1999. Dwyer, C. M.; Lawrence, A. B.: Does the behaviour of the lamb influence the expression of maternal behaviour in the sheep? Behaviour. 136:367–389 Dwyer, C. M.; Lawrence, A. B.; Brown, H. E.; Simm, G.: Effect of ewe and lamb genotype on gestation length, lambing ease, and neonatal behaviour of lambs. Reprod. Fertil. Dev. 8:1123–1129; 1996. Dwyer, C. M.; McLean, K. A.; Deans, L. A.; Chirnside, J.; Calvert, S. K.; Lawrence, A. B.: Vocalisations between mother and young in sheep: Effects of breed and maternal experience. Appl. Anim. Behav. Sci. 58:105–119; 1998. Hughes, P. E.; Pearce, G. P.; Baguio, S. S.: The effects of estradiol treatment on activity, growth and survival of neonatal piglets. Anim. Prod. 54:299–304; 1992. Huntingford, F. A.: An investigation of the territorial behaviour of the three-spined stickleback (Gasterosteus aculeatus) using principal components analysis. Anim. Behav. 24:822–834; 1976. Insel, T. R.: Regional changes in brain oxytocin receptors postpartum: Time course and relationship to maternal behaviour. J. Neuroendocrinol. 2:539–545; 1990. Kendrick, K. M.; da Costa, A. P.; Broad, K. D.; Ohkura, S.; Guevara, R.; Lévy, F.; Keverne, E. B.: Neural control of maternal behaviour and olfactory recognition of offspring. Brain Res. Bull. 44:383–395; 1997. Kendrick, K. M.; da Costa, A. P.; Hinton, M. R.; Keverne, E. B.: A simple method for fostering lambs using anoestrous ewes with artificially induced lactation and maternal behaviour. Appl. Anim. Behav. Sci. 34:345–357; 1992. Kendrick, K. M.; Keverne, E. B.: Effects of intracerebroventricular naltrexone and phentolamine on central and peripheral oxytocin release and on maternal behaviour induced by vaginocervical stimulation in the ewe. Brain Res. 505:329–332; 1989. Kendrick, K. M.; Keverne, E. B.: Importance of progesterone and oestrogen priming for the induction of maternal behaviour by vaginocervical stimulation in sheep: Effects of maternal experience. Physiol. Behav. 49:745–750; 1991. Kendrick, K. M.; Keverne, E. B.; Baldwin, B. A.: Intracerebroventricular oxytocin stimulates maternal behaviour in the sheep. Neuroendocrinology 46:56–61; 1987. Keverne, E. B.: Central mechanisms underlying the neural and neuroendocrine determinants of maternal behaviour. Psychoneuroendocrinology 13:127–141; 1988. Keverne, E. B.; Kendrick, K. M.: Morphine and corticotrophinreleasing factor potentiate maternal acceptance in multiparous ewes after vaginocervical stimulation. Brain Res. 540:55–62; 1991. Keverne, E. B.; Kendrick, K. M.: Oxytocin facilitation of maternal behaviour in sheep. Annal. NY Acad. Sci. 652:83–101; 1992. Keverne, E. B.; Kendrick, K. M.: Maternal behaviour in sheep and its neuroendocrine regulation. Acta Paediatr. (Suppl.) 83:47– 56; 1994. Keverne, E. B.; Lévy, F.; Poindron, P.; Lindsay, D. R.: Vaginal stimulation: An important determinant of maternal bonding in sheep. Science 219:81–83; 1983. Krehbiel, D.; Poindron, P.; Lévy, F.; Prud’homme, M. J.: Effects of peridural anaesthesia on maternal behaviour in primiparous and multiparous parturient ewes. Physiol. Behav. 40:463–472; 1987. Le Neindre, P.; Poindron, P.; Delouis, C.: Hormonal induction of maternal behaviour in non-pregnant ewes. Physiol. Behav. 40:463–472; 1979.

453 36. Lent, P. C.: Mother–infant relationships in ungulates. In: Geist, V.; Walther, F., eds. The behaviour of ungulates and its relationship to management. Morges, Switzerland: IUCN Publication No. 24; 1974:14–55. 37. Lévy, F.; Poindron, P.: The importance of amniotic fluids for the establishment of maternal behaviour in experienced and inexperienced ewes. Anim. Behav. 35:1188–1192; 1987. 38. Lévy, F.; Kendrick, K. M.; Keverne, E. B.; Piketty, V.; Poindron, P.: Intracerebral oxytocin is important for the onset of maternal behaviour in inexperienced ewes delivered under peridural anaesthesia. Behav. Neurosci. 106:427–432; 1992. 39. Lévy, F.; Poindron, P.; Le Neindre, P.: Attraction and repulsion by amniotic fluids and their olfactory control in the ewe around parturition. Physiol. Behav. 31:687–692; 1983. 40. Lyons, D. M.; Price, E. O.; Moberg, G. P.: Individual differences in temperament of domestic dairy goats: Constancy and change. Anim. Behav. 36:1323–1333; 1988. 41. McKelvey, W. A. C.; Robinson, J. J.; Aitken, R. P.: A simplified technique for the transfer of ovine embryos by laparoscopy. Vet. Rec. 117:492–494; 1985. 42. McKelvey, W. A. C.; Robinson, J. J.; Aitken, R. P.; Henderson, G. D.: The evaluation of a laparoscopic technique in ewes. Theriogenology 24:519–535; 1985. 43. McNeilly, A. S.; Fraser, H. M.: Effect of gonadotropin-releasing hormone against induced suppression of LH and FSH on follicle growth and corpus luteum function in the ewe. J. Endocrinol. 115:273–282; 1987. 44. Mellor, D. J.; Dingwall, R. A.; King, T. J.: Successful fostering of orphan lambs by nonpregnant ewes induced to lactate using exogenous hormone treatment. N.Z. Vet. J. 41:200–204; 1993. 45. Mellor, D. J.; MacKay, J. M. K.; Williams, J. T.: Effects of oestrogen on activity and survival of lambs delivered by hysterectomy. Res. Vet. Sci. 13:399–401; 1972. 46. Nicol, M. B.; Hirst, J. J.; Walker, D.; Thorburn, G. D.: Effect of alteration of maternal plasma progesterone concentrations on fetal behavioural state in late gestation. J. Endocrinol. 152:379–386; 1997. 47. Nowak, R.: Lamb bleats: Important for the establishment of the mother–young bond? Behaviour 115:14–29; 1990. 48. Nowak, R.; Goursand, A.-P.: Development of a preference for the mother by the new-born lamb: Involvement of peripheral cholesystokinin receptors. Proc. 32nd Congress ISAE, France; 1998. 49. Nowak, R.; Murphy, T. M.; Lindsay, D. R.; Alster, P.; Andersson, R.; Uvnäs-Moberg, K.: Development of a preferential relationship with the mother by the newborn lamb: Importance of the sucking activity. Physiol. Behav. 62:681–688; 1997. 50. Numan, M.; Sheehan, T. P.: Neuroanatomical circuitry for mammalian maternal behaviour. Ann. NY Acad. Sci. 807:101–125; 1994. 51. O’Connor, C. E.; Jay, N. P.; Nicol, A. M.; Beatson, P. R.: Ewe maternal behaviour score and lamb survival. Proc. N.Z. Soc. Anim. Prod. 45:159–162; 1985. 52. O’Connor, C. E.; Lawrence, A. B.; Wood–Gush, D. G. M.: Influence of litter size and parity on maternal behaviour at parturition in Scottish Blackface sheep. Appl. Anim. Behav. Sci. 33:345–355; 1992. 53. Patterson, H. D.; Thompson, R.: Recovery of inter-block information when block sizes are unequal. Biometrika 58:545–554; 1971. 54. Poindron, P.; Le Neindre, P.: Endocrine and sensory regulation of maternal behaviour in the ewe. Adv. Study Behav. 11:75–199; 1980. 55. Poindron, P.; Lévy, F.; Krehbiel, D.: Genital, olfactory amd endocrine interactions in the development of maternal behaviour in the parturient ewe. Psychoneuroendocrinology 13:99–125; 1988. 56. Pryce, C. R.: The regulation of maternal behaviour in marmosets and tamarins. Behav. Process. 30:201–224; 1993. 57. Pryce, C. R.; Abbott, D. H.; Hodges, J. K.; Martin, R. D.: Maternal behaviour is related to prepartum urinary oestradiol levels in red-bellied tamarin monkeys. Physiol. Behav. 44:717–726; 1988. 58. Putu, I. G.; Poindron, P.; Lindsay, D. R.: A high level of nutrition in late pregnancy improves subsequent maternal behaviour of Merino ewes. Proc. Aust. Soc. Anim. Prod. 17:294–297; 1988.

454 59. Shipka, M. P.; Ford, S. P.: Relationship of circulating oestrogen and progesterone concentrations during late pregnancy and the onset phase of maternal behaviour in the ewe. Appl. Anim. Behav. Sci. 31:91–99; 1991. 60. Slee, J.; Springbett, A.: Early post-natal behaviour in lambs of ten breeds. Appl. Anim. Behav. Sci. 15:229–240; 1986. 61. Thomson, A. M.; Thomson, W.: Lambing in relation to the diet of the pregnant ewe. Br. J. Nutr. 2:290–305; 1949. 62. Wallace, J. M.; Aitken, R. P.; Cheyne, M. A.; Humblot, P.: Pregnancy-specific protein B and progesterone concentrations in relation to nutritional regimen, placental mass and pregnancy outcome in growing adolescent ewes carrying singleton fetuses. J. Reprod. Fertil. 109:53–58; 1997.

DWYER, DINGWALL AND LAWRENCE 63. Wang, M.-W.; Crombie, D. L.; Hayes, J. S.; Heap, R. B.: Aberrant maternal behaviour in mice treated with a progesterone receptor anatagonist during pregnancy. J. Endocrinol. 145:371– 377; 1995. 64. Wang, M.-W.; Heap, R. B.; Taussig, M. J.: Abnormal maternal behaviour in mice previously immunized against progesterone. J. Endocrinol. 134:257–267; 1995. 65. Webb, R.; Baxter, G.; McBride, D.; Nordblom, G. D.; Shaw, M. P. K.: The measurement of testosterone and oestradiol-17b using iodinated tracers and incorporating an affinity chromatography extraction procedure. J. Steroid Biochem. 23:1043–1051; 1985. 66. Weary, D. M.; Fraser, D.: Calling by domestic piglets: Reliable signals of need? Anim. Behav. 50:1047–1055; 1995.