- Email: [email protected]
The effect of pasture allowance fed to twin- and triplet-bearing ewes in late pregnancy on ewe and lamb behaviour and performance to weaning
Livestock Production Science 97 (2005) 253 – 266 www.elsevier.com/locate/livprodsci
The effect of pasture allowance fed to twin- and triplet-bearing ewes in late pregnancy on ewe and lamb behaviour and performance to weaning J.M. Everett-Hincks*, H.T. Blair, K.J. Stafford, N. Lopez-Villalobos, P.R. Kenyon, S.T. Morris Institute of Veterinary Animal and Biomedical Sciences, Massey University, Private Bag 11 222, Palmerston North, New Zealand Received 29 March 2004; received in revised form 21 April 2005; accepted 13 May 2005
Abstract This paper investigates whether pasture allowance before and at lambing affects ewe behaviour, lamb behaviour and subsequent twin and triplet lamb performance to weaning. The trial was carried out at Massey University Keeble Farm in New Zealand. At pregnancy scanning, 186 first cycle-mated ewes were identified as carrying twins (n = 96) or triplets (n = 90). The ewes were randomly divided into four nutritional treatments (2, 4, 6 and 8 cm post-grazing sward heights). Each treatment was replicated (n = 12 ewes in each replicate during pregnancy) on two separate blocks. The trial ewes remained on the four nutritional treatments during pregnancy and parturition. The mean lambing date was the 9 September 2002. Within 12 h of birth lambs were weighed, their sex recorded and identified to the ewe. Maternal and lamb behaviours were then recorded for a period of 5 min. Twin litter weight at birth was similar for ewes grazing 2-, 4- and 6-cm sward heights. Increasing pasture allowance from just 2 cm to 4 cm to ewes with triplet litters increased litter weight at birth by 2 kg, increased weight of lamb weaned by 8 kg and improved litter survival to tagging by 4%. Triplet lamb survival to tagging increased by 15% from the lowest pasture allowance to the highest (from 2 cm to 8 cm). Triplet litter survival was similar to twin litter survival at the highest pasture allowance (8 cm; 2000 kg DM/ha). Pasture allowance had a greater effect on lamb behaviour than ewe behaviour and triplets were more affected than twins. Feeding level did not affect Maternal Behaviour Score (MBS); however, MBS was related to a number of lamb behaviours. Ewes on higher pasture allowances were more likely to stay with their litter at the tagging site. Ewes with triplet litters on the 2cm sward height were further from their litter 5 min after tagging and high-pitch bleated significantly more than ewes on higher swards. These ewes were less likely to make contact with their lamb after tagging. Triplets born on lower pasture allowances were less likely to stand, locate their dam’s udder and follow their moving dam. Fewer triplet lambs bleated following tagging and this might be why fewer ewes made contact.
* Corresponding author. Invermay Agricultural Centre, AgResearch, Puddle Alley, Private Bag 50034, Mosgiel, New Zealand. Tel.: +64 3 489 9164; fax: +64 489 9024. E-mail address: [email protected] (J.M. Everett-Hincks). 0301-6226/$ - see front matter D 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.livprodsci.2005.05.006
254
J.M. Everett-Hincks et al. / Livestock Production Science 97 (2005) 253–266
Ewes with triplets grazing high pasture allowances in late pregnancy can achieve litter survival rates similar to those with twins. Ewes with triplets require higher pasture allowances to achieve similar production as ewes with twin litters. D 2005 Elsevier B.V. All rights reserved. Keywords: Maternal behaviour; Lamb behaviour; Lamb survival; Pasture allowance
1. Introduction Improved lambing percentage makes the biggest contribution to higher profits on New Zealand sheep farms (Geenty, 1997). Pregnancy scanning enables farmers to identify and preferentially feed ewes with multiple foetuses. However, triplet lamb survival is a problem. A ewe’s maternal behaviour and litter survival are under minimal genetic control and the major source of variation in these traits is due to temporary environmental effects (Everett-Hincks et al., 2005). Lamb survival is predominantly controlled by the environment (Lopez-Villalobos and Garrick, 1999; Morris et al., 2000; Everett-Hincks et al., 2002; Everett-Hincks et al., 2005). Therefore, a better understanding of environmental conditions, particularly feeding levels throughout pregnancy and at lambing, may allow a reduction in lamb losses seen in highly fecund ewes. In late pregnancy and early lactation, the nutritional requirements of ewes with twins are very high (Holmes, 1975). The last 4–6 weeks before lambing are critical (Scales et al., 1986). When multiple bearing ewes were offered additional feed in late pregnancy to increase live weight, lamb mortality was reduced, suggesting merit in improved pre-lamb feeding for ewes carrying more than one lamb (Scales et al., 1986). To date, doptimalT pasture feeding levels for ewes with triplets have not been adequately defined as research into ewe energy requirements in late pregnancy has focused on ewes with singles and twins (Geenty and Rattray, 1987). Inadequate feeding can result in poor maternal behaviour, an important cause of lamb death (Nowak, 1996). Dwyer et al. (2003) demonstrated that a moderate reduction in maternal nutrition in late pregnancy caused a reduction in the expression of maternal behaviours, in particular the Maternal Behaviour Score (MBS), at parturition under intensive indoor conditions. The New Zealand sheep industry is pasture based and the strength of ewe– lamb attachment for ewes with twin and triplet litters
grazing differing pasture allowances in late pregnancy is unknown. Ewe–lamb bonding and lamb survival are maximised by management practices that increase the time spent on the birth site by the ewe after parturition (Nowak, 1996). If pasture quality and quantity are good and the paddock has adequate shelter, then the ewe will stay at the birth site with her lambs for a longer period (Pollard and Littlejohn, 1999). In addition, Lindsay et al. (1990) observed that the time spent on the birth site increased when ewes with twins were on a higher plane of nutrition 6 weeks before lambing. Nowak (1996) observed a considerable improvement in twin lamb bonding with their dam and twin lamb survival when the mother remained on the birth site for a minimum of 6 h. However, Murphy et al. (1994a) suggested that it is not that the dam is on the birth site for 6 h after parturition that improves lamb survival but that the dam is together with all of her lambs. Nowak and Lindsay (1992) concluded that the ability of twin lambs to discriminate their dams from alien ewes 12 h after birth is related to survival during the neonatal period and Shillito-Walser (1978) stated that the formation of the dam–lamb bond is the result of a rapid learning process influenced by both the behaviour of the dam and the lamb. Therefore, lamb and maternal behaviour must be investigated conjointly to identify the environmental conditions that impact on the formation of a successful ewe–lamb bond. This study investigates whether pasture allowance before and at lambing affects ewe behaviour, lamb behaviour and subsequent twin and triplet lamb performance to weaning.
2. Materials and methods 2.1. Animals and measurements The animals included in the trial were monitored from pregnancy scanning to lamb weaning. At preg-
J.M. Everett-Hincks et al. / Livestock Production Science 97 (2005) 253–266
nancy scanning (17 June; about day 64 of pregnancy) 186 first cycle-mated ewes were identified as carrying twins (n = 96) or triplets (n = 90). The trial was carried out at Massey University Keeble Farm. The trial ewes were made up evenly of mixed age and two-tooth ewes with predominantly Romneys. The ewes were randomly divided into four nutritional treatments (2, 4, 6 and 8 cm post-grazing sward heights) representing a dpoorT (800 kg DM/ ha), dnormalT (1200 kg DM/ha), dgenerousT (1600 kg DM/ha) and dovergenerousT (2000 kg DM/ha) feed supply, respectively (Rattray et al., 1987). Each treatment was replicated (n = 12 ewes in each replicate during pregnancy) on two separate blocks (dPinesT and dCaravanT blocks) to ensure statistical integrity. The pasture height was measured weekly by rising plate meter method and maintained by modifying the stocking rate. The trial ewes remained on the four nutritional treatments during pregnancy and parturition and were removed after the lambs were ear-tagged. Approximately 24 h after the lamb was ear-tagged, the ewe and her lambs were transferred to one of two pastures with post-grazing sward heights of 4 and 8 cm and remained on this feed level until weaning (6 December 2002, mean lamb age 88 days). The mean lambing date was the 9 September 2002. Weather conditions were mild and consistent throughout lambing. Date and hour of birth was recorded for each ewe and her litter. Within 12 h of birth, lambs (alive and dead) were weighed, ear-tagged, their sex recorded and identified to the ewe. This procedure is referred to as dtaggingT and was performed by two people, one person measuring and the second person recording. Immediately following tagging, the lambs were placed together on the ground, the two recorders moved 20–30 m away where they recorded maternal and lamb behaviours for a period of 5 min. Litter size was recorded at pregnancy scanning, birth, tagging and weaning. Litter survival was measured from birth to tagging and tagging to weaning and was calculated by dividing the latter litter size by the former litter size and recorded as a percentage. Litter survival from birth to weaning was also calculated. Litter weight at birth and weight of lamb weaned were recorded for each ewe. Ewes that lost an entire litter were included in the analysis as rearing 0 kg of lamb to weaning.
255
Lamb survival was recorded at tagging and weaning. Lamb survival was recorded as a binary trait, lambs alive at each of these stages was given a score of d1,T whereas lambs not present at tagging or weaning were given a score of d0T. Litter survival and lamb survival are related but were analysed separately in this study. Litter survival measures the dam’s ability to rear all of her lambs born in a particular parity, while lamb survival is measuring an individual lamb’s ability to survive and is influenced by the dam. Lamb growth rate (grams/day) was calculated by subtracting birth weight from weight at weaning and dividing it by the number of days between the two measurements. 2.2. Maternal behaviour A Maternal Behaviour Score (MBS) described by O’Connor et al. (1985) was scored on a 5-point scale (low, 1 to high, 5) based on the distance a ewe retreated from her lambs when the shepherd tagged them. The following ewe behaviours were recorded immediately after tagging. The time to first contact after tagging between the ewe and the lamb was recorded (ewe–lamb contact). Lambs that had no contact with their ewe 5 min after tagging were recorded as d0T and those that did have ewe contact were recorded as d1.T Ewe bleats were counted and categorised as low-pitch bleats (rumble calls) and high-pitch bleats. Low bleats made with the mouth closed generally occur when the ewe is in close proximity to the lamb. High ewe bleats made with the mouth open and are generally made by ewes when their lambs are separated from them (Shillito, 1972). Ewes that moved further than 5 m from the tagging site were recorded as a binary trait (d0T ewe moved away and d1T ewe stayed). The maximum distance of the ewe from any one of her lambs was estimated in meters at the end of the 5-min recording period. For each ewe and her litter, the date and time of tagging was recorded as well as the approximate age of the lambs in hours at tagging. 2.3. Lamb behaviour At tagging, each lamb was identified by a colour spray mark and its behaviour recorded for 5 min. Each
256
J.M. Everett-Hincks et al. / Livestock Production Science 97 (2005) 253–266
lamb was placed on its side on the ground immediately after tagging. The time from the end of tagging until the lamb stood was recorded in minutes. In some instances, the lamb did not stand during the 5 min of observation and was recorded as d0,T whereas a lamb that stood during the 5-min post-tagging period was given a score of d1T. Individual ewe–lamb contact time was recorded (as described in Section 2.2). Lamb bleating while separated from its dam was recorded as a binary trait, d0T for lambs that did not bleat and d1T for those that did. If the ewe moved more than 5 m away from the tagging site, the time (minutes) taken for the lamb to follow was recorded. A lamb that failed to follow its moving dam was recorded as d0,T whereas a lamb that followed was given a d1T. The time taken for each lamb to locate the ewe’s udder and to suck were also recorded. Lambs that did not perform these dbehavioursT during the 5 min following tagging were given a score of d0,T whereas those that did were given a score of d1T. 2.4. Statistical analyses The trial was set up as a split plot design and tested for the fixed effect of pasture allowance in late pregnancy as the big plot and litter size at birth as the small plot. Continuously dependent maternal behaviours, lamb behaviours at tagging and lamb measurements and weights were analysed using the MIXED procedure (SAS, 2002) and lamb traits were adjusted for the random maternal effect as shown by fitting the following linear model: Yijklm ¼ l þ Bi þ Pj þ Lk þ PLjk þ BPij þ bXijklm þ dm þ eijklm where Yijklm is the response of the lth animal of the kth litter size from the mth dam of the jth pasture allowance in the ith block, l is the general mean, B i is the random effect of the ith block (i = Pines block or Caravan block), P j is the fixed effect of the jth pasture allowance ( j = 2, 4, 6 or 8 cm sward height) in late pregnancy, L k is the fixed effect of the kth litter size at birth (k = twin or triplet), PL jk is the fixed interaction effect of pasture allowance j and litter size at birth k, BP ij is the random interaction effect of pasture allowance j and block i (main plot error), X ijklm is the
covariate tested (e.g., day of birth, ewe age, live weight, condition score), b is the regression of Y on the covariate tested (e.g., day of birth), d is the random effect of dam m where lamb l was born to dam m and provides the maternal effect and litter effects common for the lambs born in the same litter. e ijklm is the random residual effect unique to Yijklm . The fixed effect of pasture allowance from tagging to weaning (4 cm or 8 cm) was added to the model to test differences in performance from tagging to weaning, such as lamb survival and weight at weaning and only reported if statistically significant at P b 0.05. To help decide the pathways involved in the differences in ewe and lamb performance and ewe and lamb behaviour, different covariates were tested and retained in the model if statistically significant at P b 0.05. Results are given as least square means and their standard errors. Multiple comparisons between least squares means for twin and triplet litters, sward height pasture allowance and combinations between these fixed effects were performed and only reported if statistically significant at P b 0.05. PROC CATMOD (SAS, 2002) was used to analyse categorical dependent behaviour traits (i.e., 0 = did not stand, 1 = did stand) and lamb survival at tagging and weaning (0 = dead and 1 = alive). Least squares means for the fixed effects of litter size at birth, pasture allowance and the interaction between pasture allowance and litter size were derived. The CONTRAST statement was used to determine if differences in fixed effects were significant at P b 0.05. The LIFETEST procedure (SAS, 2002) was used to compute and plot a survival distribution function (SDF) of post-tagging time behaviours for twin and triplet lambs on 2, 4, 6 and 8 cm sward heights. These behaviour observations are right-censored due to the termination of the experiment (i.e., 5 min after tagging). But LIFETEST correctly uses the censored observations as well as the non-censored. The SDF evaluated at t (i.e., 5 min) is the probability that a lamb from a specified sward height and litter size will have a dbehaviourT time exceeding t; that is, S(t) = Pr(T N t), where S(t) denotes the survival distribution function and T is the behaviour time of a randomly selected lamb. The cumulative distributive functions for each of the post-tagging behaviour traits are graphically plotted and are defined as 1 S(t), that is, the probability that a behaviour time does not exceed time
J.M. Everett-Hincks et al. / Livestock Production Science 97 (2005) 253–266
257
(t) (i.e., 5 min). To determine whether the curves plotted are homogenous across sward heights ( P N 0.05), PROC LIFETEST provides two rank tests (log-rank and Wilcoxon). The Wilcoxon test places more weight on early times and the log-rank test places more weight on larger times. The association between covariates and the behaviour time variable were investigated using the Wilcoxon test and reported if significant at P b 0.05. The covariate tests are pooled across sward height treatments and it is not possible to calculate the directional effect the covariate is having on the dependent variable. Pearson correlation coefficients (r) were computed using SAS (PROC CORR) for continuously independent ewe and lamb behaviour and production traits and were reported if significant at P b 0.05.
MBS high bleated more (r = 0.54, P b 0.0001). A greater proportion of ewes grazing the 8-cm sward stayed with their litter after tagging compared with ewes on the 2-cm sward ( P = 0.07) (Table 1). Ewes with triplet litters on the 2-cm sward were farthest from their litter at the end of the 5-min observation period compared to ewes on higher allowances and ewes with twin litters on the 2-cm sward (Table 1). Ewes with triplets high bleated significantly more (8 F 1.15 bleats) than ewes with twins (5 F 1.06 bleats) ( P b 0.05) (Table 1). High-pitch bleating frequency was similar for ewes grazing all swards ( P N 0.05) (Table 1) and there was no significant interaction between feeding level up to tagging and litter size at birth ( P N 0.05). Litter size at birth tended to influence ewe low bleat frequency; however, this was not significant ( P N 0.05).
3. Results
3.2. Dam performance
3.1. Maternal behaviour
Litter weight was significantly lighter for ewes grazing the 2-cm sward height (9.2 F 0.30 kg) in late pregnancy compared to ewes grazing the 8 cm (10.6 F 0.30 kg) sward height ( P b 0.05) (Table 2). Triplet litters weighed significantly more at tagging than twin litters regardless of pasture allowance in late
MBS was similar between ewes on differing pasture allowances (Table 1). High-pitch bleating frequency had a significant effect on MBS (b = 0.08 F 0.009, P b 0.0001), the ewes with lower
Table 1 The effect of pasture allowance fed to ewes in late pregnancy and litter size at birth on Maternal Behaviour Score (MBS) and ewe tagging behaviour traits (least squares mean F standard error) Pasture allowance levela
Trait
MBS
Ewes that stayed (proportion)
Ewe high bleats
Ewe low bleats
Maximum distance after tagging (m)
a
Twin Triplet Pb Twin Triplet Pb Twin Triplet Pb Twin Triplet Pb Twin Triplet Pb
2 cm
4 cm
6 cm
8 cm
3.4 F 0.25 2.9 F 0.29 ns 0.26 F 0.092 0.26 F 0.101 ns 4 F 2.1 11 F 2.4 * 2.6 F 0.50A 0.7 F 0.61 * 5 F 3.6 16 F 4.5B *
3.0 F 0.26 2.7 F 0.27 ns 0.45 F 0.106 0.32 F 0.107 ns 6 F 2.1 6 F 2.3 ns 1.4 F 0.51 1.7 F 0.59 ns 3 F 3.6 0 F 4.6A ns
3.1 F 0.25 3.3 F 0.29 ns 0.23 F 0.089 0.35 F 0.116 ns 7 F 2.1 7 F 2.3 ns 1.2 F 0.49B 0.8 F 0.59 ns 6 F 3.6 1 F 4.2A ns
3.4 F 0.26 3.1 F 0.26 ns 0.48 F 0.104 0.42 F 0.113 ns 2 F 2.2 7 F 2.2 ns 1.5 F 0.53 1.1 F 0.58 ns 3 F 3.7 5 F 3.8 ns
Pasture allowance from pregnancy scanning to lamb tagging. Means within the same row with different superscript differ significantly (no superscript = not significant, P N 0.05; superscripts different P b 0.05). b Mean comparison within each level of pasture allowance (ns = not significant, P N 0.05; * P b 0.05).
258
J.M. Everett-Hincks et al. / Livestock Production Science 97 (2005) 253–266
Table 2 The effect of pasture allowance fed to ewes in late pregnancy and litter size at birth on ewe lamb rearing performance (Least squares mean F standard error) Pasture allowance levela
Trait
Litter size at tagging
Litter size at weaning
Litter survival birth to tagging
Litter survival tagging to weaning
Litter survival birth to weaning
Litter weight at birth (kg)
Weight of lamb weaned (kg)
Twin Triplet Pb Twin Triplet Pb Twin Triplet Pb Twin Triplet Pb Twin Triplet Pb Twin Triplet Pb Twin Triplet Pb
2 cm
4 cm
6 cm
8 cm
1.8 F 0.13 2.4 F 0.14A ns 1.7 F 0.15 1.9 F 0.16 ns 0.92 F 0.052 0.79 F 0.056B * 0.88 F 0.056 0.78 F 0.061 ns 0.83 F 0.057 0.62 F 0.06 * 8.6 F 0.36A 9.7 F 0.38B * 44 F 3.7 43 F 3.9A ns
1.9 F 0.14 2.5 F 0.14A ns 1.8 F 0.15 2.2 F 0.16 ns 0.96 F 0.054 0.83 F 0.055B * 0.95 F 0.058 0.89 F 0.059 ns 0.91 F 0.06 0.73 F 0.06 * 9.2 F 0.37 11.5 F 0.37A ns 49 F 3.8 55 F 3.9B ns
2.0 F 0.13 2.7 F 0.15 ns 1.8 F 0.14 2.1 F 0.17 ns 0.98 F 0.052 0.89 F 0.058 ns 0.90 F 0.055 0.77 F 0.065 ns 0.88 F 0.056 0.69 F 0.07 * 9.4 F 0.35 10.5 F 0.39B ns 51 F 3.6 50 F 4.2 ns
1.8 F 0.13 2.9 F 0.14B ns 1.7 F 0.15 2.1 F 0.15 * 0.92 F 0.052 0.96 F 0.054A ns 0.85 F 0.056 0.74 F 0.058 ns 0.83 F 0.057 0.70 F 0.06 ns 9.7 F 0.36B 11.6 F 0.36A * 47 F 3.8 49 F 3.8 ns
a Pasture allowance from pregnancy scanning to lamb tagging. Means within the same row with different superscript differ significantly (no superscript = not significant, P N 0.05; superscripts different P b 0.05). b Mean comparison within each level of pasture allowance (ns = not significant, P N 0.05; * P b 0.05).
pregnancy (twins 9.2 F 0.22 kg; triplets 10.8 F 0.23 kg, P b 0.0001). Pasture allowance in late pregnancy and litter size at birth did not significantly affect weight of lamb weaned ( P N 0.05). Dam age and pasture allowance from tagging to weaning also did not affect weight of lamb weaned ( P N 0.05). Days from tagging to weaning significantly influenced weight of lamb weaned (b = 0.86 F 0.278, P b 0.01). Twin litter survival from birth to tagging was similar regardless of pasture allowance in late pregnancy (Table 2). Triplet litter survival was significantly greater for ewes grazing the 8-cm sward compared to ewes on the 2-cm and 4-cm swards (Table 2). Overall triplet litter survival (86 F 4%) was lower than twin litter survival (94 F 4%) to tagging ( P b 0.05). Pasture allowance in late pregnancy and from tagging to weaning did not have a significant effect on litter survival from tagging to weaning ( P N 0.05). Litter survival from tagging to weaning was signifi-
cantly higher for twin (90 F 3%) than triplet litters (79 F 3%) ( P b 0.05). Pasture allowance in late pregnancy did not significantly affect overall litter survival from birth to weaning ( P N 0.05) (Table 2). Twin litter survival to weaning was significantly higher than triplet litter survival (twins 86 F 3%; triplets 68 F 3%) ( P b 0.001). Ewes giving birth to triplet litters on the highest pasture allowance had significantly larger litters at tagging than ewes giving birth to twin litters ( P b 0.05). Litter size at weaning was similar despite litter size at birth and feeding level in late pregnancy (Table 2). 3.3. Lamb behaviour Behaviours of individual lambs were recorded. Sward allowance up to and at lambing had a significant effect on lamb behaviours observed (Table 3). The proportion of lambs bleating before ewe contact
J.M. Everett-Hincks et al. / Livestock Production Science 97 (2005) 253–266
259
Table 3 The effect of pasture allowance in late pregnancy and litter size at birth on lamb behaviour in the 5-min observation period immediately following tagging (least squares mean F standard error) Pasture allowance levela
Trait (proportion of lambs)
2 cm Lambs that bleated before ewe contact Lambs to make ewe contact Lambs that stood Lambs that located ewe udder Lambs that suck from ewe Lambs that followed moving ewe
Twin Triplet Twin Triplet Twin Triplet Twin Triplet Twin Triplet Twin Triplet
4 cm A
0.90 F 0.05 0.73 F 0.07A 0.91 F 0.04 0.63 F 0.07A 0.91 F 0.04 0.63 F 0.07 0.72 F 0.07 0.35 F 0.07 0.37 F 0.07 0.20 F 0.06 0.77 F 0.07 0.49 F 0.08
6 cm A
0.88 F 0.05 0.83 F 0.06A 0.93 F 0.04 0.69 F 0.06 0.91 F 0.04 0.67 F 0.06 0.66 F 0.07 0.35 F 0.06 0.27 F 0.07 0.18 F 0.05 0.88 F 0.07 0.43 F 0.08
8 cm B
0.65 F 0.07 0.80 F 0.06A 0.89 F 0.05 0.85 F 0.05 0.89 F 0.05 0.85 F 0.05 0.64 F 0.07 0.62 F 0.07 0.27 F 0.07 0.19 F 0.05 0.86 F 0.06 0.65 F 0.08
0.82 F 0.06 0.55 F 0.07B 0.93 F 0.03 0.87 F 0.05B 0.82 F 0.06 0.81 F 0.05 0.69 F 0.07 0.51 F 0.07 0.29 F 0.07 0.16 F 0.05 0.61 F 0.10 0.75 F 0.08
a
Pasture allowance from pregnancy scanning to lamb tagging. Means within the same row with different superscript differ significantly (no superscript = not significant, P N 0.05; superscripts different P b 0.05).
was affected by the interaction between pasture allowance and litter size ( P b 0.01). Triplet lambs born to ewes grazing the 8-cm sward were less likely to bleat before ewe contact than lambs born to ewes grazing lower swards (Table 3). Ewe–lamb contact was significantly affected by the interaction between pasture allowance and litter size ( P b 0.001) (Table 3). The proportion of ewes to make contact with twin lambs was similar regardless of pasture allowance. However, fewer ewes with triplet litters made lamb contact if grazing lower swards. The proportion of lambs that stood was not significantly affected the interaction between pasture allow-
ance and litter size ( P b 0.05). Fewer triplets stood at lower swards and this was significantly different to the proportion of twin lambs that stood ( P b 0.05) (Table 3). A lamb’s ability to locate its dam’s udder after tagging was significantly affected by the interaction between pasture allowance and litter size ( P b 0.05). As pasture allowance increased the number of triplet lambs that located their dam’s udder increased ( P b 0.10) (Table 3). Fewer triplet lambs sucked from their dam after tagging compared to twin lambs ( P b 0.01). Pasture allowance did not significantly effect the proportion of lambs that sucked from their dams ( P N 0.05) (Table 3).
Table 4 The effect of pasture allowance from pregnancy scanning to lamb tagging on lamb behaviour times after tagging for twins and triplets (Wilcoxon mean F standard error) Trait
Litter size at birth
2 cm
4 cm
6 cm
8 cm
P Wilcoxon
P log-rank
Ewe–lamb contact
Twin Triplet Twin Triplet Twin Triplet Twin Triplet Twin Triplet
0.7 F 0.16 0.6 F 0.12 0.5 F 0.09 0.6 F 0.11 1.9 F 0.22 2.3 F 0.29 2.0 F 0.20 2.2 F 0.22 2.5 F 0.24 2.2 F 0.42
0.6 F 0.07 1.2 F 0.19 0.5 F 0.09 1 F 0.16 1.9 F 0.24 2.4 F 0.30 1.8 F 0.20 1.8 F 0.24 2.8 F 0.33 2.8 F 0.34
0.6 F 0.12 1.0 F 0.18 0.7 F 0.13 0.8 F 0.14 2.4 F 0.27 2.3 F 0.25 1.9 F 0.21 2.6 F 0.29 3.0 F 0.39 2.8 F 0.45
0.4 F 0.04 0.9 F 0.16 0.4 F 0.11 0.9 F 0.18 1.7 F 0.23 2.3 F 0.28 2.4 F 0.33 2.8 F 0.27 2.5 F 0.39 3.7 F 0.37
ns * ns ns ns ** * ns ns ns
ns *** ns * ns ** * ns ns ns
Lamb to stand Lamb locate udder Follow dam Lamb to suckle
Times are presented in minutes. ns=not significant at P b 0.05, * P b 0.05, ** P b 0.01, *** P b 0.001.
260
J.M. Everett-Hincks et al. / Livestock Production Science 97 (2005) 253–266
a
b 90%
80% 70% 60% 50% 40% 30% 2cm
20%
4cm
10%
Proportion of triplet lambs
100%
90%
Proportion of twin lambs
100%
80% 70% 60% 50% 40% 30%
8cm
0% 0
1
2
3
4
2cm
20%
4cm
10%
6cm
6cm 8cm
0% 0
5
1
2
3
4
5
Time (minutes)
Time (minutes)
Fig. 1. (a and b) The effect of pasture allowance from pregnancy scanning to lamb tagging on the proportion of lambs to have ewe contact within 5 min from tagging. Graphs of estimated dbehaviourT functions for twin lambs (a) and for triplets (b).
Approximately 80% of all twin lambs made ewe contact in 1 min and this did not differ significantly across sward heights (Fig. 1a, Table 4). Approximately 50% of triplet lambs made ewe contact in 1 min and this did differ across sward heights ( P b 0.05) (Fig. 1b, Table 4). MBS, high ewe bleats and day of tagging were all related to ewe–lamb contact time for twin ( P b 0.001) and triplet lambs ( P b 0.001). Ewe–lamb contact time after tagging was quicker if the ewe had a higher MBS and high bleated fewer times. Fifty-five percent of the triplet lambs on the 2-cm sward stood in 1 min compared with 65% on the 8-cm sward (Fig. 2b), whereas 80% of twin lambs born on the 2-cm pasture sward height had stood by the same time (Fig. 2a). Factors significantly affecting the time for twin and triplet lambs to stand included MBS, ewe high bleat frequency, whether the lamb was vocal or
The proportion of lambs to follow their dam was significantly affected by the interaction between the main effects ( P b 0.001) (Table 3). The proportion of triplet lambs that followed their dam increased as pasture allowance increased ( P b 0.01). Fewer triplets followed their dam compared to twin born lambs at all sward heights except the highest. Times related to behaviour after tagging were analysed using the LIFETEST procedure and the comparison of cumulative distributions for 2-, 4-, 6and 8-cm swards are reported separately for twins and triplets (Table 4). Triplet lamb behaviour differed significantly across sward heights for the majority of time-related traits, whereas twin behaviour was similar (Table 4). Lamb birth weight was not significantly related to twin and triplet lamb post-tagging behaviour times ( P N 0.05).
a
b 90%
80% 70% 60% 50% 40% 30% 2cm
20%
4cm
10%
6cm 8cm
0% 0
1
2
3
Time (minutes)
4
5
Proportion of triplet lambs
100%
90%
Proportion of twin lambs
100%
80% 70% 60% 50% 40% 30% 2cm
20%
4cm
10%
6cm 8cm
0% 0
1
2
3
4
5
Time (minutes)
Fig. 2. (a and b) The effect of pasture allowance from pregnancy scanning to lamb tagging on the proportion of lambs to stand within 5 min from tagging. Graphs of estimated dbehaviourT functions for twins (a) and for triplets (b).
J.M. Everett-Hincks et al. / Livestock Production Science 97 (2005) 253–266
a
b
100%
100% 90%
80% 70% 60% 50% 40% 30% 2cm
20%
4cm
10%
Proportion of population
Proportion of twin lambs
90%
8cm
0
1
2
3
4
2cm 4cm 6cm 8cm
70% 60% 50% 40% 30% 20% 10%
6cm
0%
80%
261
0% 0
5
1
Time (minutes)
2
3
4
5
Time (minutes)
Fig. 3. (a and b) The effect of pasture allowance from pregnancy scanning to lamb tagging on the proportion of lambs to locate their dam’s udder within 5 min from tagging. Graphs of estimated dbehaviourT functions for twin lambs (a) and for triplets (b).
triplets (Fig. 4b, Table 4, P N 0.05). There was no significant correlation between ewe low bleat frequency and time for lamb to follow. Approximately 30% of all twin lambs sucked from their dam (Fig. 5a), whereas less than 20% of triplet lambs did (Fig. 5b). Lambs that bleated after tagging were closer to their dam (1.2 F 0.85 m) at the end of the 5-min observation period than lambs that did not bleat (6.3 F 1.44 m) (b = 5.11 F1.536, P b 0.01). Other factors significantly affecting lamb–dam distance 5 min after tagging included ewe high bleat frequency (b = 0.61 F 0.078, P b 0.0001), time of day lambs were ear-tagged (b = 1.11 F 0.44, P b 0.01) and lamb age at tagging (b = 0.24 F 0.097, P b 0.05). Pasture allowance did not significantly affect twin lamb distance from its dam 5 min after tagging (twins 4.7 F 1.12 m). However, triplet lambs born to ewes on the 2-cm
not and lamb age at tagging (all P b 0.01). Younger lambs (determined by the number of hours between birth and tagging) took longer to stand (r = 0.19; P b 0.001) and the longer the lamb took to stand the more its ewe high bleated (r = 0.48; P b 0.0001). Lambs that were quick to stand bleated more. Within 2 min, approximately 40% of all twin lambs across all sward heights had located their ewe’s udder (Fig. 3a), whereas only 20% of all triplet lambs had located the udder (Fig. 3b). Ewe high bleat frequency was related to the time taken for the udder to be located by the lamb after tagging ( P b 0.01). The longer the lamb took to locate the ewe’s udder the more the ewe high bleated (r = 0.31, P b 0.0001). The distribution curves for twin lambs following their dam were different over time ( P b 0.05) (Fig. 4a, Table 4). Ewe low bleating frequency was related to follow time for twin lambs ( P b 0.05) and not for
a
b 90%
80% 70% 60% 50% 40% 30% 2cm
20%
4cm
10%
6cm 8cm
0% 0
1
2
3
Time (minutes)
4
5
Proportion of triplet lambs
100%
90%
Proportion of twin lambs
100%
80%
2cm 4cm 6cm 8cm
70% 60% 50% 40% 30% 20% 10% 0% 0
1
2
3
4
5
Time (minutes)
Fig. 4. (a and b) The effect of pasture allowance from pregnancy scanning to lamb tagging on the proportion of lambs to follow their moving dam within 5 min from tagging. Graphs of estimated dbehaviourT functions for twin lambs (a) and for triplets (b).
262
J.M. Everett-Hincks et al. / Livestock Production Science 97 (2005) 253–266
a
b 100% 90%
4cm 6cm
80%
8cm
70% 60% 50% 40% 30% 20%
Proportion of triplet lambs
Proportion of twin lambs
100%
2cm
2cm
90%
4cm
80%
6cm 8cm
70% 60% 50% 40% 30% 20% 10%
10%
0%
0% 0
1
2
3
4
5
0
1
2
Time (minutes)
3
4
5
Time (minutes)
Fig. 5. (a and b) The effect of pasture allowance from pregnancy scanning to lamb tagging on the proportion of lambs to suckle within 5 min from tagging. Graphs of estimated behaviour functions for twin lambs (a) and for triplets (b).
sward were further from their dam (8 F 2.1 m) than triplets born to ewes allocated higher pasture allowances (triplets 2.9 F 1.08 m) ( P b 0.05). 3.4. Lamb performance The interaction between pasture allowance and litter size at birth significantly affected lamb birth weight ( P b 0.05), where twin and triplet lambs were significantly different weights from the 6-cm sward height treatment compared to the lower sward heights (Table 5). Twin lambs (4.5 F 0.06 kg) were heavier than triplet lambs (3.7 F 0.06 kg) ( P b 0.0001). Pasture allowance in late pregnancy did not affect lamb birth weight ( P N 0.05).
Pasture allowance from scanning to tagging did not affect overall lamb survival over the same period ( P N 0.05). Twin lambs had higher survival than triplets ( P b 0.01) and the interaction between pasture allowance and litter size at birth was significant at P b 0.10. Twin and triplet lamb survival were not significantly different at the 8-cm sward and triplet lamb survival was lowest for lambs born to ewes allocated the 2-cm sward and increased as pasture allowance increased (Fig. 6). Ewe pasture allowance from scanning to tagging (2-, 4-, 6- and 8-cm sward height) and from tagging to weaning (4- and 8-cm sward height) did not significantly affect lamb weaning weight and growth rate to weaning ( P N 0.05) (Table 5). Twin lambs were heavi-
Table 5 The effect of pasture allowance fed to ewes in late pregnancy and litter size at birth on lamb performance traits (least squares mean F standard error) Pasture allowance levela
Trait
2 cm Birth weight (kg)
Weaning weight (kg)
Growth rate to weaning (grams/day)
a
Twin Triplet Pb Twin Triplet Pb Twin Triplet Pb
4.5 F 0.13 3.7 F 0.13A * 26 F 0.83 25.5 F 0.88 ns 251 F 9.6 246 F 10.1 ns
4 cm
6 cm A
4.4 F 0.12 3.8 F 0.11A * 26.2 F 0.61 26 F 0.68 ns 251 F 7.2 251 F 8.1 ns
8 cm B
4.7 F 0.12 3.5 F 0.12BA * 27.9 F 0.68 25.2 F 0.69 ns 270 F 7.8 240 F 8.2 *
4.6 F 0.12 3.9 F 0.11A * 26.6 F 0.68 25 F 0.68 ns 257 F 7.94 240 F 8.1 ns
Pasture allowance from pregnancy scanning to lamb tagging. Means within the same row with different superscript differ significantly (no superscript = not significant, P N 0.05; superscript different P b 0.05). b Mean comparison within each level of pasture allowance (ns = not significant, P N 0.05; * P b 0.05).
J.M. Everett-Hincks et al. / Livestock Production Science 97 (2005) 253–266
263
100%
Lamb survival
95%
90%
85%
80%
75%
70% Twins 2cm
Triplets 2cm
Twins 4cm
Triplets 4cm
Twins 6cm
Triplets 6cm
Twins 8cm
Triplets 8cm
Pasture Allowance (Sward Height)
Fig. 6. The effect of pasture allowance from pregnancy scanning to lamb tagging and litter size at birth on lamb survival at tagging (standard errors included).
er at weaning and grew faster (26.7 F 0.36 kg and 257 F 4.2 g/day) than triplet lambs (25.4 F 0.37 kg and 244 F 4.36 g/day) ( P b 0.05). Factors which significantly influenced lamb weaning weight and growth rate to weaning included ewe weight change over the same period, where ewes that gained more weight produced lambs with lower weaning weights (b = 0.01 F 0.004, P b 0.01; r = 0.27, P b 0.001) and growth rates (b = 0.15 F 0.044, P b 0.001; r = 0.21; P b 0.001). Lamb birth weight significantly affected weaning weight (b = 2.39 F 0.295, P b 0.0001) and growth rate (b = 15.73 F 3!.322, P b 0.0001). Days from tagging to weaning significantly affected lamb weaning weight (b = 0.29 F 0.049, P b 0.0001). Time to ewe–lamb contact after tagging significantly affected lamb growth rate to weaning (b = 10.1 F 4.474, P b 0.05). Lambs grew at a slower
rate to weaning if their dam took longer to make contact after tagging. Time for lamb to stand after tagging significantly affected lamb growth rate (b = 9.34 F 4.559, P b 0.05). Lambs that were slower to stand after tagging grew faster to weaning. Correlations were performed to determine if there were any relationships between ewe and lamb behaviours and individual lamb performance (Table 6). Lambs born to ewes with higher MBS at tagging had quicker ewe–lamb contact times, stood and located the udder sooner and were significantly closer to their dams 5 min after tagging than lambs born to ewes with low MBS (Table 6). The longer it took for ewe–lamb contact, the longer it took for the lamb to stand, locate the udder and suck. The ewe also high bleated a greater number of times. As ewe high bleat frequency increased, the greater the distance from their lamb 5 min after tagging (Table 6).
Table 6 Correlations between lamb behaviour and performance traits across all litter sizes at birth
Birth weight Body length Growth rate Ewe high bleats MBS
Time for ewe contact
Time to stand
Time to locate udder
Time to suckle
Time to follow
Distance from ewe
0.13* 0.11* 0.14* 0.74**** 0.50****
ns ns ns 0.48**** 0.23****
0.15* 0.15* ns 0.31**** 0.18**
ns 0.23* ns 0.26* ns
0.22** 0.20** ns ns ns
ns ns ns 0.38**** 0.32****
ns = not significant ( P N 0.05). * P b 0.05. ** P b 0.01. **** P b 0.0001.
264
J.M. Everett-Hincks et al. / Livestock Production Science 97 (2005) 253–266
4. Discussion Pasture allowance during late pregnancy and lambing affected ewe and lamb performance and behaviour. Ewes with triplets were most affected. Twin litter weight at birth was similar for ewes grazing 2-, 4- and 6-cm swards. However, improvements in triplet litter weight at birth, weight of lamb weaned, litter survival to tagging and weaning were observed with increasing pasture allowance. Triplet lamb and litter survival to tagging increased by 15% from the lowest pasture allowance to the highest. Similar benefits have been reported elsewhere. When multiple bearing ewes were offered additional feed in late pregnancy, lamb mortality was reduced for single and twin lambs (Scales et al., 1986). A 10-kg increase in ewe live weight during the last 6 weeks of pregnancy resulted in a birth weight increase of 0.46 kg in single and 0.52 kg in twin lambs (Scales et al., 1986). Moore et al. (1986) reported that lambs born from high feed allowance ewes had higher rectal temperatures at birth than those born from low allowance ewes indicating the lambs poorly fed in utero tend to achieve high metabolic rates more slowly than those which receive a good nutrient supply (McCutcheon et al., 1981). An increase in lamb birth weight was associated with increased lamb vigour, with increased rectal temperature only for male lambs (Moore et al., 1986). This study shows that lamb birth weight was affected by ewe pasture allowance in late pregnancy; however, lamb birth weight was not significantly related to lamb behaviour after tagging. This contradicts the finding of Astroshi and Osterberg (1979) who studied prolific sheep and concluded that lamb birth weight influences behaviour but litter size does not, except indirectly through its affect on birth weight and the smaller the lamb, the less active it is. However, Fahmy et al. (1996) found that Finn lambs, weighing half that of Suffolk lambs, were quicker to reach the udder after birth. Hight and Jury (1970) reported that lambs born to ewes fed well in late pregnancy are better equipped to survive conditions predisposing to exposure and starvation as they have more energy stored as brown fat reserves and maintain their suckling drive for longer than lambs born to poorly fed ewes. The effect of lamb birth weight on lamb behaviour is inconclusive and it is highly likely
that a complex of physiological and genetic factors underpin lamb behaviour, not excluding the effect of maternal behaviour. Maternal behaviour influences lamb behaviour and is critical to lamb survival (O’Connor et al., 1985). However, according to Smith et al. (1966) bineptitude of the lambs retards acceptance by the motherQ and mother–young acceptance is more likely if the newborn lamb does the following; stands soon after birth, sucks soon after standing, follows the mother closely and moves to the mother if separated (Alexander, 1988). This study showed that triplet lambs from ewes on the lower pasture allowance were less likely to stand, locate their dam’s udder and follow their dam. Fewer triplet lambs bleated following tagging and before ewe contact and this may be why fewer ewes made contact. Lindsay et al. (1990) observed that increased vocal behaviour of lambs was strongly associated with earlier recognition of their mother and better bonding. In this study, lambs were closer to their dam 5 min after tagging if they bleated when separated. Tagging temporarily separates a ewe and her litter. Ewes react differently to the separation and depending on the strength of the bond formed with her litter, the fear for the shepherd, the visibility of lambs versus flock lambs, the proximity of the flock and the ewe’s ddriveT to be with the flock, the ewe will stay close to her litter or desert them in favour for the flock (Everett-Hincks et al., 2005). Ewe feeding level did not affect MBS but ewes on higher pasture allowances were more likely to stay with their litter at the tagging site during the 5-min observation period after tagging. Ewes with triplet litters on the 2-cm sward were a greater distance from their litter 5 min after tagging, high-pitch bleated significantly more and were less likely to make contact with their lamb 5 min after tagging. Putu et al. (1988) observed that nutrition during late pregnancy did not affect lamb desertion in single bearing ewes, but in the twin bearing ewes the low level of nutrition resulted in a higher proportion of permanent desertions of at least one of their twin lambs (19.2%) compared with ewes on a high level of nutrition (4.3%). The maternal environment in utero and at birth affected birth weights and lamb survival to tagging but did not continue to impair lamb performance to
J.M. Everett-Hincks et al. / Livestock Production Science 97 (2005) 253–266
weaning as lamb growth and weaning weights were similar between twins and triplets and between feeding levels. Napolitano et al. (1995) observed that average daily weight gain was lower in lambs that had been separated earlier from the ewe. This study showed that lambs that were slower to stand after tagging grew faster. Lambs grew an additional 10 g/ day for every minute they stayed lying down after tagging. Lambs that had quick ewe contact after tagging grew faster to weaning. Ewes that stayed close to their litter during tagging and did not return to the flock were quick to make contact with their lambs. These ewes are more attentive to their lambs and this may explain why their lambs grow faster. A lamb that takes longer to stand after tagging is not as easily stressed as a lamb that stands quickly and flees. These dcontentT lambs have faster growth rates to weaning. In this study, some lambs had sucked prior to tagging as their stomachs appeared full and this probably reduced their drive to suck following tagging and they appeared more dcontent.T Murphy et al. (1994b) reported that animals of quiet temperament grow faster and are better producers than animals that are restless, nervous or aggressive. O’Connor (1996) suggested that the ewe–lamb relationship may be weaker for ewes and lambs that maintain close contact as the lamb ages; alternatively, lambs may stay closer to ewes with lower levels of milk production especially where there is competition for milk from littermates. Lamb behaviour is more adversely affected than ewe behaviour by pasture allowance from scanning to tagging and triplets are more affected than twins. It is not dpoorT maternal behaviour but inadequate lamb behaviour that is a limiting factor in triplet lamb survival on the lower pasture allowances. This is supported by O’Connor and Lawrence (1992) who also found that it is lamb behaviour, not maternal behaviour that is the major factor limiting lamb survival. Shillito-Walser (1978) also appreciated the importance of effective lamb behaviour, concluding that the formation of the maternal–offspring bond in sheep is the result of a rapid learning process influenced by the behaviour of both the ewe and the lamb. In conclusion, this study showed that triplet lamb survival was similar to twin lamb survival when pasture allowance was not restricted in late pregnancy and ewes with triplets require higher pasture allowances to achieve similar production as ewes with twin
265
litters. Therefore, it is recommended that ewes carrying triplets are identified, classified as priority stock and preferentially fed throughout pregnancy. Further investigation is required of the physiological mechanisms that determine lamb behaviour in larger litters resulting from changes in the maternal environment.
Acknowledgements Financial assistance was provided by Meat and Wool Innovation Limited and an AGMARDT Scholarship. The authors also thank Dean Burnham and Graham Poole from Massey University for their involvement in the project.
References Alexander, G., 1988. What makes a good mother? Components and comparative aspects of maternal behaviour in ungulates. Proceedings of the Australian Society of Animal Production 17, 25 – 41. Astroshi, F., Osterberg, S., 1979. The behaviour of Finnsheep during and shortly after lambing. Acta Agriculturae Scandinavica 29, 258 – 262. Dwyer, C.M., Lawrence, A.B., Bishop, S.C., Lewis, M., 2003. Ewe–lamb bonding behaviours at birth are affected by maternal undernutrition in pregnancy. British Journal of Nutrition 89, 123 – 136. Everett-Hincks, J.M., Lopez-Villalobos, N., Blair, H.T., Stafford, K.J., 2002. Genetic variation in maternal behaviour score and lamb survival. 7th World Congress on Genetics Applied to Livestock Production, August 19–23, Montpellier, France. Session 14. Behaviour and Welfare. Communication, p. 14. Everett-Hincks, J.M., Lopez-Villalobos, N., Blair, H.T., Stafford, K.J., 2005. The effect of maternal behaviour score on lamb and litter survival. Livestock Production Science 93, 51–61. Fahmy, M.H., Robert, S., Castonguay, F., 1996. Ewe and lamb behaviour at parturition in prolific and non-prolific sheep. Canadian Journal of Animal Science, 9 – 15. Geenty, K.G. 1997 b200 by 2000. A Guide to Improved Lambing PercentageQ. Wools of New Zealand and the New Zealand Meat Producers Board. Geenty, K.G., Rattray, P.V., 1987. The energy requirements of grazing sheep and cattle. Feeding Livestock on Pasture, vol. 3, pp. 39 – 53. Hight, G.K., Jury, K.E., 1970. Hill country sheep production. II Lamb mortality and birthweights in Romney and Border Leicester Romney flocks. New Zealand Journal of Agricultural Research 13, 735 – 752. Holmes, R.J., 1975. Relationship of maternal behaviour to neonatal loss in prolific sheep. Proceedings of the New Zealand Veterinary Association Sheep Society 5, 67 – 75.
266
J.M. Everett-Hincks et al. / Livestock Production Science 97 (2005) 253–266
Lindsay, D.R., Nowak, R., Gede Putu, I., McNeill, D., 1990. Behavioural interactions between the ewe and her young at parturition: a vital step for the lamb. School of Agriculture (Animal Science), Reproductive Physiology of Merino Sheep—Concepts and Consequences (The University of Western Australia, Perth) 15, 191 – 204. Lopez-Villalobos, N., Garrick, D.J., 1999. Genetic parameter estimates for lamb survival in Romney sheep. Proceedings of the New Zealand Society of Animal Production 59, 121 – 124. McCutcheon, S.N., Holmes, C.W., McDonald, M.F., 1981. The starvation-exposure syndrome and neonatal lamb mortality. A review. Proceedings of the New Zealand Society of Animal Production 41, 209 – 217. Moore, R.W., Millar, C.M., Lynch, P.R., Dow, B.W., 1986. The effect of prenatal nutrition and type of birth and rearing of lambs on vigour, temperature and weight at birth and weight and survival at weaning. Proceedings of the New Zealand Society of Animal Production 46, 259 – 262. Morris, C.A., Hickey, S.M., Clarke, J.N., 2000. Genetic and environmental factors affecting lamb survival at birth and through to weaning. New Zealand Journal of Agricultural Research 43, 515 – 524. Murphy, P.M., Lindsay, D.R., Purvis, I.W., 1994a. The importance of the birth site on the survival of merino lambs. Proceedings of the Australian Society of Animal Production 20, 251 – 254. Murphy, P.M., Purvis, I.W., Lindsay, D.R., Le Neindre, P., Orgeur, P., Poindron, P., 1994b. Measures of temperament are highly repeatable in Merino sheep and some are related to maternal behaviour. Proceedings of the Australian Society of Animal Production 20, 247 – 250. Napolitano, F., Marino, V., De Rosa, G., Capparelli, R., Bordi, A., 1995. Influence of artificial rearing on behavioural and immune response of lambs. Applied Animal Behaviour Science 45, 245 – 253. Nowak, R., 1996. Neonatal survival: contributions from behavioural studies in sheep. Applied Animal Behaviour Science 49, 61 – 72.
Nowak, R.F., Lindsay, D.R., 1992. Discrimination of Merino ewes by their newborn lambs: important for survival? Applied Animal Behaviour Science 34, 61 – 74. O’Connor, C.E., 1996. Ewe maternal behaviour score and lamb growth: ten years on. Proceedings of the New Zealand Society of Animal Production 56, 107 – 109. O’Connor, C.E., Lawrence, A.B., 1992. Relationship between lamb vigour and ewe behaviour at parturition. British Society of Animal Production 54, 361 – 366. O’Connor, C.E., Jay, N.P., Nicol, A.M., Beatson, P.R., 1985. Ewe maternal behaviour score and lamb survival. Proceedings of the New Zealand Society of Animal Production 45, 159 – 162. Pollard, J.C., Littlejohn, R.P., 1999. Shelter for lambing in southern New Zealand: II. Sheltering behaviour and effects of productivity. New Zealand Journal of Agricultural Research 42, 171 – 177. Putu, I.G., Poindron, P., Lindsay, D.R., 1988. Early disturbance of Merino ewes from the birth site increases lamb separations and mortality. Proceedings from the Australian Society of Animal Production 17, 298 – 301. Rattray, P.V., Thompson, K.F., Hawker, H., Sumner, R.M.W., 1987. Pastures for sheep production. Livestock Feeding on Pasture, vol. 7, pp. 89 – 103. SAS, 2002. SAS/STAT Software. Release 8.1. SAS Institute, Incorporated, Cary, NC. Scales, G.H., Burton, R.N., Moss, R.A., 1986. Lamb mortality, birth weight and nutrition in late pregnancy. New Zealand Journal of Agricultural Research 29, 75 – 82. Shillito, E., 1972. Vocalisation in sheep. Journal of Physiology 226, 45 – 46. Shillito-Walser, E., 1978. The development of the maternaloffspring bond in sheep. Applied Animal Ethology 4, 293 (Abstract). Smith, F.V., Van-Toller, C., Boyes, T., 1966. The dcritical periodT in the attachment of ewes and lambs and ewes. Animal Behaviour 14, 120 – 125.
The effect of pasture allowance fed to twin- and triplet-bearing ewes in late pregnancy on ewe and lamb behaviour and performance to weaning J.M. Everett-Hincks*, H.T. Blair, K.J. Stafford, N. Lopez-Villalobos, P.R. Kenyon, S.T. Morris Institute of Veterinary Animal and Biomedical Sciences, Massey University, Private Bag 11 222, Palmerston North, New Zealand Received 29 March 2004; received in revised form 21 April 2005; accepted 13 May 2005
Abstract This paper investigates whether pasture allowance before and at lambing affects ewe behaviour, lamb behaviour and subsequent twin and triplet lamb performance to weaning. The trial was carried out at Massey University Keeble Farm in New Zealand. At pregnancy scanning, 186 first cycle-mated ewes were identified as carrying twins (n = 96) or triplets (n = 90). The ewes were randomly divided into four nutritional treatments (2, 4, 6 and 8 cm post-grazing sward heights). Each treatment was replicated (n = 12 ewes in each replicate during pregnancy) on two separate blocks. The trial ewes remained on the four nutritional treatments during pregnancy and parturition. The mean lambing date was the 9 September 2002. Within 12 h of birth lambs were weighed, their sex recorded and identified to the ewe. Maternal and lamb behaviours were then recorded for a period of 5 min. Twin litter weight at birth was similar for ewes grazing 2-, 4- and 6-cm sward heights. Increasing pasture allowance from just 2 cm to 4 cm to ewes with triplet litters increased litter weight at birth by 2 kg, increased weight of lamb weaned by 8 kg and improved litter survival to tagging by 4%. Triplet lamb survival to tagging increased by 15% from the lowest pasture allowance to the highest (from 2 cm to 8 cm). Triplet litter survival was similar to twin litter survival at the highest pasture allowance (8 cm; 2000 kg DM/ha). Pasture allowance had a greater effect on lamb behaviour than ewe behaviour and triplets were more affected than twins. Feeding level did not affect Maternal Behaviour Score (MBS); however, MBS was related to a number of lamb behaviours. Ewes on higher pasture allowances were more likely to stay with their litter at the tagging site. Ewes with triplet litters on the 2cm sward height were further from their litter 5 min after tagging and high-pitch bleated significantly more than ewes on higher swards. These ewes were less likely to make contact with their lamb after tagging. Triplets born on lower pasture allowances were less likely to stand, locate their dam’s udder and follow their moving dam. Fewer triplet lambs bleated following tagging and this might be why fewer ewes made contact.
* Corresponding author. Invermay Agricultural Centre, AgResearch, Puddle Alley, Private Bag 50034, Mosgiel, New Zealand. Tel.: +64 3 489 9164; fax: +64 489 9024. E-mail address: [email protected] (J.M. Everett-Hincks). 0301-6226/$ - see front matter D 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.livprodsci.2005.05.006
254
J.M. Everett-Hincks et al. / Livestock Production Science 97 (2005) 253–266
Ewes with triplets grazing high pasture allowances in late pregnancy can achieve litter survival rates similar to those with twins. Ewes with triplets require higher pasture allowances to achieve similar production as ewes with twin litters. D 2005 Elsevier B.V. All rights reserved. Keywords: Maternal behaviour; Lamb behaviour; Lamb survival; Pasture allowance
1. Introduction Improved lambing percentage makes the biggest contribution to higher profits on New Zealand sheep farms (Geenty, 1997). Pregnancy scanning enables farmers to identify and preferentially feed ewes with multiple foetuses. However, triplet lamb survival is a problem. A ewe’s maternal behaviour and litter survival are under minimal genetic control and the major source of variation in these traits is due to temporary environmental effects (Everett-Hincks et al., 2005). Lamb survival is predominantly controlled by the environment (Lopez-Villalobos and Garrick, 1999; Morris et al., 2000; Everett-Hincks et al., 2002; Everett-Hincks et al., 2005). Therefore, a better understanding of environmental conditions, particularly feeding levels throughout pregnancy and at lambing, may allow a reduction in lamb losses seen in highly fecund ewes. In late pregnancy and early lactation, the nutritional requirements of ewes with twins are very high (Holmes, 1975). The last 4–6 weeks before lambing are critical (Scales et al., 1986). When multiple bearing ewes were offered additional feed in late pregnancy to increase live weight, lamb mortality was reduced, suggesting merit in improved pre-lamb feeding for ewes carrying more than one lamb (Scales et al., 1986). To date, doptimalT pasture feeding levels for ewes with triplets have not been adequately defined as research into ewe energy requirements in late pregnancy has focused on ewes with singles and twins (Geenty and Rattray, 1987). Inadequate feeding can result in poor maternal behaviour, an important cause of lamb death (Nowak, 1996). Dwyer et al. (2003) demonstrated that a moderate reduction in maternal nutrition in late pregnancy caused a reduction in the expression of maternal behaviours, in particular the Maternal Behaviour Score (MBS), at parturition under intensive indoor conditions. The New Zealand sheep industry is pasture based and the strength of ewe– lamb attachment for ewes with twin and triplet litters
grazing differing pasture allowances in late pregnancy is unknown. Ewe–lamb bonding and lamb survival are maximised by management practices that increase the time spent on the birth site by the ewe after parturition (Nowak, 1996). If pasture quality and quantity are good and the paddock has adequate shelter, then the ewe will stay at the birth site with her lambs for a longer period (Pollard and Littlejohn, 1999). In addition, Lindsay et al. (1990) observed that the time spent on the birth site increased when ewes with twins were on a higher plane of nutrition 6 weeks before lambing. Nowak (1996) observed a considerable improvement in twin lamb bonding with their dam and twin lamb survival when the mother remained on the birth site for a minimum of 6 h. However, Murphy et al. (1994a) suggested that it is not that the dam is on the birth site for 6 h after parturition that improves lamb survival but that the dam is together with all of her lambs. Nowak and Lindsay (1992) concluded that the ability of twin lambs to discriminate their dams from alien ewes 12 h after birth is related to survival during the neonatal period and Shillito-Walser (1978) stated that the formation of the dam–lamb bond is the result of a rapid learning process influenced by both the behaviour of the dam and the lamb. Therefore, lamb and maternal behaviour must be investigated conjointly to identify the environmental conditions that impact on the formation of a successful ewe–lamb bond. This study investigates whether pasture allowance before and at lambing affects ewe behaviour, lamb behaviour and subsequent twin and triplet lamb performance to weaning.
2. Materials and methods 2.1. Animals and measurements The animals included in the trial were monitored from pregnancy scanning to lamb weaning. At preg-
J.M. Everett-Hincks et al. / Livestock Production Science 97 (2005) 253–266
nancy scanning (17 June; about day 64 of pregnancy) 186 first cycle-mated ewes were identified as carrying twins (n = 96) or triplets (n = 90). The trial was carried out at Massey University Keeble Farm. The trial ewes were made up evenly of mixed age and two-tooth ewes with predominantly Romneys. The ewes were randomly divided into four nutritional treatments (2, 4, 6 and 8 cm post-grazing sward heights) representing a dpoorT (800 kg DM/ ha), dnormalT (1200 kg DM/ha), dgenerousT (1600 kg DM/ha) and dovergenerousT (2000 kg DM/ha) feed supply, respectively (Rattray et al., 1987). Each treatment was replicated (n = 12 ewes in each replicate during pregnancy) on two separate blocks (dPinesT and dCaravanT blocks) to ensure statistical integrity. The pasture height was measured weekly by rising plate meter method and maintained by modifying the stocking rate. The trial ewes remained on the four nutritional treatments during pregnancy and parturition and were removed after the lambs were ear-tagged. Approximately 24 h after the lamb was ear-tagged, the ewe and her lambs were transferred to one of two pastures with post-grazing sward heights of 4 and 8 cm and remained on this feed level until weaning (6 December 2002, mean lamb age 88 days). The mean lambing date was the 9 September 2002. Weather conditions were mild and consistent throughout lambing. Date and hour of birth was recorded for each ewe and her litter. Within 12 h of birth, lambs (alive and dead) were weighed, ear-tagged, their sex recorded and identified to the ewe. This procedure is referred to as dtaggingT and was performed by two people, one person measuring and the second person recording. Immediately following tagging, the lambs were placed together on the ground, the two recorders moved 20–30 m away where they recorded maternal and lamb behaviours for a period of 5 min. Litter size was recorded at pregnancy scanning, birth, tagging and weaning. Litter survival was measured from birth to tagging and tagging to weaning and was calculated by dividing the latter litter size by the former litter size and recorded as a percentage. Litter survival from birth to weaning was also calculated. Litter weight at birth and weight of lamb weaned were recorded for each ewe. Ewes that lost an entire litter were included in the analysis as rearing 0 kg of lamb to weaning.
255
Lamb survival was recorded at tagging and weaning. Lamb survival was recorded as a binary trait, lambs alive at each of these stages was given a score of d1,T whereas lambs not present at tagging or weaning were given a score of d0T. Litter survival and lamb survival are related but were analysed separately in this study. Litter survival measures the dam’s ability to rear all of her lambs born in a particular parity, while lamb survival is measuring an individual lamb’s ability to survive and is influenced by the dam. Lamb growth rate (grams/day) was calculated by subtracting birth weight from weight at weaning and dividing it by the number of days between the two measurements. 2.2. Maternal behaviour A Maternal Behaviour Score (MBS) described by O’Connor et al. (1985) was scored on a 5-point scale (low, 1 to high, 5) based on the distance a ewe retreated from her lambs when the shepherd tagged them. The following ewe behaviours were recorded immediately after tagging. The time to first contact after tagging between the ewe and the lamb was recorded (ewe–lamb contact). Lambs that had no contact with their ewe 5 min after tagging were recorded as d0T and those that did have ewe contact were recorded as d1.T Ewe bleats were counted and categorised as low-pitch bleats (rumble calls) and high-pitch bleats. Low bleats made with the mouth closed generally occur when the ewe is in close proximity to the lamb. High ewe bleats made with the mouth open and are generally made by ewes when their lambs are separated from them (Shillito, 1972). Ewes that moved further than 5 m from the tagging site were recorded as a binary trait (d0T ewe moved away and d1T ewe stayed). The maximum distance of the ewe from any one of her lambs was estimated in meters at the end of the 5-min recording period. For each ewe and her litter, the date and time of tagging was recorded as well as the approximate age of the lambs in hours at tagging. 2.3. Lamb behaviour At tagging, each lamb was identified by a colour spray mark and its behaviour recorded for 5 min. Each
256
J.M. Everett-Hincks et al. / Livestock Production Science 97 (2005) 253–266
lamb was placed on its side on the ground immediately after tagging. The time from the end of tagging until the lamb stood was recorded in minutes. In some instances, the lamb did not stand during the 5 min of observation and was recorded as d0,T whereas a lamb that stood during the 5-min post-tagging period was given a score of d1T. Individual ewe–lamb contact time was recorded (as described in Section 2.2). Lamb bleating while separated from its dam was recorded as a binary trait, d0T for lambs that did not bleat and d1T for those that did. If the ewe moved more than 5 m away from the tagging site, the time (minutes) taken for the lamb to follow was recorded. A lamb that failed to follow its moving dam was recorded as d0,T whereas a lamb that followed was given a d1T. The time taken for each lamb to locate the ewe’s udder and to suck were also recorded. Lambs that did not perform these dbehavioursT during the 5 min following tagging were given a score of d0,T whereas those that did were given a score of d1T. 2.4. Statistical analyses The trial was set up as a split plot design and tested for the fixed effect of pasture allowance in late pregnancy as the big plot and litter size at birth as the small plot. Continuously dependent maternal behaviours, lamb behaviours at tagging and lamb measurements and weights were analysed using the MIXED procedure (SAS, 2002) and lamb traits were adjusted for the random maternal effect as shown by fitting the following linear model: Yijklm ¼ l þ Bi þ Pj þ Lk þ PLjk þ BPij þ bXijklm þ dm þ eijklm where Yijklm is the response of the lth animal of the kth litter size from the mth dam of the jth pasture allowance in the ith block, l is the general mean, B i is the random effect of the ith block (i = Pines block or Caravan block), P j is the fixed effect of the jth pasture allowance ( j = 2, 4, 6 or 8 cm sward height) in late pregnancy, L k is the fixed effect of the kth litter size at birth (k = twin or triplet), PL jk is the fixed interaction effect of pasture allowance j and litter size at birth k, BP ij is the random interaction effect of pasture allowance j and block i (main plot error), X ijklm is the
covariate tested (e.g., day of birth, ewe age, live weight, condition score), b is the regression of Y on the covariate tested (e.g., day of birth), d is the random effect of dam m where lamb l was born to dam m and provides the maternal effect and litter effects common for the lambs born in the same litter. e ijklm is the random residual effect unique to Yijklm . The fixed effect of pasture allowance from tagging to weaning (4 cm or 8 cm) was added to the model to test differences in performance from tagging to weaning, such as lamb survival and weight at weaning and only reported if statistically significant at P b 0.05. To help decide the pathways involved in the differences in ewe and lamb performance and ewe and lamb behaviour, different covariates were tested and retained in the model if statistically significant at P b 0.05. Results are given as least square means and their standard errors. Multiple comparisons between least squares means for twin and triplet litters, sward height pasture allowance and combinations between these fixed effects were performed and only reported if statistically significant at P b 0.05. PROC CATMOD (SAS, 2002) was used to analyse categorical dependent behaviour traits (i.e., 0 = did not stand, 1 = did stand) and lamb survival at tagging and weaning (0 = dead and 1 = alive). Least squares means for the fixed effects of litter size at birth, pasture allowance and the interaction between pasture allowance and litter size were derived. The CONTRAST statement was used to determine if differences in fixed effects were significant at P b 0.05. The LIFETEST procedure (SAS, 2002) was used to compute and plot a survival distribution function (SDF) of post-tagging time behaviours for twin and triplet lambs on 2, 4, 6 and 8 cm sward heights. These behaviour observations are right-censored due to the termination of the experiment (i.e., 5 min after tagging). But LIFETEST correctly uses the censored observations as well as the non-censored. The SDF evaluated at t (i.e., 5 min) is the probability that a lamb from a specified sward height and litter size will have a dbehaviourT time exceeding t; that is, S(t) = Pr(T N t), where S(t) denotes the survival distribution function and T is the behaviour time of a randomly selected lamb. The cumulative distributive functions for each of the post-tagging behaviour traits are graphically plotted and are defined as 1 S(t), that is, the probability that a behaviour time does not exceed time
J.M. Everett-Hincks et al. / Livestock Production Science 97 (2005) 253–266
257
(t) (i.e., 5 min). To determine whether the curves plotted are homogenous across sward heights ( P N 0.05), PROC LIFETEST provides two rank tests (log-rank and Wilcoxon). The Wilcoxon test places more weight on early times and the log-rank test places more weight on larger times. The association between covariates and the behaviour time variable were investigated using the Wilcoxon test and reported if significant at P b 0.05. The covariate tests are pooled across sward height treatments and it is not possible to calculate the directional effect the covariate is having on the dependent variable. Pearson correlation coefficients (r) were computed using SAS (PROC CORR) for continuously independent ewe and lamb behaviour and production traits and were reported if significant at P b 0.05.
MBS high bleated more (r = 0.54, P b 0.0001). A greater proportion of ewes grazing the 8-cm sward stayed with their litter after tagging compared with ewes on the 2-cm sward ( P = 0.07) (Table 1). Ewes with triplet litters on the 2-cm sward were farthest from their litter at the end of the 5-min observation period compared to ewes on higher allowances and ewes with twin litters on the 2-cm sward (Table 1). Ewes with triplets high bleated significantly more (8 F 1.15 bleats) than ewes with twins (5 F 1.06 bleats) ( P b 0.05) (Table 1). High-pitch bleating frequency was similar for ewes grazing all swards ( P N 0.05) (Table 1) and there was no significant interaction between feeding level up to tagging and litter size at birth ( P N 0.05). Litter size at birth tended to influence ewe low bleat frequency; however, this was not significant ( P N 0.05).
3. Results
3.2. Dam performance
3.1. Maternal behaviour
Litter weight was significantly lighter for ewes grazing the 2-cm sward height (9.2 F 0.30 kg) in late pregnancy compared to ewes grazing the 8 cm (10.6 F 0.30 kg) sward height ( P b 0.05) (Table 2). Triplet litters weighed significantly more at tagging than twin litters regardless of pasture allowance in late
MBS was similar between ewes on differing pasture allowances (Table 1). High-pitch bleating frequency had a significant effect on MBS (b = 0.08 F 0.009, P b 0.0001), the ewes with lower
Table 1 The effect of pasture allowance fed to ewes in late pregnancy and litter size at birth on Maternal Behaviour Score (MBS) and ewe tagging behaviour traits (least squares mean F standard error) Pasture allowance levela
Trait
MBS
Ewes that stayed (proportion)
Ewe high bleats
Ewe low bleats
Maximum distance after tagging (m)
a
Twin Triplet Pb Twin Triplet Pb Twin Triplet Pb Twin Triplet Pb Twin Triplet Pb
2 cm
4 cm
6 cm
8 cm
3.4 F 0.25 2.9 F 0.29 ns 0.26 F 0.092 0.26 F 0.101 ns 4 F 2.1 11 F 2.4 * 2.6 F 0.50A 0.7 F 0.61 * 5 F 3.6 16 F 4.5B *
3.0 F 0.26 2.7 F 0.27 ns 0.45 F 0.106 0.32 F 0.107 ns 6 F 2.1 6 F 2.3 ns 1.4 F 0.51 1.7 F 0.59 ns 3 F 3.6 0 F 4.6A ns
3.1 F 0.25 3.3 F 0.29 ns 0.23 F 0.089 0.35 F 0.116 ns 7 F 2.1 7 F 2.3 ns 1.2 F 0.49B 0.8 F 0.59 ns 6 F 3.6 1 F 4.2A ns
3.4 F 0.26 3.1 F 0.26 ns 0.48 F 0.104 0.42 F 0.113 ns 2 F 2.2 7 F 2.2 ns 1.5 F 0.53 1.1 F 0.58 ns 3 F 3.7 5 F 3.8 ns
Pasture allowance from pregnancy scanning to lamb tagging. Means within the same row with different superscript differ significantly (no superscript = not significant, P N 0.05; superscripts different P b 0.05). b Mean comparison within each level of pasture allowance (ns = not significant, P N 0.05; * P b 0.05).
258
J.M. Everett-Hincks et al. / Livestock Production Science 97 (2005) 253–266
Table 2 The effect of pasture allowance fed to ewes in late pregnancy and litter size at birth on ewe lamb rearing performance (Least squares mean F standard error) Pasture allowance levela
Trait
Litter size at tagging
Litter size at weaning
Litter survival birth to tagging
Litter survival tagging to weaning
Litter survival birth to weaning
Litter weight at birth (kg)
Weight of lamb weaned (kg)
Twin Triplet Pb Twin Triplet Pb Twin Triplet Pb Twin Triplet Pb Twin Triplet Pb Twin Triplet Pb Twin Triplet Pb
2 cm
4 cm
6 cm
8 cm
1.8 F 0.13 2.4 F 0.14A ns 1.7 F 0.15 1.9 F 0.16 ns 0.92 F 0.052 0.79 F 0.056B * 0.88 F 0.056 0.78 F 0.061 ns 0.83 F 0.057 0.62 F 0.06 * 8.6 F 0.36A 9.7 F 0.38B * 44 F 3.7 43 F 3.9A ns
1.9 F 0.14 2.5 F 0.14A ns 1.8 F 0.15 2.2 F 0.16 ns 0.96 F 0.054 0.83 F 0.055B * 0.95 F 0.058 0.89 F 0.059 ns 0.91 F 0.06 0.73 F 0.06 * 9.2 F 0.37 11.5 F 0.37A ns 49 F 3.8 55 F 3.9B ns
2.0 F 0.13 2.7 F 0.15 ns 1.8 F 0.14 2.1 F 0.17 ns 0.98 F 0.052 0.89 F 0.058 ns 0.90 F 0.055 0.77 F 0.065 ns 0.88 F 0.056 0.69 F 0.07 * 9.4 F 0.35 10.5 F 0.39B ns 51 F 3.6 50 F 4.2 ns
1.8 F 0.13 2.9 F 0.14B ns 1.7 F 0.15 2.1 F 0.15 * 0.92 F 0.052 0.96 F 0.054A ns 0.85 F 0.056 0.74 F 0.058 ns 0.83 F 0.057 0.70 F 0.06 ns 9.7 F 0.36B 11.6 F 0.36A * 47 F 3.8 49 F 3.8 ns
a Pasture allowance from pregnancy scanning to lamb tagging. Means within the same row with different superscript differ significantly (no superscript = not significant, P N 0.05; superscripts different P b 0.05). b Mean comparison within each level of pasture allowance (ns = not significant, P N 0.05; * P b 0.05).
pregnancy (twins 9.2 F 0.22 kg; triplets 10.8 F 0.23 kg, P b 0.0001). Pasture allowance in late pregnancy and litter size at birth did not significantly affect weight of lamb weaned ( P N 0.05). Dam age and pasture allowance from tagging to weaning also did not affect weight of lamb weaned ( P N 0.05). Days from tagging to weaning significantly influenced weight of lamb weaned (b = 0.86 F 0.278, P b 0.01). Twin litter survival from birth to tagging was similar regardless of pasture allowance in late pregnancy (Table 2). Triplet litter survival was significantly greater for ewes grazing the 8-cm sward compared to ewes on the 2-cm and 4-cm swards (Table 2). Overall triplet litter survival (86 F 4%) was lower than twin litter survival (94 F 4%) to tagging ( P b 0.05). Pasture allowance in late pregnancy and from tagging to weaning did not have a significant effect on litter survival from tagging to weaning ( P N 0.05). Litter survival from tagging to weaning was signifi-
cantly higher for twin (90 F 3%) than triplet litters (79 F 3%) ( P b 0.05). Pasture allowance in late pregnancy did not significantly affect overall litter survival from birth to weaning ( P N 0.05) (Table 2). Twin litter survival to weaning was significantly higher than triplet litter survival (twins 86 F 3%; triplets 68 F 3%) ( P b 0.001). Ewes giving birth to triplet litters on the highest pasture allowance had significantly larger litters at tagging than ewes giving birth to twin litters ( P b 0.05). Litter size at weaning was similar despite litter size at birth and feeding level in late pregnancy (Table 2). 3.3. Lamb behaviour Behaviours of individual lambs were recorded. Sward allowance up to and at lambing had a significant effect on lamb behaviours observed (Table 3). The proportion of lambs bleating before ewe contact
J.M. Everett-Hincks et al. / Livestock Production Science 97 (2005) 253–266
259
Table 3 The effect of pasture allowance in late pregnancy and litter size at birth on lamb behaviour in the 5-min observation period immediately following tagging (least squares mean F standard error) Pasture allowance levela
Trait (proportion of lambs)
2 cm Lambs that bleated before ewe contact Lambs to make ewe contact Lambs that stood Lambs that located ewe udder Lambs that suck from ewe Lambs that followed moving ewe
Twin Triplet Twin Triplet Twin Triplet Twin Triplet Twin Triplet Twin Triplet
4 cm A
0.90 F 0.05 0.73 F 0.07A 0.91 F 0.04 0.63 F 0.07A 0.91 F 0.04 0.63 F 0.07 0.72 F 0.07 0.35 F 0.07 0.37 F 0.07 0.20 F 0.06 0.77 F 0.07 0.49 F 0.08
6 cm A
0.88 F 0.05 0.83 F 0.06A 0.93 F 0.04 0.69 F 0.06 0.91 F 0.04 0.67 F 0.06 0.66 F 0.07 0.35 F 0.06 0.27 F 0.07 0.18 F 0.05 0.88 F 0.07 0.43 F 0.08
8 cm B
0.65 F 0.07 0.80 F 0.06A 0.89 F 0.05 0.85 F 0.05 0.89 F 0.05 0.85 F 0.05 0.64 F 0.07 0.62 F 0.07 0.27 F 0.07 0.19 F 0.05 0.86 F 0.06 0.65 F 0.08
0.82 F 0.06 0.55 F 0.07B 0.93 F 0.03 0.87 F 0.05B 0.82 F 0.06 0.81 F 0.05 0.69 F 0.07 0.51 F 0.07 0.29 F 0.07 0.16 F 0.05 0.61 F 0.10 0.75 F 0.08
a
Pasture allowance from pregnancy scanning to lamb tagging. Means within the same row with different superscript differ significantly (no superscript = not significant, P N 0.05; superscripts different P b 0.05).
was affected by the interaction between pasture allowance and litter size ( P b 0.01). Triplet lambs born to ewes grazing the 8-cm sward were less likely to bleat before ewe contact than lambs born to ewes grazing lower swards (Table 3). Ewe–lamb contact was significantly affected by the interaction between pasture allowance and litter size ( P b 0.001) (Table 3). The proportion of ewes to make contact with twin lambs was similar regardless of pasture allowance. However, fewer ewes with triplet litters made lamb contact if grazing lower swards. The proportion of lambs that stood was not significantly affected the interaction between pasture allow-
ance and litter size ( P b 0.05). Fewer triplets stood at lower swards and this was significantly different to the proportion of twin lambs that stood ( P b 0.05) (Table 3). A lamb’s ability to locate its dam’s udder after tagging was significantly affected by the interaction between pasture allowance and litter size ( P b 0.05). As pasture allowance increased the number of triplet lambs that located their dam’s udder increased ( P b 0.10) (Table 3). Fewer triplet lambs sucked from their dam after tagging compared to twin lambs ( P b 0.01). Pasture allowance did not significantly effect the proportion of lambs that sucked from their dams ( P N 0.05) (Table 3).
Table 4 The effect of pasture allowance from pregnancy scanning to lamb tagging on lamb behaviour times after tagging for twins and triplets (Wilcoxon mean F standard error) Trait
Litter size at birth
2 cm
4 cm
6 cm
8 cm
P Wilcoxon
P log-rank
Ewe–lamb contact
Twin Triplet Twin Triplet Twin Triplet Twin Triplet Twin Triplet
0.7 F 0.16 0.6 F 0.12 0.5 F 0.09 0.6 F 0.11 1.9 F 0.22 2.3 F 0.29 2.0 F 0.20 2.2 F 0.22 2.5 F 0.24 2.2 F 0.42
0.6 F 0.07 1.2 F 0.19 0.5 F 0.09 1 F 0.16 1.9 F 0.24 2.4 F 0.30 1.8 F 0.20 1.8 F 0.24 2.8 F 0.33 2.8 F 0.34
0.6 F 0.12 1.0 F 0.18 0.7 F 0.13 0.8 F 0.14 2.4 F 0.27 2.3 F 0.25 1.9 F 0.21 2.6 F 0.29 3.0 F 0.39 2.8 F 0.45
0.4 F 0.04 0.9 F 0.16 0.4 F 0.11 0.9 F 0.18 1.7 F 0.23 2.3 F 0.28 2.4 F 0.33 2.8 F 0.27 2.5 F 0.39 3.7 F 0.37
ns * ns ns ns ** * ns ns ns
ns *** ns * ns ** * ns ns ns
Lamb to stand Lamb locate udder Follow dam Lamb to suckle
Times are presented in minutes. ns=not significant at P b 0.05, * P b 0.05, ** P b 0.01, *** P b 0.001.
260
J.M. Everett-Hincks et al. / Livestock Production Science 97 (2005) 253–266
a
b 90%
80% 70% 60% 50% 40% 30% 2cm
20%
4cm
10%
Proportion of triplet lambs
100%
90%
Proportion of twin lambs
100%
80% 70% 60% 50% 40% 30%
8cm
0% 0
1
2
3
4
2cm
20%
4cm
10%
6cm
6cm 8cm
0% 0
5
1
2
3
4
5
Time (minutes)
Time (minutes)
Fig. 1. (a and b) The effect of pasture allowance from pregnancy scanning to lamb tagging on the proportion of lambs to have ewe contact within 5 min from tagging. Graphs of estimated dbehaviourT functions for twin lambs (a) and for triplets (b).
Approximately 80% of all twin lambs made ewe contact in 1 min and this did not differ significantly across sward heights (Fig. 1a, Table 4). Approximately 50% of triplet lambs made ewe contact in 1 min and this did differ across sward heights ( P b 0.05) (Fig. 1b, Table 4). MBS, high ewe bleats and day of tagging were all related to ewe–lamb contact time for twin ( P b 0.001) and triplet lambs ( P b 0.001). Ewe–lamb contact time after tagging was quicker if the ewe had a higher MBS and high bleated fewer times. Fifty-five percent of the triplet lambs on the 2-cm sward stood in 1 min compared with 65% on the 8-cm sward (Fig. 2b), whereas 80% of twin lambs born on the 2-cm pasture sward height had stood by the same time (Fig. 2a). Factors significantly affecting the time for twin and triplet lambs to stand included MBS, ewe high bleat frequency, whether the lamb was vocal or
The proportion of lambs to follow their dam was significantly affected by the interaction between the main effects ( P b 0.001) (Table 3). The proportion of triplet lambs that followed their dam increased as pasture allowance increased ( P b 0.01). Fewer triplets followed their dam compared to twin born lambs at all sward heights except the highest. Times related to behaviour after tagging were analysed using the LIFETEST procedure and the comparison of cumulative distributions for 2-, 4-, 6and 8-cm swards are reported separately for twins and triplets (Table 4). Triplet lamb behaviour differed significantly across sward heights for the majority of time-related traits, whereas twin behaviour was similar (Table 4). Lamb birth weight was not significantly related to twin and triplet lamb post-tagging behaviour times ( P N 0.05).
a
b 90%
80% 70% 60% 50% 40% 30% 2cm
20%
4cm
10%
6cm 8cm
0% 0
1
2
3
Time (minutes)
4
5
Proportion of triplet lambs
100%
90%
Proportion of twin lambs
100%
80% 70% 60% 50% 40% 30% 2cm
20%
4cm
10%
6cm 8cm
0% 0
1
2
3
4
5
Time (minutes)
Fig. 2. (a and b) The effect of pasture allowance from pregnancy scanning to lamb tagging on the proportion of lambs to stand within 5 min from tagging. Graphs of estimated dbehaviourT functions for twins (a) and for triplets (b).
J.M. Everett-Hincks et al. / Livestock Production Science 97 (2005) 253–266
a
b
100%
100% 90%
80% 70% 60% 50% 40% 30% 2cm
20%
4cm
10%
Proportion of population
Proportion of twin lambs
90%
8cm
0
1
2
3
4
2cm 4cm 6cm 8cm
70% 60% 50% 40% 30% 20% 10%
6cm
0%
80%
261
0% 0
5
1
Time (minutes)
2
3
4
5
Time (minutes)
Fig. 3. (a and b) The effect of pasture allowance from pregnancy scanning to lamb tagging on the proportion of lambs to locate their dam’s udder within 5 min from tagging. Graphs of estimated dbehaviourT functions for twin lambs (a) and for triplets (b).
triplets (Fig. 4b, Table 4, P N 0.05). There was no significant correlation between ewe low bleat frequency and time for lamb to follow. Approximately 30% of all twin lambs sucked from their dam (Fig. 5a), whereas less than 20% of triplet lambs did (Fig. 5b). Lambs that bleated after tagging were closer to their dam (1.2 F 0.85 m) at the end of the 5-min observation period than lambs that did not bleat (6.3 F 1.44 m) (b = 5.11 F1.536, P b 0.01). Other factors significantly affecting lamb–dam distance 5 min after tagging included ewe high bleat frequency (b = 0.61 F 0.078, P b 0.0001), time of day lambs were ear-tagged (b = 1.11 F 0.44, P b 0.01) and lamb age at tagging (b = 0.24 F 0.097, P b 0.05). Pasture allowance did not significantly affect twin lamb distance from its dam 5 min after tagging (twins 4.7 F 1.12 m). However, triplet lambs born to ewes on the 2-cm
not and lamb age at tagging (all P b 0.01). Younger lambs (determined by the number of hours between birth and tagging) took longer to stand (r = 0.19; P b 0.001) and the longer the lamb took to stand the more its ewe high bleated (r = 0.48; P b 0.0001). Lambs that were quick to stand bleated more. Within 2 min, approximately 40% of all twin lambs across all sward heights had located their ewe’s udder (Fig. 3a), whereas only 20% of all triplet lambs had located the udder (Fig. 3b). Ewe high bleat frequency was related to the time taken for the udder to be located by the lamb after tagging ( P b 0.01). The longer the lamb took to locate the ewe’s udder the more the ewe high bleated (r = 0.31, P b 0.0001). The distribution curves for twin lambs following their dam were different over time ( P b 0.05) (Fig. 4a, Table 4). Ewe low bleating frequency was related to follow time for twin lambs ( P b 0.05) and not for
a
b 90%
80% 70% 60% 50% 40% 30% 2cm
20%
4cm
10%
6cm 8cm
0% 0
1
2
3
Time (minutes)
4
5
Proportion of triplet lambs
100%
90%
Proportion of twin lambs
100%
80%
2cm 4cm 6cm 8cm
70% 60% 50% 40% 30% 20% 10% 0% 0
1
2
3
4
5
Time (minutes)
Fig. 4. (a and b) The effect of pasture allowance from pregnancy scanning to lamb tagging on the proportion of lambs to follow their moving dam within 5 min from tagging. Graphs of estimated dbehaviourT functions for twin lambs (a) and for triplets (b).
262
J.M. Everett-Hincks et al. / Livestock Production Science 97 (2005) 253–266
a
b 100% 90%
4cm 6cm
80%
8cm
70% 60% 50% 40% 30% 20%
Proportion of triplet lambs
Proportion of twin lambs
100%
2cm
2cm
90%
4cm
80%
6cm 8cm
70% 60% 50% 40% 30% 20% 10%
10%
0%
0% 0
1
2
3
4
5
0
1
2
Time (minutes)
3
4
5
Time (minutes)
Fig. 5. (a and b) The effect of pasture allowance from pregnancy scanning to lamb tagging on the proportion of lambs to suckle within 5 min from tagging. Graphs of estimated behaviour functions for twin lambs (a) and for triplets (b).
sward were further from their dam (8 F 2.1 m) than triplets born to ewes allocated higher pasture allowances (triplets 2.9 F 1.08 m) ( P b 0.05). 3.4. Lamb performance The interaction between pasture allowance and litter size at birth significantly affected lamb birth weight ( P b 0.05), where twin and triplet lambs were significantly different weights from the 6-cm sward height treatment compared to the lower sward heights (Table 5). Twin lambs (4.5 F 0.06 kg) were heavier than triplet lambs (3.7 F 0.06 kg) ( P b 0.0001). Pasture allowance in late pregnancy did not affect lamb birth weight ( P N 0.05).
Pasture allowance from scanning to tagging did not affect overall lamb survival over the same period ( P N 0.05). Twin lambs had higher survival than triplets ( P b 0.01) and the interaction between pasture allowance and litter size at birth was significant at P b 0.10. Twin and triplet lamb survival were not significantly different at the 8-cm sward and triplet lamb survival was lowest for lambs born to ewes allocated the 2-cm sward and increased as pasture allowance increased (Fig. 6). Ewe pasture allowance from scanning to tagging (2-, 4-, 6- and 8-cm sward height) and from tagging to weaning (4- and 8-cm sward height) did not significantly affect lamb weaning weight and growth rate to weaning ( P N 0.05) (Table 5). Twin lambs were heavi-
Table 5 The effect of pasture allowance fed to ewes in late pregnancy and litter size at birth on lamb performance traits (least squares mean F standard error) Pasture allowance levela
Trait
2 cm Birth weight (kg)
Weaning weight (kg)
Growth rate to weaning (grams/day)
a
Twin Triplet Pb Twin Triplet Pb Twin Triplet Pb
4.5 F 0.13 3.7 F 0.13A * 26 F 0.83 25.5 F 0.88 ns 251 F 9.6 246 F 10.1 ns
4 cm
6 cm A
4.4 F 0.12 3.8 F 0.11A * 26.2 F 0.61 26 F 0.68 ns 251 F 7.2 251 F 8.1 ns
8 cm B
4.7 F 0.12 3.5 F 0.12BA * 27.9 F 0.68 25.2 F 0.69 ns 270 F 7.8 240 F 8.2 *
4.6 F 0.12 3.9 F 0.11A * 26.6 F 0.68 25 F 0.68 ns 257 F 7.94 240 F 8.1 ns
Pasture allowance from pregnancy scanning to lamb tagging. Means within the same row with different superscript differ significantly (no superscript = not significant, P N 0.05; superscript different P b 0.05). b Mean comparison within each level of pasture allowance (ns = not significant, P N 0.05; * P b 0.05).
J.M. Everett-Hincks et al. / Livestock Production Science 97 (2005) 253–266
263
100%
Lamb survival
95%
90%
85%
80%
75%
70% Twins 2cm
Triplets 2cm
Twins 4cm
Triplets 4cm
Twins 6cm
Triplets 6cm
Twins 8cm
Triplets 8cm
Pasture Allowance (Sward Height)
Fig. 6. The effect of pasture allowance from pregnancy scanning to lamb tagging and litter size at birth on lamb survival at tagging (standard errors included).
er at weaning and grew faster (26.7 F 0.36 kg and 257 F 4.2 g/day) than triplet lambs (25.4 F 0.37 kg and 244 F 4.36 g/day) ( P b 0.05). Factors which significantly influenced lamb weaning weight and growth rate to weaning included ewe weight change over the same period, where ewes that gained more weight produced lambs with lower weaning weights (b = 0.01 F 0.004, P b 0.01; r = 0.27, P b 0.001) and growth rates (b = 0.15 F 0.044, P b 0.001; r = 0.21; P b 0.001). Lamb birth weight significantly affected weaning weight (b = 2.39 F 0.295, P b 0.0001) and growth rate (b = 15.73 F 3!.322, P b 0.0001). Days from tagging to weaning significantly affected lamb weaning weight (b = 0.29 F 0.049, P b 0.0001). Time to ewe–lamb contact after tagging significantly affected lamb growth rate to weaning (b = 10.1 F 4.474, P b 0.05). Lambs grew at a slower
rate to weaning if their dam took longer to make contact after tagging. Time for lamb to stand after tagging significantly affected lamb growth rate (b = 9.34 F 4.559, P b 0.05). Lambs that were slower to stand after tagging grew faster to weaning. Correlations were performed to determine if there were any relationships between ewe and lamb behaviours and individual lamb performance (Table 6). Lambs born to ewes with higher MBS at tagging had quicker ewe–lamb contact times, stood and located the udder sooner and were significantly closer to their dams 5 min after tagging than lambs born to ewes with low MBS (Table 6). The longer it took for ewe–lamb contact, the longer it took for the lamb to stand, locate the udder and suck. The ewe also high bleated a greater number of times. As ewe high bleat frequency increased, the greater the distance from their lamb 5 min after tagging (Table 6).
Table 6 Correlations between lamb behaviour and performance traits across all litter sizes at birth
Birth weight Body length Growth rate Ewe high bleats MBS
Time for ewe contact
Time to stand
Time to locate udder
Time to suckle
Time to follow
Distance from ewe
0.13* 0.11* 0.14* 0.74**** 0.50****
ns ns ns 0.48**** 0.23****
0.15* 0.15* ns 0.31**** 0.18**
ns 0.23* ns 0.26* ns
0.22** 0.20** ns ns ns
ns ns ns 0.38**** 0.32****
ns = not significant ( P N 0.05). * P b 0.05. ** P b 0.01. **** P b 0.0001.
264
J.M. Everett-Hincks et al. / Livestock Production Science 97 (2005) 253–266
4. Discussion Pasture allowance during late pregnancy and lambing affected ewe and lamb performance and behaviour. Ewes with triplets were most affected. Twin litter weight at birth was similar for ewes grazing 2-, 4- and 6-cm swards. However, improvements in triplet litter weight at birth, weight of lamb weaned, litter survival to tagging and weaning were observed with increasing pasture allowance. Triplet lamb and litter survival to tagging increased by 15% from the lowest pasture allowance to the highest. Similar benefits have been reported elsewhere. When multiple bearing ewes were offered additional feed in late pregnancy, lamb mortality was reduced for single and twin lambs (Scales et al., 1986). A 10-kg increase in ewe live weight during the last 6 weeks of pregnancy resulted in a birth weight increase of 0.46 kg in single and 0.52 kg in twin lambs (Scales et al., 1986). Moore et al. (1986) reported that lambs born from high feed allowance ewes had higher rectal temperatures at birth than those born from low allowance ewes indicating the lambs poorly fed in utero tend to achieve high metabolic rates more slowly than those which receive a good nutrient supply (McCutcheon et al., 1981). An increase in lamb birth weight was associated with increased lamb vigour, with increased rectal temperature only for male lambs (Moore et al., 1986). This study shows that lamb birth weight was affected by ewe pasture allowance in late pregnancy; however, lamb birth weight was not significantly related to lamb behaviour after tagging. This contradicts the finding of Astroshi and Osterberg (1979) who studied prolific sheep and concluded that lamb birth weight influences behaviour but litter size does not, except indirectly through its affect on birth weight and the smaller the lamb, the less active it is. However, Fahmy et al. (1996) found that Finn lambs, weighing half that of Suffolk lambs, were quicker to reach the udder after birth. Hight and Jury (1970) reported that lambs born to ewes fed well in late pregnancy are better equipped to survive conditions predisposing to exposure and starvation as they have more energy stored as brown fat reserves and maintain their suckling drive for longer than lambs born to poorly fed ewes. The effect of lamb birth weight on lamb behaviour is inconclusive and it is highly likely
that a complex of physiological and genetic factors underpin lamb behaviour, not excluding the effect of maternal behaviour. Maternal behaviour influences lamb behaviour and is critical to lamb survival (O’Connor et al., 1985). However, according to Smith et al. (1966) bineptitude of the lambs retards acceptance by the motherQ and mother–young acceptance is more likely if the newborn lamb does the following; stands soon after birth, sucks soon after standing, follows the mother closely and moves to the mother if separated (Alexander, 1988). This study showed that triplet lambs from ewes on the lower pasture allowance were less likely to stand, locate their dam’s udder and follow their dam. Fewer triplet lambs bleated following tagging and before ewe contact and this may be why fewer ewes made contact. Lindsay et al. (1990) observed that increased vocal behaviour of lambs was strongly associated with earlier recognition of their mother and better bonding. In this study, lambs were closer to their dam 5 min after tagging if they bleated when separated. Tagging temporarily separates a ewe and her litter. Ewes react differently to the separation and depending on the strength of the bond formed with her litter, the fear for the shepherd, the visibility of lambs versus flock lambs, the proximity of the flock and the ewe’s ddriveT to be with the flock, the ewe will stay close to her litter or desert them in favour for the flock (Everett-Hincks et al., 2005). Ewe feeding level did not affect MBS but ewes on higher pasture allowances were more likely to stay with their litter at the tagging site during the 5-min observation period after tagging. Ewes with triplet litters on the 2-cm sward were a greater distance from their litter 5 min after tagging, high-pitch bleated significantly more and were less likely to make contact with their lamb 5 min after tagging. Putu et al. (1988) observed that nutrition during late pregnancy did not affect lamb desertion in single bearing ewes, but in the twin bearing ewes the low level of nutrition resulted in a higher proportion of permanent desertions of at least one of their twin lambs (19.2%) compared with ewes on a high level of nutrition (4.3%). The maternal environment in utero and at birth affected birth weights and lamb survival to tagging but did not continue to impair lamb performance to
J.M. Everett-Hincks et al. / Livestock Production Science 97 (2005) 253–266
weaning as lamb growth and weaning weights were similar between twins and triplets and between feeding levels. Napolitano et al. (1995) observed that average daily weight gain was lower in lambs that had been separated earlier from the ewe. This study showed that lambs that were slower to stand after tagging grew faster. Lambs grew an additional 10 g/ day for every minute they stayed lying down after tagging. Lambs that had quick ewe contact after tagging grew faster to weaning. Ewes that stayed close to their litter during tagging and did not return to the flock were quick to make contact with their lambs. These ewes are more attentive to their lambs and this may explain why their lambs grow faster. A lamb that takes longer to stand after tagging is not as easily stressed as a lamb that stands quickly and flees. These dcontentT lambs have faster growth rates to weaning. In this study, some lambs had sucked prior to tagging as their stomachs appeared full and this probably reduced their drive to suck following tagging and they appeared more dcontent.T Murphy et al. (1994b) reported that animals of quiet temperament grow faster and are better producers than animals that are restless, nervous or aggressive. O’Connor (1996) suggested that the ewe–lamb relationship may be weaker for ewes and lambs that maintain close contact as the lamb ages; alternatively, lambs may stay closer to ewes with lower levels of milk production especially where there is competition for milk from littermates. Lamb behaviour is more adversely affected than ewe behaviour by pasture allowance from scanning to tagging and triplets are more affected than twins. It is not dpoorT maternal behaviour but inadequate lamb behaviour that is a limiting factor in triplet lamb survival on the lower pasture allowances. This is supported by O’Connor and Lawrence (1992) who also found that it is lamb behaviour, not maternal behaviour that is the major factor limiting lamb survival. Shillito-Walser (1978) also appreciated the importance of effective lamb behaviour, concluding that the formation of the maternal–offspring bond in sheep is the result of a rapid learning process influenced by the behaviour of both the ewe and the lamb. In conclusion, this study showed that triplet lamb survival was similar to twin lamb survival when pasture allowance was not restricted in late pregnancy and ewes with triplets require higher pasture allowances to achieve similar production as ewes with twin
265
litters. Therefore, it is recommended that ewes carrying triplets are identified, classified as priority stock and preferentially fed throughout pregnancy. Further investigation is required of the physiological mechanisms that determine lamb behaviour in larger litters resulting from changes in the maternal environment.
Acknowledgements Financial assistance was provided by Meat and Wool Innovation Limited and an AGMARDT Scholarship. The authors also thank Dean Burnham and Graham Poole from Massey University for their involvement in the project.
References Alexander, G., 1988. What makes a good mother? Components and comparative aspects of maternal behaviour in ungulates. Proceedings of the Australian Society of Animal Production 17, 25 – 41. Astroshi, F., Osterberg, S., 1979. The behaviour of Finnsheep during and shortly after lambing. Acta Agriculturae Scandinavica 29, 258 – 262. Dwyer, C.M., Lawrence, A.B., Bishop, S.C., Lewis, M., 2003. Ewe–lamb bonding behaviours at birth are affected by maternal undernutrition in pregnancy. British Journal of Nutrition 89, 123 – 136. Everett-Hincks, J.M., Lopez-Villalobos, N., Blair, H.T., Stafford, K.J., 2002. Genetic variation in maternal behaviour score and lamb survival. 7th World Congress on Genetics Applied to Livestock Production, August 19–23, Montpellier, France. Session 14. Behaviour and Welfare. Communication, p. 14. Everett-Hincks, J.M., Lopez-Villalobos, N., Blair, H.T., Stafford, K.J., 2005. The effect of maternal behaviour score on lamb and litter survival. Livestock Production Science 93, 51–61. Fahmy, M.H., Robert, S., Castonguay, F., 1996. Ewe and lamb behaviour at parturition in prolific and non-prolific sheep. Canadian Journal of Animal Science, 9 – 15. Geenty, K.G. 1997 b200 by 2000. A Guide to Improved Lambing PercentageQ. Wools of New Zealand and the New Zealand Meat Producers Board. Geenty, K.G., Rattray, P.V., 1987. The energy requirements of grazing sheep and cattle. Feeding Livestock on Pasture, vol. 3, pp. 39 – 53. Hight, G.K., Jury, K.E., 1970. Hill country sheep production. II Lamb mortality and birthweights in Romney and Border Leicester Romney flocks. New Zealand Journal of Agricultural Research 13, 735 – 752. Holmes, R.J., 1975. Relationship of maternal behaviour to neonatal loss in prolific sheep. Proceedings of the New Zealand Veterinary Association Sheep Society 5, 67 – 75.
266
J.M. Everett-Hincks et al. / Livestock Production Science 97 (2005) 253–266
Lindsay, D.R., Nowak, R., Gede Putu, I., McNeill, D., 1990. Behavioural interactions between the ewe and her young at parturition: a vital step for the lamb. School of Agriculture (Animal Science), Reproductive Physiology of Merino Sheep—Concepts and Consequences (The University of Western Australia, Perth) 15, 191 – 204. Lopez-Villalobos, N., Garrick, D.J., 1999. Genetic parameter estimates for lamb survival in Romney sheep. Proceedings of the New Zealand Society of Animal Production 59, 121 – 124. McCutcheon, S.N., Holmes, C.W., McDonald, M.F., 1981. The starvation-exposure syndrome and neonatal lamb mortality. A review. Proceedings of the New Zealand Society of Animal Production 41, 209 – 217. Moore, R.W., Millar, C.M., Lynch, P.R., Dow, B.W., 1986. The effect of prenatal nutrition and type of birth and rearing of lambs on vigour, temperature and weight at birth and weight and survival at weaning. Proceedings of the New Zealand Society of Animal Production 46, 259 – 262. Morris, C.A., Hickey, S.M., Clarke, J.N., 2000. Genetic and environmental factors affecting lamb survival at birth and through to weaning. New Zealand Journal of Agricultural Research 43, 515 – 524. Murphy, P.M., Lindsay, D.R., Purvis, I.W., 1994a. The importance of the birth site on the survival of merino lambs. Proceedings of the Australian Society of Animal Production 20, 251 – 254. Murphy, P.M., Purvis, I.W., Lindsay, D.R., Le Neindre, P., Orgeur, P., Poindron, P., 1994b. Measures of temperament are highly repeatable in Merino sheep and some are related to maternal behaviour. Proceedings of the Australian Society of Animal Production 20, 247 – 250. Napolitano, F., Marino, V., De Rosa, G., Capparelli, R., Bordi, A., 1995. Influence of artificial rearing on behavioural and immune response of lambs. Applied Animal Behaviour Science 45, 245 – 253. Nowak, R., 1996. Neonatal survival: contributions from behavioural studies in sheep. Applied Animal Behaviour Science 49, 61 – 72.
Nowak, R.F., Lindsay, D.R., 1992. Discrimination of Merino ewes by their newborn lambs: important for survival? Applied Animal Behaviour Science 34, 61 – 74. O’Connor, C.E., 1996. Ewe maternal behaviour score and lamb growth: ten years on. Proceedings of the New Zealand Society of Animal Production 56, 107 – 109. O’Connor, C.E., Lawrence, A.B., 1992. Relationship between lamb vigour and ewe behaviour at parturition. British Society of Animal Production 54, 361 – 366. O’Connor, C.E., Jay, N.P., Nicol, A.M., Beatson, P.R., 1985. Ewe maternal behaviour score and lamb survival. Proceedings of the New Zealand Society of Animal Production 45, 159 – 162. Pollard, J.C., Littlejohn, R.P., 1999. Shelter for lambing in southern New Zealand: II. Sheltering behaviour and effects of productivity. New Zealand Journal of Agricultural Research 42, 171 – 177. Putu, I.G., Poindron, P., Lindsay, D.R., 1988. Early disturbance of Merino ewes from the birth site increases lamb separations and mortality. Proceedings from the Australian Society of Animal Production 17, 298 – 301. Rattray, P.V., Thompson, K.F., Hawker, H., Sumner, R.M.W., 1987. Pastures for sheep production. Livestock Feeding on Pasture, vol. 7, pp. 89 – 103. SAS, 2002. SAS/STAT Software. Release 8.1. SAS Institute, Incorporated, Cary, NC. Scales, G.H., Burton, R.N., Moss, R.A., 1986. Lamb mortality, birth weight and nutrition in late pregnancy. New Zealand Journal of Agricultural Research 29, 75 – 82. Shillito, E., 1972. Vocalisation in sheep. Journal of Physiology 226, 45 – 46. Shillito-Walser, E., 1978. The development of the maternaloffspring bond in sheep. Applied Animal Ethology 4, 293 (Abstract). Smith, F.V., Van-Toller, C., Boyes, T., 1966. The dcritical periodT in the attachment of ewes and lambs and ewes. Animal Behaviour 14, 120 – 125.