The Expression of Maternal Behavior in Chinese Hu Sheep and its Effects on Lamb Survival and Body Weight

Among mammals, the responsibility for nursing is generally undertaken by the female animals, which is called maternal behavior. Maternal behavior in livestock refers to the behaviors related to delivery and nursing, which are expressed by the females before and after the birth of young stock, including the selection of parturition place, nesting, delivery and cleaning the newborn animals, as well as recognizing, lactating, nurturing and protecting the newborns. These behaviors can improve the survival conditions for offspring and help them to adapt to the environment.
        
Individual variation in the expression of maternal behavior can affect offspring development and survival in sheep farming (Dwyer, 2008a). Typical maternal behaviors of the ewe at lambing have been outlined including grooming or licking, sucking, following, low-pitched bleat, head-up and udder-refusal (Nowak et al., 2000; Nowak and Poindron, 2006; Pickup and Dwyer, 2011; Dodd et al., 2012; Brown et al., 2016). In recent years, the litter size is one of the most important traits in sheep production particularly for ewes, suggesting targets for selection where the aim is to increase prolificacy in breeding projects. Favorable maternal behaviors are crucial for sheep production, because the death of lambs frequently occurs within the first week after birth. Many studies indicated that 50% of the lambs died within 24 hours after birth, 30% of the individuals died within 1 to 3 days, which had a certain correlation with the weak maternal behavior of ewes (Huffman et al., 1985; Nowak et al., 2000; Sawalha et al., 2007; Dwyer, 2013). Maternal behaviour score were collected at lamb tagging from 32 Australian sheep breeds over several years in a variety of environments, showing that there were small favourable genetic correlations between maternal behaviour score and various body weights (Brown et al., 2016).
        
Hu sheep are one of the famous sheep breeds in China, particularly characterized by its high fecundity with an average litter size of more than 2.29. To determine the effects of maternal behavior expression in ewes on lamb survival and the early body weights of lambs, the expression characteristics of maternal behavior in Hu sheep were revealed through the analysis on the maternal behavior of adult ewes under stall-feeding conditions, together with the effect of litter size on birth weight and lamb survival.

Animals
 
The ewes for maternal observation experiments were provided by Suzhou Sheep Breeding Farm in China. In total 527 pregnant ewes with delivery were selected, including 246 primiparous ewes and 281 multiparous ewes. For ensuring the safety of ewes and newborn lambs, as well as the record of behavior data, the ewes were brought together in the same shed before delivery for centralized feeding and management.
        
Ewes were balanced for liveweight and condition score before being allocated to a diet group so that each of the diet group consisted of 10 sheep. Ewes of the same diet group were then allocated to a pen. At approximately mid-pregnancy (75 days), all selected ewes were housed and they were housed until one week after birth. At day 140 of gestation, ewes were placed in the lambing pen with other two ewes and the pen was 4 m × 5 m in size. After lambing, the primiparous and multiparous ewes were allocated at random to four groups balanced for litter size as single, twin, triplet and quadruplet born lambs.
 
Behavioural data collection
 
The ewes for maternal behavior observation were arranged in the fixed shed in advance and the ear mark and the located sheepfold number were recorded. The monitoring equipment were installed and debugged before the expected lambing date, the pre-test was conducted to ensure the videos could be normally stored and observed. In the experiment, 3 ewes were bred in one sheepfold. After the first ewe in each sheepfold was delivered, it was sprayed with black mark, and the lambs were pasted with a fluorescent label transversely. The lambs of the ewe with posterior delivery were pasted with a fluorescent label vertically. After the video recording time finished, the fluorescent labels were removed from the lambs. Both the 16 channel Digital Video Recorder (DVR) and camera with SONY chip were provided by Hanbang company. The video recordings were viewed and behavioural measures noted from the video materials.
        
The observation record of maternal behavior began from the 10 days before the birth of the first lamb to 3 days (72 hours) after with continuous observation (527 ewes). The behaviors of ewe to the lamb could be classified into favorable (licking, suckling and seeking lambs) and abnormal maternal behaviors (denial of suckling after delivery, butting and trampling lambs, abandoning lambs, stealing lambs). Simultaneously the ear marks for the lambs were made and the 4-stage body weights of newborn, 7-day-old, 14-day-old and 35-day-old (weaning) lambs were recorded.
        
Eight characteristic parametersof maternal behavior observed specifically in this experiment are shown in Table 1. These behaviors were recorded as described by previous reports (Dwyer and Lawrence 1998; Dwyer 2008b; Everett-Hincks and Dodds 2008; Lv et al., 2015; Lv et al., 2016).
 
Statistical analysis
 
Microsoft Excel was applied for the preliminary analysis of data and making charts. SPSS20.0 was used for the difference comparison of delivery time between primiparous and multiparous ewes. The fourfold table chi-square test in SPSS20.0 was applied for analyzing the 3 behaviors of abandonment, seeking and stealing lambs in primiparous and multiparous ewes. Independent-sample T test was used for analyzing the 4 behaviors of licking, lactating, trampling and refusing suckle in primiparous and multiparous ewes as well as the body weight data at each stage of lambs.

Maternal behavior expression
 
The expression of maternal behavior in the ewe is influenced by many factors, such as maternal experience, temperament, nutrition in pregnancy, breed and lamb behaviour (Dwyer, 2008a). Maternal behaviour depends on olfactory cues and parturition, and is facilitated by maternal experience.
        
It was observed that there was significant change in behavioural state of Hu sheep before delivery. At about 1-2 hours before delivery, the ewe was despondent and restless, walked back and forth, left the group and circled, used its fore hooves to paw the stable beds, and ingestion and rumination were basically stopped. The time from the outflow of the amniotic fluid to the termination of delivery was differed, between 0.5 hour and 2 hours. The delivery time of primiparous ewes with single and twin lambs was longer than for multiparous ewes, and there was a significant difference between them in the twin lamb group (P<0.05). It was believed that the multiparous ewes had more experience than the primiparous ewes, and experience can shorten the delivery time. It was consistent with previous research results (Dwyer, 2008b). However, the delivery time of primiparous ewes in the multiple lamb group was shorter than multiparous ewes (P>0.05) (Table 2). This was due to the larger fetus of multiparous ewe than that of primiparous ewe, or the abnormal fetal position. In addition, the litter size of many sheep breeds rarely achieved 3, therefore it was unable to be used to compare with the delivery time of Hu sheep.
        
It was indicated in Table 3 that there was statistical significance for the lamb abandonment frequency in both ewe groups, the frequency of occurrence in primiparous group was significantly higher than in the multiparous group (P<0.01). There was no significant difference in seeking behavior between primiparous and multiparous ewes (P>0.05). There was an obvious difference on the frequency of stealing lambs in both ewe groups (P<0.01), namely the primiparous ewe group more commonly exhibited the behavior of stealing lambs. This may be because primiparous ewes were more likely to initiate maternal behaviour towards other lambs before delivery.
        
Table 4 showed that there were differences in the time taken to perform four behaviors including licking, lactating, trampling and refusing to suckle between primiparous and multiparous ewes. When the T test of variances was conducted, there was no significant difference in licking and trampling behaviors between the primiparous and multiparous ewe groups (P>0.05), while there was significant difference on the denial of suckle and suckle between them (P<0.01).
        
Primiparous ewes are often less competent as mothers than are experienced ewes. They tend to have a longer labor than experienced ewes and are slower to begin grooming their lambs after birth (Morris et al., 2000; Dwyer and Lawrence, 2005). Primiparous ewes are more disturbed by the behavior of the lamb and more likely to circle, back away, or walk forward over the top of the lamb as it attempts to reach the udder, which results in delayed sucking in the lamb (Dwyer, 2003). Primiparous ewes are also more likely to show fearful behavior and aggressive behavior toward their lambs. They may fail to show maternal behavior and abandon their lamb in some cases. In a short period after delivery, some primiparous ewes did not lick the lamb initially, or only discontinuously lick them. When the lamb stood up and came close to the ewe, the ewe feared and drew back, used the posture of “head-to-head” in order to make the lamb unable to touch the udder. This kind of behavior easily reduced the feeding impulse of lamb and is harmful for establishing a relationship between mother and infant, which is regarded as the treatment of lamb abandonment. However, when compared with multiparous ewes, primiparous ewes show equivalent amounts of grooming behavior over the first 3 h after delivery and make a similar number of low-pitched bleats.
        
Housing ewes at a high stocking density can bring parturient ewes into closer contact with newborn lambs than would occur in a more natural situation, in which parturient ewes can choose seclusion, thereby leading to an increase in mismothering and lamb stealing (Dwyer, 2008a). Twenty percent of pre-parturient Merino ewes may be attracted to lambs of other ewes and interest ranges from brief inspection, grooming, suckling and even stealing of the newborn (Arnold, 1975). Unfortunately, the alien is often abandoned when the ewe delivers her own lamb. There were great differences in the ewe’s behavior in terms of stealing lambs of Hu sheep. Some ewes had maternal behaviors to the other lambs before delivery, when their biological lambs were born, the foster lambs were abandoned, and the ewes refused to suckle them. Some ewes abandoned their biological lambs after delivery and only suckled the foster lambs. An individual ewe could take care of the biological and foster lambs simultaneously, but the amount of colostrum for the biological lambs was limited. In this experiment, all the above phenomena occurred in Hu sheep. The results showed that there were significant differences in the occurrence rate of stealing lambs between the primiparous and multiparous ewe groups. The primiparous ewe group had higher stealing rates than multiparous ewe group and the implications of this were that primiparous ewe group exhibited poor maternal behavior.
 
Lamb survival
 
The difference of survival within time periods and cumulative survival was shown in Table 5. Lamb survival within 1 to 7 days of multiparous ewes was greater than primiparous ewes (P<0.05), but the same trend was that survival within 8 to 14 days and 15 to 35 days were higher than 1 to 7 days. For cumulative lamb survival, there was a downward trend from birth to weaning age, but no significant difference was observed between primiparous ewes and multiparous ewes at each stage (P>0.05). Lamb mortality in the early postnatal period (between 1 and 7 days) of primiparous ewes was a little greater than multiparous ewes (P<0.05). But the situation was exactly the opposite in the later postnatal period (between 8 and 35 days). Almost the same lamb survival of multiparous and primiparous ewes arised at weaning age, indicating that lamb survival was largely outside the control of the ewe beyond 7 days. Through human intervention after delivery, the abnormal maternal behavior of the ewes was corrected in time and the ewes were able to take good care of the lambs. Hence, lamb survival was effectively improved.
Lamb survival is an important issue in highly fecund sheep breeds. There is a broad variety in lamb mortality rates between sheep breeds in different countries. The average lamb mortality rate is 9%-30% from birth to weaning in Australian Merino sheep (Hatcher et al., 2010; Plush et al., 2011; Hinch and Brien, 2014). Most postpartum lamb loss occurs within the first three days of life and is largely caused by starvation, exposure to cold conditions and mismothering (rejection of the newborn lamb) from the ewe under extensive conditions. Lamb survival is influenced not only by the environment but also by ewe and lamb factors. The key ewe and lamb factors affecting lamb survival have been summarized by previous detailed reports (Sawalha et al., 2007; Dwyer, 2008a; Brien et al., 2014; Simitzis et al., 2016; Moraes et al., 2016). Maternal behavior affects lamb survival and vigor during the neonatal period. Our data confirm that a breed of ewe with higher lamb survival show differences in the quantity and quality of maternal behaviour expression throughout the lactation period under housed conditions. Lamb survival of multiparous ewes was greater than that of primiparous ewes (about 92%), but no significant difference between them at each stage. Primiparous ewes were often less competent as mothers than were experienced ewes, and the mortality of their lambs was higher. Lamb mortality in the later postnatal period (between 7 and 35 days) was about one-half that of the early postnatal period, indicating that within 7 days lamb survival may be affected by the maternal behavior expression of ewe.

Lamb weight

Body weight of lambs at different age stages influenced the development of sheep production enterprises and birth weight is an economically important trait in any sheep breed because it affects on growth rate before weaning period (Radwan et al., 2018). In this study, the comparison of lamb weight at each stage between primiparous and multiparous ewes were shown in Table 6. Lamb birth weight of multiparous ewes were on average 0.42 kg heavier than lambs of primiparous ewes (P<0.05). It shows that the body weights of newborn, 7-day-old, 14-day-old and 35-day-old lambs in multiparous ewes were larger than the primiparous ewes. The body weights of newborn and 7-day-old lambs in multiparous ewe group were significantly higher than the primiparous ones (P<0.05), but no significant difference was observed at 14-day-old and 35-day-old (P>0.05). The similar results were founded in the Small-Tail Han sheep of China that the weight of lambs of multiparous ewes tended to beheavier than those of primiparous ewes at birth (Lv et al., 2015; Lv et al., 2016). While by weaning at 35 days of age, lamb weight of the Small-Tail Han sheep was significantly greater in lambs of multiparous ewes compared with primiparous ewes. The results of the present study indicate that the maternal behaviour of Hu sheep, like many other breeds of ewes, was also significantly influenced by parity. Multiparous ewes showed greater maternal care than primiparous ewes and seemed to be related to the increased physiological sensitivity of experienced ewes.

This study demonstrates that there was no significant difference in the behaviors of licking and trampling between the primiparous and multiparous ewe groups, while there were significant differences in the behaviors of lactating and refusing to suckle between them. There was significant difference on the body weight of newborn and 7-day-old lamb between the primiparous and multiparous ewe groups, while there was no significant difference on the body weight of 14-day-old and 35-day-old lamb. These observations revealed that ewe parity affected early lamb weight gain. It would provided evidence for further improving the selection of maternal behavior and management measures of feeding in future.

This work was supported by the National Natural Science Foundation of China (grant number 31101747) and the six talent peaks project of Jiangsu Province in China (grant number 2017-NY-100).

Authors declare no conflict of interest.