Effect of artificial or natural rearing of replacement ewe lambs on their maternal behaviors

Final Report for GNE13-065

Project Type: Graduate Student
Funds awarded in 2013: $14,890.00
Projected End Date: 12/31/2014
Grant Recipient: Cornell University
Region: Northeast
State: New York
Graduate Student:
Faculty Advisor:
Dr. Michael Thonney
Cornell University
Expand All

Project Information


Lambing time can be the most stressful time of year for sheep producers, especially if ewes are not attentive mothers, yet the biological basis of individual variation in ewe maternal care is not well understood. Maternal motivation appears to be driven by neural oxytocin, which is very difficult to measure. Circulating estradiol is much easier to measure and plays a key role in regulating the brain’s sensitivity to oxytocin. Separately, cortisol levels have been shown to influence maternal care – mothers that are less stressed at parturition groomed their lambs more (Dwyer, Gilbert, and Lawrence 2004).

This project investigated the role of circulating estradiol, progesterone and cortisol at breeding time and at lambing time in four groups of ewes from the Cornell sheep flock. Two of the ewe groups had been reared artificially, two reared by ewes; within these ewes, two groups were inexperienced mothers, while two groups had reared lambs before. In the fall, ewes were synchronized for breeding, and estradiol and progesterone measured over 30 hours. Ewes were then exposed to a dog as a source of stress, and cortisol levels were measured. Ewes did show increased cortisol in response to the dog, but this response and baseline cortisol did not differ by group. Multiparous ewes produced significantly more estradiol at breeding time than primiparous ewes; progesterone levels were negligible. At this time, ewes were resynchronized and bred to lamb in the spring. Ewes lambed under video surveillance, while the same hormones were measured over several days prior to and one day post-parturition. Multiparous ewes again produced more estradiol, and levels of all hormones were affected by litter size. All groups showed good maternal care with no differences between rearing styles or experience levels, nor correlated with any measured hormone levels: this is in contrast to what has been reported previously. Further research is certainly needed to explain these discrepancies.


The purpose of this project is to better understand the biology behind individual variation in maternal behavior so that small ruminant farmers may rely more on the maternal skill of their dams, saving labor, money, and stress. Lambing and kidding time are undeniably the most stressful and labor-intensive portions of a small ruminant farmer’s year. This is especially true when dams fail to raise all of their own lambs or kids without intervention.
One of the easiest ways to increase profit in a small ruminant production system is to increase the number of offspring born annually per dam. Unfortunately, many ewes and does are not up for the challenge of raising three or more offspring, and so in such systems, a large number of lambs or kids may be reared artificially, increasing costs for farmers. Additionally, artificially reared lambs and kids tend to have lower weaning weights, an increased number of health issues, and our preliminary data suggests that artificially-reared ewe lambs kept for breeding are significantly more likely to have their own lambs removed and reared artificially. These issues are particularly true in lamb production systems where ewes are expected to rear their own lambs – in these cases, when a lamb is removed from its dam and reared artificially, there has usually been some additional stress such as illness, rejection or inadequate milk supply.  Lambs reared artificially by plan, such as in milking systems, don’t suffer the same stress and set-backs.
A growing body of scientific literature examines the impact of the maternal care a newborn receives on the animal’s future development and expression of stress responses and maternal behavior. Rat pups that fail to receive maternal grooming lack an epigenetic marker, making these individuals less likely to properly groom their own offspring (Champagne and Curley 2009). Young animals of many species, including humans and sheep, that do not receive proper maternal care and grooming, or that undergo significant early life stress, have been shown to have elevated stress responses . In sheep this has only been demonstrated in young lambs, but in other species these stress responses appear to persist throughout the life of the animal.
Stress response appears to have a direct effect on maternal behavior in sheep  found that ewes showing heightened cortisol responses shortly after giving birth also displayed decreased maternal grooming. In sheep, maternal grooming in the first few hours of life is very important for development of the ewe-lamb bond. The amount of grooming immediately after labor has also been correlated with maternal attentiveness later in the lactation period. Tied to this expression of maternal behavior and linked to stress responses is the mother’s circulating estradiol level; animals with higher estradiol levels show lowered cortisol responses and increased maternal grooming shortly after birth (Dwyer, Gilbert and Lawrence 2004, Dwyer 2013).
From a biological perspective, this project examines the relationship between stress response (acute cortisol levels), circulating estradiol, and maternal care. Does the heightened stress response observed in maternally deprived individuals of other species persist into adulthood in sheep? Do ewes that show increased stress responses outside of pregnancy also show increased stress responses to the experience of parturition? Can circulating estrogen levels outside of pregnancy predict stress levels outside of pregnancy and does this correlate with hormonal signals at parturition? The mechanisms controlling maternal behavior and bonding in goats appear to be very similar to those in sheep. Thus, we anticipate that our results will have application to both sheep and goat farmers. Are artificially reared ewes and does doomed to be poor mothers, not to be kept for replacement breeding stock?
If artificially-reared dams are prone to more extreme stress reactions and will make worse mothers, then farmers should be advised to cut their losses early and select replacement breeding stock not from artificially reared lambs or kids, no matter how healthy, but from female offspring exposed to ample maternal care. If dams have insufficient milk to raise all their offspring, it would be better to 1) foster future replacement females onto dams that are just birthing using VCS (vaginal cervical stimulation) techniques, ensuring that the newborns receive substantial and immediate grooming and maternal attention or 2) to leave them with their original dams for the first few critical days of life while supplementing their milk supply so that they receive plenty of maternal attention prior to removing them completely for artificial rearing.  These two options would be better than 1) fostering them onto protesting dams that birthed earlier and must be restrained by stanchioning and/or hobbling to accept the fostered offspring, or 2) removing them immediately for artificial rearing.
From a management perspective, this project examines the viability of a source of replacement dams in highly productive sheep flocks and between flocks with very different management styles. In addition, the hormonal measurements obtained in the project may provide ways to predict maternal abilities of replacement ewes and does that were reared naturally, giving insight into the role of individual temperament. The ultimate objective of the project is to make small ruminant farming less stressful for both the animals and the farmer.

Bickell, S. L., R. Nowak, P. Poindron, and D. Ferguson. 2010. “Maternal Behaviour at Parturition in Outdoor Conditions Differs Only Moderately between Single-Bearing Ewes Selected for Their Calm or Nervous Temperament.” Animal Production 50:675–82. Retrieved (http://www.publish.csiro.au/?paper=AN09118).
Dwyer, Cathy M., Colin L. Gilbert, and Alistair B. Lawrence. 2004. “Prepartum Plasma Estradiol and Postpartum Cortisol, but Not Oxytocin, Are Associated with Interindividual and Breed Differences in the Expression of Maternal Behaviour in Sheep.” Hormones and behavior 46(5):529–43. Retrieved March 26, 2013 (http://www.ncbi.nlm.nih.gov/pubmed/15555494).

Project Objectives:

  1. To examine the relationship between circulating estradiol levels, estradiol-to-progesterone ratios (E2:P4) and cortisol responses to acute stress in non-pregnant ewes for nulliparous (not previously lambed) and multiparous ewes artificially-reared or ewe-reared as lambs.

The most interesting result to come from this objective was the finding that artificially reared ewes showed decreased cortisol levels around parturition – they were less stressed by the entire process. This was likely because these ewes had been socialized to humans when they were young lambs. The lambing and sample-collection setup for this project involved a lot of human contact for the ewes. After weaning, they had received no different husbandry than any other ewes in the study.
The results for circulating estradiol and progesterone did not differ by artificial or natural rearing style nor by parity, though they did follow the expected patterns previously reported in the literature. Estradiol-to-progesterone ratios were not informative, and instead progesterone was analyzed independently. Cortisol levels did not differ by experience as was predicted – if anything, there was a trend toward higher cortisol levels in experienced ewes. (ANOVA, P=0.0771)

  1. To confirm the relationship between immediate maternal care in parturient ewes with circulating estradiol and E2:P4 levels, and acute cortisol response to parturition.

Multiparous ewes did show increased levels of circulating estradiol before and after parturition compared to primiparous ewes, but this did not translate to increased maternal attention. No differences were found between groups for cortisol levels, nor did cortisol levels correspond with the amount of time a ewe spent grooming her lambs. This is in contrast with what has been reported in the literature.

  1. To compare the levels of maternal care for artificially-reared and ewe-reared ewes within primiparous (first time lambers) and multiparous ewe groups and to determine if these relationships are related to concentrations of estradiol, E2:P4 levels, and cortisol both prior to breeding and at parturition.

Unexcitingly, but happily for sheep producers, there were no differences in maternal care between any groups, either by rearing style or by level of maternal experience. All of the ewes were very good mothers, regardless of estradiol or cortisol levels. No ewes in the study showed negative or aggressive behaviors toward their lambs. Perhaps there is not enough genetic variation within the Cornell flock to yield much individual variation in behavior, or perhaps the management system was too effective, setting all ewes up for success with no challenges. On the positive side, it appears that even ewes that got a rough start to life and that have never lambed before are perfectly able to be effective mothers.


Click linked name(s) to expand
  • Dr. Michael Thonney


Materials and methods:

Four groups of twelve Dorset-Finnsheep crossbred and Dorset ewes from the Cornell sheep farm were used in this examination of hormones and behavior at breeding and lambing time. Two groups were bred as ewe lambs to lamb for the first time at about a year of age (primiparous) and two groups will be experienced at raising lambs (multiparous). One group of primiparous and one group of multiparous were reared artificially as orphan lambs. This gives the factorial experiment design shown in Table 1.

In the fall of 2013, ewes were synchronized using progesterone-releasing intravaginal devices (CIDRs) followed by an injection of prostglandin F2a (PGF). At this time, indwelling catheters were inserted in the jugular veins and secured with tape and veterinary wrap. 18 hours after PGF administration, blood samples were drawn every 4 hours for 30 hours or until behavioral estrous. Not all ewes showed signs of estrous. These samples were stored as plasma and analyzed by radioimmunoassay to measure estradiol and progesterone. Peak estradiol levels were used in analysis; progesterone levels were negligible.

Forty-eight hours after PGF administration, ewes were exposed to an unfamiliar dog for 60 seconds. The dog was separated from the ewes by a wire panel and hidden from sight by a barrier until the 60-second exposure began. The dog lay down, sat up and caught a ball in front of each ewe. These ewes were not accustomed to dogs, so this display was sufficient to produce a stress response. Ten minutes after dog exposure, a blood samples was taken. Fifty minutes later (sixty minutes after dog exposure), a second blood sample was taken, after the animals had recovered from the stressful dog experience. These two samples, plus a “baseline” sample taken at the same time of day during the previous 30 hours of sampling, were stored as plasma and analyzed by radioimmunoassay to measure cortisol levels. After the 60-minute blood sample, catheters were removed and ewes were returned to group housing on-campus, with pasture access and bedded pens.

Originally, I had hoped to video-record the dog exposure and score the ewes for measures of fear such as vocalizations and escape attempts. Unfortunately, my makeshift testing arena proved too easily thwarted by frightened ewes, so members of the research team had to stand near the sides and attempt to prevent the ewes from escaping. Also, technical difficulties resulted in the loss of 25% of the videos – but they would not have been consistent for data collection anyway, due to our interventions. This setup did at least achieve the desired effect of stressing the ewes and raising a measurable cortisol response.

Ten days after the initial PGF treatment, ewes were implanted with a new progesterone CIDR, to resynchronize for breeding. The CIDR remained in place for 7 days, at which point it was removed and the ewes placed with a vasectomized ram overnight. 18 hours after CIDR removal, ewes were placed with two (presumably) fertile rams. Unfortunately, at least one of these rams was not as fertile as anticipated, and there were a few difficulties getting all the ewes bred on time. The main repercussion of this was an extended lambing period in the spring, and difficulty predicting exact lambing times. This is why some ewes have many more lambing-time data points than others.

During pregnancy, ewes were managed according to standard Cornell sheep farm procedures. They were kept with the remainder of the flock until approximately 3-4 weeks prior to expected lambing, at which time there were brought into the barn for sorting, vaccination against Clostridium perfringens types C, D and T, and transported to the research facility where lambing would take place.

Three days prior to the expected lambing date, an indwelling jugular catheter was inserted and secured as in the Fall for breeding-time sample collection. Blood samples were taken every 6 hours: at midnight, noon, 6am and 6pm. Sample collecting was paused at the first sign of labor and continued 120 minutes after the birth of the last lamb, with the final sample collected 24 hours after the birth of the last lamb. Unexpected issues here resulting in the loss of some samples include ewes that pulled out their own catheters by rubbing on objects or kicking at their necks, ewes that pulled out or chewed on other ewes’ catheters by reaching through panels, and ewes that were too fat for proper catheter placement – if you can’t find the jugular through the fat rolls, catheter placement becomes very difficult.

The original proposal included the collection of audio data. This would have been very useful, but it did not come to fruition for several reasons. The quality of audio recording equipment that I was able to include in the grant was not high enough for reliable collection. Even if the resulting audio files had been decipherable, many of them could not be recovered. The ewes found the microphones very interesting, and many chewed on them or pulled them off their mountings. I still think that audio data at parturition is very important; in the future I recommend investing in hard-wired quality microphones and recording equipment, and hanging the microphones well out of ewes’ reach.

The ewes were video-recorded for behavior analysis; once ewes were catheterized, they were placed into individual 8-by-4-foot pens made of wire panels. A surveillance system was focused on each of these pens, recording around the clock. These videos were used for behavioral analysis of maternal care and attention.

Behavior was analyzed beginning at the time of the birth of the final lamb from each ewe. Using the video analysis program JWatcher, students logged every behavior performed by the ewes during the observation period. Each ewe was observed for the 60 minutes following the birth of their final lamb, and then for an additional 60 minutes during the following three hours, with one ten-minute observation occurring during each 30-minute block for hours 2, 3, and 4. These behavior data have been analyzed for the presence of any aggression toward the lambs, and for the percentage of time spent grooming by the ewes. Much further analysis is possible with the data generated here.

Research results and discussion:


Breeding time (Fall) and Dog-induced stress challenge
Hormone analysis:

Cortisol concentration at ten minutes after exposure to the dog was significantly increased compared to cortisol at baseline (ANOVA P<0.001), while cortisol sixty minutes after dog exposure was significantly lower over all groups (ANOVA P=0.0165)(Figure 1.) There were no effects of rearing, parity nor any combination of rearing and parity on cortisol levels at baseline, ten minutes or sixty minutes after dog stress.
Cortisol, estradiol and progesterone levels at parturition and during the 24 hours prior to parturition were not significantly correlated with any cortisol measurements at breeding time with one exception: cortisol levels 60 minutes after dog exposure were correlated with average cortisol over the 24 hours prior to parturition (ANOVA, P=0.0169). Ewes with higher cortisol levels after dog exposure and recovery time were likely to have increased cortisol levels at lambing time.

Progesterone and Estradiol
Progesterone levels in the Fall were negligible, remaining undetectable in the vast majority of samples and ewes. Estradiol levels show a curve, climbing from undetectable concentrations to an average of 4.0 pg/ml, over the course of 8-12 hours. Each ewe displayed a very different time-course for this estradiol rise, meaning that the full curve was not always detected. For this reason, analysis for this measure uses peak estradiol rather than a curvilinear approach.
Fall peak estradiol was significantly higher in multiparous ewes (ANOVA, P=0.0002)(Figure 2). There was a trend approaching significance such that ewes with higher estradiol concentrations at breeding time have higher concentrations of estradiol at lambing time (ANOVA, P=0.0559). Fall peak estradiol was not correlated with any other hormonal measures nor with the amount of time ewes spent grooming their lambs.

Lambing time (Spring) Results

Hormone analysis
A summary of all hormone levels measured around parturition is in Figure 3. Estradiol levels by experimental group are presented in Figure 4.
Estradiol levels were relatively stable from the start of sampling (as early as 120 hours prior to parturition, depending on the ewe), until around 48 hours prior to parturition, at which point estradiol levels began to rise sharply with peak values reached in the samples taken just prior to parturition. By two hours after parturition, estradiol levels had begun to decrease. Progesterone levels decreased from the start of sampling, nearing zero in the samples taken just prior to and after parturition. Progesterone concentrations showed much wider individual variation than did estradiol concentrations.
Cortisol levels tended to increase as parturition approached, peaking just prior to parturition and then decreasing again. Many ewes had negligible cortisol levels more than 24 hours prior to- and 6 hours post-parturtition.

Regression models were fit to the hormone measurements made prior to parturition. Rearing (natural or artificial) and parity (primiparous or multiparous) were included as fixed effects. Hours from lambing (how long prior to parturition the sample was drawn), and Litter size were fit as continuous variables for each hormone measurement. The interactions of Rearing and Parity, Litter size and Parity, and Litter size and Rearing were included in the initial model. A backward step-down regression model was used, excluding effects with P-values over 0.1. The same model was used to evaluate estradiol, progesterone, and cortisol from hours -60 to 0.

The final model for estradiol included significant effects of Hours from lambing, Parity (P<0.01), and Litter size (P<0.05). The final model for progesterone included significant effects of hours from lambing, litter size and the interaction between litter size and parity (P<0.01). Multiparous ewes and ewes with larger litters had higher concentrations of circulating estradiol and progesterone (Figure 4).

The final model for cortisol included significant effects of hours from lambing, Litter size, Rearing, the interaction between Litter size and Rearing (p<0.01), and the interaction between Litter size and and Parity (p<0.05). Naturally reared ewes had higher cortisol levels prior to parturition. Ewes that carried three or four lambs on average had higher cortisol levels than ewes that carried one or two lambs.

Multiparous ewes had larger litters.  The average litter size for primiparous ewes was 1.5 lambs, compared to 2.3 lambs for multiparous ewes (ANOVA P<0.001).  Primiparous ewes were more likely to need assistance (Chi square P=0.05). When days from first breeding to lambing is considered, artificially reared ewes showed a trend for longer times to lambing: artificially-reared ewes on average lambed 154 days after breeding, while naturally-reared ewes lambed an average of 148 days after breeding. (P-0.06).

Behavioral Analysis
No ewes rejected their lambs nor showed aggression toward their lambs. No significant effects on time spent grooming were found for any rearing style, experience level or combination thereof (Figure 5). No correlations were found between time spent grooming and concentrations of any measured hormone at parturition or during the 12 hours prior to parturition.  More in-depth analysis of behavioral responses is warranted.


For estradiol, progesterone and cortisol alike, Hours from lambing was always a significant effect. This is because the variable Hours from lambing describes the curve made as each hormone decreases or increases as the time of parturition nears. This is expected to be significant since these hormones obviously and expectedly change as parturition approaches. For estradiol and progesterone, placental size and fetal weight affect hormone production, so litter size is an expected effect.

An interesting effect is that estradiol appears to be increased in multiparous ewes, even when litter size is considered. Previous literature suggests that the effect of parity on estradiol is only due to the proclivity of multiparous ewes to have larger litter sizes. This was true in the current study, as only one primiparous ewe carried triplets One possibility is that multiparous ewes grow larger or better developed placentas, providing more substrate for steroid hormone production.

The effect of rearing on cortisol was opposite than predicted. Naturally reared ewes had increased cortisol compared to artificially reared ewes. This may be due to the AR ewes being more well-socialized to people; this was a very intensive lambing setup, and the sampling protocol necessitated handling the ewes regularly. It is interesting that individual cortisol at 60 minutes after dog exposure was correlated with individual cortisol around parturition. This suggests some role of temperament, but not one determined by experience or rearing, and not a difference affecting maternal attentiveness, which is in line with the results presented by Bickell (2010).

My results conflict with some of the literature, at least at first glance. Dwyer (2008) found that the increased circulating estradiol in multiparous ewes was due entirely to the effect of litter size. In this project, the effect appears to be larger than is accounted for by litter size alone. However, Dywer’s methods ceased blood sampling approximately one week pre-parturition, whereas the sampling in the current project did not begin until only a few days prior to parturition. Multiparous ewes might experience a greater amount of placental growth during this period, allowing for increased estradiol production. Similar increases were not seen in progesterone, which is consistent because the late-pregnancy placenta would be producing less progesterone as parturition approaches.

The current findings of no difference in maternal behavior, estradiol or progesterone between rearing types supports the results in Bickell et al (2010 and 2011), in which sheep from flocks selected for active or calm temperaments showed only modest differences in maternal behavior, peripartum estradiol or progesterone. The sheep used by Bickell et al were all of the same breed, merinos. Dwyer (2004) found differences in maternal behavior, estradiol and cortisol between Scottish Blackface and Suffolk ewes at the time of parturition. This suggests that the divergence between breeds is large enough to show measurable differences in these hormones and behaviors, but that individual variation within breeds or genetically similar crossbred animals is not strong enough to yield such data. This may be due to the nature of individual variation – there’s a lot of “noise” in biological data because every animal is different – so trends need to be distinct to be visible. This doesn’t mean that individual variation isn’t worth studying, but perhaps experimental designs need to be more challenging (and data collection precise) so as to make the differences more apparent.

Bickell, S. L., Nowak, R., Poindron, P., & Ferguson, D. (2010). Maternal behaviour at parturition in outdoor conditions differs only moderately between single-bearing ewes selected for their calm or nervous temperament. Animal Production, 50, 675–682. Retrieved from http://www.publish.csiro.au/?paper=AN09118
Bickell, S., Poindron, P., Nowak, R., Ferguson, D., Blackberry, M., & Blache, D. (2011). Maternal behaviour and peripartum levels of oestradiol and progesterone show little difference in Merino ewes selected for calm or nervous temperament under indoor housing conditions. Animal, 5(04), 608–614. http://doi.org/10.1017/S175173111000217X
Dwyer, C. M., Gilbert, C. L., & Lawrence, A. B. (2004). Prepartum plasma estradiol and postpartum cortisol, but not oxytocin, are associated with interindividual and breed differences in the expression of maternal behaviour in sheep. Hormones and Behavior, 46(5), 529–43. http://doi.org/10.1016/j.yhbeh.2004.05.011
Dwyer, C. M. (2008). Individual variation in the expression of maternal behaviour: a review of the neuroendocrine mechanisms in the sheep. Journal of Neuroendocrinology, 20(4), 526–34. http://doi.org/10.1111/j.1365-2826.2008.01657.x

Research conclusions:

This project found no differences between the maternal attentiveness of artificially reared and naturally reared ewes. This is good for flexibility in sheep farming management. In flocks with prolific ewes, there are often a number of lambs that end up being reared artificially. These results support the ability of such lambs to mature into perfectly capable mothers. For another example, artificially reared lambs from milking sheep could be transferred to a lamb-based production system and successfully rear their own lambs.
The lower cortisol levels of artificially reared ewes did not affect the maternal attention that ewes gave to their lambs, but in a more reactive sheep breed these distinctions could be larger.From a welfare view, these results showing lower cortisol in artificially reared ewes support those producers who opt to socialize their ewes as young lambs, particularly if those ewes will later be managed in an “intensive” or hands-on manner during lambing.
Extension audiences – the ones I have met in the Central NY region, at the least – are always very eager to learn not only about lambing best practices and techniques, but also welfare considerations and the “experiences” of their animals. Behavior and stress measurements are the scientific window to this. Hopefully this research will help sheep producers make informed decisions about the management of their ewes at lambing time.

Participation Summary

Education & Outreach Activities and Participation Summary

Participation Summary

Education/outreach description:

The results from this project have not yet been formally presented to farmers, but over the duration of the study I have given several lectures on lambing management and biology, including topics that I have become more familiar with as a result of this work. The time I spent raising the artificially reared primiparous group and lambing all four groups under near-constant surveillance have given me a great additional wealth of experience to share with the farmers and students. I tend to lecture on introductory and intermediate management topics, such as “Preparing for lambing and kidding”. Even without having presented my experimental results, I’ve been able to discuss events and issues that I encountered during my intensive lambing and lamb-raising experiences.

In the future, I plan to present the results of this research to both farmer and student groups. These results will also be submitted for publication in an appropriate research journal.

Project Outcomes

Project outcomes:

The project in its current state has not had any measurable impacts on farm economics, but the knowledge generated will help sheep producers to make the best management decisions for their farm. As sheep dairies grow in popularity, the number of lambs reared artificially is likely to increase. Knowing that these lambs can be sold or purchased as breeding stock for a management system in which they will rear their own offspring is reassuring.

Farmer Adoption

While I haven’t directly presented the results of this project to farmers, I have informally discussed the project through venues such as the Catskills Winter Agriculture Conference. The farmers I spoke to were happy to hear that all of my ewes turned out to be good mothers, even if they had a challenging start to life. In this case, my lack of a substantial difference between groups ends up being reassuring to farmers – yes, your good management practices will help your ewes succeed, even if they has a stressful start to life.

Assessment of Project Approach and Areas of Further Study:

Areas needing additional study

There is so much potential for further work in this area! One step would be to compare the common “types” of artificially reared lambs, which my study did not discriminate between. Some healthy lambs may be separated from the ewe immediately and fed colostrum from a bottle. Others may be left with the ewe for a day, then removed to artificial rearing after they have drank enough colostrum. These practices are not uncommon in milking systems. Other lambs are removed from the ewe because they become sick. These “recovery” lambs often have delayed or stunted growth, and can have received any amount of maternal care before depending on when they became sick. How do these differences in the origin of artificially reared lambs affect the potential of these lambs?

Another aspect of maternal bonding that deserves attention is lamb behavior. Ewes respond to lamb activity – how much of the ewe’s maternal response is linked to the lamb’s behavior? Are there ewe temperament differences such that some ewes are more responsive to “dopey” lambs, or that other ewes are made nervous by particularly energetic lambs?

Auditory information, which I intended to include in this project but was unable, is worth a great amount of further research. Ewes and lambs are very vocal creatures, as anybody who has been present during weaning time can attest. It has already been shown that lambs can discriminate their dam’s low-pitched rumbles. High-pitched vocalizations are accepted as a measure of stress. Further study of vocalizations during labor and bonding, and during periods of stress, will be very informative.

Any opinions, findings, conclusions, or recommendations expressed in this publication are those of the author(s) and do not necessarily reflect the view of the U.S. Department of Agriculture or SARE.