Review of sheep body condition score in relation to production characteristics

Abstract

Body condition scoring of sheep was first developed as a technique in the 1960s. Unlike live weight, it circumvents the issues of skeletal size, breed and physiological state and is not influenced by gut fill or the length and wetness of the fleece. This review outlines the use of the technique and the relationships between body condition score and other physical measures. In addition, it summarises the literature, across a range of breeds and environments, on the effects of body condition score on reproductive and lactational performance, and the growth and survival of the offspring to weaning. We have proposed that while the relationship between body condition and production traits is positive, it is unlikely to be linear. Where appropriate, the review outlines areas that would benefit from further research. Finally, it outlines what a suitable body condition score profile might be for a ewe over the entire breeding cycle.

Keywords:

• ewes
• rams
• fatness
• condition
• production
• reproduction

Introduction

Although all nutrients are ultimately derived from food intake, the pool of disposable nutrients in an animal includes those that are stored in body tissues, such as adipose tissue and intramuscular fat, that can be mobilised if required (Rauw 2008). It is apparent that when nutrient supply is limiting, or when feed demand is high, animals, including sheep, will utilise their body reserves in an attempt to meet their requirements. An animal's total body nutrient reserves are difficult to measure or assess. Reserves located in the abdominal cavity, such as glycogen in the liver and intra-abdominal fat, cannot be assessed easily on live animals without specific equipment. However, the subcutaneous and muscle reserves along the backbone can be assessed using the body condition scoring (BCS) technique, which is described below, and forms the basis for this review. The BCS technique has the potential to be a useful management tool for producers to lift animal performance.

In this review, we first describe the methods used to assess BCS and how this assessment relates to live weight (LW) and direct fat deposition measurements. We then discuss and compare studies that have investigated the relationships between BCS and reproductive success, and examine the impact of BCS on every step of the reproduction process in ewes. The role of BCS in milk production and lamb growth is also reviewed. Further, we briefly discuss the relationships between BCS and ram performance and animal welfare. The importance of animal genetic make-up and the dynamic aspects of the impact of BCS, and the interplay between nutrient intake and BCS are illustrated throughout the review. We conclude by reflecting on what an optimum BCS range might be throughout the production cycle.

Body condition score: the technique

Body condition scoring is a practical and easily learned technique making it an ideal management tool. It has an advantage over live weight measurement in that it requires no specialised equipment. Furthermore, BCS, in comparison with live weight, circumvents the issues of skeletal size between and within breeds and physiological state (i.e. pregnancy) and is not influenced by gut fill or the length and wetness of the fleece (Jefferies 1961; Adalsteinsson 1979; Russel 1984a, b; Gonzalez et al. 1997; Esmailizadeh et al. 2009).

The BCS of an animal is assessed by the palpation of the lumbar region, specifically on and around the backbone (spinous and transverse processes) in the loin area, immediately behind the last rib and above the kidneys to examine the degree of sharpness or roundness (Jefferies 1961; Russel et al. 1969; Russel 1984a, b). In simple terms, it is a means of subjectively assessing the degree of fatness or condition of a live animal (Russel 1984b). The technique was first published by Jefferies (1961) and was based on a 0 to 5 scale, including only whole units (Table 1). Later, Russel et al. (1969) introduced the concept of 0.5 and 0.25 units. The original purpose of the BCS technique proposed by Jefferies (1961) was to: (1) control the condition/nutrition of sheep, so that available food supplies were utilised more efficiently; (2) detect small differences in body condition not noticeable by outside appearance; (3) allow farmers to be immediately aware of major losses in body condition; (4) follow trends in nutrition and body weight.

Table 1 Description of the BCS technique and an illustration of the vertebra and ribs and approximate muscle and fat distribution.

Grade	Description	Illustration
Score 1	The spinous processes are prominent and sharp. The transverse processes are also sharp, with fingers passing easily under the end of this process. The eye muscle areas are shallow with little to no fat cover.
Score 2	The spinous processes are smooth but still prominent. The individual processes can still be felt but only as fine corrugations. The transverse processes are smooth and rounded. However, it is still possible to pass the fingers under the ends of the processes with some pressure. The eye muscle areas are of moderate depth, but have sparse fat cover.
Score 3	The spinous processes are smooth and rounded, and individual bones can only be felt with some pressure applied. The transverse processes are also smooth and are well covered. Firm pressure is required to feel over the ends. Eye muscle area is full and covered by a moderate degree of fat.
Score 4	With pressure applied, the spinous processes can just be detected, although the ends of the transverse processes cannot. Eye muscle areas are full with a thick covering of fat.
Score 5	Even with firm pressure applied, the spinous processes cannot be detected. Due to a high level of fat adjacent to the spinous process, a depression directly above where the spinous processes would normally be felt may be present. It is not possible to detect the transverse processes. The eye muscle areas are very full with very thick fat cover. It is possible to have significant deposits of fat cover over the rump and tail.

Adapted from Jefferies (1961); Russel et al. (1969), Russel (1984a, b).

Repeatability within and between assessors

Due to the subjective nature of BCS, repeatability within and between assessors has the potential to be a significant issue that can limit the potential use and effectiveness of this technique. Between studies, the repeatability of the BCS technique within assessor has varied from ‘low’ to ‘high’ (Table 2). Combined, the data suggest that inexperienced assessors can have difficulty achieving consistency between assessments (Everitt 1962; Yates & Gleeson 1975), whereas experienced assessors appear to be able to achieve high levels of consistency, even when assessing ewes to 0.25 units (Teixeira et al. 1989; van Burgel et al. 2011; Phythian et al. 2012). Yates & Gleeson (1975) stated that assessors found the later stages of pregnancy particularly difficult to assess. This may suggest that changes in the shape of ewes in late pregnancy can influence the ability of the assessor to accurately determine BCS. It appears, to date, that studies have not attempted to determine how the change in ewe shape that is associated with the later stages of pregnancy affects the accuracy of the technique. This warrants investigation.

Table 2 Repeatability of the BCS technique between and within assessors.

Reference	Scale system used and smallest units	Assessor experience level	Repeatability between assessors	Repeatability within assessors
Everitt (1962)	1–10, whole units	Inexperienced	Variation between and within assessor over time, values not stated
Russel et al. (1969)	0–5, 0.25 units	Not stated	>70% absolute agreement, <20% differed by 0.5 BCS units and <10% by 1.0	>80%, <15% of observations varied by 0.5 BCS units and <5% by 1.0 units
Milligan & Broadbent (1974)	0–5, units not stated	Not stated		r = 0.49–0.67
Yates & Gleeson (1975)	0–5, 0.25 units	Inexperienced	r^Footnotea = 0.05–0.27	r = 0.16–0.44
Evans (1978)	0–5, 0.5 units	Not stated	r = 0.81	r = 0.88
Teixeira et al. (1989)	0–5, 0.25 units	Experienced	80%	90%
Calavas et al. (1998)	0–5, 0.25 units			k^Footnoteb = 0.6–1.0
Shands et al. (2009)	0–5, 0.25, 0.5 units	Mixed skill level	r = 0.73–0.89	r = 0.64–0.84
van Burgel et al. (2011)	0–5, 0.25, 0.5 units	Experienced		0.2 unit difference in BCS mean of animals tested between assessments
Phythian et al. (2012)	0–5, 1.0 and 0.5 units	Experienced	k = 0.4–0.6	k = 0.6–0.7

^a r = correlation.

^b k = weighted Kappa analysis: <0.4 (poor level agreement); 0.40–0.75 (fair-good); > 0.75 (excellent).

Phythian et al. (2012) reported that a short period of recalibration of assessors appeared to improve consistency. This statement, combined with the apparent greater consistency of experienced operators, indicates that either assessors would benefit from a standardised ‘tool’ for calibration (see later section) and/or that the BCS technique benefits from the assessor regularly utilising it.

If BCS is to be utilised as a technique to develop and help implement management guidelines across farms, there needs to be consistency between assessments of individual animals. On-farm, the variation could be reduced by having two different assessors provide an estimate for each ewe (Evans 1978). Although there is some merit in the idea, it is unlikely to be a practical solution. Similarly, Calavas et al. (1998) provided evidence that the BCS technique was easy for a single person to master for their own flock, but that for across flocks and assessor comparisons, the technique could be improved. A potential means to reduce between-operator bias is a greater use of ‘condition score’ models (score 1–5, in 0.5 units; Van Burgel et al. 2011) such as those developed by ‘Lifetimewool’, allowing for assessor calibration and training (Curnow et al. 2011). These models have been found to be an effective tool for communicating the value of BCS, when supported by a worksheet that gives descriptors of each score (Curnow et al. 2011). ‘The tactile nature of the models allowed the audience to immediately get a “feel” for condition scoring, without the need for real sheep and the provision of a handout served to consolidate the take-home message on how easy it was to score sheep’ (Curnow et al. 2011). Alternatively, van Burgel et al. (2004) showed that it was possible to create calibration equations to adjust BCS values recorded by different assessors.

Future studies may wish to compare repeatability both within and between assessors, before and after retraining, utilising both experienced and novice operators across a wide range of BCSs. It is likely that greater variability would be shown with novice or less experienced assessors and that these assessors would benefit the most from retraining. It is also likely to be worth determining how often assessors should calibrate themselves to ensure consistency.

Relationship between BCS and physical attributes

BCS and live weight

The live weight of a ewe is a combination of both body size and condition, and therefore live weight per se may not necessarily be a good indicator of an animal's body condition (Ducker & Boyd 1977). A particular live weight could be a large-framed animal in poor condition or a smaller-framed animal in very good condition. Despite that potential confounding, there are well-established positive relationships between BCS and live weight (Table 3). Although the increase in live weight required to raise ewe BCS by 1.0 unit varies, the majority of studies suggest a linear relationship exists between live weight and BCS. In contrast, Teixeira et al. (1989) with Rasa Aragonesa ewes reported a curvilinear relationship with increases in live weight of 7, 10, 12 and 16 kg required to increase BCS by 1.0 unit over the BCS range of 1.0–2.0, 2.0–3.0, 3.0–4.0 and 4.0–5.0, respectively.

Table 3 Reported change required in live weight per unit of BCS across various breeds and sheep classes.

Reference	Timing of measurement^b	Breed^c	Live weight change per unit of BCS (kg)
Jefferies (1961)			6.8
Russel et al. (1969)			10.6
Geisler & Fenlon (1979)	Breeding	Eight breeds	3.3–7.8
Hossamo et al. (1986)	Breeding	Awassi	5.8
Teixeira et al. (1989)	Dry	Rasa Aragonesa	7.0–16
Sanson et al. (1993)			5.1
Frutos et al. (1997)			5.6
Kenyon et al. (2004a)	Breeding	Romney	7.3
Kenyon et al. (2004b)	Breeding	Romney	7.9
Kenyon et al. (2004b)	Breeding	Romney composite	4.8
Freer et al. (2007)^a	Dry	Polwarth × SA Merino	6.3
	Dry (maiden)	Polwarth × SA Merino	7.3
	Dry	Saxon Merino	5.6
	Dry (maiden)	Saxon Merino	7.0
	Lactating	SA Merino	5.0
	Lactating	Saxon Merino	5.5
	Wethers	Saxon Merino	7.0, 10.0
	Weaners, wethers	Saxon Merino	9.3
	Weaners, ewes	Saxon Merino	7.0
Van Burgel et al. (2011)			9.2

^a Adapted from Freer et al. (2007).

^b Time within production system measured and age if stated.

^c Breed if stated.

A simple multiplier of mature weight may be used to determine the change in live weight required to increase BCS by 1.0 unit. Freer et al. (2007) reported a mean multiplier of 0.15 across a number of breeds; however, it is likely that a standard multiplier is not suitable across breeds. Reported multipliers are 0.11, 0.12 and 0.13 for Karagouniko, Serres and Boutsko breeds (Zygoyiannis et al. 1997), 0.13 for Churras (Frutos et al. 1997) and 0.19 for Merinos (Van Burgel et al. 2011), respectively. There is also evidence to indicate that the relation between live weight and BCS may vary between the sexes (i.e. wethers vs. ewes) and ages (Table 3). In summary, the magnitude of the live weight difference per unit BCS differs between breeds. The likely explanations for this include differences in: (1) body size; (2) conformation; (3) standard reference weight; and (4) fat distribution throughout the body (Russel et al. 1969; Geisler & Fenlon 1979; Russel 1984b). Therefore, given this variation between breeds, individual multipliers are likely to be needed to be developed for each breed across age groups. It is apparent that studies that have examined the relationship between BCS and live weight have been based on between-animal variation, i.e. flock means. Future studies may wish to examine this relationship within-animal, by manipulating individual animal live weights.

A gain in live weight comes at an energy cost. Therefore, a gain in BCS also has an energy cost, which varies depending on the physiological state of the animal and its current BCS. It requires more energy to gain an additional unit of BCS in animals with a heavier reference weight or a high BCS (Freer et al. 2007, i.e. BCS 4.0 vs. 2.0), probably because this additional gain is predominantly body fat which is energy dense. The greater total live weight gain requirement for an additional unit of BCS in a heavy animal may partly explain the variation reported between breeds. There is also evidence to suggest that it requires less energy to gain an additional unit of BCS in lactating ewes than it does in non-lactating ewes (Freer et al. 2007), suggesting lactating ewes are more efficient at gaining BCS.

BCS and fat

Almost 100 years ago, Murray (1919) used the word ‘condition’ to describe the level of fatness of an animal. Russel et al. (1971) thought that subjective estimates of fatness were of more value in describing the condition of a sheep than was live weight alone, while Sanson et al. (1993) stated that BCS was a better estimate of energy reserves than live weight. Similarly, both Yates & Gleeson (1975) and Teixeira et al. (1989) concluded that BCS was better than live weight for determining total body fat.

In contrast, Frutos et al. (1997) reported that live weight was a better predictor of total body fat than was BCS in Churra ewes. They stated that there is variation in the sites where different breeds deposit body fat. Frutos et al. (1997) thought that the Churra might differ from other breeds in that it has a high proportion of non-carcass fat, which is often the case in breeds selected for milk production. Similarly, the Finnish Landrace breed is thought to have a greater proportion of its body fat around the internal organs compared with English breeds (Russel 1984b), and so presumably BCS might not be the best indicator of fatness in that breed. To verify this would require either detailed whole body dissection studies or use of computerised tomography (CT). If a poor relationship were found, then the use of BCS may have limited use in the Finn and composites derived from it. It could be hypothesised that BCS has limited value in estimating overall fatness levels in fat-tailed breeds of sheep (i.e. the Awassi), which are not common in temperate farming scenarios. However, as outlined later in the review, BCS is still related to ewe performance in fat-tailed ewes (Hossamo et al. 1986; Atti et al. 2001; Esmailizadeh et al. 2009; Aliyari et al. 2012).

Despite the potential problem of the BCS not accounting for fat that is deposited at sites other than the lumbar spine, BCS has been shown to be positively related to empty body chemical fat percentage (Russel et al. 1969) and the level of omental, mesenteric, pelvic and heart fat, subcutaneous and intramuscular fat, and total body fat (Yates & Gleeson 1975; Delfa et al. 1989; Teixeira et al. 1989; Frutos et al. 1997; Glimp et al. 1998). BCS is correlated with both intermuscular fat in the lumbar region (Frutos et al. 1997) and the depth of the fat above the longissimus dorsi muscle (Delfa et al. 1989). Similarly, there is a positive exponential relationship between BCS and eye muscle fat depth at the C site, although the absolute relationship differed between studies (Van Burgel et al. 2011). Furthermore, BCS is correlated with fat above the eye muscle (Verbeek et al. 2012). Although there are valid reasons to suppose that the BCS technique could underestimate fat deposition in breeds that lay down fat at sites other than the lumbar spine, the results of the studies mentioned here seem to suggest that fat deposition at any site in the body does correlate with an increase in BCS. Interestingly, BCS accounted for more of the variation in the percentage of lipid and protein in the carcass than live weight (Sanson et al. 1993). Low BCS animals have leaner carcasses than animals of higher BCS (Glimp et al. 1998). Future studies may wish to consider what percentage of total body fat is accounted for by the BCS technique. This would require either dissection studies or the use of CT and would ideally use a range of breeds, age groups and physical states (i.e. pregnant vs. non pregnant). This basic knowledge would help to interpret the ability of the animal to buffer against nutritional shortages.

BCS and GR

Soft tissue depth at 110 mm from the mid-line of the 12th rib (GR) is used to estimate both the level of fatness of a carcass and the lean meat yield (Kirton & Johnson 1977; Kirton et al. 1978). Van Burgel et al. (2011) reported that ewe BCS was positively related to GR tissue depth, although at low GR depths the relationship was not always as strong. Similarly, both Shands et al. (2009) and Verbeek et al. (2012) reported a positive relationship between BCS and GR. However, because BCS measurement is based around tissue depth over the extremities of the backbone, a very high correlation with GR should not be expected, due to the site of measurement of the latter (Shands et al. 2009).

BCS and fat score

Fat score is a subjective assessment of the fat over the ribs using a scale of 1–5 (1 = cannot feel any tissue over ribs, 5 = ribs barely felt [Moxham & Brownlie 1976; Shands et al. 2009]). Fat scoring involves the fingers of the assessor being run over the skin of a lamb at the GR site to estimate GR depth, or level of ‘fatness’ (Moxham & Brownlie 1976; Shands et al. 2009, van Burgel et al. 2011). It has also more recently been used with mature stock to aid in feed management decisions (Shands et al. 2009). BCS and fat score are positively related, although the fat score had a higher correlation with GR tissue depth than did BCS (Shands et al. 2009). Van Burgel et al. (2011) suggested that BCS was a better technique than fat scoring, especially for poorer condition animals, as an alternative for live weight when determining the nutritional status of a ewe. They stated that at BCS below 2.5, by definition, animals would have a fat score of 1, which would limit fat scoring as a tool for determining those animals of the greatest nutritional need and under conditions of compromised welfare.

BCS and muscle depth

The body condition score has been found to be related to measurements of muscle depth. Positive correlations have been described between BCS and the depth of the longissimus dorsi muscle in mature Rasa Aragonesa ewes (Delfa et al. 1989), the depth of the eye muscle at the C site in mature Merino ewes (Van Burgel et al. 2011) and the eye muscle width and depth in mature Coopworth cross ewes (Verbeek et al. 2012).

Heritability of BCS and genetic and phenotypic relationships with BCS

The heritability of BCS is in the range of 0.16–0.30 with a repeatability of 0.27–0.41 (Everett-Hincks & Cullen 2009; Shackell et al. 2011). Genetic correlations between LW and BCS in the range of 0.58–0.75 and phenotypic correlations between 0.53–0.65 have also been reported (Shackell et al. 2011). Everett-Hincks & Cullen (2009) reported that the genetic correlation between mid-pregnancy BCS and litter survival in an across flock and breed analysis was moderate to high (0.39–0.73), although the phenotypic correlations were very low. The genetic correlations suggest that selecting for higher BCS is a potential means of improving lamb survival. This indicates that farmers could incorporate BCS in their breeding programmes if they wished to alter the BCS of future generations as an indirect means of improving productive performance. It would also be beneficial for farmers if a greater number of phenotypic and genetic correlations between BCS and other production traits could be determined. This knowledge would help determine if it were more economic for farmers to select for increased BCS or to provide improved nutrition.

BCS technique use by industry

Farmers may accept BCS as a management tool if they understand and accept the benefits it will provide to their production system. Further, if the technique is advocated and utilised by those providing training and education to farmers, it is more likely to be accepted by farmers. The ‘Lifetimewool’ programme in Australia has developed BCS models, ewe production cycle BCS profile guidelines, and work sheets for farmers based on research findings (Curnow et al. 2011; Jones et al. 2011). The awareness and use of these tools by extension practitioners, consultants and farmers was increased after the completion of this programme (Curnow et al. 2011; Jones et al. 2011; Trompf et al. 2011). In addition, farmers had a greater understanding of the importance of BCS targets and the key messages to achieve high performance rates (Jones et al. 2011; Trompf et al. 2011). Combined, these studies indicate that BCS is a technique that farmers will utilise if they are given the appropriate training and provided with evidence that clearly indicates the advantages of using this tool.

Interestingly, even though a considerable proportion of producers indicated that they monitored the condition of their sheep (96%), only 7% used hands-on assessment to determine the BCS of a ewe, while the remainder of farmers used either visual assessment only (41%) or a combination of visual and hands-on assessment (52%) (Jones et al. 2011). These findings indicate that farmers still have difficulty with the concept that each ewe needs to be assessed by hand, especially if it is not directly off-shears, to get a true understanding of the BCS of an individual. Further work is likely required to persuade farmers of the need to assess BCS with their hands only.

In New Zealand, recent data suggest that approximately 40% of farmers have utilised BCS (Corner et al. 2013). Beef + Lamb NZ has recently released a BCS manual (Anon. 2013) which aims to inform farmers of the benefit of using BCS, and outlines how to use the technique. This manual will be followed up with a series of workshops.

The preceding review of the literature suggests that the BCS technique is correlated with measures of the fat that an animal is carrying, and also correlates with measures of skeletal muscle deposition, thus providing a gauge of the condition of an animal. BCS is a trait that can be selected for and current evidence suggests that farmers have a degree of willingness to use the technique. The remainder of the review focuses on the relationships between BCS and production performance, questioning whether sheep in better condition have improved reproductive performance, and whether the gain in production performance is linear with changes in BCS. These sections predominantly focus on the mature animal, and review studies that have examined the relationships in individual animals.

Effect of BCS on ewe reproductive traits

In lamb production systems, the total number and the weight of lambs weaned have a significant impact on profitability (Morel & Kenyon 2006; Young et al. 2010). The energy balance of a ewe is an important factor in determining the number and weight of lambs weaned (Scaramuzzi et al. 2006). Therefore, it might be expected that ewes of lower BCS will display reduced reproductive performance in comparison with those of greater BCS. The end point from a reproductive perspective for farmers is often the numbers of lambs weaned per ewe presented for breeding. Kleemann et al. (2006) in an across-flock analysis with Merino ewes, Newton et al. (1980) with Masham ewes, and Saul et al. (2011) with Merino ewes all reported that BCS at breeding and/or mid-pregnancy was positively correlated with the number of lambs weaned per ewe exposed to the ram. Vatankhah et al. (2012), with Lori-Bakhtiari ewes, reported that BCS at breeding had a positive effect on the number of lambs weaned per ewe, but only to a BCS of 3.0 before it plateaued. The ovulation rate of a ewe sets the potential for the number of lambs that can potentially be weaned per ewe per year.

Ovulation rate is the major determinant of variation in the number of lambs weaned between farms (Knight et al. 1975; Lindsay et al. 1975; Kelly 1982; Knight 1990). Post-ovulation, the aim should be to minimise reproductive wastage. Reproductive loss can occur through: (1) failure to conceive; (2) ova loss; (3) embryonic and fetal loss; and (4) peri- and post-natal lamb loss. The following section discusses the impacts of BCS on each of these reproductive parameters and others, with the overall aim of outlining how BCS influences the number of lambs weaned per ewe exposed to the ram. In addition, in each of these sections, the breed of the ewe is outlined, as this is likely to influence the results observed. Further tables have been used to outline individual study findings, reducing the need for laborious summaries in the text.

Breeding season

There is enough evidence to suggest that BCS can influence the response of a ewe to seasonal cues. Scottish Blackface ewes, with lower BCS, had a delayed start to the breeding season (Gunn & Doney 1975, Table 4), while late in the breeding season, Masham ewes with higher BCS were more likely to display oestrus (Newton et al. 1980). Rasa Aragonesa ewes with higher BCS had a longer breeding season, predominately due to a later onset of seasonal anoestrous (Forcada et al. 1992), and a shorter total seasonal anoestrous period (Rondon et al. 1996). Combined, these studies indicate that ewes with higher BCS display longer breeding seasons. However, the effects of BCS on the length of breeding season are relatively minor, and it is unlikely that the manipulation of BCS could be used to shift the timing of the breeding season significantly.

Table 4 Summary of studies examining the relationship between BCS and breeding season, ovulation rate and conception rate.

Reference	When BCS recorded and range tested	Nutritional treatment(s) during examination period^a	BCS and length of breeding season relationship	BCS and ovulation rate relationship	BCS and conception rate relationship
Gunn et al. (1969)	Breeding, 1.5 and 3.0	Low, maintenance, high		+
Bastiman (1972)	Breeding, 2.5 to 3.5	N/S		+	+
Gunn et al. (1972)	Breeding, 1.5 and 3.0	Fed to maintain BCS		+
Gunn & Doney (1975)	Breeding, 1.0 to 3.0	Low, maintenance, high		+
Gunn & Doney (1979)	Breeding, 2.0 and 3.0	Fed to maintain BCS		+
Newton et al. (1980)	Breeding, 2.0 and 4.0	Fed to maintain BCS	+ late in breeding season
Knight & Hockey (1982)	Pre-breeding	Commercial conditions		+
Rhind et al. (1984a)	Breeding, 1.8 and 2.8	Fed to maintain BCS		+
Rhind et al. (1984b)	Pre-breeding, 2.5–3.0 and 3.25–3.75	Fed to maintain BCS		+
Rhind & McNeilly (1986)	Pre-breeding, 1.8 and 2.9	Fed to maintain BCS		+
McNeilly et al. (1987)	Pre-breeding, 1.8 and 2.9	Fed to maintain BCS		+
Gunn et al. (1988)	Breeding, ≤1.5, 1.75–2.0, 2.25–2.5 and ≥2.75	Low, high		+ and + to 2.25–2.5 in two differing breeds
Gunn et al. (1991a)	Pre-breeding, ≤2.25, 2.5 and ≥2.75	Low, high		+ and + to BCS 2.5 in two differing breeds	+ and + to BCS 2.5 in two differing breeds
Gunn et al. (1991b)	Pre-breeding, ≤2.25, 2.50–2.75 and ≥3.0	Low, maintenance		+ to BCS 2.5–2.75	+ to 2.5–2.75 then –
Forcada et al. (1992)	Breeding, ≤2.5 and ≥2.75	Fed to maintain BCS	+	+
Rondon et al. (1996)	Breeding, ≤2.5 and ≥2.75	High	+
Kleemann & Walker (2005)	Breeding,	Commercial conditions		+
Sejian et al. (2009)	Pre-breeding, 2.5, 3.0–3.5 and 4.0	Fed to maintain BCS			+ to BCS 3.0–3.5 then –

^a Unless otherwise stated there are no interactions between nutritional treatments and BCS.

N/S, not stated; +, positive relationship; −, negative relationship.

Ovulation rate

The static and dynamic effects of nutrition on ovulation rate are well established in sheep, such that ewes of greater live weight and/or those offered higher levels of nutrition prior to breeding are more likely to be multiple-bearing (Smith 1991; Scaramuzzi et al. 2006). The data also suggest that the relationship between live weight and ovulation rate is not linear but instead curvilinear, such that for each additional kg of live weight, the relative gain in ovulation rate diminishes (Smith 1991). Further, ewes of heavier live weight are less likely to respond, in terms of an increase in ovulation rate, to improved levels of nutrition (dynamic effect) prior to breeding, than are lighter ewes (Smith 1991). Therefore, it might be expected that as BCS increases, the relative gain in ovulation rate might also be reduced and that ewes of higher BCS will be less responsive to improved nutrition in comparison with ewes of low BCS. These points will be addressed in the following sections where appropriate.

Across studies, a general positive relationship between BCS and ovulation rate occurs (Table 4). In studies where only two BCS categories have been compared, the group with the higher BCS has displayed a higher ovulation rate (Gunn et al. 1972 [Scottish Blackface]; Gunn & Doney 1979 [Cheviot]; Rhind et al. 1984a, b [Scottish Blackface and Scottish Blackface crossbred ewes]; Rhind & McNeilly 1986 [Scottish Blackface]; McNeilly et al. 1987 [Scottish Blackface]; Forcada et al. 1992 [Rasa Aragonesa]; Vinoles et al. 2002 [Polwarth]). An across-flock analysis in Merinos showed a positive relationship between mean flock BCS and flock ovulation rate (Kleemann & Walker 2005).

Ovulation rate increased in Cheviot ewes up to a BCS of 2.5–2.75, with no further increase in ovulation rate at higher BCS (Gunn et al. 1991a). Similarly, ovulation rate peaked at BCS 2.5 in Cheviot ewes, but continued to increase with BCS in Welsh Mountain ewes (Gunn et al. 1991b), suggesting differences in BCS optima between the breeds. Similarly, the ovulation rate in Beulah ewes increased with BCS, but in Cheviot ewes it increased only to BCS 2.25–2.5 (Gunn et al. 1988). These data lead to the conclusion that, similar to the static effect, for some breeds at least there is likely to be a BCS level above which ovulation rate does not increase further.

Russel (1984b) postulated that the effect of BCS on ovulation might be diminished in highly fecund breeds such as the Finn and Finn crosses. At lower BCS, there was no difference in ovulation rate between North and South Country Cheviot ewes, but at higher BCS, North Country Cheviots produced more multiple ovulations than South Country Cheviot ewes (Gunn & Doney 1979).

Increasing the level of feeding had a positive effect on ovulation rate in Scottish Blackface ewes with low BCS, but had no effect in ewes with higher BCS (Gunn et al. 1969). This suggests that there is a threshold BCS, above which feed intake has no effect on ovulation rate and below which, feed intake is an important determinant of ovulation rate. However, in a similar study, Gunn & Doney (1975) found no interaction between the level of food intake prior to mating and BCS for ovulation rate of Scottish Blackface ewes, contradicting the earlier findings of Gunn et al. (1969). Similarly, Gunn et al. (1988) reported no interaction between feeding level and BCS for ovulation rate in Beulah ewes. However, the same authors reported that, in Cheviot ewes, a low level of feeding reduced the positive effect of high BCS on ovulation rate.

Gunn et al. (1988) reported that the ovulation response to immunisation against androstenedione (Fecundin®) was greater in lower BCS Beulah and Cheviot ewes than in ewes of higher BCS. They concluded that under conditions of high BCS, there was little advantage from using Fecundin®. This finding also supports the concept of the dynamic effect in which ewes of greater live weight are less likely to respond (Smith 1991; Scaramuzzi et al. 2006).

Conception rates and returns to service

Due to the design of many studies, it is not possible to determine if a ‘return to service’ was caused by a lack of fertilisation (conception) or early embryonic failure (which will be discussed in a later section). The effects of BCS on conception rates and returns to service have been combined in this section of the review and individual results are summarised in Table 4. Gritstone and Masham ewes of BCS 2.5 displayed lower conception rates than those of higher BCS (Bastiman 1972). Scottish Blackface ewes with BCS of 1.5 were more likely to return to service than those with a BCS of 3.0 (Gunn et al. 1972), whereas Romney ewes with BCS of 2.0 or less and Romney composite ewes with BCS of 1.5 were more likely to return to service than ewes of higher BCS (Kenyon et al. 2004a). Combined, these data suggest that there is an effect of BCS on return to service, but that there may be genotype differences for the minimum BCS and the rate of return to service. In support of this hypothesis, the conception rate of Welsh Mountain ewes increased with BCS, but in Cheviot ewes peaked at a BCS of 2.5 (Gunn et al. 1991a).

While the data presented above suggest that there is a minimum BCS above which the return to service rate decreases, there is also some evidence that there is an upper limit to BCS above which conception rates can decrease. The conception rate was higher in Malpura ewes with BCS of 3.0–3.5 than it was in ewes of both lower and higher BCS (Sejian et al. 2009). In Cheviot ewes, conception increased with BCS up to 2.5–2.75, but declined above that level (Gunn et al. 1991b). Similarly, the return to service rate of Romney composite ewes with BCS 3.5–4.0 did not differ from those of BCS 1.5, whereas the return to service rate was lower in ewes with BCS of 2.0–3.0 compared with BCS 1.5 (Kenyon et al. 2004b).

In Cheviot ewes, low BCS, in combination with poor levels of feeding, reduced the conception rate, while at higher BCS, the level of feeding had no influence on conception rate (Gunn et al. 1991b). The implication of those findings is that some of the gain in conception rate in moving a ewe from low to moderate BCS can be achieved by short-term nutrition even before BCS increases because there is a high sensitivity to pre-breeding nutrition below a BCS of 2.25, at least in Cheviot ewes (Gunn et al. 1991b).

Ova loss and embryo mortality

In most studies where ovulation rate was measured, the percentage of corpora lutea without viable embryos (also termed ova loss) has been used as a proxy measure of embryo mortality (Gunn et al. 1972; Gunn & Doney 1975, 1979; Rhind et al. 1984a, b) and therefore, in this section, ova loss and embryo mortality studies have been combined and may help to explain the results observed in the previous section. Although some studies have found no effect of BCS on embryo mortality, for example in Merino and Merino cross ewes across a range of BCS (Cumming et al. 1975), and when comparing Scottish Blackface ewes with BCS of 1.5–1.75 with those of 2.75–3.0 (Rhind et al. 1984a), most other studies report that embryo mortality does vary with BCS. For example, Scottish Blackface ewes with low BCS (1.5) had higher embryo mortality in the early stages of pregnancy than ewes with higher BCS (3.0), but that BCS had no effect after day 26 of pregnancy (Gunn et al. 1972). Similarly, Gunn & Doney (1975) compared Scottish Blackface ewes across a range of BCS (1.5–3.0) and found that embryo loss decreased as BCS at breeding increased. Supplemental feeding had no effect on embryo loss in animals with a BCS of 1.5, but at moderate BCS (2.5), high levels of feeding reduced embryo loss. Interestingly, in another study, embryo survival was lower in Scottish Blackface cross ewes of high BCS versus moderate BCS (3.4 vs. 2.7, Rhind et al. 1984b). A recent study showed that housed Ossimi ewes of BCS of 1.5 and 4.0 were more likely to lose pregnancies than ewes of the intermediate BCS (Abdel-Mageed 2009). Combined, these studies suggest that both low and high BCS can be detrimental to embryo survival, which is not unexpected given the similar relationship previously observed regarding conception and return to service rates. Relevant to the increased embryo mortality in ewes of high BCS are the findings of Parr (1992), who found that ewes fed well above maintenance displayed lower progesterone concentrations and were less likely to maintain their pregnancy. They proposed that the mechanism was increased liver blood flow, leading to increased clearance of progesterone from the circulation (Parr 1992).

There is also evidence of differences between genotypes in the relationship between BCS and ova loss. Ewes of lower BCS had higher rates of ova loss in North Country Cheviots, but there was no effect of BCS on embryo loss in South Country Cheviots ewes (BCS 2.0 vs. 3.0, Gunn & Doney 1979). However the results are somewhat confounded by the differences in ovulation rates between the breeds and this needs to be considered.

Pregnancy rates

Individual studies have examined the relationship between BCS and either barrenness, fertility, pregnancy rate or lambing rate. For the purpose of simplicity in this section, we have combined these findings into pregnant/non-pregnant. In studies under commercial feeding conditions, where it appears that a range of BCS was examined, a positive relationship between BCS and pregnancy rate has been reported in Manchega (Molina et al. 1994), Barbarine (Atti et al. 2001) and Merino (Kleemann & Walker 2005) breeds. Similarly, Romney and Kivircik ewes of BCS of 2.0 and 2.5, respectively, at breeding, were more likely to get pregnant than ewes of lower BCS (Kenyon et al. 2004b; Yilmaz et al. 2011). Analysing the same issue another way, pregnant ewes had a higher mean BCS than their non-pregnant counterparts in Kurdi (Esmailizadeh et al. 2009) and in various breeds (Gonzalez et al. 1997). In manipulative studies, Gunn & Doney (1979) reported that when Cheviot ewes were fed to maintain a set BCS, those ewes with a BCS of 3.0 had higher pregnancy rates at first mating than ewes with a BCS of 2.0.

These studies show that the pregnancy rate is improved when ewes have a BCS that exceeds a minimum threshold. The question then becomes whether further increases in BCS are worth pursuing. In the studies reported above with Romney and Kivircik ewes, there was no further improvement in pregnancy rate above BCS 2.5 (Kenyon et al. 2004b; Yilmaz et al. 2011). In Chokla ewes, the pregnancy rate increased as BCS went from 2.5 to 3.0, but did not increase any further in ewes of BCS 3.5 (Maurya et al. 2009). In Malpura ewes fed to maintain BCS, pregnancy rate increased up to BCS 3.0–3.5, before declining in ewes with a BCS of 4.0 (Sejian et al. 2009). However, the number of Ossimi lambs per ewe mated increased when the BCS of the ewe increased up to 2.5, then plateaued before beginning to decline above BCS 3.5 (Abdel-Mageed 2009).

The previous section indicates that both low and high BCS have the potential to negatively influence ovulation rate, ova and embryonic loss, conception rates and returns to service, all of which are components of whether a ewe will be pregnant. Therefore, it may not be surprising to suggest that a similar relationship would be observed for pregnancy rate. This may help to explain why no difference in pregnancy rates was observed between Masham ewes fed to maintain BCS of 2.0 or BCS of 4.0 (Newton et al. 1980).

Number of embryos/fetuses per ewe

A positive relationship between flock BCS and the number of fetuses per Merino ewe was demonstrated using an across-flock comparison (Kleemann & Walker 2005, Table 5). However, in Beulah and Welsh Mountain ewes, there was a positive relationship between the number of embryos per ewe and BCS, but not in Cheviots (Gunn et al. 1998, 1991a). Gunn et al. (1991a) reported that the number of embryos in Welsh Mountain ewes increased with BCS, but in Cheviot ewes it peaked at BCS 2.5. These results suggested that the BCS objectives may need to be different for different breeds, with some breeds having a lower desirable BCS than other breeds. In support of that conclusion, there was an increase in the number of fetuses per ewe at low BCS in both Romney and Romney composites, but there was no increase in fetuses per ewe when the Romney ewes exceeded a BCS of 3.0, while in Romney composites there was no further increase above a BCS of 2.0.

Table 5 Summary of studies examining the relationship between BCS and the number of embryos/fetuses, number of lambs born and lamb survival.

Reference	When BCS recorded and range tested	Nutritional treatment(s) during examination period^a	BCS and number of embryos/fetuses per ewe relationship	BCS and number of lambs born relationship	BCS and lamb survival relationship
Gunn et al. (1969)	Breeding, 1.5 and 3.0	Low, maintenance, high		+
Pollott & Kilkenny (1976)	Breeding, BCS range not stated	N/S		+
Adalsteinsson (1979)	Pre-breeding, 2.0 to 4.0	Commercial conditions		+ to BCS 3.0–3.5
Newton et al. (1980)	Breeding, 2.0 and 4.0	Fed to maintain BCS		+
Gunn et al. (1983)	Pre-breeding, ≤2.25, 2.5–2.75, ≥3.0	Low, high		BCS 2.5–2.75 greater than BCS ≤2.25 and ≥3.0
Rhind et al. (1984b)	Breeding, 2.75, 3.0, 3.25, ≥3.5Pre-breeding, 2.5–3.0 and 3.25–3.75	N/SFed to maintain BCS	–	–
Gunn et al. (1988)	Breeding, ≤1.5, 1.75–2.0, 2.25–2.5 and ≥2.75	Low, high	+ in one breed, NR in second breed
Gunn et al. (1991b)	Pre-breeding, ≤2.25, 2.50–2.75 & ≥3.0	Maintenance, high		BCS 2.50–2.75 greater than ≤2.25 and ≥3.0
Gunn et al. (1991a)	Pre-breeding, ≤2.25, 2.5 and ≥2.75	Low, high	In high BCS + to 2.5, no effect low feeding
Al-Sabbagh et al. (1995)	Pre-lambing, BCS 2.5, 3.0, 3.5	High			NR
Gonzalez et al. (1997)	Breeding, 2.0, 2.5, 3.0, 3.5 and 4.0	Commercial conditions		+
Litherland et al. (1999)	Pre-lambing, 1.5 and 2.5	Low, high			+ in one of two studies
Atti et al. (2001)	Pre-breeding, BCS range not stated	Commercial conditions		+ to BCS 3.5–4.0
Kenyon et al. (2004b)	Breeding, 1.5 to 4.0	Commercial conditions	+ to BCS 2.0 in one breed and + to BCS 3.0 in second breed
Oregui et al. (2004)	Breeding, ≤1.75, 2.0–2.25, 2.5–2.75, 3.0–3.25, ≥3.5	Commercial conditions		+ to BCS 2.5–2.75
Kleemann & Walker (2005)	Breeding, BCS range not stated	Commercial conditions	+	+	+
Rozeboom et al. (2007)	Pre-lambing, 1.5 to 3.5	N/S		NR
Everett-Hincks & Dodds (2007)	Mid-pregnancy, BCS range not stated	Commercial conditions			+
Abdel-Mageed (2009)	Pre-breeding	Maintenance		+ to BCS 2.5 then – after for BCS 4.0
Kenyon et al. (2011)	Mid-pregnancy, ≤2.0, 2.5 and ≥3.0	Medium, high			BCS 2.5 lower than ≤2.0
Oldham et al. (2011)	Day 100 of pregnancy, 2.0 and 3.0	Various feeding levels			NR
Kenyon et al. (2012a)	Mid-pregnancy, 2.0, 2.5 and 3.0	Medium, high			BCS 2.5 lower than 2.0
Kenyon et al. (2012b)	Mid-pregnancy, 2.0, 2.5 and 3.0	Medium, high
Aliyari et al. (2012)	Pre-breeding, 2.0, 2.5, 3.0 and 3.5 +	Ad libitum		NR

^a Unless otherwise stated there are no interactions between nutritional treatments and BCS.

NR, no relationship or no effect; N/S, not stated; +, positive relationship; −, negative relationship.

Number of lambs born per ewe

There have been reports that the number of lambs born per ewe is independent of ewe BCS in Merino (McInnes & Smith 1966), Suffolk (Rozeboom et al. 2007), Afshari (Aliyari et al. 2012) and various other breeds (Geisler & Fenlon 1979, Table 5). In contrast, other studies report a positive effect of BCS on the number of lambs born in Scottish Black Face (Gunn et al. 1969), Masham (Newton et al. 1980) and various other breeds (Pollott & Kilkenny 1976).

While it is conceivable that the variance between studies may be due to differences in breed, another possible explanation is that the positive relationship between BCS and the number of lambs born is not linear and instead curvilinear. If this is the case, the outcome reported in studies depends on the ewe BCS range examined. A plateau has been reported in the relationship between ewe BCS and lambs born in Icelandic (Adalsteinsson 1979), Latxa (Oregui et al. 2004) and Lori-Bakhtiari (Vatankhah et al. 2012) breeds. To complicate matters further, it has been reported that high BCS is associated with a decline in the number of lambs in Scottish Black Face (Rhind et al. 1984b) and Afshari (Aliyari et al. 2012) breeds, whereas the lambing rate in Cheviot ewes increased up to a BCS of 2.50–2.75 before decreasing, especially when the ewes were grazing lower pasture covers (Gunn et al. 1991b). This apparent decrease in lambing rate at higher BCS has also been observed in other breeds. For example, in Barbary ewes, the number of lambs born per ewe increased up to a BCS of approximately 3.5, before beginning to decline (Atti et al. 2001). Similarly in Ossimi ewes, there was a positive relationship between BCS and lambing rate up to BCS 2.5, before a plateau and then a decline above BCS 4.0 (Abdel-Mageed 2009).

Combined, these studies suggest a curvilinear relationship between BCS and the number of lambs born, indicating there may be a tipping point for each breed above which there is a negative effect of BCS on lambing rate. Gunn et al. (1991b) concluded that the lower reproductive performance of ewes with a BCS of 3.0 and above was related to an increase in ova and embryo loss (see earlier sections) and may be associated with reduced feed intake. However, it is also important to note that Pollott & Kilkenny (1976) concluded that BCS accounted for only a very small percentage (R² = 2%) of the total variation in number of lambs born.

Effect of ewe BCS on the lamb until weaning

Fetal growth and birth weight

In the majority of studies, BCS has had no effect on lamb birth weight, but there is large variation between studies. It is probable that this variation is due to differences in the timing of BCS measurement, the number of fetuses per ewe and maternal nutrition (see Table 6). Fetal growth and lamb birth weight have not been related to BCS measured pre-breeding in Scottish Halfbred and Afshari ewes (Hossamo et al. 1986; Aliyari et al. 2012, respectively), in mid-pregnancy with Scottish Halfbred, Romney and Coopworth cross ewes (Gibb & Treacher 1982; Kenyon et al. 2011, 2012a; Verbeek et al. 2012, respectively) and in late pregnancy with Scottish Halfbred and Polypay ewes (Gibb & Treacher 1980; Al-Sabbagh et al. 1995, respectively). In Merino ewes, BCS at breeding had no impact on fetal or placental size (McNeill et al. 1997a, b). Similarly, BCS had no effect on fetal weight in early gestation, even under conditions where Welsh Mountain ewes were fed less than maintenance (Osgerby et al. 2003).

Table 6 Summary of studies examining the relationship between BCS and lamb birth and weaning weight and lamb growth to weaning.

Reference	When BCS recorded and range tested	Nutritional treatment(s) during examination period^a	BCS and lamb birth weight relationship	BCS and lamb growth relationship	BCS and lamb weaning weight relationship
Gibb & Treacher (1980)	Pre-lambing, 2.4 and 3.2	Low, High	NR	+
Gibb & Treacher (1982)	Day 90 pregnancy, 2.6 and 3.3	Low, High in pregnancy, high in lactation	NR	NR
Wilson et al. (1985)	Pre-lambing, 1.0 to 3.5	N/S		+ to BCS 1.5
Hossamo et al. (1986)	Pre-breeding and pre-breeding, 1.0-3.5	Commercial	NR/ +	NR	NR
Molina et al. (1991)	Pre-lambing, <2.5, 2.5–3.0, >3.0		+		+ to BCS > 3.0
Al-Sabbagh et al. (1995)	Pre-lambing, 2.5, 3.0, 3.5	High	NR		NR
Atti et al. (1995)	Pre-lambing, <2 and >3	Maintenance		+
Litherland et al. (1999)	Pre-lambing, 1.5 and 2.5	Low, high		NR	NR
Kenyon et al. (2004a)	Breeding, 1.5 to 4.0	Commercial conditions	BCS 3.5–4.0 < 3.0	+
Alvarez et al. (2007)	Pre-breeding	Commercial conditions		+ for twins, + and NR for singletons
Everett-Hincks & Dodds (2007)	Mid-pregnancy, BCS range not stated	Commercial conditions	+
Maurya et al. (2009)	Breeding, 2.5, 3.0 and 3.5	Commercial conditions	+
Sejian et al. (2009)	Pre-breeding, 2.5, 3.0–3.5 and 4.0	Fed to maintain BCS	+		+ to BCS 3.0–3.5
Kenyon et al. (2011)	Mid-pregnancy, ≤2.0, 2.5 and ≥3.0	Medium, high	NR		BCS ≤2.0 lower than 2.5
Oldham et al. (2011)	Day 100 of pregnancy, 2.0 and 3.0	Various feeding levels	+ in two of four studies
Kenyon et al. (2012a)	Mid-pregnancy, 2.0, 2.5 and 3.0	Medium, high	NR		+ to BCS2.5
Kenyon et al. (2012b)	Mid-pregnancy, 2.0, 2.5 and 3.0	Medium, high	NR		+ to BCS2.5
Verbeek et al. (2012b)	BCS mid pregnancy, 2.0, 2.9, 3.7	Fed to maintain BCS	NR		NR
Aliyari et al. (2012)	Pre-breeding, 2.0, 2.5, 3.0 and 3.5 +	Ad libitum	NR		NR

^a Unless otherwise stated there are no interactions between nutritional treatments and BCS.

NR, no relationship or no effect; N/S, not stated; +, positive relationship.

On the other hand, there are numerous studies that have reported a positive relationship between BCS and lamb birth weight. These include studies where BCS was measured at breeding (Maurya et al. 2009; Sejian et al. 2009), mid-pregnancy (Everett-Hincks & Dodds 2007) and in late pregnancy (Hossamo et al. 1986; Molina et al. 1991). Also, in two of four studies, Merino ewes with a BCS of 3.0 at day 100 of gestation gave birth to heavier lambs than ewes with a BCS of 2.0 at day 100 (Oldham et al. 2011). In contrast, Romney ewes with a BCS of 3.5–4.0 at breeding gave birth to lambs lighter than those born to ewes with a BCS of 3.0 at breeding (Kenyon et al. 2004a).

In late pregnancy, the nutritional demand for the ewe, especially the multiple-bearing ewe, is significantly increased (Nicol & Brookes 2007). Under conditions where the ewe cannot meet the increased nutritional demand via intake, she must utilise body reserves. Therefore, it might be expected that the impact of BCS on fetal growth and lamb birth weight would be greatest in late pregnancy, especially in situations where ewe nutrition is limited.

Lamb survival

Ewe BCS has been reported to have either no effect on lamb survival to weaning (Al-Sabbagh et al. 1995; Oldham et al. 2011) or a positive effect (Litherland et al. 1999; Everett-Hincks & Dodds 2007, Table 5). It should be noted that in the study of Al-Sabbagh et al. (1995), a low number of lambs probably limited the ability of that study to detect an effect on lamb survival. In contrast, a positive curvilinear relationship between BCS and singleton lamb survival, with a diminishing response as BCS increased above 3.0 has been reported in Merino ewes (Kleemann & Walker 2005). In twins, the relationship remained linear. Combined, these studies lead to the conclusion that BCS could be used as a management tool to ensure adequate lamb survival. However, recently it has been reported that under ad libitum pastoral feeding conditions, multiple-born lamb survival rates were lower in lambs born to ewes with a starting BCS of 2.5, compared with those born to ewes with starting BCS 2.0 (Kenyon et al. 2011, 2012a). Intriguingly, the survival of lambs born to the BCS 2.0 group did not differ from those born to ewes with a BCS of 3.0, and lamb birth weight was not affected by BCS at all. The reasons behind these findings are unknown.

A lack of vigour in lambs born to ewes of low BCS is a potential mechanism to explain lower lamb survival in lambs born to ewes of low BCS. In fact, lambs born to Merino ewes of low body condition took longer to suckle post birth and suckled for a shorter period than lambs born to ewes of better condition, these behaviours being linked to lamb vigour (Banchero & Quintans 2007).

Lamb growth to weaning and live weight at weaning

Ewe BCS has been reported to have either no influence on lamb growth to weaning (Gibb & Treacher 1982; Litherland et al. 1999; Thompson et al. 2011) or weaning weight (Al-Sabbagh et al. 1995; Litherland et al. 1999; Aliyari et al. 2012; Verbeek et al. 2012), or a positive effect on lamb growth (Gibb & Treacher 1980; Kenyon et al. 2004a, Kenyon et al. 2011; Mathias-Davis et al. 2013) and weaning weight (Molina et al. 1991) (see Table 6). Again this variation between studies may be due to differences in the timing of the BCS measurement, the absolute levels of BCS being compared, the feeding level and feed quality being offered, and the number of lambs born and reared per ewe. In the following paragraphs, we attempt to synthesise these studies.

In Awassi ewes, the pre-breeding BCS had no effect on lamb growth or weaning weight, while the BCS pre-lambing had a positive effect on both outcomes (Hossamo et al. 1986), a finding that suggests that the timing of BCS measurement is important. In Corriedale ewes, the BCS measured pre-breeding was related to twin but not to singleton lamb growth (Alvarez et al. 2007). A greater reliance on body reserves to meet the metabolic needs for lactation is conceivable in a twin-bearing ewe, which is likely to explain this relationship (also see later milk section). Gibb & Treacher (1980) reported that ewe BCS affected lamb growth, particularly when ewe intake of herbage was reduced by high stock density. In contrast, in a similar study when the ewe was provided with a larger nutritional allowance in lactation, there was no effect of ewe BCS on lamb live weight (Gibb & Treacher 1982).

Lambs born to Malpura ewes, with a BCS of 2.5 at breeding, were lighter at weaning than those from ewes with a BCS of 3.0–3.5 or 4.0, although the lamb weights of the latter two groups did not differ at weaning (Sejian et al. 2009). Similarly, under unrestricted feeding conditions, the weight at weaning of multiple-born Romney lambs increased with ewe BCS in mid-pregnancy to a BCS of 2.5, with no further gain above this (Kenyon et al. 2011, 2012a, b). In a study in which the breed was not stated, ewes with a BCS of 3.0–3.5 at pregnancy scanning or at weaning, weaned a larger total weight of lambs than ewes with BCS > 3.5 (Mathias-Davis et al. 2011). However, with triplet-rearing ewes, those with a BCS of 3.0 or higher weaned a higher total weight of lambs than ewes with a BCS of less than 3.0. Similarly, Everett-Hincks et al. (2013) reported BCS of triplet-bearing ewes had a positive effect on their offsprings’ performance to weaning. While these combined results show there is a general positive relationship between ewe BCS and lamb growth, they also suggest that there may be a plateau above which no further gains are made, and that the number of lambs a ewe rears might affect the relationship observed.

Milk production

Given the general positive effect of BCS on lamb growth outlined above, it may be expected that ewes of low BCS would produce less milk than ewes of higher BCS. Similar to the effects of BCS on other performance traits, the reports relating BCS to milk production have been inconsistent. There have been reports of no effect on milk production of BCS measured in Scottish Halfbred ewes at breeding and in mid-pregnancy (Hossamo et al. 1986; Gibb & Treacher 1982, respectively) and in Latxa ewes in late pregnancy (Oregui et al. 2004). In contrast, a positive effect of BCS on milk production has been reported when BCS was measured in late pregnancy in Awassi (Hossamo et al. 1986) and Scottish Halfbred (Gibb & Treacher 1980) breeds. In early lactation, up to a third of the milk produced by a ewe is achieved by mobilising body fat and protein reserves (Cannas 2002). Given that BCS is likely to change in a given ewe during pregnancy, it is not too surprising that Hossamo et al. (1986) reported that pre-lambing BCS had a positive effect on milk yield and lactation length, whereas BCS measured pre-breeding had no effect.

Another confounder between studies is differences in feeding levels. Under conditions in which ewes lost live weight during lactation, ewes of a higher BCS tended to produce more milk than those of low BCS (Gibb & Treacher 1980). When lactating ewes had access to adequate herbage, BCS had no impact on milk production (Gibb & Treacher 1982). At low levels of feeding, the milk production of ewes with less body fat was more affected than it was in ewes with greater body fat (Cannas 2002). At higher feeding levels, there was little effect of body fat on milk production, mirroring the findings of Gibb & Treacher (1980, 1982).

In contrast to the general picture emerging from these studies, at maintenance feeding, the milk production of Sarda ewes of higher BCS in mid-lactation was less than it was in ewes of lower BCS (Pulina et al. 2012). When feed was restricted, milk production was similarly negatively affected in both ewe BCS groups. Cannas (2002) wondered if there was the potential for very fat ewes to have low milk production, due to high quantities of visceral fat compressing the rumen, thereby reducing intake. If that is the case, it could help explain the results of Pulina et al. (2012). Alternatively, it could be hypothesised that ewes with high BCS in mid-lactation are those that are metabolically programmed not to partition body reserves towards milk production, explaining the lower production.

Ewes with a BCS of 2.5–3.5 tended to produce more colostrum than ewes with a higher or lower BCS (Al-Sabbagh 2009). However, in both Polypay ewes (Al-Sabbagh et al. 1995) and Suffolk ewes (Rozeboom et al. 2007), BCS has no effect on colostrum IgG concentrations. Combined, these few data suggest that BCS may have little influence on colostrum production.

Ram performance

There is sparse literature on the impact of BCS on the reproductive performance of rams. Early reports suggested that either obesity or long-term under-nutrition (and therefore low BCS) had deleterious effects on sperm quality (McKenzie & Phillips 1934) and fertility (Mori 1959). More recently, in a semi-arid environment, rams with a BCS of 3.0 had the highest reproductive performance, with poorer performance observed at very low and high BCS, due to reduced sperm motility and an inhibition of sexual behaviour (Maurya et al. 2010). This report contrasts with earlier recommendations that rams should be of BCS 4.0 just prior to the mating season and maintained at not less than BCS 3.0 for the remainder of the year (Jefferies 1961). It has been demonstrated that a short-term increase in the plane of nutrition (to at least 150% of maintenance requirements) increases body size, BCS, testicular size and sperm production (Martin et al. 2010). It is not clear if the increase in sperm production was related directly to the increase in nutrition or to the consequent increase in BCS (Blache et al. 2003).

Animal welfare and BCS

Although this review focuses mainly on the relationships between BCS and productivity it would be inappropriate to ignore potential impacts on animal welfare (AW). It has been suggested that BCS is good indicator of animal welfare (Morgan-Davies et al. 2008; Phythian et al. 2011). The Sheep Code of Welfare in New Zealand recommends that all adult sheep should be in the BCS range of 3.0–4.0 at all times (Anon. 2010). Further, the welfare code in New Zealand indicates that any sheep with a BCS of 1.0 or below requires urgent remedial action to improve its condition, or that it should be destroyed humanely. Interestingly, it is not clear, from reading these guidelines, the decision process behind them and the source of the material used to develop these. Below, we have attempted to briefly outline the potential issues with using BCS as a welfare indicator and some of the known relationships between BCS and AW.

Like any other AW indicator, the use of BCS to define AW has limitations (Keeling et al. 2011). First, AW could be considered satisfactory if the health, production and reproduction of an animal are not compromised. As described in this paper, BCS is generally positively related to production capacity, although the relationship may not be linear for all traits. Thus simply suggesting higher BCS means greater AW may not be correct.

Second, AW can be assessed based on the five freedoms (hunger and thirst, health, environmental, behaviour, mental; [Farm Animal Welfare Council 1992]). The data of Sibbald & Rhind (1997) indicate greater BCS is associated with reduced intake, which does not simply suggest they are not hungry. The results of Caldeira et al. (2007) propose that metabolic imbalance can occur with a BCS below 2.0 or above 4.0, which may be suggestive of reduced health outcomes for ewes in these ranges. The survival of Scottish Blackface ewes within the flock has been shown to increase up to BCS 2.5, but above this level no further gain is apparent (Morgan-Davies et al. 2008; Annett et al. 2011). On one of three farms, periodontal disease was associated with lower BCS (Orr & Chalmers 1988) indicating poor teeth can have implications for a ewe's ability to maintain body condition. Loss of body condition in pregnancy has been associated with increased concentrations of beta-hydroxybutyrate, an indicator of metabolic stress (Everett-Hincks et al. 2004). This potentially indicates regular measuring of BCS could be used as a tool to determine the apparent health status of a ewe in pregnancy.

At extreme BCS, animals can have problems expressing normal behaviour, for example, obese rams might not be able to mount a female, or will injure her when mounting, because of the weight difference between the two sexes (Mori 1959). While these briefly outlined studies do suggest AW issues can occur at both low and high BCS, the data are not that easily interpreted. Further information is required before clear conclusions can be drawn. It would be of benefit for more data to be collected examining the impact of BCS on ewe and ram health status, survival rates and therefore longevity in large numbers of animals under commercial grazing conditions.

Potential trade-off between greater BCS and financial efficiency

In this review, we have outlined across some breeds at least, that it is unlikely that a simple linear relationship between BCS and productive performance exists. Instead, we suggest that the relationship is curvilinear for many traits. If this relationship is accepted, then it would hold that above some threshold, the gain in production for each additional unit gain in BCS would diminish. In fact, for some traits there seems to be a point above which no further gain occurs, and at very high BCS there seems to be a point above which a decrease in some production traits can occur. Therefore, a farmer must consider if it is efficient to simply aim for the highest BCS possible. If we assume a standard reference weight of 70 kg, a non-pregnant ewe has a theoretical maintenance requirement of approximately 3605 MJ ME/yr (Nicol & Brookes 2007). The theoretical maintenance requirement of a ewe either 7 kg lighter or heavier, which is equivalent to 1.0 unit change in BCS of a Romney ewe (Kenyon et al. 2004a, b), is 3241 and 3965 MJ ME/yr, respectively. Thus in very simple terms, any gain in production that is achieved by increasing BCS by 1.0 unit (feeding up a 70 kg ewe to a 77 kg ewe) must be large enough to offset the additional 360 MJ ME/yr required for maintenance, plus any other associated increase in nutritional demand with the associated increase in performance (i.e. the additional cost of bearing and rearing multiples above that of a singleton). Therefore, the economic benefit of an additional increase in BCS will depend on the cost and availability of extra feed sources and the value of the additional product produced. This will likely need to be determined on a case by case basis.

Is there such a thing as the optimal BCS?

Based on the data reviewed here, it is unlikely that a single BCS could be termed optimal from either a biological or economic perspective. The BCS that is associated with the highest productivity varies depending on physiological state (non-pregnant vs. pregnant vs. lactating), the number of offspring a ewe is bearing and rearing, the age of the animal, its sex, its breed (genotype), the production system, and the level and quality of the feed on offer. Below, we have attempted to outline briefly what might be considered a suitable BCS range for a ewe across the production cycle based on the data reviewed here and those of previous publications (Russel 1984a, b, 1985; Cannas 2002; Curnow et al. 2011; Hocking-Edwards et al. 2011), although, as stated above, this will likely depend on a number of variables.

We suggest that, at breeding, a ewe should have a BCS in the range of 2.5–3.5. BCS should be maintained in early pregnancy. In mid- to late pregnancy, ewes will likely lose body condition due to the demands of pregnancy. However, ewes can lose 0.5 to 1.0 units of BCS, depending on their starting point, with minimal impacts on productivity (Russel 1984a, b, 1985). Ideally, ewes would be of BCS 2.5–3.0 at lambing, with an absolute minimum of 2.0. This minimum is important, because of the likely further loss of body condition during lactation. At weaning, ewes should not have a BCS below 2.0 and should not have lost more than 1.0 unit of BCS in lactation (Cannas 2002). Post-weaning, ewes need to be managed appropriately to ensure that the target breeding BCS is met. For highly fecund ewes, BCS at the various stages of production should probably be at the upper end of the recommended target range.

It is difficult to determine the economic impact of a given ewe BCS profile across the whole production cycle due to the potential interactions between many of the factors listed earlier in this section. However, what is clear is that at low BCS, animal performance will be reduced as will potential income. Conversely, consideration must be given to the potential cost of the extra feed required to achieve very high BCS when the additional lift in animal performance may be small and in some cases reduced. Therefore, before aiming for very high ewe BCS, farmers should undertake a detailed cost-benefit analysis, as indicated earlier.

Managing to an average, a minimum or within a range

If it is agreed that a curvilinear relationship exists between BCS and production traits, then the biggest gain can be achieved by reducing the number of ewes with the lowest BCS in a flock, or ensuring that all individuals are above a target minimum. Simply aiming for an average target flock BCS will result in a number of individuals being below the target BCS. With Romney ewes, Kenyon et al. (2004a) justified a ‘minima’ approach over an ‘average’ approach, based on the following logic. Assuming a standard deviation of 0.52 for BCS (Kenyon et al. 2004a), if two groups of ewes with average CS of 3.0 and 3.5 are compared, one would expect to find 50% of ewes in the BCS 3.0 group with a BCS below 3.0, but only 34% of ewes below BCS 3.0 in the BCS 3.5 group. If performance is reduced below a BCS of 3.0, then the higher overall reproductive performance of the BCS 3.5 group of ewes is likely due to fewer individual ewes with a BCS below 3.0 in that group, rather than enhanced reproductive performance per se of having more ewes above a BCS of 3.5 (Kenyon et al. 2004b). Thus, having an average target flock BCS is not the best approach and instead a minimum BCS approach, which animals should not be below, may be a better option.

A potential problem with setting a target minimum BCS is that when feeding the whole flock, with the aim of lifting the ewes below the minimum to the specified target, then due to the variation in BCS within a flock, some ewes will end up well above the minimum. Because of the curvilinear plateauing of the relationship between BCS and productivity, those ewes pushed well above the minimum will likely be heavier and thus less efficient. Furthermore, assuming that there is actually a point for many of the production traits where very high BCS actually results in reduced productivity, then there is a risk of reducing profitability, and possibly welfare, even further. The optimal strategy in economic terms is to reduce the variation in BCS across the flock. One way to achieve that is to use a split flock management system, where ewes are regularly assessed for BCS, and those below the target are offered improved nutrition to gain BCS, while those at or above the target are fed to maintain condition.

While it might be appropriate to randomly weigh a sample of ewes within a flock to obtain an average live weight, this is not a suitable approach for BCS. If BCS is to be used for nutritional management decisions, all ewes need to be individually assessed and managed appropriately. Failure to assess each ewe will not identify those that would benefit from improved levels of nutrition. Conversely, there are stages in the production cycle when there is no benefit from offering ewes of high BCS a level of nutrition to gain BCS; in fact this is likely to be an inefficient use of herbage.

It is recommended that ewes have their BCS assessed at three stages of the production system. First, ewes should have their BCS determined at the weaning of their lambs, allowing ewes of poor BCS to be identified and managed appropriately to gain BCS before rebreeding. Second, in mid-pregnancy, at least multiple-bearing ewes should have their BCS assessed, and ewes of poor BCS offered suitable levels of nutrition to allow for BCS to be gained before the period of increase nutritional need in late pregnancy and lactation. Lastly, ewes should be assessed just prior to lambing, and ewes of lower BCS offered higher herbage allowances in lactation.

Conclusion

The BCS technique has been used in the sheep industry for more than 50 years. It is widely accepted that BCS has many advantages over live weight for determining the condition and well-being of an animal. However, further work is still required to reduce the variability that occurs between assessors and to improve farmers’ use of the technique. It seems that there is enough evidence to suggest that, in some breeds at least, while the relationship between BCS and productivity is positive, it may not be linear and is instead curvilinear. If this is the case, the shape of this relationship means that the biggest gains can be made by reducing the number of low BCS animals within the flock. The optimal BCS is likely to be within the BCS range of 2.5–3.0.

Acknowledgements

The authors would like to acknowledge the input of AN Thompson with helping formulate the discussion of the effects of BCS on ewe performance.