Analysis of the Non-Genetic Factors Affecting the Growth of Segureño Sheep

Abstract

A study was conducted to evaluate the effects of non-genetic factors on the growth behaviour of Segureña sheep breed. Growth related data (early weaning weight, late weaning weight and weight at 80 days of age) were taken from 59,704 lambs belonging to historical data from National Association of Segureño Sheep Breeders (ANCOS) during a period of 11 years. Statistical analyses were performed by using the multifactorial analysis of variance of IBM SPSS Statistics v.19 software. The model included non-genetic factors – lamb sex (S), birth season (N), herd (H), birth year (A) and birth type (P) – as main effects, the dam’s age at lambing and the lamb’s age when weighed as covariables, and the interactions between these factors. Results showed that all weights at developmental stages were significantly (P<0.001) affected by all factors, except for A and the covariable age of dams at lambing on lambs aged 80 days. Double interactions H×A, H×P and H×N were significant (P<0.001) for all variables, as well as the triple interaction H×A×P. Non-genetic factors have a very important role in the development and growth of the Segureña sheep breed, at different ages or growth stages, therefore a correction is necessary to increase the accuracy of direct selection on lamb weight at early weaning, late weaning and at slaughtering (80 days of age).

Key Words:

• Growth
• Interactions
• Multifactorial analysis
• Sex

Introduction

The growth described as a change in volume, size or shape with the passing of time is a very important characteristic of living organisms, especially in production animals as meat sheep. Recording a battery of weights during the animal’s life is difficult to interpret without a way to summarise them; for this reason it is essential to investigate the best traits to be used in sheep meat breeding (Sarti et al., 2001). The use of growth equations provides a good way to summarise information contained in these data into some parameters with biological interpretation (Brown et al., 1972; Fitzhugh, 1976).

Not only is information on live weight important for selection purposes, but the degree of maturity and growth rate throughout the animal’s life are also relevant due to their association with other characteristics and as they can help to optimise production. Studies on sheep growth have generally been based on live weights over a relevant economic time period, becoming subjects of major interest among animal scientists and producers (Bathaei and Leroy, 1998). Knowing which factors influence the growth curve may help management and improvement programmes (McManus et al., 2003).

The Segureña sheep breed represents one of the three pillars of the Spanish sheep meat production of autochthonous breeds. Both in the past and present, it has helped establish the rural population by maintaining centenary livestock farming activities where, until now, transhumant practices adapted perfectly to the glens and paths that flow through most of the country (Hernández, 2004). Mainly exploited in extensive and semi-extensive conditions, these animals are one of the components that balance the ecosystem of the regions they inhabit, thereby making them a mainstay of environmental and social sustainability. Despite this difficult and disadvantaged environment, this breed reached the most competitive and highest production returns (León et al., 2007). Its general appearance is agile and graceful, with a medium-size body (males reach 90 kg and females 60 kg) and a proportional head. Their profile is slightly convex and presents overall elongated shapes. One of the characteristics of the Segureña breed is its good sexual precocity where, in a well-fed group, the first delivery can occur at between twelve and fourteen months. Ewe prolificacy is generally high, with some selected herds reaching 175 animals born from 100 parturitions, although the usual numbers are around 135 or 140 animals per 100 parturitions (Hernández, 2004).

Lambs are suckled by the dams during their first weeks of life. Currently, reproduction remains closely tied to commercialisation and lambs are weaned early, with an interval between births of about eight months, i.e. three parturitions every two years. The lifespan of a breeding ewe is approximately 7 years, during which it can have approximately 10 parturitions.

The lambs of this breed are of high quality for human consumption and are slaughtered when their weight is between 24 and 30 kg. The slaughterhouse return rate is about 51% of harvested meat, owing to a lightweight lambskin that represents 8% of the total live lamb.

In 1997, after the implementation and development of weight control in lambs and the morphological assessment programme, the National Association of Segureño Sheep Breeders (ANCOS) received official support from the Regional Autonomous Governments of Andalucia and Murcia to initiate the corresponding selection and genetic improvement scheme for this breed, with the assistance of the University of Córdoba.

Several authors (Strizke and Whiteman, 1982; Dimsoski et al., 1999; Macedo and Arredondo, 2008; Gbangboche et al., 2011; Momoh et al., 2013) have identified sex and birth type as two factors that have major influence on the growth of sheep, variables that significantly affect the performance of production systems, whose purpose is to obtain the greatest economic profit in the shortest time possible (Tron et al., 2003). Due to market demands, lambs are born year-round, leading several researchers (Strizke and Whiteman, 1982; Souza and Bianchini Sobrinho, 1994; Tron et al., 2003; Hinojosa-Cuéllar et al., 2012) to study the effect of birth season on lamb weights in order to determine if there is in fact any season with better results for lamb weight.

The herd has an effect on lamb growth that is caused by a range of components with overlapping influences, resulting in direct and indirect effects on the animals’ development. Indirect effects are usually the most important for ruminant, such as ambient temperature and factors that control the moisture level in soil which, by affecting plant growth, ultimately influence the quantity and quality of available nutrients. A large part of the same variety of factors determine the microclimate of many microorganisms and vectors, which play an important role in the dynamics of parasitic organisms and therefore in disease levels. Cold, wet and windy conditions, which are most extreme in high altitude areas, cause acute stress and lead to hypothermia in neonates, or pneumonia in suckling and immature lambs. These factors generally do not persist long enough to affect growth in the long term (Charles, 1985).

Due to the lack of information on growth patterns in Segureño lambs, the aim of this study was to characterise and evaluate the effects of the main non-genetic factors, such as sex, birth season, herd, birth year and birth type, on the growth of Segureño lambs, and their interactions on weight and growth parameters in the Segureño sheep breed. This knowledge is crucial when designing the best models for genetic analysis with a minimal computational cost (Lambe et al., 2006).

Materials and methods

The growth data used in this study belongs to historical data from ANCOS collected over the past 11 years in 41 herds. Values where the average weight data was±2*standard deviation in each age were excluded, resulting in a total of 59,704 lambs with 129,377 body weight.

The animals are raised during the day in natural pasture, depending on the hours of daylight and therefore the time of year and weather conditions; sheep are supplemented in the last month of gestation and during lactation (usually cereals such as barley) and during the dry season all animals are supplemented. Usually animals are sheltered overnight. Births occur year-round, but mainly in three seasons: January (replacement), May (replacement/sale) and August (sale). Lambs are left with dams until aged 45 days, from this age they are sent progressively for fattening aged around 60 days until reaching slaughter age.

Information considered age and weight at three developmental stages: P1) early weaning (stage considered between 16 and 35 days), with 20,851 observations; P2) late weaning (stage considered between 36 and 55 days), with 17,872 observations; and P3) final weight (stage considered between 56 and 80 days), with 14,216 observations.

Lambs from triple or higher birth type were clustered on one level denominated triple or higher, due to their low frequency, thus the birth type classification was established on three levels: single, double and triple or higher.

Birth dates were grouped into four seasons: from March 21 to June 20 (spring); from 21 June to 22 September (summer); from September 23 to December 20 (autumn); from December 21 through March 20 (winter).

As a preliminary study, a statistical analysis using the univariate general linear model from the statistical package IBM SPSS Statistics v.19 was performed to analyse the volume of variance explained by the pure effect of the isolated factors on each variable (P1, P2 and P3). Statistic model, using the non-linear regression from IBM SPSS Statistics v. 19, included 5 fixed effect factors: sex (2 levels), birth season (4 levels), herd (38 levels for P1, 35 levels for P2 and 24 levels for P3), year (11 levels) and birth type (3 levels) and 2 covariates: lamb’s age at weighing and age of dam at lambing. The model fitted was:

Y=μ+S+N+H+A+P+SN+SH+SA+SP+NH+NA+NP+HA+HP+AP+SNH+SNA+SNP+SHA+SHP+SAP+NHA+NHP+NAP+HAP+SNHA+SNHP+SHAP+SNHAP+E+M+ε

where Y=weight of a lamb at age A;

• µ=mean;

• S=sex (1,2);

• N=birth season (1..4);

• H=herd (1..38);

• A=year (1..11);

• P=birth type (1..3);

• E=lamb’s age (days);

• M=dam’s age at lambing (year);

SN, SH, SA, SP, NH, NA, NP, HA, HP, AP, SNH, SNA, SNP, SHA, SHP, SAP, NHA, NHP, NAP, HAP, SNHA, SNHP, SHAP, NHAP, SNHAP=interactions between factors; ε=residual error.

The dam’s age at lambing, in years, and the lamb’s age at weighing, in days, were used as covariates to correct the phenotype observation of weaning weight. This is because lambs are not all born on the same day but are weighed together, which means they are weighed at different ages. Comparison of means was performed a posteriori by Tukey test, setting P<0.05 to identify significant differences between treatments.

Non-significant interactions (P>0.05) were removed from the model in each developmental phase, and the data were re-analysed with a model including only the main factors, covariates, and significant interactions.

Results and discussion

Weight recording is essential in meat breeding, because these traits constitute important selection criteria in the improvement programmes. These data are not only useful in genetic purposes, they are also important to develop economic prognostics and evaluations, and to make prevision on feeding, reproduction and other farming activities. Weight recording is expensive and complex and for this reason some successful alternative has been proposed in meat sheep by using simple characters highly related to the lamb growth (Sarti et al., 2003).

In the present study an important amount of data belonging to the meat recording programme of the Segureña sheep breed has been used to develop a deep study of the physiological growing performance of the breed, taking into account the most important non genetic sources of variation.

As shown in Table 1, the sex effect on lambs weight increases as the animal develops (1.4, 2.8 and 5.2% for P1, P2 and P3 respectively), showing the impact of sexual dimorphism, as found in the study in Navajo sheep performed by Eltawill et al. (1970) and in contrast to the finding from Sarti et al. (2003), in Appenninica and Merinizzata sheep breeds; regarding birth type, the effect decreases as the lamb develops (9.8, 7.4 and 7.3% for P1, P2 and P3 respectively) (Table 1).

The analysis of variance was used to evaluate the influence of several non-genetic sources of variation and interactions between these factors on weight at different developmental stages, with a view to accesses to the physiological growth of the Segureña lambs under the influence of different non genetic factors and their interactions. General results of the ANOVAs are shown in Table 2. The main effect of the factor year and age of dam at lambing covariable are non-significant at P3 (P>0.05). The interaction between all five factors is non-significant (P>0.05) for all developmental stages while some of double, triple and quadruple interactions are significant (P<0.05) for the three variables under study. The reduced determinative coefficient of the model that includes only the significant main effects, covariables and interactions, ranges between 44.3 and 48.9%.

Sex

Sexual dimorphism was evident in the breed: lamb weights significantly different between males and females (P<0.001) as has been pointed out in Table 2. This fact has been also reported by Macedo and Arredondo (2008), Baneh and Hafezian (2009), among other authors. As shown by numerous authors (Strizke and Whiteman, 1982; Dimsoski et al., 1999; Rodríguez et al., 1999; González et al., 2002; McManus et al., 2003; Ulutas et al., 2010; Gbangboche et al., 2011) males presented a higher average weight compared to females. All the literature consulted reflected a significant effect of the sex, with higher weight recorded for males when compared to females. However, other studies performed in different races and under different production systems indicated no differences attributed to sex for preweaning growth (González et al., 2002) and some authors even found higher growth rates at 30 and 60 days in females, with no difference in weaning weight for lambs of both sexes (Gbangboche et al., 2006b).

This tendency in body weight may be attributable to different physiological functions in both sexes, mainly of hormonal nature, which tend to become more pronounced as the animals approach maturity. This effect of sex on postnatal growth is related to testosterone production, a steroid hormone whose anabolic effects act as growth promoter (Macedo and Arredondo, 2008). Regarding the influence of sex on lamb weight at different developmental stages, in general terms, it is estimated that in male lambs this is 5.07, 6.60 and 8.33% higher than in females, at P1, P2 and P3, respectively. Similar values were found by Macedo and Arredondo (2008).

The effect of double interactions between sex and other factors was only significant with the herd factor in variable P3 (P≤0.001) and with the birth type factor in variables P1 and P3 (P<0.05), an interaction that was also identified by other authors (Assan and Mazuka, 2005; Akhtar et al., 2012) (Table 2).

Birth season

Birth season is another factor that must be considered in the development of animals and, consequently, in their reproductive management (Hernández, 2004; Quesada et al., 2002). Because there are four seasons, food production suffers large variations throughout the year, affecting both the quantity and quality of food, which influences lamb lactation by altering their physical condition, a situation evidenced by several studies (Baneh and Hafezian, 2009; Momoh et al., 2013).

The birth season effect was significant (P<0.001) for all developmental stages (Table 2), according to others authors (Akhtar et al., 2012; Hinojosa-Cuéllar et al., 2012). The lowest average weights were reached by lambs born in summer (8.73±1.77 kg), and the highest values were presented by lambs born in winter (9.10±1.77 kg) (Appendix), in agreement with the results found by Rodríguez et al. (1999), Quesada et al. (2002), Hernández (2004) and Momoh et al. (2013).

Interaction between birth season and herd was significant (P<0.001) in all variables and the interaction between birth season and birth year was significant (P<0.05) in variables P1 and P3 (Table 2).

Herd

One important performance factor on lambs is the physical environment and climate, as was also observed by Charles (1985), Sarti et al. (2003), Kittelsen (2008) and Lavvaf and Noshary (2008). In his study on the effect of herd and management in the Ancient Norse breed, Kittelsen (2008) found significant differences among animal weights in different locations, although found no differences among the weights due to different types of management. The herd can have significant effect due to differences in management and environmental conditions (Baneh and Hafezian, 2009).

The effect of the herd factor was significant (P<0.001) on lamb weights at different developmental stages. The double interactions effect between herd and birth season, birth year and birth type were significant (P<0.001), while the double interaction between herd and sex was only significant (P≤0,001) for older lambs (Table 2).

Using the Tukey test, and under the herd effect, average weights were grouped into 18 homogeneous subsets at P1. This number dropped to 14 at P2 and the most developed lambs were grouped together into 11 homogeneous subsets; this can be due to a smaller number of herds being observed. There are herds that stand out due to the superiority of their average weight at all developmental stages (e.g. herd B with mean weights of 10.43, 15.29 and 20.54 kg for P1, P2 and P3, respectively), while others stand out due to the inferiority of these values (e.g. herd HS with mean weights of 7.76, 12.41 and 17.53 kg for P1, P2 and P3, respectively) (Appendix). Bela and Haile (2009) also found variations in the growth characteristics of lambs reared in different locations. The pronounced differences between the herds studied show that diversity in management (health and nutrition) has a great influence on lamb weights. The same results were achieved by Baneh and Hafezian (2009).

Table 1. Determinative coefficient of the effect of each factor on weight of Segureña breed lambs, at three developmental stages, obtained with the univariate general linear model.

Factor	Developmental stage
	P1	P2	P3
Sex	1.4	2.8	5.2
Birth season	1.0	1.3	0.8
Herd	9.2	6.8	7.0
Year	2.2	3.1	1.8
Birth type	9.8	7.4	7.3

P1, early weaning between 16 and 35 days; P2, late weaning between 36 and 55 days; P3, final weight between 56 and 80 days. Values are expressed as percentage.

Table 2. Determinative coefficients of significant level of factors, interactions, covariates and models obtained with ANalysis Of VAriance for weight at the three developmental stages.

Source	Developmental stage
	P1	P2	P3
SEX	<0.001	<0.001	<0.001
B_SEASON	0.002	<0.001	<0.001
HERD	<0.001	<0.001	<0.001
YEAR	0.001	<0.001	-
B_TYPE	<0.001	<0.001	<0.001
AGE	<0.001	<0.001	<0.001
AGEDAM	<0.001	0.006	-
HERD * YEAR	<0.001	<0.001	<0.001
HERD * SEX	-	-	0.001
HERD * B_TYPE	<0.001	<0.001	<0.001
HERD * B_SEASON	<0.001	<0.001	<0.001
B_SEASON * YEAR	<0.001	-	0.003
YEAR * B_TYPE	<0.001	-	0.002
SEX * B_TYPE	0.012	-	0.015
HERD * YEAR * B_TYPE	<0.001	<0.001	<0.001
HERD * B_SEASON * YEAR	<0.001	0.003	<0.001
HERD * B_SEASON * B_TYPE	0.002	0.033	0.003
B_SEASON * YEAR * B_TYPE	-	-	0.015
HERD * B_SEASON * YEAR * SEX	-	-	0.031
HERD * B_SEASON * SEX * B_TYPE	0.031	0.016	-
B_SEASON * YEAR * SEX * B_TYPE	-	0.049	0.032
R^2reduced	0.489	0.456	0.443

P1, early weaning between 16 and 35 days; P2, late weaning between 36 and 55 days; P3, final weight between 56 and 80 days.

Table 3. Average weights and number of observations by birth type and sex at three developmental stages of Segureña breed lambs.

		P			Sex			P×Sex
							S		D		T
		S	D	T	M	F	M	F	M	F	M	F
P1	N	10042	9925	884	10464	10387	4909	5133	5089	4836	466	418
Average weight	9.50^a	8.04^b	7.89^c	9.12^a	8.68^b	9.82^d	9.19^a	8.55^e	8.22^c	8.00^c	7.77^c
P2	N	8415	8645	812	8925	8947	4152	4263	4375	4270	398	414
Average weight	14.51^a	13.18^b	12.37^c	14.21^a	13.33^b	15.06	13.97	13.53	12.81	12.71	12.05
P3	N	5754	7625	837	6910	7306	3018	2736	3856	3769	432	405
Average weight	20.12^a	18.56^b	17.12^c	19.89^a	18.36^b	19.30^a	21.03^c	17.85^b	19.28^d	16.48^d	17.81^e

P, birth type; S, single birth type; D, double birth type; T, triple or higher birth type; M, male; F, female; P1, early weaning between 16 and 35 days; P2, late weaning between 36 and 55 days; P3, final weight between 56 and 80 days; N, number of observations.

^a-eMeans with different letters within each factor or interaction are statistically different (P≤0.05).

Birth year

The birth year can cause variations in lamb weights at different ages due to the effect of climate conditions (precipitation, humidity and temperature), environmental conditions and management. Climate and environmental changes affect the quality and quantity of pasture forages, which also affects the provision of food and other animal needs (Momoh et al., 2013). Differences in nutrition (especially during pregnancy), management and hygiene in the various years, are some reasons for the effect of birth year on body weight in different ages (Baneh and Hafezian, 2009).

Birth year affected (P<0.001) body weight at weaning but was not significant (P>0.05) on the body weight of older lambs, unlike Akhtar et al. (2012) who, in their study on Buchi sheep, concluded that the effect of birth year was only significant on more developed animals. Double interactions were significant (P<0.001) between birth year and herd, in agreement with Baneh and Hafezian (2009). The effect of the interactions between birth year and birth season and between birth year and birth type were significant (P<0.05) for P1 and P3 variables (Table 2). Unlike the results found by Akhtar et al. (2012), the interaction effect was not significant (P>0.05) between birth year and sex in all three variables (Table 2), indicating that these factors are independent. The interaction between birth year and birth season demonstrates that seasons are not equal throughout the years, and different effects were obtained with different combinations of birth year and seasons. The average body weight for birth year and birth season ranged between 7.67±1.19 kg (spring year 4) and 10.23±1.81 kg (autumn year 6) at P1 and between 18.07±3.35 kg (summer year 11) and 21.32±3.26 kg (spring year 2) in more developed lambs.

Using the Tukey test, under the birth year effect, average weights were grouped into 6 homogeneous subsets at P1 and into 5 at P2 (Appendix).

Birth type

Although Quesada et al. (2002) found no differences in weaning weights among lambs from different birth types, and other authors (González et al., 2002) concluded that after weaning, lambs from multiple births reach daily weight gains that are higher than those of single birth lambs, several studies in different breeds (Eltawil et al., 1970; Robinson et al., 1977; Dimsoski et al., 1999; Rodríguez et al., 1999; Hernández, 2004; Hinojosa-Cuéllar et al., 2012, Momoh et al., 2013) show that birth type affects the birth weight of lambs from birth to 80 days, with single born lambs reaching the highest weights. Thus, although in general terms birth type does affect ovine growth, there may be particular variations for each breed and each production system.

Birth type had significant effect on weight (P<0.001) at every developmental stage (Appendix). The double interaction effect between birth type and herd was significant (P<0.001) at all developmental stages (Table 2), in agreement with the results found by Baneh and Hafezian (2009). Interaction effects were significant (P<0.05) between birth type and birth year and between birth type and sex for P1 and P3 variables (Table 3). In their studies on ovine, Hussain et al. (2013) and Akhtar et al. (2012) obtained a non-significant interaction between birth type and sex, unlike the findings in this study. The effect of interactions between birth type and birth season was non-significant (P>0.05) for any of the variables (Table 2).

Single birth lambs were heavier by 18.16, 10.09 and 8.41% when compared with twins at P1, P2 and P3, respectively. Differences became more pronounced when comparing single birth lambs with triple or higher birth, with percentage values of 20.41, 17.30 and 17.52% (Appendix). Most studies confirm the superiority of the body weight of single birth lambs over multiple birth lambs (Dickson-Urdaneta et al., 2004; Gbangboche et al., 2006a; Hinojosa-Cuéllar et al., 2012; Ramirez-Tello et al., 2013). This also occurs in the study by Bela and Haile (2009), where the effect of birth type on growth decreases with the lamb’s age.

Conclusions

Applying univariate models to isolated variation factors, the determinative coefficient for each is an indicator of the variation explained by each factor. Gathering with the significance inferred for the several factors with the multifactorial models, we can appreciate how sexual dimorphism is evident at the earliest weights but is extreme at the final weight. Therefore, the decision for consignment to the slaughterhouse in either sex should be based on weight and never age.

The most important effects were herd and birth type: the first is justified by great differences in the management of herds; the second case is due to the difference between the dam’s lactogenesis capacity, meaning that this factor’s influence is greater when the lamb is more dependent and decreases when the lamb reaches late weaning age.

In most meat sheep breeding programmes the genetic models enclose the interactive effects of the combined herd-year-season factors. In the present study these factors are also studied isolated, in order to access to their direct effects upon the traits. The influence of the birth year was relatively small and uncontrolled, like the birth season.

The variance explained by the multifactorial model ranged between 44 and 49%, therefore we consider it very efficient for the three variables in study and find that it provides very important information for the design of genetic analysis models in order to obtain maximum orthogonality at a lower computational cost.

The herd factor proved highly interactive: both double and triple interactions, in which this factor was involved were, in the most part, clearly significant. This derives, mainly, from its great responsibility in explaining the variance. The opposite occurred with the birth season factor: this is why, despite being polioestrus animals, the concentration of births in certain times is due to commercial reasons of market demand and not to a better response to growth in certain seasons.

Birth type significantly interacted with other factors, due to the large volume of variance that it explains. The influence of this factor on variation decreases with age but, contrary to what is described in the literature, in this case compensatory growth does not neutralise the weight differences found in single births when compared with multiple births. This leads us to question genetic improvement based solely on prolificacy, as multiple birth lambs grow significantly less and thus reach market weight later. The situation is most extreme among female lambs from triple or higher births, compared with single birth males.

Acknowledgements

The authors thank the National Association of Segureño Sheep Breeders (ANCOS). Special thanks go to Mr. José Puntas and Mr. Miguel Serrano for all their support in the field work.

APPENDIX

Appendix Table 1. Basic statistical information of each factor that affects weight at the three developmental stages of Segureña breed lambs.

F^a ctor	P1				P2					P3
	N	MW, kg	SD	CV	N	MW, kg	SD	CV	N	MW, kg	SD	CV
Sex		^***				^***				^***
M^a le	10,464	9.12^a	1.89	20.72	8925	14.21^a	2.79	19.62	6910	19.89^a	3.5	17.7
Female	10,387	8.68^b	1.7	19.55	8947	13.33^b	2.42	18.13	7306	18.36^b	2.9	16
Birth season		^**				^***				^***
Spring	3540	9.05^a	1.8	19.92	2614	13.94^a	2.63	18.86	2220	18.96^a	3.3	17.2
Summer	10,186	8.73^b	1.77	20.24	9140	13.51^b	2.61	19.33	6821	18.87^a	3.3	17.6
Autumn	2919	9.03^a	1.96	21.74	2342	13.77^c	2.71	19.67	2144	19.23^b	3.4	17.6
Winter	4206	9.10^a	1.77	19.43	3776	14.26^d	2.62	18.4	3031	19.64^c	3.3	16.6
Herd		^***				^***				^***
AF	231	8.54^cdef	1.9	22.3	-	-	-	-	-	-	-	-
AN	234	9.26^klmnopq	1.88	20.3	269	14.21^hijkl	2.51	17.67	-	-	-	-
AT	2301	8.42^cde	1.6	18.97	1802	13.33^cde	2.51	18.84	1957	18.51^bcd	3.3	17.6
B	393	10.43^r	1.98	18.97	433	15.29^m	2.7	17.64	332	20.54^ik	2.8	13.8
CO	318	8.98^fghijklmno	1.73	19.28	297	14.09^ghijk	2.31	16.38	-	-	-	-
DR	556	9.21^jklmnopq	1.68	18.2	506	13.68^defghi	2.25	16.46	375	18.33^abc	2.8	15.4
FP	446	9.29^lmnopq	1.72	18.56	368	14.83^lm	2.75	18.56	315	19.73^ghi	3.2	16
FR	-	-	-	-	402	13.71^defghi	2.76	20.11	364	20.56^k	3.4	16.7
FV	496	9.14^ijklmnopq	1.97	21.55	322	14.32^ijkl	2.81	19.64	-	-	-	-
GH	661	9.01^ghijklmno	1.79	19.82	527	13.33^cde	2.45	18.41	554	17.66^a	3	17
GI	360	8.68^defghi	1.7	19.57	290	13.00^abc	2.65	20.4	233	18.87^bcdef	3.4	17.8
GO	280	8.98^fghijklmno	1.66	18.5	260	13.39^cdef	2.43	18.13	319	18.89^bcdef	3.3	17.4
GT	1340	9.04^hijklmnopq	1.71	18.86	1037	13.78^efghij	2.42	17.56	810	20.66^k	3.2	15.6
HC	189	8.83^efghijkl	1.53	17.33	-	-	-	-	218	18.19^ab	3.5	19.2
HM	-	-	-	-	149	14.10^ghijk	2.74	19.41	-	-	-	-
HR	553	8.20^abc	1.73	21.16	527	12.57^ab	2.78	22.08	448	17.62^a	3.1	17.8
HS	864	7.76^a	1.55	20.01	824	12.41^a	2.51	20.23	766	17.53^a	3.0	17
JF	889	8.70^defghij	1.63	18.8	699	13.55^cdefg	2.51	18.54	805	19.00^cdefg	3.2	16.8
JJ	343	8.80^efghijk	1.51	17.15	304	13.77^efghij	2.45	17.8	-	-	-	-
JR	430	9.10j^klmnop	1.76	19.37	372	14.39^jkl	2.57	17.84	-	-	-	-
LH	237	9.64^q	1.67	17.34	351	14.27^ijkl	2.69	18.83	-	-	-	-
LM	1438	9.06^hijklmnop	1.76	19.4	1077	13.83^efghij	2.61	18.88	1154	19.25^defgh	3.3	17
MM	390	8.99^fghijklmno	1.7	18.94	346	14.13^ghijk	2.52	17.81	-	-	-	-
NA	745	10.18^r	1.92	18.86	426	16.16^n	2.75	17.03	-	-	-	-
NI	260	9.39^opq	1.6	17.04	250	13.84^efghij	2.4	17.32	-	-	-	-
NN	344	9.01^ghijklmno	1.66	18.47	360	13.74^defghij	2.31	16.84	-	-	-	-
PV	372	8.58^cdefgh	1.69	19.65	370	13.29^cde	2.44	18.37	-	-	-	-
R	269	8.84^efghijkl	1.75	19.77	267	14.06^ghijk	2.62	18.61	-	-	-	-
RJ	200	9.10^jklmnop	1.89	20.8	-	-	-	-	-	-	-	-
RM	375	10.08^r	1.78	17.67	-	-	-	-	198	18.93^bcdef	3.0	16
RR	585	8.57^cdefg	1.94	22.58	794	13.81^efghij	2.78	20.1	959	19.61^fgh	3.7	19
SB	281	8.68^defghi	1.84	21.2	235	13.52^cdefg	2.6	19.21	207	19.76^ghij	3.2	16.4
SC	414	8.83^efghijkl	1.95	22.09	336	13.48^cdefg	2.7	20	440	18.74^bcde	3.2	16.8
TO	-	-	-	-	180	13.10^bcd	2.43	18.56	283	19.04^cdefg	3.3	17.1
VC	845	8.45^cde	1.63	19.25	870	13.37^cdef	2.42	18.13	866	19.00^cdefg	2.9	15.3
VJ	1004	9.07^hijklmnop	1.66	18.34	809	14.02^fghij	2.6	18.55	800	19.19^defgh	3.1	16.3
VS	194	7.83^a b	1.61	20.59	-	-	-	-	-	-	-	-
XP	204	9.34^mnopq	2.03	21.79	-	-	-	-	-	-	-	-
Z	921	9.37^nopq	1.78	18.95	860	14.69^klm	2.53	17.19	719	20.51^ijk	3.1	15
ZA	167	9.56^pq	1.78	18.59	187	13.78^efghij	2.35	17.05	403	19.96^hijk	3.3	16.5
ZF	722	8.26^bcd	1.61	19.45	766	13.56^cdefgh	2.61	19.26	691	19.33^efgh	3.3	17.1
Year		^***				^***				ns
1	1331	8.74^b	1.77	20.28	963	12.93^a	2.43	18.78	625	19.44	3.2	16.5
2	1118	9.08^e	1.71	18.78	972	14.20^d	2.5	17.62	785	20.36	3.5	17
3	1871	9.03^de	1.74	19.25	664	14.89^e	2.76	18.52	216	19.38	3.2	16.7
4	1037	9.00^bcde	1.84	20.46	813	13.95^cd	2.61	18.67	589	19.33	3.3	16.8
5	1911	8.89^bcd	1.72	19.35	1699	13.58^b	2.59	19.09	1310	18.46	3.3	17.7
6	2485	9.39^f	1.86	19.83	2275	14.12^cd	2.64	18.72	1991	19.42	3.2	16.2
7	2316	9.00^cde	1.78	19.74	2279	14.16^d	2.59	18.27	2176	19.05	3.3	17.5
8	1870	8.81^bc	1.74	19.81	1739	13.84b^c	2.5	18.07	1391	19.30	3.0	15.7
9	1440	8.83^b	1.84	20.87	1620	13.99^cd	2.76	19.74	1246	19.20	3.4	17.8
10	1596	9.00^bcde	1.76	19.57	1467	13.60^b	2.49	18.29	1741	18.93	3.4	17.9
11	3876	8.49^a	1.84	21.73	3381	13.14^a	2.67	20.32	2146	18.57	3.4	18.3
Birth type		^***				^***				^***
Simple	10,042	9.50^a	1.86	19.55	8415	14.51^a	2.65	18.29	5754	20.12^a	3.32	16.50
Double	9925	8.04^b	1.58	18.88	8645	13.18^b	2.47	18.71	7625	18.56^b	3.14	16.91
Triple or higher	884	7.89^c	1.38	17.49	812	12.37^c	2.24	18.09	837	17.12^c	2.88	16.84

P1, early weaning between 16 and 35 days; P2, late weaning between 36 and 55 days; P3, final weight between 56 and 80 days; N, number of observations; MW, mean weight; SD, standard deviation; CV, coefficient of variation (%).

^a-rMeans with different letters within each column and within each factor are statistically different.

**P≤0.05;

***P≤0.001; ns, not significant.