Estimates of Direct and Maternal Genetic Effects on Birth and Weaning Weights in Salem Black Goats from India

Simple Summary

Salem Black goats, a popular meat breed in southern India, are essential for rural families, supplying both meat and cash to ensure food security. This study investigated the genetic and non-genetic factors influencing growth traits in Salem Black using data from 930 kids collected for a period of 16 years from 2004 to 2019. Phenotypic and genetic parameters for birth weight (BW), weaning weight (WW), and average daily gain (ADG) were estimated using multiple-trait animal models. These findings imply that genetic improvement in Salem Black goats should prioritize direct selection for growth traits, whereas management measures that improve early-life environmental conditions will outperform selection for maternal genetic merit. For smallholder farmers, increased nutrition, health care, and doe management during kidding can result in immediate gains in kid development and survival, supplementing longer-term genetic selection efforts.

Abstract

Growth traits in goats are impacted by both genetic and non-genetic variables; as such, it is critical to separate direct and maternal effects for reliable genetic assessment. This study determined the phenotypic and genetic characteristics for birth weight (BW), weaning weight (WW), and average daily gain (ADG) in Salem Black goats. The MTDFREML software was used to evaluate data from 930 kids, the progeny of 147 bucks and 804 does, gathered between 2004 and 2019. Three models were compared: Model 1 (direct genetic effects only), Model 2 (adding permanent environmental effects), and Model 3 (adding maternal genetic effects and direct–maternal covariance). The overall mean was 2.21 kg for BW, 9.23 kg for WW, and 78.27 g/day for ADG. The direct heritability estimates for BW, WW, and ADG were 0.06–0.22, 0.13–0.40, and 0.11–0.16 across models, respectively, whereas maternal heritability ranged from 0.01 to 0.13. The study revealed maternal genetic effects on birth weight but their total contribution to growth trait variation is modest, and model parsimony suggests that maternal effects on WW and ADG be excluded. Breeding programs should therefore prioritize direct selection for growth performance, whereas management strategies such as enhanced nutrition, targeted kidding, and support for first-parity do provide more rapid and effective avenues to improve pre-weaning developmental outcomes. Hence, a pragmatic approach that combines direct genetic selection and environmental optimization would result in greater genetic gain and support long-term meat production in Salem Black goats.

1. Introduction

Goats (Capra hircus) are essential for smallholder farmers in developing nations, particularly India, where they support meat production, rural livelihoods, and nutritional security [1,2]. Their capacity to adapt to severe environments, minimal input requirements, and high reproductive rates make them ideal for sustainable livestock systems in resource-constrained settings [3]. Goats’ tolerance to harsh temperatures, combined with their cheap investment requirements and high reproductive efficiency, makes them an important component of sustainable livestock production systems in resource-limited environments [3]. India, with a population of approximately 148 million goats, has a diverse range of indigenous varieties, each particularly adapted to its native environment [4]. Among these, the Salem Black goat stands out as an important meat-type breed endemic to Tamil Nadu’s tropical climate. They are specifically raised in the Salem, Namakkal, and Dharmapuri areas, which have a hot, dry climate similar to India’s tropical conditions [5]. Characterization studies have shown that this breed is well-adapted, with a predominantly black coat, a strong frame, and body weight averages of 35.4 kg and 26.5 kg for adult bucks and does [5]. Despite its economic importance and apparent adaptability, a thorough assessment of its growth performance under these particular tropical circumstances is required. While the breed is typically handled in semi-intensive systems based on grazing and agricultural leftovers [6], its productivity is regulated by a complex interaction of genetic and environmental factors. Performance parameters like BW, WW and ADG indicate not only genetic potential but also an animal’s capacity to flourish in its production environment. Understanding these characteristics is critical for designing successful genetic improvement and management strategies. Therefore, the present study was conducted to quantitatively evaluate the phenotypic and genetic parameters for key growth traits in Salem Black goats, thereby providing a scientific basis for enhancing productivity and sustainability under the tropical climatic conditions of India.

Growth traits like BW and WW are important indicators of goat productivity and survival. According to evaluations of small ruminant genetics, body weight is a primary determinant of neonatal survival, with lower weights consistently linked to increased mortality risks due to less vigor and resistance [7]. The WW, typically measured at around three months, reflects the cumulative pre-weaning growth performance and is a key economic driver as it directly influences the animal’s market value and the overall efficiency of the production system [7,8]. The manifestation of these features is influenced by a complex interaction of non-genetic and genetic variables. Environmental factors such as the kid’s gender, birth type (single vs. multiple), and the dam’s parity are well-documented causes of variation [8]. Furthermore, genetic influences are dual in nature, encompassing both the direct genetic effect of the kid’s own genotype on its growth and the maternal genetic effect, which encompasses the dam’s genotype through her ability to provide a favorable uterine environment and adequate milk production [9]. These maternal impacts, which include intrauterine nutrition and lactation performance, are especially important for goat growth features before weaning. While the importance of partitioning genetic variance into direct and maternal components has been established in livestock breeding for decades [9], and studies on breeds such as the Jamunapari goat in India have demonstrated significant maternal influences on body weights [10], such comprehensive genetic analyses are still underexplored in many South Indian breeds. This is a crucial information gap for the economically significant Salem Black goat. As stressed in sheep genetic studies, omitting maternal effects can result in significantly inaccurate estimations of breeding values. When maternal effects are present but not included in the assessment model, direct heritability can be highly skewed, resulting in suboptimal selection decisions and eventually slowing the rate of genetic advancement in breeding programs [11]. As a result, a detailed understanding of both direct and maternal genetic contributions is required for developing an efficient genetic improvement plan for the Salem Black goat.

Despite the recognized economic importance of Salem Black goats, extensive research into their genetic factors is remarkably sparse. Preliminary investigations and work on comparable tropical breeds suggest that BW and WW have moderate to low heritability, with maternal influences playing a key role [8,12]. However, a solid comprehension of the genetic makeup of growth in this specific breed is missing, particularly in terms of the level of direct maternal genetic covariance, which is crucial for predicting response to selection. The goals of this study were twofold: first, to estimate the impact of key non-genetic factors (such as sex, birth type, dam parity, and year of birth) on BW, WW, and average daily gain (ADG) and, second, to quantify the variance components and genetic parameters for direct and maternal genetic effects on these traits in a Salem Black goat herd reared in typical tropical conditions in India. We explicitly expect that both direct and maternal genetic factors have a considerable impact on BW, WW, and ADG, with maternal effects being stronger during the early pre-weaning stage. The study’s findings are expected to have substantial practical applications. A more efficient breeding value calculation system will be made possible by this research’s precise estimates of heritability and genetic correlations. This will enable the application of a balanced selection technique that enhances the dams’ maternal capacity and the young one’s natural growth potential at the same time. In the end, this research will yield the fundamental information required to create a long-term genetic development program that will boost farmer earnings, increase production, and guarantee the preservation and improved use of the Salem Black goat breed.

2. Materials and Methods

2.1. Study Location and Agro-Climatic Context

2.1.1. Geographical Location

The study was conducted on a herd of Salem Black goats raised under a semi-intensive management system at the Mecheri Sheep Research Station (MSRS), Pottaneri, Tamil Nadu, India. This location is roughly 200 m (650 feet) above mean sea level (MSL), with extensive pedigree and production data records available for analysis over nearly two decades from 2002 to 2022.

2.1.2. Climatic Conditions and Seasonal Impact on Forage

The semi-arid tropical climate in which this research station operates has a direct impact on pasture availability, as well as feed quantity and quality for Salem Black goats. During the rainy season (October–December), high humidity (>80%) and abundant rainfall (usually 450–500 mm) result in lush, high-quality pastures with quick growth of grasses and legumes, providing ideal nutrient density and abundant green fodder. The winter season provides moderate climatic stress, with temperatures ranging from 18 to 22 °C and relative humidity levels ranging from 55 to 65%, and pasture fields are dominated by the residual moisture-fed grasses such as Cenchrus ciliaris and leguminous fodder such as Stylosanthes, which have a moderate protein content and digestibility. In contrast, the summer months present serious pastoral challenges with nearly no natural pasture growth due to high temperatures ranging from 30 to 38 °C, low rainfall, and dry soils. During this time, the station relies solely on grown and saved fodder such as CO-4, cowpea, sorghum hay, and silage to meet nutritional needs, ensuring that Salem Black goats remain productive despite the seasonal forage gap. The tract received annual rainfall of 800 to 950 mm. The goats were given regulated feeding regimens to suit their nutritional needs in addition to grazing during different seasons. These comprised green fodder and concentrate combinations according to age, physiological status (such as pregnancy or breastfeeding), and overall health.

2.2. Herd Management Practices

2.2.1. Grazing and Feeding Regimen

The Salem Black goats’ daily foraging routine consists of both grazing and browsing. Although they have access to natural pastures of grasses like Cenchrus ciliaris, the length of time they spend foraging in ecosystems with tree fodder suggests that browsing on available trees and shrubs is a regular activity. Additionally, in the “garden land ecosystem” that MSRS Pottaneri has identified as suitable for Salem Black goats, the following can be found: Neem (Azadirachta indica), whose tender leaves are occasionally used as a herbal dewormer; Kodukapuli (Pithecellobium dulce); Vaagai (Albizia lebbeck); Subabul (Leucaena leucocephala), a high-protein fodder that should be fed cautiously due to its mimosin content; and Agathi (Sesbania grandiflora), a perennial legume that is highly palatable to goats. Supplementation with 300 g per day during the “flushing” period, which is three weeks before and after mating, and the “steaming-up” period, which is four weeks before and after kidding, has been proven to improve body weight gain, body condition score, and child birth weight.

2.2.2. Housing and Animal Classification

Goats were kept according to their age, gender, health status, and reproductive stage, such as pregnant or lactating does, and this systematic segregation is typical procedure at the research station to reduce stress and manage the herd efficiently.

2.3. Reproductive and Health Management

A controlled breeding system was implemented to manage seasonal reproductive cycles. Natural mating was scheduled and the does were mated with selected bucks with two services per estrus cycle to increase conception rates. Each kid was given a unique identifying number via an ear tattoo, and a detailed record was kept for each kid, including their body weight (BW), gender, and date of birth. The kids were kept in pens with their mothers until they were 90 days or three months of age. At that point, they were progressively introduced to short-duration grazing until they were six months old. Additionally, a proactive health management strategy was put in place to maintain the health and productivity of the herd, with strategic deworming carried out twice a year to control internal and external parasites, specifically before the monsoon in May–June and after it had passed in September–October. This timing is crucial in tropical climates to combat the increased parasite load associated with the rainy season.

2.4. Data Records

The study included 930 data records, comprising 147 bucks and 804 does from the Salem Black goat population, resulting in a comprehensive dataset spanning multiple generations and giving a solid platform for genetic parameter assessment. The variables studied are birth weight (BW), weaning weight (WW), and average daily gain (ADG), all of which are important growth features that have a direct impact on the herd’s economic viability and productivity. A digital weighing scale with an accuracy of ±10 g was used to record the kids’ birth weight within 24 h of kidding, providing exact baseline data before any postnatal environmental influences could significantly impact body weight. Weaning weight was measured on the same day (i.e., 90 days of age, or three months) as per the station’s established management practice; this consistency in weaning age is crucial for getting comparable growth data across all individuals. The ADG was calculated using the following formula: ADG = (Weaning Weight − Birth Weight)/Age in days. This gives a standardized measure of growth rate from birth to weaning that accounts for the exact number of days each kid remained with its doe. To ensure accuracy in growth rate estimation for the upcoming genetic study, ADG was calculated using individual performance records rather than population means, preserving the true phenotypic variance required for distinguishing genetic and environmental effects in subsequent quantitative genetic analyses. This individual-based technique avoids the loss of information that occurs when utilizing averaged results, improving the reliability of heritability estimations and breeding value projections. Table 1 describes the data format used in the investigation, including a full breakdown of dataset size and the computational framework used for genetic evaluation. The table shows the total number of data, the distribution of animals by sex, the dimensions of the association matrix, and the computing intensity needed for three distinct statistical models.

Table 1. Structure of data used in analysis.

Observations	Number
No. of records	930
No. of bucks	147
No. of does	804
Animals in relationship matrix (A−1)	981
Model 1
Mixed-model equations (MMEs)	3006
No. of iterations	2500
Model 2
Mixed-model equations (MMEs)	3912
No. of iterations	4155
Model 3
Mixed-model equations (MMEs)	6855
No. of iterations	8670

Table 1 explains the data structure used in the analysis.

2.5. Statistical Analysis

The BW, WW and ADG were analyzed using a mixed model, which included the main effects of sex, month and year of birth, type of birth, parity, and weight of doe as covariates, as well as the random effects of bucks and does within bucks and residual, as described in [13]. This comprehensive statistical technique allows for the partitioning of observed phenotypic variance into genetic and environmental components. The weight of the doe was included as a linear covariate to account for doe size and body condition during kidding, which have a direct impact on fetal growth via uterine capacity and placental efficiency, as well as milk production potential during lactation. Bucks’ random effects were used to estimate the sire variance component, which accounts for one-quarter of the additive genetic variance, while does within bucks were fitted as a random effect to account for maternal genetic and permanent environmental effects shared by offspring from the same dam, with the residual random effect capturing the remaining unexplained environmental variation. The following mixed linear model was used.

Y_ijklmno = µ + S_i + d_ij + sek + M_l + Y_m + t_n + p_o+ bx + bx²+ e_ijlmnop

where

• Y_ijklmnop = the performance variable;
• µ = overall mean;
• S_i = random effect of the Ith bucks;
• d_ij= random effect of the jth does within the ith bucks;
• sek = fixed effect of the kth sex;
• ml = fixed effect of the lth month of birth;
• _ym = fixed effect of year of calving, l = 2004 …, and 2019;
• t_n = fixed effect of the n the type of birth, m = 1, 2 or 3;
• p_o = fixed effect of the parity, o = 1, 2 …… and 6;
• b_x = partial linear regression coefficient of BW, WW and ADG on body weight of doe;
• b_x² = partial quadratic regression coefficient of BW, WW and ADG on body weight of doe;
• e_ijklmnop = random error with mean zero and variance o²e.

Estimation of Variance Components and Genetic Parameters

All variables were examined using the Multiple-Trait Animal Model (MTAM), as described in [14], a sophisticated statistical approach that analyzes multiple traits at the same time to account for genetic and environmental correlations between traits, improving the accuracy of genetic parameter estimates and breeding value predictions when compared to single-trait analyses. Three multi-trait animal models were used, representing a progressive increase in complexity and biological realism, with Model 1 incorporating the primary effects of sex, month and year of birth, type of birth, parity, and doe weight as covariates, as well as random effects of animals and error, to provide a basic framework that partitions phenotypic variance into additive genetic and residual environmental components. Using matrix notation, animal Model 1 was as follows:

Y = Xb + Zg + e

where

• Y = observation vectors of animals;
• b = fixed-effect vectors (sex, month and year of birth, type of birth and parity and weight of doe as covariate); g = animal genetic vector; e = error effect vectors; and Z and W are incidence matrices.

Model 2 is similar to Model 1 but includes the permanent environmental effect, which accounts for non-genetic factors that cause long-term similarities between repeated records or offspring of the same individual, such as the doe’s maternal ability, health status, or nutritional history, which consistently influence all of her progeny. Using matrix notation, animal Model 2 was as follows:

Y = Xb + Zg + Wp + e

where

• Y = observation vectors of animals;
• b = fixed-effect vectors (sex, month and year of birth, type of birth and parity and weight of doe as covariate); g = animal genetic vector; p = permanent effect vector; e = error effect vectors and Z and W were incidence matrices.

Model 3 is the most complete biological model that acknowledges that an individual’s phenotype is influenced not only by its own genes (direct genetic effects) but also by the genes of its dam expressed through her maternal ability (maternal genetic effects). It incorporates the fixed effects of sex, month and year of birth, type of birth, parity, and weight of doe as a covariate, as well as the random effects of animals, maternal genetic effects, permanent environmental effects, and errors. The following Model 3 was employed in matrix notation:

$$\mathrm{Y}=\mathrm{Xb}+\mathrm{Zu}+{\mathrm{W}}_{\mathrm{m}}+{\mathrm{S}}_{\mathrm{pe}}+\mathrm{e}$$

where

• Y = vector of observations;
• Xb = vector of main effects (sex, month and year of birth type of birth and parity and weight of doe as covariate);
• Zu = vector of random animal effects;
• Wm = vector of random maternal (indirect) genetic effects;
• Sp_e = vector of permanent environmental effects;
• e = vector of random residual effects.
• X, Z, W and S are incidence matrices relating records to fixed, animal, maternal genetic and permanent environmental effects, respectively.

It is assumed that

$$\mathrm{var}\left[\begin{array}{c}\mathrm{u}\\ \mathrm{m}\\ \mathrm{p}\mathrm{e}\\ \mathrm{e}\end{array}\right]=\left[\begin{array}{cccc}{\mathrm{g}}_{11}\mathrm{A}& {\mathrm{g}}_{12}\mathrm{A}& \mathrm{O}& \mathrm{O}\\ {\mathrm{g}}_{21}\mathrm{A}& {\mathrm{g}}_{22}\mathrm{A}& \mathrm{O}& \mathrm{O}\\ \mathrm{O}& \mathrm{O}& \mathrm{I}{\mathsf{\sigma }}^2\mathrm{P}\mathrm{e}& \mathrm{O}\\ \mathrm{O}& \mathrm{O}& \mathrm{O}& \mathrm{I}{\mathsf{\sigma }}^2\mathrm{e}\end{array}\right]$$

where

• g₁₁ = additive genetic variability for direct effects;
• g₂₁ = additive genetic variability for maternal effects;
• g₁₂ = additive genetic co-variability between direct and maternal effects;
• O²p_e = variance due to permanent environmental effects;
• O²_e = residual error variance.

In accordance with [14], we included estimates of direct heritability (h²_d), which measures the percentage of phenotypic variance attributable to direct additive genetic effects; maternal heritability (h²_m), which measures the percentage of phenotypic variance attributable to maternal genetic effects; genetic correlation (r_g), which measures the strength and direction of the genetic relationship between direct and maternal effects or between different traits; and phenotypic correlations (r_p), which measure the overall observed relationship between traits that include both genetic and environmental components. The Multiple-Trait Derivative-Free Restricted Maximum Likelihood (MTDFREML) software Version 1 [14], which implements restricted maximum likelihood algorithms that provide unbiased estimates of variance components by accounting for the loss of degrees of freedom due to fixed effects in the model, was used for all genetic parameter estimation. The initial mixed model included sire (buck) and dam-within-sire effects, resulting in a fundamental pedigree-based approach that used paternal half-sib families to estimate variance components. However, this model was limited in its ability to completely exploit all genetic ties in the community. Animal Model 1 includes the direct additive genetic influence, which represented the individual animal’s own genetic merit for growth traits. Animal Model 2 introduced the dam’s permanent environmental effect, which accounts for non-genetic factors that consistently influence all progeny of a given doe, such as her maternal behavior, milk production capacity, and uterine environment, preventing these environmental influences from being incorrectly attributed to genetic variance. Furthermore, Animal Model 3 included the maternal genetic effect, allowing estimation of its covariance with the direct genetic effect, which is critical for understanding the genetic relationship between an individual’s genes for growth and its dam’s genes for maternal ability, as negative covariance between these effects can indicate genetic antagonisms that complicate selection programs. In order to account for potential non-linear relationships between maternal size and offspring growth, such as diminishing returns or optimal body condition ranges, beyond which additional weight does not confer further benefits or may even become detrimental, the analysis included the fixed effects of sex, month and year of birth, type of birth, dam parity, and doe weight fitted as both a linear and quadratic covariate. The log-likelihood value is a key indicator of how well a statistical model fits the observed data given its parameterization. For each attribute, the model with the lowest −2 log L (negative twice the log-likelihood) was deemed ideal since it reduces deviation from a perfect match and converts the likelihood into a deviance statistic with a known distribution for hypothesis testing. This criterion was used to ensure that the chosen model strikes the ideal balance between goodness-of-fit and parsimony by penalizing superfluous parameters that do not significantly improve fit. It does this by directly comparing how well various models explain the observed data while taking their complexity into account. When evaluating whether additional parameters, such as maternal genetic effects, produced statistically significant increases in model fit, the use of −2 log L also made formal likelihood ratio comparisons comparing nested models easier [13].

3. Results

Least-Squares Means of Production Traits

Table 2 displays the unadjusted means, standard error (SE), standard deviations (SDs), and coefficient of variation (CV%) for the various variables evaluated in Salem Black goats, based on 930 records collected between 2002 and 2022, providing a foundational overview of the breed’s growth performance under the semi-intensive management system at the Mecheri Sheep Research Station. The average body weight (BW) was 2.21 ± 0.014 kg, with a standard deviation of 0.432 kg and a coefficient of variation of 19.56%. This indicates moderate variability in birth weight due to factors such as birth type, doe parity, and mother nutrition throughout pregnancy. This level of variance is typical for goat populations and provides adequate genetic variety for selection programs targeted at enhancing birth weight while avoiding dystocia. The mean weaning weight (WW) was 9.23 ± 0.084 kg, with a higher standard deviation of 2.56 kg and a coefficient of variation of 27.69%, indicating greater variability in weaning weight compared to birth weight. This is expected because weaning weight accumulates the effects of additional environmental factors such as dam milk production, competition among litter mates, and health status during the suckling period, thereby amplifying the expression of genetic potential. The Salem Black breed’s average daily gain (ADG) from birth to weaning was 78.27 ± 0.890 g, with a standard deviation of 27.10 g and a maximum coefficient of variation of 34.60%. This high variability suggests substantial opportunities for genetic improvement through selection, as ADG is a key trait influencing the age at which animals reach market weight and the overall efficiency of meat production.

Table 2. Unadjusted average (±SE) body weights of Salem Black goats.

Variables	Mean ± SE	Standard Deviation	Coefficient of Variation (%)
BW, kg	2.21 ± 0.014	0.432	19.56
WW, kg	9.23 ± 0.084	2.56	27.69
ADG, g	78.27 ± 0.890	27.10	34.60
Number of records = 930

kg = kilogram; g = gram; BW = birth weight; WW = weaning weight; ADG = average daily gain from birth to weaning.

Table 3 shows the least squares means for the variables influencing several Salem Black goat characteristics, with adjusted means that account for the data’s non-orthogonal nature and allow for unbiased comparisons across different levels of fixed effects by keeping other factors constant at their average levels. The birth weight (BW), weaning weight (WW), and average daily gain (ADG) of Salem goats were significantly affected by the following factors, according to least squares means: sex, season of birth, year of birth, type of birth, and parity. Male offspring weighed 2.28 ± 0.02 kg at birth compared to 2.10 ± 0.02 kg for females, attained weaning weights of 9.87 ± 0.10 kg compared to 8.70 ± 0.10 kg for females, and achieved average daily gains of 84.01 ± 1.12 g compared to 73.11 ± 1.10 g for females. These results are indicative of sexual dimorphism in mammals. According to seasonal fluctuation, kids born in seasons two and three—summer and south-west monsoon, respectively—performed better in terms of growth than kids born in other seasons. Summer-born kids achieved the highest weaning weight of 10.02 ± 0.18 kg and ADG of 87.86 ± 1.97 g, while south-west monsoon-born kids also demonstrated strong performance with weaning weight of 9.66 ± 0.16 kg and ADG of 82.40 ± 1.69 g. This is probably because the late gestation and early lactation periods coincide with optimal milk production in does and subsequent child growth. The highest weaning weight and ADG were recorded in 2016–18 (10.81 ± 0.18 kg and 94.80 ± 1.92 g, respectively) and 2019–22 (10.03 ± 0.17 kg and 87.09 ± 1.79 g, respectively), while the lowest performance was recorded in 2007–09 (8.41 ± 0.19 kg and ADG of 70.77 ± 2.07 g). Single-born kids had the highest birth weight (2.36 ± 0.02 kg) compared to twins (2.08 ± 0.02 kg) and triplets (1.85 ± 0.06 kg), indicating the fierce competition for uterine space and nutrients during gestation. Interestingly, the type of birth did not significantly affect weaning weight or ADG in the least squares analysis, suggesting that compensatory growth may occur in multiples or that management techniques like supplemental feeding help offset the initial disadvantage by weaning.

Table 3. Least- squares means (±SE) for body weight of Salem Black goats.

Effects	N	Birth Weight (kg)	Weaning Weight (kg)	ADG (g)
Overall mean	920	2.19 ± 0.02	9.29 ± 0.07	78.62 ± 0.79
Sex		**	**	**
Females	477	2.10 ± 0.02 a	8.70 ± 0.10 a	73.11 ± 1.10 a
Male	453	2.28 ± 0.02 b	9.87 ± 0.10 b	84.01 ± 1.12 b
Period		**	**	**
2004–06	85	2.30 ± 0.05 c	8.32 ± 0.23 a	66.70 ± 2.49 a
2007–09	121	2.04 ± 0.04 a	8.41 ± 0.19 a	70.77 ± 2.07 b
2010–12	245	2.26 ± 0.03 bc	8.29 ± 0.16 b	66.98 ± 1.76 a
2013–15	177	2.17 ± 0.03 bc	9.43 ± 0.16 bc	80.06 ± 1.77 c
2016–18	136	2.26 ± 0.04 bc	10.81 ± 0.18 c	94.80 ± 1.92 e
2019–22	166	2.13 ± 0.03 a	10.03 ± 0.17 d	87.09 ± 1.79 d
Season		*	**	**
Winter	301	2.18 ± 0.03 a	8.60 ± 0.14 ab	70.94 ± 1.47 ab
Summer	128	2.11 ± 0.04 a	10.02 ± 0.18 c	87.86 ± 1.97 c
South-west monsoon	173	2.21 ± 0.03 ab	9.66 ± 0.16 bc	82.40 ± 1.69 bc
North-east monsoon	328	2.24 ± 0.02 ab	9.15 ± 0.12 a	76.40 ± 1.33 a
Type of birth		**	NS	NS
Single	452	2.36 ± 0.02 b	9.55 ± 0.11	79.51 ± 1.19
Triplets	32	1.85 ± 0.06 a	9.16 ± 0.34	80.28 ± 3.68
Twins	446	2.08 ± 0.02 c	9.08 ± 0.10	77.65 ± 1.10
Parity		NS	*	*
1.00	285	2.12 ± 0.03	9.19 ± 0.15 c	78.25 ± 1.62 b
2.00	180	2.20 ± 0.03	8.89 ± 0.16 ab	73.74 ± 1.74 a
3.00	149	2.18 ± 0.03	9.23 ± 0.17 ab	78.02 ± 1.85 a
4.00	113	2.15 ± 0.04	8.91 ± 0.19 a	74.96 ± 2.03 a
5.00	82	2.26 ± 0.05	9.70 ± 0.22 c	82.52 ± 2.43 b
6.00	121	2.29 ± 0.04	10.15 ± 0.19 bc	87.16 ± 2.12 ab

^abcde Means within each column bearing different superscript letters differ significantly (p < 0.05). kg = kilogram; g = gram; ADG = average daily gain from birth to weaning. * Significant (p < 0.05), ** significant (p < 0.01), NS—Non significant.

The parity effects were evident, with parity 4 exhibiting the lowest weaning weight (8.91 ± 0.19 kg) and ADG (74.96 ± 2.03 g), while parities 1 and 5 were associated with better weaning weight (9.19 ± 0.15 kg and 9.70 ± 0.22 kg, respectively) and ADG (78.25 ± 1.62 g and 82.52 ± 2.43 g, respectively), reflecting the complex relationship between maternal age, physiological maturity, and reproductive performance. These effects were confirmed by ANOVA, as shown in Table 4, where sex, birth type, month and year, and parity are shown to significantly influence BW, WW, and ADG (p < 0.05 to p < 0.01). This indicates that these factors must be taken into account in genetic evaluation models in order to obtain objective estimates of breeding values and genetic parameters. Regression analysis revealed that doe weight had a positive linear effect on WW and ADG but not on BW. This suggests that larger does produce more milk and provide better maternal care, which improves postnatal growth. However, fetal growth is buffered against maternal size variation to prevent dystocia, indicating an evolutionary adaptation that keeps birth weight within a narrow physiological range despite variations in maternal body condition.

Table 4. Analysis of variance for factors affecting body weight in Salem Black Salem goats.

		Variables
		Birth Weight	Weaning Weight	Average Daily Weight Gain
Source of Variation	d.f	F-Values	F-Values	F-Values
Between Bucks	146	1.01 **	2.60 **	2.64 **
Between Does: Bucks	657	1.15 **	1.62 **	1.61 **
Between sex	2	47.16 **	46.47 **	32.18 **
Between month of birth	3	5.02 **	5.54 **	6.39 **
Between year of birth	5	7.94 **	25.25 **	23.06 **
Between type of birth	2	89.38 **	23.34 **	11.11 **
Between parity	5	3.11 *	71.19 *	8.12 **
Regression
Weight of doe, linear	1	0.00 ns	3.85 *	4.45 *
Weight of doe, quadratic	1	1.46 ns	0.16 ns	0.43 ns
Errors, MS	108	0.134	4.65	554

* p < 0.05, ** p < 0.01; ns—not significant; MS—mean squares.

These effects were confirmed by ANOVA, as shown in Table 4, where sex, birth type, month of birth, year of birth, and parity significantly are shown to influence birth weight (BW), weaning weight (WW), and average daily gain (ADG) (p < 0.05 to p < 0.01). The F-values quantified the relative contribution of each factor to phenotypic variation in growth traits. Larger does improve postnatal growth through better milk production and maternal care, while fetal growth is protected against maternal size variation to reduce the risk of dystocia. This evolutionary adaptation keeps birth weight within a limited physiological range. Doe weight had a positive linear effect on WW and ADG but not on BW. A straightforward linear association between mother size and offspring growth within the observed range of body weights in this population was confirmed by the quadratic term for doe weight being non-significant for all attributes, with no indication of decreasing benefits at extreme levels.

A thorough assessment of the underlying genetic architecture of growth traits in the Salem Black goat population is made possible by the progressive inclusion of direct genetic, permanent environmental, and maternal genetic effects. Table 5 displays the variance component and heritability estimates for birth weight (BW), weaning weight (WW), and average daily gain (ADG) using three alternative animal models. Direct heritability was low for BW (0.06 ± 0.03) but moderate for WW (0.13 ± 0.03) and ADG (0.16 ± 0.02) in Model 1, which included only direct additive genetic effects and residual variance. This suggests that selection for improved postnatal growth may be more effective in this population than selection for birth weight. The genetic correlations between features ranged from 0.85 ± 0.13 for BW × WW to unity (1.00 ± 0.11) for BW × ADG and WW × ADG, demonstrating a considerable shared genetic influence. These three growth factors are predominantly controlled by the same genes. Genetic correlations approaching one indicate that BW, WW, and ADG are basically the same trait from a genetic standpoint, and that selection for any of these qualities would result in correlated genetic improvements in the others.

Table 5. Variance components and heritability estimates for body weights for Salem goats.

Model 1	o2a	o2m	σam	o2e	σ2pe	o2p	h2d	h2m
BW	0.187			2.86		3.05	0.06 ± 0.03
WW	0.040			2.62		0.30	0.13 ± 0.03
ADG	1.150			8.59		9.75	0.16 ± 0.02
rg BW × WW	0.85 ± 0.13
rg BW × ADG	1.00 ± 0.11
rg WW × ADG	1.00 ± 0.11
−2 log L	70,614.00
Model 2
BW	27.38			72.45	6.81	106.64	0.26 ± 0.10
WW	110.77			147.50	18.07	276.33	0.40 ± 0.16
ADG	57.11			171.64	4.60	233.35	0.24 ± 0.05
rg BW × WW
rg BW × ADG
rg WW × ADG
−2 log L	16,064.61
Model 3
BW	1.28	0.75	−1.50	4.50	0.77	5.80	0.22 ± 0.09	0.13 ± 0.06
WW	36.33	2.56	−3.38	83.65	1.66	120.82	0.30 ± 0.10	0.02 ± 0.05
ADG	27.54	1.72	0.79	209.31	20.99	260.35	0.11 ± 0.06	0.01 ± 0.01
rg BW × WW	0.70 ± 0.001
rg BW × ADG	0.01 ± 0.001
rg WW × ADG	0.15 ± 0.001
rm BW × WW	0.30 ± 0.09
rm BW × ADG	0.26 ± 0.10
rm WW × ADG	0.10 ± 0.09
−2 log L	16,062.50

σ²_a: Direct additive genetic variance, σ²_m: maternal genetic variance, σ_am: direct–maternal genetic covariance, σ²_e: residual environmental variance, σ²_pe: permanent environmental variance, σ²_p: phenotypic variance, h²_d: direct heritability, h²_m: maternal heritability, r_g: additive genetic correlation, r_m: maternal genetic correlation, −2logL: twice the negative log-likelihood. BW = birth weight; WW = weaning weight; ADG = average daily gain from birth to weaning.

The heritability estimates increased significantly when permanent environmental influences were incorporated into Model 2, with moderate-to-high values for BW at 0.26 ± 0.10, WW at 0.40 ± 0.16, and ADG at 0.24 ± 0.05. This suggests that taking environmental factors into account improved variance partitioning and produced more realistic estimates of additive genetic contributions by preventing permanent environmental variance from being incorrectly absorbed into the residual or genetic components. Model 3 included additive direct genetic, maternal genetic, permanent environmental, and residual influences. Direct heritability estimates were highest for WW at 0.30 ± 0.10, moderate for BW at 0.22 ± 0.09, and low for ADG at 0.11 ± 0.06. Maternal heritability was significant only for BW at 0.13 ± 0.06, but negligible for WW at 0.02 ± 0.05 and ADG at 0.01. The birth weight and weaning weight had negative additive–maternal covariances, with values of −1.50 for BW and −3.38 for WW, indicating a potential antagonistic relationship between direct and maternal genetic effects on these parameters. Genes that promote higher growth rates in the offspring may be negatively associated with genes that enhance maternal ability, creating a genetic conflict that can complicate selection programs if both traits are to be improved.

Model 3 did not outperform Model 2 in terms of fit based on the likelihood ratio test (LRT), despite having a lower −2 log L value (16,062.50) than Model 2 (16,064.61), indicating a slightly better absolute fit to the data. This was because, when compared to the chi-square distribution with the appropriate degrees of freedom, the decrease in −2logL was insufficient to justify the additional parameters estimated in the more complex model. Model 3 was only utilized for birth weight because of the significant maternal heritability estimate of 0.13 ± 0.06 for this trait, indicating that maternal genetic influences are trait-specific and not generally significant across all development variables in Salem Black goats. Model 2 was found to be the best fit for weaning weight and average daily gain. By incorporating just those characteristics that significantly enhance model performance, Model 2 achieves the best trade-off between explaining observed variation and preserving parsimony, according to the likelihood ratio tests. Despite the inclusion of maternal effects in Model 3, the analysis reveals that these effects are not statistically justified for the majority of features because the improvement in model fit was insignificant in comparison to the additional complexity. Maternal genetic effects did not provide any additional benefit (Model 2 vs. 3: p > 0.05), suggesting that permanent environmental effects sufficiently capture maternal influences without the need to separate them into genetic and environmental components. However, adding permanent environmental effects significantly improved the model (Model 1 vs. 2: p < 0.001), indicating that accounting for non-genetic factors that create lasting similarities among offspring of the same dam is essential for accurate variance component estimation. Therefore, a statistically valid and biologically understandable framework for genetic evaluation and selection decisions in the Salem Black goat population is provided by the model that combines direct genetic and persistent environmental impacts to adequately represent the variance in these traits.

4. Discussion

4.1. Non-Genetic Factors Influencing Growth Traits in Salem Black Goats

The Salem Black goats had mean BW, WW, and ADG of 2.21 ± 0.014 kg, 9.23 ± 0.084 kg, and 78.27 ± 0.890 g/day, respectively. These findings suggest that semi-intensive conditions in Tamil Nadu, India, have a high potential for early growth. The values are similar to those reported for other tropical and indigenous goat breeds, such as Zaraibi [8] and Jakhrana goats [15], although slightly higher than Barbari and Assam Hill goats [16,17]. However, they remain lower than larger breed like Jamunapari [10,18]. These comparisons highlight the Salem Black breed’s intermediate body size and adaptability, with differences primarily due to breed genetics, feeding regime, and environmental management. The coefficients of variation (CV%) for BW, WW, and ADG ranged from 19.56 to 34.60%, demonstrating moderate-to-high variability between individuals. The significantly lower CV for BW indicates greater genetic control and less environmental influence at birth, whereas larger variance in WW and ADG reflects postnatal environmental effects such as nutrition and maternal care. Similar variable patterns have been shown in Jamunapari goats [10,18], highlighting the importance of management in postnatal growth. Sex strongly influenced all development traits, with males having higher BW, WW, and ADG than females. This is consistent with previous research on Barbari, Zaraibi, and Markhoz goats [8,16,19], and is linked to testosterone-enhanced muscular and skeletal development. Although some authors [8,20] revealed non-significant sex effects, such inconsistencies are most likely due to uniform management or small sample numbers. Season and year of birth also had a major impact.

Kids born in the winter and spring had higher WW and ADG than those born in hotter or wetter times. Seasonal variation is likely due to variations in feed supply, ambient temperature, and disease pressure. Similar trends have been observed in Barbari, Jamunapari and Zaraibi goats [16,18,21]. Birth type had a significant impact on performance, with single-born kids outperforming twins and triplets on all growth parameters. This is consistent with results in several breeds [8,22,23,24,25], and can be related to decreased intrauterine and postnatal competition for nutrition and milk. The persistent advantage of singletons demonstrates the trade-off between litter size and kid growth in meat-type goats. Parity effects followed a predictable pattern, with middle- to late-parity offspring (especially fifth and sixth) having higher BW and WW than first-parity kids. This is most likely owing to improved maternal maturity, udder development, and milk output, as shown in Barbari and Cashmere goats [16,24]. Finally, regression analysis demonstrated a strong positive relationship between doe body weight and kid WW and ADG, implying that bigger dams had faster-growing offspring, probably due to superior prenatal nutrition and increased milk production.

4.2. Correlations, Variance Components, and Heritability Estimates

The analysis of variance components and heritability estimates for key growth traits—birth weight (BW), weaning weight (WW), and average daily gain (ADG)—in Salem Black goats sheds light on the genetic and environmental factors influencing early growth performance in the indigenous Indian breed. Salem Black goats are native to Tamil Nadu and are well-adapted to tropical climates, with a reputation for resilience, efficient feed consumption, and high meat quality [5,6]. Understanding the genetic architecture of growth traits in this population is critical for developing efficient selection strategies to increase productivity in smallholder and semi-intensive systems. Three hierarchical linear mixed models were used to estimate genetic parameters, gradually integrating new factors to provide a more nuanced variance partitioning. Model 1 was a basic univariate animal model including only direct additive genetic effects (σ²a) and residuals (σ²e), excluding maternal and permanent environmental components. This resulted in low phenotypic variances (σ²p = 3.05 for BW, 0.30 for WW, and 9.75 for ADG) and a poor fit, as indicated by the high −2logL value (70,614). The inflated genetic correlations (r_g = 0.85–1.00) likely stemmed from unpartitioned maternal effects, consistent with biases observed in other studies [26,27,28]. Model 2 extended this by adding permanent environmental variance (σ²pe) to capture persistent non-genetic maternal influences, such as uterine capacity and dam nutrition. This inclusion substantially increased phenotypic variances (106.64 for BW, 276.33 for WW, and 233.35 for ADG) and improved model fit (−2logL = 16,064.61), supporting our hypothesis that environmental factors at the dam level significantly shape early growth. These findings are consistent with those in other Indian and African goat breeds, where maternal environment accounts for a significant part of variation [29,30]. Model 3, which was the most comprehensive, included maternal genetic effects, direct–maternal covariance and σ²_pe, in addition to σ²_a and σ²_e. Negative covariances for BW and WW showed antagonistic interactions, possibly due to trade-offs between maternal resource allocation and fetal growth potential [31].

Phenotypic variances were moderate, with genetic correlations falling within biologically plausible ranges. The progressive decline in −2logL across models underscores the importance of multi-trait structures and maternal components in avoiding biased estimates, as noted in sheep populations [32]. Direct heritability (h²_d) estimates varied across models, reflecting the impact of unaccounted effects in simpler formulations. These estimations are within the ranges recorded for Salem Black goats under farm settings [33] and equivalent to low-to-moderate values in other Indian breeds, such as Assam Hill {17], Beetal [17,34], Jamunapari [35], Sirohi [36], and Mahabubnagar [37]. Variations can result from breed variances, environmental contexts, sample sizes, inbreeding levels, or model assumptions. The LRT found substantial maternal genetic and environmental impacts on BW, WW, and ADG due to uterine environment, pregnancy nutrition, mother behavior, litter size, or milk yield [32].

Although there was a negative direct–maternal genetic covariance for birth weight, maternal genetic effects were non-significant for weaning weight and average daily increase, and Model 3 did not significantly outperform Model 2. Thus, maternal genetic selection is not statistically justifiable for the majority of growth traits. Genetic improvement should instead focus on direct selection for WW and ADG. For BW, where maternal effects are strong, selection response will be sluggish due to poor heredity and antagonistic covariance; as such, BW should receive minimal selection emphasis. Management interventions, such as ideal kidding seasons, lactation supplementation, and support for first-parity does, provide immediate and significant benefits while also improving long-term environmental consequences. A community-based recording and genetic evaluation system is suggested to improve selection intensity and accuracy. Genomic tools may be included later, when costs fall. This combination technique of direct genetic selection and enhanced husbandry is the most viable and statistically justifiable pathway for long-term growth improvement in Salem Black goats.

4.3. Implications for Conservation of Salem Goats in India

The Salem goat, an indigenous species largely raised in the southern Indian states of Tamil Nadu and Kerala, plays an important role in smallholder farmers’ livelihoods, contributing significantly to meat production, rural employment, and biodiversity conservation. Salem goats, one of India’s 37 registered goat breeds, are adapted to semi-arid and tropical regions; however, they suffer problems such as crossbreeding with alien breeds, habitat loss, and climate unpredictability, all of which reduce genetic diversity. This study’s findings on growth parameters (BW, WW, ADG) provide actionable information for conservation initiatives. Moderate variability in BW (CV 19.56%) and larger variability in WW (CV 27.69%) and ADG (CV 34.60%) highlight the breed’s potential for genetic improvement through selective breeding, which could increase output without sacrificing adaptability. For example, the significant influence of non-genetic factors such as season of birth, year of birth, type of birth, and parity (Table 3 and Table 4) emphasizes the need for targeted management interventions, such as optimizing kidding seasons to favor monsoon/post-monsoon periods (seasons 2 and 3) for better growth, which is consistent with India’s variable agro-climatic zones. This could reduce kid mortality and enhance weaning rates, which are critical for supporting populations estimated at 1–2 million people.

Heritability estimates, particularly from Model 3, indicate substantial additive genetic variance for BW, suggesting effective selection for heavier births to bolster early survival—a key bottleneck in goat conservation where pre-weaning mortality can exceed 20 per cent. In order to counteract the negative additive–maternal covariance that may otherwise impede progress, the significant maternal heritability highlights the importance of maintaining dam lines with substantial maternal effects. In terms of conservation, these characteristics help India’s National Goat Mission create breed-specific breeding strategies, like in situ conservation through community cooperatives or village-based nucleus herds. Furthermore, nutritional supplementation programs for does are recommended by the positive linear regression of dam weight on WW and ADG, which enhances flock survival and progeny performance. All things considered, these results support a comprehensive strategy that combines environmental management with genetic selection for growth, which may boost Salem goat productivity and strengthen their economic feasibility to encourage farmer-led conservation in the face of India’s rising chevon demand.

5. Conclusions

This study confirmed that it is possible for Salem Black goats to respond to direct selection by finding moderate direct heritability for both weaning weight (WW) and average daily gain (ADG). Maternal genetic selection is not justified because maternal impacts on post-birth development features were insignificant. Management interventions such as lactation supplementation, promoting first-parity does, and managing kidding seasons can yield immediate and economic benefits. To increase the precision and intensity of selection, a community-based genetic examination and recording system is advised.