Livestock play an important role in Indian economy and provides regular source of income and thus help smoothen household consumption, particularly during the drought period (Rani et al., 2025). Dairy sector has slowly emerged as lifeline of livestock sector as well as agriculture and allied sectors in India (Kaur and Singla, 2018). The average milk productivity of buffaloes in India is much higher (6.9 kg/day/animal) than indigenous cattle (3.9 Kg/day/animal) which revealed the importance of buffalo as compared to cattle (BAHS, 2017). India has the largest buffalo population of the world having 109.8 million buffaloes. Total buffalo population is 13.7 million in Rajasthan (Livestock Census, 2019). Surti buffalo population was estimated as 0.38 million in south eastern part of Rajasthan (BAHS, 2017). Surti buffalo are incredibly important for the livelihood in the southern eastern region of Rajasthan because of its characteristics like better adaptability to the hilly region, higher disease resistance, low cost of maintenance, high milk fat yield and higher reproductive efficiency. Genetic analysis of economic traits of Surti buffalo is imperative requirement for promotion of rural livelihood and to provide the nutritional security for tribes in south eastern region of Rajasthan. Profitability of farmers is required for promotion of rural livelihood and it is based on evaluation of production and reproductive traits in population. The earlier studies conducted in cattle and buffaloes have revealed quite large predictability by the use of test-day milk yields because of high association between test-day milk yields and first lactation milk yield. Therefore, the present study was carried out to find out genetic and non-genetic factor affecting on different test day milk records and first lactation milk yield and their inheritance pattern. At the present time most of the countries are using monthly test-day milk yield records using random regression model (RRM) as an alternative of 305-day milk yield for genetic evaluation of dairy animals. The RRM is used for the analysis of longitudinal data on individual over time measured on trajectory. The estimates of genetic parameters for milk production and reproduction traits help in preparing the breeding programme for buffalo improvement and it also helpful to predict the response to selection and to choose breeding plan for future improvement in dairy animal. Estimation of genetic parameters for production traits is important to know the genetic basis for production performance of the animal and to access how much improvement is possible in the next generations. Hence, the present investigation was carried out to know the effect of various genetic factors on 305 day milk yield  and reproduction traits of Surti buffaloes.

The data on production and reproduction performance records on Surti buffaloes herd maintained at livestock research station (LRS), Vallabhnagar for ICAR network project on buffalo improvement were utilized for the present study. The farm is located in western part of India and situated at 582 m above mean sea level on 24°65*N latitude and 74.02*E longitudes, which characterized with semi-arid climate with undulated topography having average rainfall of 586 MM and CV 31% (Singh et al., 2013) Similarly, the temperature ranges from 2.3°C to 42.3°C. A total record of 649 parity belongs to 203 animals maintained at  network project on buffalo improvement (NPBI) surti unit of livestock research station (LRS), Vallabhnagar (Udaipur) spread over a period from 1993 to 2018 were analyzed to study the effects on genetic and non-genetic factors on performance traits in Surti buffaloes (Table 1). The model used examine the effect of non-genetic factors was as follows:
 

Y_ijkl = µ + P_i + S_j + P_k + e_ijkl

       
Y_ijkl= Observation of the l^th individual in i^th season, j^th period and k^th pairty. 
µ= Overall population mean.
S_j= Effect of j^th season of calving.
P_i= Effect of i^th period of calving.
P_k= Effect of k^th pairty of individual.
e_ijkl = Random error, NID.


       
The heritability values were estimated by using restricted maximum likelihood (REML) procedure as proposed by Patterson and Thompson (1971) fitted an individual animal model, sire model and repeatability model. Best linear unbiased prediction estimates of breeding values of individual animals were estimated by fitting animal model using WOMBAT statistical analysis software (Henderson, 1975).
 

y = xb + zu + e

 
Where,
Y= Vector of observations for i^th trait (i = 1, 2, 3).
b= Vector of observations of unknown i^th fixed effects (season, period and parity).
u= Vector of observations of unknown i^th random effect (sire).
e= Vector of random error.
       
Matrices for fixed (X) and random (Z) effects were as follows:

 
y_i= Vector of observations for i^th traits (i= 1 2 3).
b_i= Vector of fixed effect of period (1,2,…5) and season and pairty.
µ_i = Vector of random additive genetic effect of animal for i^th traits.
x_i and z_i= Design matrices for fixed and random animal effects respectively.
       
With E_(y) = x_b and variance covariance structure as given by:

 
Where,
Var(u)= G
Var (e)= R
Cov (u, e)= 0
Var(y)= V=   ZGZ+R
T= Matrix (3×3) of additive genetic variance and covariance.
A= Numerator relationship matrix.
E= Residual co variance matrix.
*= Direct product operator.
       
The univariate animal repeatability model was also run using WOMBAT for genetic evaluation of all lactation data. When there are more than one record on an animal for a trait, then the genetic evaluation and breeding value prediction can be done by the repeatability model. The repeatability model not only estimates the breeding value of an animal but also derives its permanent environmental effects. The repeatability is:
 

y = X_b + Z_a + W_pe + e

 
Where,
p_e= Vector of permanent environmental effects and non-genetic effects.
W= Incidence matrix relating records to permanent environmental effects.
       
The permanent environmental effects and residual effects are assumed to be normally independently distributed with means zero and variance σ²pe and σ²e, respectively. Therefore:

       
The phenotypic structure for three observations of an individual under this model is:
 

Mean performance
 
The analysis of variance showed that mean squares for non-genetic factors were significant for 305 DMY and PY traits. This indicates that existence of high degree of variability in the season and parity of calving to be incorporated during the formulation of breeding programme for Surti buffalo and that also reflected in the broad ranges observed for performance traits (Table 3). The overall least square means along with standard errors for 305 days or less milk yield (305DMY), Lactation length (LL), Peak yield (PY) and age at first calving (AFC) have been presented in Table 2. Season period and parity of calving had significant effect on production traits except lactation length which was mainly influenced by period of calving. The buffaloes that calved in winter season showed higher value for 305DMY, peak yield and Lactation length and lower AFC. The highest value of total milk yield, 305 day milk yield and peak yield were observed during the first period (1993 -1997). All milk production traits were found highest in 5^th parity and lowest in first parity of animal except Lactation length which was lowest presented in 3^rd parity of animal. The buffaloes that calved fourth period (2008-2012) had the lowest age at first calving. Production performance over the parity may be due to gradual growth of mammary tissue and associated physiological changes in the body of the animal. Decrease in milk production for higher Parities (5 and 6) may be due to mammary tissue damage in older animals.




 
Genetic parameter
 
The REML estimates of variance components and genetic parameter along with standard errors for 305 days or less milk yield, lactation length, peak yield and age at first calving has been presented in Table 4. Simple univariate animal model partitioned the total phenotypic variance (σ2p=81741.1) into additive and residual variance for age at first calving. It showed heritability (0.38±0.04) for AFC. Simple univariate sire model reduced the phenotypic variance (σ²_p=83155) with additive variance σ²_a (15337) and residual variance (σ²_e=67819). In animal model, heritability for AFC was found to be higher than sire model. Based on log L value, sire model showed better fitting than to animal model. Genetic parameter showed moderate estimates of heritability and repeatability for production performance traits. The heritability estimate of AFC was higher than the estimate of sire model and it was estimated moderate heritability as 0.38±0.04 using animal model. Therefore, genetic improvement can be achieved through selection of studied traits in performance of surti buffalo.


       
Finding of the present study showed similar estimates of heritability for 305DMY in surti buffalo by Patel (1994); Pathodiya (1997). Estimates of heritability were also in consonance with Singh et al., (2011) in nili-ravi buffaloes for peak yield. Estimates of repeatability was found to be more or less conformity with the results observed by Galsar et al. (2016) 0.12±0.04 in Mehasana buffalo. Similar estimates of heritability for AFC were also reported by Kothari (2004) and Rathod et al. (2018) and it was found as 0.36±0.15 and 0.23±0.15 in surti buffalo. Contrary to this, low heritability estimates was reported by Galsar et al., (2016) in Mehsana buffalo, Thevamanoharan et al. (2002) in Nili-Ravi. Higher estimates of heritability was reported by Rana et al., (2021) for first lactation 305DMY and peak yield in Murrah buffalo and by Rathod et al., (2018) for LL and 305 DMY in surti buffalo. Lower estimates of heritability for AFC was also observed Sujit and Sadan (2000) and Warade et al. (2005) in buffalo.

The present study was undertaken on production and fertility records of surti  buffaloes maintained at NPBI unit of  Livestock Research Station,Vallabhnagar over  a  period  of  25 years  (1993-2018). The  data  were  classified  according  to  season,  period,  and  parity of calving to  study  the  effect  of   genetic factors. The buffaloes calved in winter season showed better performance for 305DMY, for lactation length, peak yield and age at first calving. The variation in performance traits among herds is mainly due to differences in feed resources and environmental conditions of the farm. The parity of animal had significant effect on 305 day milk yield, peak yield and non-significant effect on lactation length. An increasing trend was observed for all production traits from first parity to fifth parity then decreasing trend was observed from fifth to sixth parity of buffalo. Therefore, increase in parity of animal gradual increase production performance of surti buffalo. Production performance over the parity was observed due to gradual growth of mammary tissue and associated physiological changes in the body of the animal. Decrease in milk production during higher parities may be due to mammary tissue damage in older animals. The estimates of genetic parameters for milk production and reproduction traits help in preparing the breeding programme for buffalo improvement and it also helpful to predict the response to selection and to choose breeding plan for future improvement in dairy animal. The moderate to higher estimates of heritability of the traits  like age at first calving, 305DMY and  lactation length peak yield  revealed  that these traits can be used to evolve  multi trait selection criteria for buffalo evaluation. The moderate to higher estimates of heritability of age at first calving (AFC), 305 DMY, lactation length and peak yield revealed that these  traits can be used as multi trait selection criteria for genetic  improvement of surti buffalo in South eastern part of Rajasthan. This study can be used to select the genetically superior Surti buffalo for production and reproduction traits.

Authors declare that there is no conflict of interest regarding the publication of the article.