INTRODUCTION
The white-tailed deer (Odocoileus virginianus) is the most important species involved in diversified livestock systems and wildlife hunting activities in northeast Mexico. As a natural consequence, in northern regions, wildlife hunting activities have become very attractive as an alternative source of income for cattle producers attempting to diversify their livestock systems. In the last decade, some ranchers have shifted from extensive to semi-extensive production systems, which include artificial insemination, fawn monitoring and nursing (mainly for breeding purposes) and the local dissemination of animals, which promotes sustainable reintroduction in response to extractive actions. The purpose of a semi-extensive system is to produce large numbers of trophy-quality animals to be released into natural conditions after weaning. According to the General Wildlife Law of Mexico (DOF 2000), diversified livestock systems as an assumed sustainable system, must have rules to prevent the hunting of females and young bucks within preserved areas, and genetic variability is maintained through the regular introduction of breeding males and germplasm from the appropriate subspecies into captive populations.
One advantage of these systems is the possibility of recording data and of eventually implementing breeding strategies for genetic improvement. Birth weight (BW), which is the first indicator of the productive life history of animals, is one of the easiest traits to record and serves as an implicit indicator of performance at older ages. Adverse conditions during the early stages of development affect growth and a variety of related traits, including metabolism, immunocompetence and subsequent sexual attractiveness in adulthood (Lindström Reference Lindström1999). In red deer, BW has been associated with total lifetime reproductive success in males (Kruuk et al. Reference Kruuk, Clutton-Brock, Rose and Guinness1999), and some evidence link low BW with the probability of death or disease in later stages of life (Sieber et al. Reference Sieber, Robert, Schybli, Sager, Miserez, Engels and Ryser-Degiorgis2010).
Both genetic and environmental factors and their interactions promote early developmental traits in mammals. In white-tailed deer, there are a few studies focused on understanding environmental or non-genetic and genetic effects on various important traits (Williams et al. Reference Williams, Krueger and Harmel1994; Schultz & Johnson Reference Schultz and Johnson1995; Blaylock Reference Blaylock2008; Monteith et al. Reference Monteith, Schmitz, Jenks, Delger and Bowyer2009). For example, Monteith et al. (Reference Monteith, Schmitz, Jenks, Delger and Bowyer2009) identified the influence of the maternal and grandmaternal condition during gestation on the subsequent growth of offspring and noted the significance of nutrition during the gestation of white-tailed deer. Similarly, a few studies have analysed genetic variability and its effect on economic traits in deer (e.g. Williams et al. Reference Williams, Krueger and Harmel1994).
The use of mixed models has been proposed as an unbiased way to analyse phenotypic data to isolate direct and maternal genetic variance components and explain how genes influence phenotypes (Kruuk & Hadfield Reference Kruuk and Hadfield2007). Animal models involving mixed model equations have been extensively employed in genetic evaluations of commercial traits in cattle, beginning in the last century (e.g. Golden et al. Reference Golden, Garrick and Benyshek2009).
As key components of genetic change and/or evolution, genetic and non-genetic factors need to be distinguished. In natural populations, environmental effects may confound estimates of the effect of genetic variation on phenotypes. Combining long-term data on certain traits in wild populations subjected to breeding schemes in statistical models can effectively separate genetic and environmental components, and such information can then be used for management or breeding purposes (Kruuk & Hadfield Reference Kruuk and Hadfield2007).
In the present study, was analysed the data available from a captive population of white-tailed deer to describe the main sources of non-genetic and genetic variation that affect BW.
MATERIALS AND METHODS
Data source
The analysis was performed using the production records from a hunting ranch in Nuevo Laredo, Tamaulipas, Mexico. This region is characterized as exhibiting a semi-arid climate with a mean annual of rainfall of 40 cm. The vegetation has been classified as Tamaulipan thorn scrub of the south Texas plain (Fulbright & Ortega-S Reference Fulbright and Ortega-S2006).
The database contained 765 BW records obtained from 2005 to 2011 and information on the genealogy of 1052 fawns, 64 bucks and 275 does. The transition to semi-extensive animal management began in 1999 with the construction of management pens and the confinement of native animals that were already present in the area. Natural mating was employed until 2005. Then, from 2006 to the present, assisted reproduction has been performed, first through intra-cervical insemination and later via laparoscopic insemination. Reproductive management is typically scheduled for the first week of December, and fawning begins approximately in the last week of June. The insemination success rate according the ranch owner is approximately 50%, and pregnancy is assured through the natural mating of non-pregnant does. Although no data on age could be obtained, age range of the does was estimated to be between 2·5 and 6 years.
Pregnant does were maintained in enclosed areas of 0·5 ha to monitor the gestational period and fawning. The feeding management procedures included natural grass and shrub grazing supplemented with a balanced pelleted commercial feed. Minerals and vitamins were provided, and water was available to the animals as all times. Birth weight was recorded 1 or 2 days after fawning using a mechanical spring weighing scale, verifying the birth type and doe identification. When necessary, paternity test verification was requested from DNA Solutions Inc. Laboratory, because most of the animals on the ranch are registered in the North American Deer Registry of the North American Deer Farmers Association (NADeFA) and the Texas Deer Association (TDA) in the USA.
Data analyses
The data were first analysed to identify the significant non-genetic factors affecting BW using the model Y ijk=μ+S i+A j+T k+ε ijk, where Y ijk is the BW; μ is the overall mean; S i is the ith effect of the sex of the fawn (male or female); A j is the jth effect of the year (2005,…, 2011); T k is the kth effect of birth type (single, twin or triple); and ε ijk is the random error. Analyses were performed using the generalized linear model (GLM) procedure in SAS software (2001), and least-square means (LSM) were subsequently estimated and compared using the PDIFF method in the same software. First-degree interactions were excluded, as they had no significant effect on BW.
Variance component estimation
The original database was edited to fit an animal model in MTDFREML genetic evaluation software (Boldman et al. Reference Boldman, Kriese, Van Vleck, Van Tassell and Kachman1995). Several models were fitted, as follows:
$$\eqalign{y & = X\beta + Zd + Wm + Mc + e,\;{\mathop{\rm cov}} _{dm} \ne 0 \cr y & = X\beta + Zd + Wm + e,\;{\mathop{\rm cov}} _{dm} \ne 0 \cr y & = X\beta + Zd + Wm + e,\;{\mathop{\rm cov}} _{dm} = 0 \cr y & = X\beta + Zd + Mc + e \cr y & = X\beta + Zd + e,} $$where y is the phenotypic trait vector for BW; X, Z, W and M is known matrices of incidence related to their respective fixed (sex, year and type of birth type) and random (genetic and correlated effects) vectors; β is the vector of fixed effects; d is the vector of random direct additive effects; m is the vector of random maternal additive effects; covdm is the covariance between the direct and maternal genetic effects; c is the vector of random common maternal effects; and e is the vector of random residual effects. A likelihood ratio test (Sorensen Reference Sorensen2004), was performed subsequently to determine the most suitable model for BW.
The best model selected included only direct genetic and common maternal effects (P<0·01). The following variance components were estimated: direct genetic variance (σ 2d), common maternal variance (σ 2c), the variance of residual environmental effects (σ 2e) and phenotypic variance (σ 2p). From the variance components, the direct heritability (h 2), proportion of common maternal effects relative to phenotypic variance (c 2) and proportion of environmental effects relative to phenotypic variance (e 2) were estimated and discussed.
RESULTS
The overall mean BW was 2·7 kg. Three non-genetic factors were included in the analyses to examine their effect on the BW of white-tailed fawns. All three factors had a highly significant effect on the examined trait (P<0·001). The LSM for these sources of variation are shown in Table 1. BW fluctuated in each evaluated year, but no pattern in this factor was identified. Birth type significantly affected BW, as BW was higher in single births compared with twin and triplet births (Table 1). Compared with the LSM obtained for single births of 2·97±0·043 kg, the deviations for twin and triple fawning were −0·26 and −0·64 kg, respectively. Similarly, the sex of the white-tailed fawns corresponded to the observed BW dimorphism, as males were approximately 0·20 kg heavier than females at birth (P<0·001). No interactions between the factors were identified.
Table 1. Least-square means±s.e. for birth weight (BW) in white-tailed deer by sex, year and birth type
C, Within a factor, means without a common superscript are significantly different (P<0·001).
The best model selected by likelihood ratio testing included random direct genetic effects and common environmental effects of does (P<0·01). The variance components and genetic parameters are reported in Table 2. The direct genetic effects showed a moderate effect size (0·28±0·126), and interestingly, the common maternal environmental effects also explained a proportion of the BW phenotype (0·15±0·062). The residual variance explained more than half of the phenotypic variance in BW (Table 2).
Table 2. Variance components and genetic parameters for the birth weight of captive white-tailed deer (Odocoileus virginianus)
σ 2d, direct genetic variance; σ 2c,maternal permanent variance; σ 2e, environmental variance; σ 2p, phenotypic variance; h 2, direct heritability; c 2,proportion of maternal environmental effects relative to phenotypic variance; e 2, environmental effects relative to phenotypic variance.
DISCUSSION
Although there is theoretically an ideal BW in mammalian species, there is a strong genetic component accounting for the variation in this trait (Gardner et al. Reference Gardner, Buttery, Daniel and Symonds2007). As BW extremes are associated with negative conditions (e.g., low BW with neonatal mortality and high BW with dystocia), understanding these and other factors such as environmental ones, will improve our ability to manage the main sources of variation related to white-tailed deer farming.
Non-genetic factors
Knowledge of factors affecting variation in BW is especially important given the positive relationships of this trait with neonatal and adult health (Lindström Reference Lindström1999; Lumma & Clutton-Brock Reference Lumma and Clutton-Brock2002; Gardner et al. Reference Gardner, Buttery, Daniel and Symonds2007), fertility and secondary sexual traits (Kruuk et al. Reference Kruuk, Clutton-Brock, Rose and Guinness1999; Lindström Reference Lindström1999; Schmidt et al. Reference Schmidt, Stien, Albon and Guinnes2001) and body mass in mature stages in white-tailed deer (Schultz & Johnson Reference Schultz and Johnson1995) and, in larger ungulate species (e.g., cattle).
The year of birth had a significant effect on the analysed trait. This factor has been associated with BW variation and with climatic factors conditioning food availability (e.g. Schmidt et al. Reference Schmidt, Stien, Albon and Guinnes2001). Some studies performed in northern latitudes have related BW variation to differences in spring climate favouring years with warm springs (Adams Reference Adams2005; Nussey et al. Reference Nussey, Clutton-Brook, Albon, Pemberton and Kruuk2005). Although the direct relationship of yearly variability with BW in natural populations is clear, in completely captive systems, less yearly variation would be expected. While in the present case, winter severity is not expected to be a determinant factor based on temperature extremes, precipitation is a key factor conditioning late winter and spring grass and shrub availability, and the observed differences may be attributed to the variation in this parameter affecting feed availability for the maintenance and gestational requirements of does.
Birth type (i.e., litter size) also had an important effect on BW, as expected, with a direct negative relationship being indicated between litter size and BW. This finding is in agreement with what has been reported in Roe deer (Andersen & Linnell Reference Andersen and Linnell1997), sheep (Gardner et al. Reference Gardner, Buttery, Daniel and Symonds2007), goats (Soundararajan & Sivakumar Reference Soundararajan and Sivakumar2011) and, previously, in white-tailed deer (Blaylock Reference Blaylock2008).
The sex of the fawns explained less of the variation in BW compared with the other factors. Proportionally, female fawns displayed a weight that was 93% of that in males at birth. This finding confirmed the sexual dimorphism observed in white-tailed deer for this trait (Sauer Reference Sauer and Halls1984). The mean values recorded in the present study were slightly lower than the BWs of captive male and female white-tailed deer recorded in Mississippi, although the pattern of divergence between the sexes was maintained, with BWs of 3·18 and 3·06 kg being found for males and females, respectively (Jacobson Reference Jacobson, Miller and Marchinton1995). Blaylock (Reference Blaylock2008) reported same trend in wild white-tailed deer from Mississippi, although with lower means being obtained, observing an average female BW that was 88% of male birth mass. Physiologically, this trend may have a sex-specific origin in foetal growth because of the endocrine effects of androgens and anti-Mullerian hormone (Gardner et al. Reference Gardner, Buttery, Daniel and Symonds2007).
Genetic effects
There are a few reports on genetic parameters estimates in deer species for BW compared with published studies performed in other domestic animals where have found that BW is moderately to highly heritable (Koots et al. Reference Koots, Gibson and Wilton1994; Van Vleck et al. Reference Van Vleck, Snowder and Hanford2003; Hanford et al. Reference Hanford, Van Vleck and Snowder2005, Reference Hanford, Van Vleck and Snowder2006; Borg et al. Reference Borg, Notter and Kott2009; Gowane et al. Reference Gowane, Chopra, Prakash and Arora2010; Shokrollahi & Baneh Reference Shokrollahi and Baneh2012). For example, for different breeds of sheep and beef cattle have been estimated direct heritabilities from 0·17 to 0·39 and 0·30 to 0·45, respectively.
Conversely, in white-tailed deer, what little evidence there is indicates that BW in deer species displays low heritability (Williams et al. Reference Williams, Krueger and Harmel1994; Delgadillo et al. Reference Delgadillo, López, Montaldo, Berruecos, Luna and Vásquez2008; Ramírez-Valverde et al. Reference Ramírez-Valverde, Sánchez-Cervantes, García-Muñiz, Núñez-Domínguez and Lemus-Ramírez2011). Estimates of BW heritability conducted in a captive white-tailed deer population in Texas ranged from 0·00 to 0·17, depending on the method of estimation (Williams et al. Reference Williams, Krueger and Harmel1994). Similarly, Delgadillo et al. (Reference Delgadillo, López, Montaldo, Berruecos, Luna and Vásquez2008) reported direct heritabilities of 0·11–0·33, depending on the statistical model used for estimation, in a red deer population in Mexico. Furthermore, in a recent analysis of a red deer population, Ramirez-Valverde et al. (Reference Ramírez-Valverde, Sánchez-Cervantes, García-Muñiz, Núñez-Domínguez and Lemus-Ramírez2011) reported an h 2 of 0·00 under captive conditions based on using a model including only direct effects.
Here the h 2 estimated had moderate influence on BW, suggesting possibility for genetic change if used as selection criterion. Similarly, McManus & Hamilton (Reference Mcmanus and Hamilton1991) estimated moderate heritabilities for BW in red deer from Scotland (0·27±0·04). They explained that, although BW is not used as primary selection criterion, its utility may be important in case of calving difficulty occurrence or for indirect genetic change of positive correlated traits such as weaning weight and mature weights.
The estimated c 2 value suggested some influence (i.e., 15%) of common maternal effects on the BW phenotype. Hanford et al. (Reference Hanford, Van Vleck and Snowder2005, Reference Hanford, Van Vleck and Snowder2006) reported lower estimates of this parameter (c 2=0·07 and 0·10, respectively) in Polipay and Rambouillet sheep, but Gowane et al. (Reference Gowane, Chopra, Prakash and Arora2010) reported similar permanent effects on BW in Malpura sheep. Even when reports define this effect as permanent for BW, it must be considered as a special case of a maternal effect due to common maternal effects on progeny (Kruuk & Hadfield Reference Kruuk and Hadfield2007).
This important effect is mostly conditioned by different factors related to the mothering ability and capacity during the gestational period (Mrode Reference Mrode1996; Matika et al. Reference Matika, Van Wyk, Erasmus and Baker2003; Adams Reference Adams2005; Nussey et al. Reference Nussey, Clutton-Brook, Albon, Pemberton and Kruuk2005; Gardner et al. Reference Gardner, Buttery, Daniel and Symonds2007; Räsänen & Kruuk Reference Räsänen and Kruuk2007). The resemblance of progeny from the same doe could be because of maternal gestational conditions, which are closely tied to passive immunity, cytoplasmic inheritance, nutrition and maternal constraints on foetal growth (Bradford Reference Bradford1972; Cundiff Reference Cundiff1972; Gardner et al. Reference Gardner, Buttery, Daniel and Symonds2007).
Therefore, these common maternal effects can be explained further by the physiological nature of multiple fawning conditioned by both environmental variance and, to a certain degree, by the genetic basis of maternal ability. As discussed above, common effects may partly explain the effects of the maternal intrauterine environment (Matika et al. Reference Matika, Van Wyk, Erasmus and Baker2003). However, the genetic basis of maternal ability can be affected by ovulation rate (Bradford Reference Bradford1972), which is a trait that is subject to large genetic effects in some cases (Davis Reference Davis2005), and by the cytoplasmic inheritance-influencing size because zygote development is impacted by the gamete size (Reinhold Reference Reinhold2002; Van Vleck et al. Reference Van Vleck, Snowder and Hanford2003). Consequently, differences in maternal ability between fawns would cause environmental variance among families. Therefore, for BW in white-tailed deer, it would be very important to consider common maternal effects in determining whether selection is based on a doe's ability alone or on a combination of breeding parameters related to direct genetic effects and the performance of does from the same herd (Mrode Reference Mrode1996).
Determining the real genetic sources of variation also relies on the possibility of unravelling confounding environmental effects from direct genetic effects based on good data and effective statistical models (Kruuk & Hadfield Reference Kruuk and Hadfield2007). As noted by Bijma (Reference Bijma2006), data structure can act as a strong limiting factor in adjusting complex models and determining unbiased genetic estimators. Accordingly, the model analysed in the present study was assumed to exhibit some constraints, such as the failure to include certain doe-related factors (e.g., doe body condition and age). Were assumed these factors to have a minimum effect because of the regular management regime and age range under the captive conditions of the examined deer. However, the data analysis allowed the adjustment and selection of best-fitted model for all of the important genetic and correlated effects in the BW analysis, as suggested by Kruuk & Hadfield (Reference Kruuk and Hadfield2007).
Based on the present results, it can be concluded that sex, year and birth type are important sources of variation that are significantly related to BW in white-tailed fawns under farmed conditions, suggesting the need to pay attention to these factors in the management of white-tailed deer breeding farms. Additive direct genetic effects have a moderate influence on BW. However, maternal environmental effects displayed some importance as a source of variation in the BW of white-tailed fawns. Under the present study conditions, when changes in the breeding regime are considered, genetic evaluations including the examination of direct and environmental maternal effects might be taken into account to achieve sustained genetic progress in BW and positive correlated economic traits.
The authors would like to thank the Instituto Politécnico Nacional through SIP20090633-20100513 and SIP20110434 project financing, also to Asociación Nacional de Criadores de Cervidos de Mexico A.G. and the Treviño Martinez brothers from San Vicente Ranch for their genuine interest and support of this collaborative research.
