- Email: [email protected]
Genetic variation and correlated changes in reproductive performance of a red tilapia line selected for improved growth over three generations
Animal Reproduction Science xxx (xxxx) xxx–xxx
Contents lists available at ScienceDirect
Animal Reproduction Science journal homepage: www.elsevier.com/locate/anireprosci
Genetic variation and correlated changes in reproductive performance of a red tilapia line selected for improved growth over three generations ⁎
Ngo Phu Thoaa, , Azhar Hamzahb, Nguyen Hong Nguyenc a b c
Research Institute for Aquaculture No.1, Tu Son, Bac Ninh, Viet Nam National Prawn Fry Production & Research Center, Kg. Pulau Sayak, 08500 Kota Kuala Muda, Kedah, Malaysia University of the Sunshine Coast, Maroochydore, QLD 4558, Australia
AR TI CLE I NF O
AB S T R A CT
Keywords: Genetics Fecundity Fitness related traits Selection and genetic improvement
The present study examines genetic variation and correlated changes in reproductive performance traits in a red tilapia (Oreochromis spp.) population selected over three generations for improved growth. A total of 328 breeding females (offspring of 111 sires and 118 dams) had measurements of body weight prior to spawning (WBS), number of fry at hatching (NFH), total fry weight (TFW) and number of dead fry (NDF) or mortality of fry including unhatched eggs at hatching (MFH). Restricted maximum likelihood (REML) analysis in a multi-trait model showed that there are heritable genetic components for all traits studied. The heritability for WBS was very high (0.80). The estimates for traits related to fecundity (NFH, TFW) and survival (NDF) were low and they were associated with high standard errors. Genetic correlations of WBS with other reproductive performance traits (NFH, TFW and NDF) were generally positive. However, NFH was negatively correlated genetically with TFW. As expected, body measurements during growth stage exhibited strong positive genetic correlations with WBS. The genetic correlations between body traits and reproductive performance (NFH, TFW, NDF) were not significant. Correlated responses in reproductive traits were measured as changes in least squares means between generations or spawning years. Except for WBS that increased with the selection programs, the phenotypic changes in other reproductive traits observed were not statistically significant (P > 0.05). It is concluded that the selection program for red tilapia has resulted in very little changes in reproductive performance of the animals after three generations. However, periodic monitoring of genetic changes in fecundity and fitness related traits such as NDF or MFH should be made in selective breeding programs for red tilapia.
1. Introduction Body weight has been the sole selection criterion in majority of genetic improvement programs for farmed aquaculture species because it is the primary determinant of animal performance and market price (Nguyen, 2016). However, reproductive performance and fitness related traits are also important to hatcheries to maximize revenue and return per unit of production. It is, therefore, crucial to understand the consequences of selecting for increased body weight on other traits of commercial importance, including reproductive performance.
⁎
Corresponding author. E-mail address: [email protected] (N.P. Thoa).
http://dx.doi.org/10.1016/j.anireprosci.2017.07.003 Received 15 March 2017; Received in revised form 26 May 2017; Accepted 4 July 2017 0378-4320/ © 2017 Published by Elsevier B.V.
Please cite this article as: Thoa, N.P., Animal Reproduction Science (2017), http://dx.doi.org/10.1016/j.anireprosci.2017.07.003
Animal Reproduction Science xxx (xxxx) xxx–xxx
N.P. Thoa et al.
Several studies in model species or farmed animals have shown that selection for high growth rate might not have negative effects on reproductive traits (Bünger et al., 2005), whereas breeding objectives placing emphasis on high efficiency of meat production might result in a reduction in fertility and litter size of the animals as well as increase the incidence of genetic defects (Rauw et al., 1998). In aquatic animal species, the genetic relationships between body traits and other economically important traits have been examined in unselected populations (Gall and Neira, 2004; Gjerde, 1986; Kristjánsson and Arnason, 2014; Su et al., 2002). Whereas the heritability and correlations have enabled the prediction of the consequences of selection for increased body weight on traits of economic importance, there have been limited selection experiments to measure the realised response in reproductive traits to selection for improved growth rate. To date, only one study was conducted to examine correlated responses in reproductive traits to the long term selection program for genetically improved farmed tilapia (GIFT) strain (Hamzah et al., 2016). A selective breeding program for a red tilapia population was conducted since 2010 to improve production performance, survival and body colour in Malaysia. Substantial direct response was achieved for body weight (the main selection criterion) after three generations (years) of selection (one generation per year). Although selection was practised solely on growth, correlated responses in other dimensions of body measurements, body colour and survival were also achieved (Nguyen, unpublished results). Nevertheless, genetic changes in reproductive traits as an indirect consequence of short term selection for enhanced growth are still unknown. In this present paper, we hypothesise that there have been no significant impacts on reproductive performance traits of a red tilapia population which has underwent three generations of selection from 2008 to 2012. In addition, we examined quantitative genetic variation in reproductive characters and their relationships with growth performance. The results obtained from the present study are expected to assist with the review and monitoring of genetic progress achieved in the genetic programs. If unfavourable changes in reproductive traits occurred, the breeding objectives for red tilapia would need to be refined to sustain long term response to selection in a preferred direction. 2. Materials and methods 2.1. Origin of the red tilapia strain Data were collected from three generations of selection for improved growth in a population of red tilapia (Oreochromis spp.) at Aquaculture Extension Centre of Department of Fisheries, Malaysia. A full detailed account of the population, family production, selection procedures and management of the animals are given in Nguyen et al. (2017). In brief, the base population (G0) was established in 2009 from a full 3 × 3 diallele cross involving three strains which originated from hatcheries in Malaysia, Taiwan and Thailand. The selected line was formed by within- family sampling from 103 families (four to 10 females and males per family). They came from 103 families produced from the full diallele cross in 2008. The line was selected for high breeding values for body weight at harvest. A combined between and within family selection was practised in the selected line. The average proportion of selected animals was about 3% in females and 2.3% in males. Note, however, that selection was on breeding values but not by truncation (due to the inability of some selected breeders to reproduce or due to mortality we had to resort to selecting lower ranking fish; also, the number of selected individuals contributed by each family was restricted to avoid later inbreeding). In generations G0 (year 2009), G1 (year 2010), G2 (year 2011) and G3 (year 2012) they were the progenies of 32–50 dams and 28–51 sires for the selected line. Matings were made among genetically unrelated broodstock based on their estimated breeding values (EBVs) and their relationship to other animals in the pedigree to produce full-sibs and (paternal) half-sibs families. Over three generations, a total of 328 breeding females (offspring of 111 sires and 116 dams) successfully produced progenies used for the analyses in this study (Table 1). The breeding scheme is given in Nguyen (2016). 2.2. Family production, rearing to harvest Family (full- and half-sibs) production was normally completed within 30–45 days. Mating was conducted in hapas, following the mating design prepared from annual routine genetic evaluation and mate allocation analyses. After 7 days of mating, only fertilized eggs at eyed stage were collected from the mouth of the female and immediately transferred to hatching jars for artificial incubation. Fry often hatched after about 5–7 days. Soon after yolk-sac absorption, the hatched fry of each family were transferred from the incubators to the nursery hapas (1 m3 with 2 mm mesh size), stocked at a density of 200 fry per m3. At least three nursery hapa replicates for each family were maintained in the same pond to reduce environmental differences between families. When the fingerlings reached an average weight of 5–10 g, about 100–150 individuals per family were randomly sampled and physically Table 1 Number of breeding female, sire and dam in the population. Spawning years
Population
Breeding female
Sire
Dam
2009 2010 2011 2012 Total
Base Selection Selection Selection
110 78 65 75 328
51 28 32 111
50 32 34 116
2
Animal Reproduction Science xxx (xxxx) xxx–xxx
N.P. Thoa et al.
Table 2 Basic statistics of reproduction and growth characteristics of female red tilapia. Category
Traits
Unit
N
Mean
SD
CV
Min
Max
Reproduction
WBS NFH TFW NDF
g no. kg no.
258 318 318 308
237.4 658.5 0.9 14.8
60.2 384.7 1.6 48.0
25.4 58.4 176.8 324.9
101.3 65 0.1 0
397.8 2120 13 748
Growth at harvest
WT LG DP WD
g cm cm cm
216 216 216 216
272.2 18.5 8.0 3.5
113.0 2.5 1.3 0.5
41.5 13.6 16.8 14.7
96.7 12.4 5.2 2.2
616 25.4 11.5 4.8
WBS = Weight before spawning, NFH = Number of fry at hatching, TFW = Total fry weight, NDF = Number of dead fry. WT = Weight, LG = Length, DP = Depth and WD = Width.
identified, using PIT (passive integrated transponder) tags. After tagging, the fingerlings were pooled and conditioned for about three days in fibreglass tanks without feeding. Then representatives of each family were sent to communal grow-out testing in freshwater earthen pond (1000 m2) and cages (3 m × 3 m × 3 m). The initial stocking density in cage was 5 fish per square meter of surface water while the density in each pond was 2–3 fish per square meter of surface water. In both testing environments, the same diet (32% protein content), feeding rate (3–5% of total body mass), culture and management practices were applied. Water quality parameters (temperature, pH, dissolved oxygen and total ammonia) were monitored once a week. The same management procedures were practised in all generations during the course of the selection program for red tilapia from stocking to harvest. At harvest, fish were measured of body traits (live weight, standard length, body width and depth). Body weight data collected at harvest were then used in the computation of estimated breeding values (EBVs) for all individuals in the pedigree. The best animals (highest EBV for body weight) were selected to become parents of the succeeding generations. In each generation, reproductive performance parameters of individual breeding females were recorded. The same breeding schemes regarding the family production, genetic evaluation and selection process were repeated in subsequent generations 2009–2012. 2.3. Measurements of reproductive traits Table 2 shows data on the reproductive traits measured in this study. Measurements of the traits were carried out during fry production. Females’ body weights before spawning (WBS) were recorded as live weight of a ready to spawn female (by visual assessment of its genital papillae condition) prior to mating. The total number of live fry at hatching (NFH) per spawning per female was recorded by counting of fry collected from each hatching jar assigned for each female breeder. The total fry weight (TFW) after hatching was measured by using a digital balance. The traits under analyses were body weight of female breeders before spawning (WBS), number of fry alive at hatching per spawning per female (NFH), total fry weight per spawning per female (TFW) and number of dead fry at hatching (NDF). The number of dead fry (NDF) at hatching included the number of eggs that were not hatched during artificial incubation. NDF was also expressed as a ratio trait (called as mortality of fry at hatching, MFH) to examine its relationship with female weight. However, this ratio trait (MFH) was not included in genetic analysis. NFH and TFW were recorded in all spawning seasons (2008–2011), whereas WBS, WAS and NDF were from three recent generations (2009–2011). 2.4. Statistical analyses 2.4.1. Genetic parameters The original scale was used to analyse data on WBS, TFW and NDF traits whereas the NFH was square root transformed in all analyses to improve the distribution of residuals. Variance (genetic and environmental) components for all the traits were estimated by restricted maximum likelihood method, applied to a standard animal model using ASReml (Gilmour et al., 2009). In matrix notation, the mixed model is written as follows: y = Xβ + Zu + e
(1)
where y is the vector of observations for reproductive traits, β is the vector of the fixed effect that included spawning season or generation (1–3), and the linear covariate of female body weight prior to spawning within line for NFH, TFW and NDF, and u is the vector of the random animal additive genetic effects ∼ (0, Aσa2 ) where A is the additive genetic (numerator) relationship matrix among the animals, and e is the vector of residual effects ∼ (0, Iσe2 ). X and Z are incidence matrices relating observations to the fixed effects and the additive genetic effect of the individual animal, respectively. The remaining effects are assumed to be distributed as var(e) = R = Iσe2 , where I is an identity matrix. The expectations of all random effects are zero, cov (u,e) = 0 and thus var (y) = ZGZ’ + WIW’ + R. In the preliminary analyses, the likelihood ratio test showed that the random effects of common effects to full-sibs and mating hapas nested within spawning season (or generation) were not significant for NFH, TFW and NDF, and they were removed from the final model for these traits. These effects on WBS and body traits at harvest were also not significant (P > 0.05). Hence, model 1 was used to analyse all traits included in this study. 3
Animal Reproduction Science xxx (xxxx) xxx–xxx
N.P. Thoa et al.
Heritabilities (h2) for all traits were estimated from a single trait model [Eq. (1)] and calculated as: h2 = additive genetic variance and the residual variance analyses, using model 1.
(σe2 ).
σˆa2 σˆa2 + σe2
where σa2 is the
Phenotypic and genetic correlations were estimated from a series of bivariate
2.4.2. Correlated selection responses The same models as described above (Eq. (1)) were used to estimate correlated responses in reproduction traits (WBS, NFH, TFW and NDF) to the selection program for improved growth in red tilapia. Furthermore, reproductive performance traits (i.e. NFH, TFW and NDF) were also analysed with a model that female body weight prior to spawning (WBS) within line was not fitted as a covariate. The correlated changes in all reproduction traits were measured as differences between least square means between generations of selection or spawning season. 2.5. Ethics statement The study and the protocol were reviewed and approved by the animal ethics committee of the Department of Fisheries, Malaysia (the approval number ANE070312). 3. Results 3.1. Basic statistics The study included a total of 328 breeding females (offspring of 116 sires and 118 dams) that had reproductive performance records collected from 2008 to 2012 (Table 1). Basic statistics including means, standard deviations, coefficients of variation, and minimum and maximum values for the reproductive traits in all generations are shown in Table 2. In the present study, fecundity is defined as the total number of fry collected at hatching from a female breeder in each spawning. When fecundity was expressed as the number of fry produced per gram of female body weight, it ranged from 2.5 to 2.8. For NDF, the coefficient of variation was large due to the limited data records for this trait and due to the high mortality that was observed in some families during incubation. 3.2. Prediction of fecundity based on female body weight prior to spawning The relationship between NFH and female body weight prior to spawning (WBS) are essentially linear (TNF = 461.1 × WBS + 0.895). However, the relationship between total fry weight (TFW) and WBS was not significant (P > 0.05). The linear regression coefficients indicated a 0.90 (s.e. 0.02) g increase in the numer of fry at hatching (NFH) and a 0.63 (s.e. 0.14) kg increase in total fry weight (TFW) for every gram increase in WBS (Fig. 1). Conversely, with increased NFH, individual fry weight (TFW divided by TNF) decreased by −1.78 × 10−6 (5.17 × 10−7) kg (results not presented). 3.3. Relationships between fecundity and total fry weight and fry mortality There was a proportional reduction of 0.0025% in mortality rate at hatching for every fry increase per spawn until the number of fry (NFH) equalled around 500 (Fig. 2A), after which there was a proportionate decrease in mortality of 0–0.1% per unit increase in
Fig. 1. Relationship between number of fry at hatching (NFH) and total fry weight (TFW, g), TFW = 0.63 × NFH + 0.006.
4
Animal Reproduction Science xxx (xxxx) xxx–xxx
N.P. Thoa et al.
Fig. 2. A) Relationship between morality of fry at hatching (MFH, %) and total number of fry (NFH), MFH = 0.413 × NFH – 2.533 × 10−5, and B) Relationship between morality of fry at hatching (MFH) and total fry weight (TFW, g), MFH = 0.043 × TFW – 0.004.
NFH. The mortality at hatching decreased rapidly with increasing total fry weight, with mortality of about 0.02% for fry with greater weight than 10 g (Fig. 2B).
3.4. Heritabilities for traits Except for the heritability (h2) for average body weight of females prior to spawning (WBS) that was fairly high (h2 = 0.80), these estimates for other reproductive traits recorded in the present study were low and not significantly different from zero (Table 3). This was due to a large proportion of variance explained by the environmental effects. The large standard errors associated with the h2 estimates for reproductive traits were likely a result of the limited sample size of breeding females recorded, although the data were collected over four generations including the base population from 2008 to 2012.
3.5. Correlations among reproductive traits Genetic and phenotypic correlations among reproductive traits are both favourable and unfavourable (Table 4). The genetic correlations between WBS and fecundity related traits (NDF and TFW) were moderate and positive (0.13–0.30), and the estimates were associated with high standard errors. On the other hand, WBS was genetically correlated negatively with fitness related traits of the fry (NFH), although magnitude of the estimates was low and not different from zero. Overall, phenotypic correlations were generally low and not significant, ranging from −0.02 to 0.16. Table 3 Variance components, heritability (h2) and common environmental effects (c2) for body traits. Category
h2
Traits
σa2
σe2
Reproduction
WBS NFH TFW NDF
2619.1 0.84 0.05 67.6
638.9 50.9 2.60 2119.6
0.80 0.02 0.02 0.03
± ± ± ±
0.16 0.11 0.12 0.12
Growth
WT LG DP WD
854.4 0.96 0.11 0.092
2679.5 1.74 0.59 0.061
0.24 0.36 0.16 0.60
± ± ± ±
0.14 0.15 0.13 0.16
σa2 = Additive genetic variance and σe2 = Residual variance. WBS = Weight before spawning, NFH = Number of fry at hatching, TFW = Total fry weight, NDF = Number of dead fry. WT = Weight, LG = Length, DP = Depth and WD = Width.
5
Animal Reproduction Science xxx (xxxx) xxx–xxx
N.P. Thoa et al.
Table 4 Phenotypic (above) and genetic (below) correlations among traits. Traits
WBS
WBS NFH TFW NDF
n.e. 0.30 ± 1.10 0.13 ± 0.84
NFH
TFW
NDF
0.16 ± 0.06
−0.03 ± 0.06
n.e. n.e.
−0.02 ± 0.06 −0.01 ± 0.06 −0.02 ± 0.06
n.e.
WBS = Weight before spawning, NFH = Number of fry at hatching, TFW = Total fry weight, NDF = Number of dead fry. WT = Weight, LG = Length, DP = Depth and WD = Width. n.e. = not estimable.
3.6. Correlations between harvest weight and reproduction traits Phenotypic (rp) and genetic correlations (rg) of body traits with reproduction traits are presented in Table 5. The genetic correlations between body traits of grow-out stocks at harvest and WBS of breeding females were close to one (0.7–0.94). There were also strong negative genetic association between body traits of the grow-out stocks and total fry weight (0.43–0.77). The genetic correlations of body traits with fecundity (NFH) were moderate and positive (0.12–0.72, respectively). The phenotypic correlations between body traits of grow-out stocks and those of females (WBS) and females’ reproductive performance (NFH and TFW) were of similar size and magnitude to those of the genetic correlations. However, both the phenotypic and genetic correlation estimates between body traits and reproductive performance (except WBS) had high standard errors and not significant different from zero. 3.7. Correlated selection responses The correlated responses in reproduction traits estimated as the changes in least squares means (LSMs) for each generation is given in Table 6. Significant differences between successive generations were observed for body weight of females prior to spawning (P < 0.01). The differences between generations or spawning seasons were not statistically significant for other reproductive traits (NFH, TFW and NDF), except that an increase in NDF in the latest generation (2012) was due to mass mortality of some families. 4. Discussions Overall, reproductive performance of the red tilapia line used in this study is comparable to other stocks reported in the literature (Eguia, 1996; Suresh and Lin, 1992; Watanabe et al., 1989). The average number of fry per gram of female body weight of the present red tilapia stock was, however, lower than those reported for Nile tilapia (2.8 vs. 3–8 fry per gram of female) (Bhujel et al., 2007; Campos-Mendoza et al., 2004; El-Sayed et al., 2003; Osure and Phelps, 2006; Siddiqui et al., 1998). Comparing fecundity among studies is relative because different strains of tilapia, size of broodstocks and breeding animals are used. Furthermore, reproductive performance of fish is largely influenced by environmental factors such as rations, testing environments and management practices. Interestingly, our results showed that the selection program for improved growth in red tilapia increased WBS. However, there were no significant changes in reproductive performance (NFH and TFW) of the animal. This is consistent with the non-significant genetic association between body traits at harvest/or female body weight at mating and fecundity (total number of fry at hatching, fry weight). The weak genetic correlations between body weight and fecundity are also reported in other aquaculture species (Arcos et al., 2004; Gall and Neira, 2004; Gjerde et al., 1994; Su et al., 1997). Additionally, it could have been because the current selection program is still at the early stage to cause any causal correlated changes in fitness related traits. We support this hypothesis on the basis of the genetic changes in the same traits measured from a long term selection program in the GIFT strain. After 6 generations of the GIFT strain in Philippines and 12 generations in Malaysia, there was also non-significant change per unit of female body weight across six measures of the reproductive traits (Hamzah et al., 2016). Selection for improved growth in farmed animals resulted in correlated improvement in birth weight in pigs and cattle (Burrow and Prayaga, 2004). However, breeding objective emphasising high efficiency of feed utilisation may impair reproductive performance of the animals (Kerr and Cameron, 1995). Except for the realised response in reproductive traits estimated in the GIFT strains, the literature is not available in other aquatic animal species to make a relative comparison with the results obtained from our present study. Table 5 Phenotypic (rp) and genetic (rg) correlations between reproductive traits and growth. Traits
WBS NFH TFW NDF
WT
LG
DP
WD
rg
rp
rg
rp
rg
rp
rg
rp
0.80 ± 0.14 0.72 ± 1.34 −0.43 ± 1.58 n.e.
0.75 ± 0.04 0.10 ± 0.07 −0.07 ± 0.08 n.e.
0.81 ± 0.14 0.74 ± 2.26 −0.72 ± 3.25 n.e.
0.69 ± 0.04 0.06 ± 0.07 −0.09 ± 0.08 n.e.
0.94 ± 0.10 0.59 ± 1.68 −0.77 ± 11.5 −0.64 ± 1.31
0.81 ± 0.03 0.12 ± 0.07 0.14 ± 0.08 −0.06 ± 0.06
0.77 ± 0.15 0.12 ± 1.13 −0.54 ± 3.77 0.07 ± 0.79
0.59 0.08 0.05 0.05
± ± ± ±
0.06 0.07 0.08 0.06
WBS = Weight before spawning, NFH = Number of fry at hatching, TFW = Total fry weight, NDF = Number of dead fry. WT = Weight, LG = Length, DP = Depth and WD = Width.
6
Animal Reproduction Science xxx (xxxx) xxx–xxx
N.P. Thoa et al.
Table 6 Least square means of reproductive traits over generations. Traits
Generation 1
WBS NFH TFW NDF
2
3
a
215.0 ± 6.38 555.2 ± 41.94 0.71 ± 0.18 a 6.61 ± 5.67 a
a
600.6 ± 49.49 0.92 ± 0.21 a 9.71 ± 6.24 a
a
271.0 ± 6.73 b 588.8 ± 43.06 a 0.94 ± 0.19 a 35.6 ± 5.43 b
WBS = Weight before spawning, NFH = Number of fry at hatching, TFW = Total fry weight, NDF = Number of dead fry. WT = Weight, LG = Length, DP = Depth and WD = Width. Means with different superscript letters in the same line are significant different (P < 0.05).
The low and non-significant heritabilities estimated for a range of reproductive traits here show that making genetic progress by direct selection for the reproductive traits, albeit possible, may be slow. With traits that are of low heritability, advanced statistical and genetic evaluation methods such as Best Linear Unbiased Prediction (BLUP) should be applied to utilise all the information available in the pedigree. Further, a large amount of data records and in-depth pedigree data should be routinely collected to obtain reliable accuracy of breeding value estimation. Also note that a zero heritability for a trait of interest does not mean that there is no genetic variation to allow scope for selection because heritability is a population parameter and as expected the reproductive traits are largely influenced by environmental factors. The low heritability for reproduction and fitness related traits have been reported in different farmed aquaculture species ranging from fish (Gima et al., 2014; Gjerde, 1986; Trọng et al., 2013a, 2013b) to crustaceans (Caballero-Zamora et al., 2015; Macbeth et al., 2007). The heritability estimates for the same traits as recorded in the present study were from 0.13 to 0.17 in the GIFT strain (Hamzah et al., 2016) or 0.03–0.25 for early survival and sexual maturity in the salinity tolerance line of Nile tilapia (Thoa et al., 2015; Thoa et al., 2016). A synthesised review from literature results together with those obtained in our present study suggest that fecundity of red tilapia can be improved through direct or indirect selection for increased growth. This is also demonstrated in the GIFT strain by Hamzah et al. (2016). However, the improvement in fecundity depends upon specific breeding objectives of genetic improvement programs to produce stocks for hatcheries or commercial production. Commercial hatcheries may prefer having highly fecund females as their main interest is selling fry to production. On the contrary, grow-out producers prefer animals that from each unit of food intake, waste least in reproduction and retain most to favour the efficient conversion of metabolic energy into growth/or protein deposition. An alternative is to develop specialised genetic lines such as a maternal line for hatcheries and a paternal line for production. This approach has been successfully implemented in well-structured animal industry and by multi-national companies (Smith, 1964). However, maintenance of multiple lines would require investments and recourses that are generally not generous in major aquaculture producing countries. There is no single simple answer in this regard. The genetic improvement of reproductive traits should be considered on the basis of its merits for individual cases. In summary, the finding of our study indicates that selection for improved growth increased female body weight before spawning. However, there were no significant differences in fecundity related traits (NFH and TFW) between generations of selection (Table 6). Although our results are generally consistent with those reported in the GIFT strain and other fishes, periodically monitoring of the selective breeding program will warrant a rigorous conclusion regarding genetic changes in reproductive characteristics to selection for high growth in this red tilapia stock. 5. Conclusions Fecundity and fitness related traits of red tilapia are lowly heritable. The genetic association for body traits in grow-out animals were retained in reproductive breeding stage in broodstock as indicated by the high and positive genetic correlations between body traits and WBS. Genetic correlations between body traits and reproductive performance, albeit positive, were associated with large standard errors and hence, the estimates were not significant. Short term selection for improved growth in the present study did not cause significant changes in reproduction productivity (NFH and TFW). Routine collection of reproductive characters should be practised to accumulate further data to obtain more reliable genetic parameter estimates as well to monitor possible genetic changes in these traits as a result of long term selection for high production performance. Conflict of interest The authors declare that they have no competing interests. Acknowledgements This study was funded by Department of Fisheries, Malaysia. We would like to thank Hoong Yip Yee and Mohd A Aziz for their kindly assistance in the management of the animals. The Aquaculture Extension Center of the Malaysian Department of Fisheries in Jitra, Kedah provided facilities and additional manpower for the study. 7
Animal Reproduction Science xxx (xxxx) xxx–xxx
N.P. Thoa et al.
References Arcos, F.G., Racotta, I.S., Ibarra, A.M., 2004. Genetic parameter estimates for reproductive traits and egg composition in Pacific white shrimp (Penaeus (Litopenaeus) vannamei). Aquaculture 236, 151–165. Bünger, L., et al., 2005. Relationships between quantitative and reproductive fitness traits in animals. Phil. Trans. R. Soc. B : Biol. Sci. 360, 1489–1502. Bhujel, R.C., Little, D.C., Hossain, A., 2007. Reproductive performance and the growth of pre-stunted and normal Nile tilapia (Oreochromis niloticus) broodfish at varying feeding rates. Aquaculture 273, 71–79. Burrow, H., Prayaga, K., 2004. Correlated responses in productive and adaptive traits and temperament following selection for growth and heat resistance in tropical beef cattle. Livest. Prod. Sci. 86, 143–161. Caballero-Zamora, A., et al., 2015. Genetic parameters for spawning and growth traits in the Pacific white shrimp (Penaeus (Litopenaeus) vannamei). Aquacult. Res. 46, 833–839. Campos-Mendoza, A., McAndrew, B., Coward, K., Bromage, N., 2004. Reproductive response of Nile tilapia (Oreochromis niloticus) to photoperiodic manipulation; effects on spawning periodicity, fecundity and egg size. Aquaculture 231, 299–314. Eguia, M.R.R., 1996. Reproductive performance of four red tilapia strains in different seed production systems. Israeli J. Aquacult. Bamidgeh 48, 10–18. El-Sayed, A.-F.M., Mansour, C.R., Ezzat, A.A., 2003. Effects of dietary protein level on spawning performance of Nile tilapia (Oreochromis niloticus) broodstock reared at different water salinities. Aquaculture 220, 619–632. Gall, G.A., Neira, R., 2004. Genetic analysis of female reproduction traits of farmed coho salmon (Oncorhyncus kisutch). Aquaculture 234, 143–154. Gima, M.E., et al., 2014. Realized heritability and response to selection for fecundity, hatching rate and fry/Kg for channel catfish females (Ictalurus punctatus) induced to ovulate and fertilized with blue catfish (Ictalurus punctatus) males for the production of hybrid catfish embryos. Aquaculture 40–421, S36–S41. Gjerde, B., Simianer, H., Refstie, T., 1994. Estimates of genetic and phenotypic parameters for body weight, growth rate and sexual maturity in Atlantic salmon. Livest. Prod. Sci. 38, 133–143. Gjerde, B., 1986. Growth and reproduction in fish and shellfish. Aquaculture 57, 37–55. Hamzah, A., et al., 2016. Genetic parameters and correlated responses in female reproductive traits in the GIFT strain. Aquacult. Res. 47, 1488–1498. Kerr, J., Cameron, N., 1995. Reproductive performance of pigs selected for components of efficient lean growth. Anim. Sci. 60, 281–290. Kristjánsson, T., Arnason, T., 2014. Strong phenotypic and genetic correlation between size and first maturity in Atlantic cod Gadus morhua L. reared in commercial conditions. Aquacult. Res. 46, 2185–2193. Macbeth, M., et al., 2007. Heritability of reproductive traits and genetic correlations with growth in the black tiger prawn Penaeus monodon reared in tanks. Aquaculture 270, 51–56. Nguyen, N.H., 2016. Genetic improvement for important farmed aquaculture species with a reference to carp, tilapia and prawns in Asia: achievements, lessons and challenges. Fish Fish. 17, 483–506. Nguyen, N.H., Hamzah, A., Ngo, P.T., 2017. Effect of genotype by environment interaction on genetic gain and genetic parameter estimates in red tilapia (Oreochromis spp.). Frontiers in Genetics 8. Osure, G.O., Phelps, R.P., 2006. Evaluation of reproductive performance and early growth of four strains of Nile tilapia (Oreochromis niloticus, L) with different histories of domestication. Aquaculture 253, 485–494. Rauw, W., Kanis, E., Noordhuizen-Stassen, E., Grommers, F., 1998. Undesirable side effects of selection for high production efficiency in farm animals: a review. Livestock Production Science 56, 15–33. Siddiqui, A., Al‐Hafedh, Y., Ali, S., 1998. Effect of dietary protein level on the reproductive performance of Nile tilapia, Oreochromis niloticus (L.). Aquaculture research 29, 349–358. Smith, C., 1964. The use of specialised sire and dam lines in selection for meat production. Anim. Prod. 6, 7–344. Su, G.-S., Liljedahl, L.-E., Gall, G.A., 1997. Genetic and environmental variation of female reproductive traits in rainbow trout (Oncorhynchus mykiss). Aquaculture 154, 115–124. Su, G.-S., Liljedahl, L.-E., Gall, G.A., 2002. Genetic correlations between body weight at different ages and with reproductive traits in rainbow trout. Aquaculture 213, 85–94. Suresh, A.V., Lin, C.K., 1992. Tilapia culture in saline waters: a review. Aquaculture 106, 201–226. Thoa, N.P., et al., 2015. Genetic variation in survival of tilapia (Oreochromis niloticus, Linnaeus, 1758) fry during the early phase of rearing in brackish water environment (5–10ppt). Aquaculture 442, 112–118. Thoa, N.P., Ninh, N.H., Knibb, W., Nguyen, N.H., 2016. Does selection in a challenging environment produce Nile tilapia genotypes that can thrive in a range of production systems? Sci. Rep. 6. Trọng, T.Q., van Arendonk, J.A., Komen, H., 2013a. Genetic parameters for reproductive traits in female Nile tilapia (Oreochromis niloticus): I. Spawning success and time to spawn. Aquaculture 416, 57–64. Trọng, T.Q., van Arendonk, J.A.M., Komen, H., 2013b. Genetic parameters for reproductive traits in female Nile tilapia (Oreochromis niloticus): II. Fecundity and fertility. Aquaculture 416–417, 72–77. Watanabe, W.O., Burnett, K.M., Olla, B.L., Wicklund, R.I., 1989. The effects of salinity on reproductive performance of Florida red tilapia. J. World Aquacult. Soc. 20, 223–229.
8
Contents lists available at ScienceDirect
Animal Reproduction Science journal homepage: www.elsevier.com/locate/anireprosci
Genetic variation and correlated changes in reproductive performance of a red tilapia line selected for improved growth over three generations ⁎
Ngo Phu Thoaa, , Azhar Hamzahb, Nguyen Hong Nguyenc a b c
Research Institute for Aquaculture No.1, Tu Son, Bac Ninh, Viet Nam National Prawn Fry Production & Research Center, Kg. Pulau Sayak, 08500 Kota Kuala Muda, Kedah, Malaysia University of the Sunshine Coast, Maroochydore, QLD 4558, Australia
AR TI CLE I NF O
AB S T R A CT
Keywords: Genetics Fecundity Fitness related traits Selection and genetic improvement
The present study examines genetic variation and correlated changes in reproductive performance traits in a red tilapia (Oreochromis spp.) population selected over three generations for improved growth. A total of 328 breeding females (offspring of 111 sires and 118 dams) had measurements of body weight prior to spawning (WBS), number of fry at hatching (NFH), total fry weight (TFW) and number of dead fry (NDF) or mortality of fry including unhatched eggs at hatching (MFH). Restricted maximum likelihood (REML) analysis in a multi-trait model showed that there are heritable genetic components for all traits studied. The heritability for WBS was very high (0.80). The estimates for traits related to fecundity (NFH, TFW) and survival (NDF) were low and they were associated with high standard errors. Genetic correlations of WBS with other reproductive performance traits (NFH, TFW and NDF) were generally positive. However, NFH was negatively correlated genetically with TFW. As expected, body measurements during growth stage exhibited strong positive genetic correlations with WBS. The genetic correlations between body traits and reproductive performance (NFH, TFW, NDF) were not significant. Correlated responses in reproductive traits were measured as changes in least squares means between generations or spawning years. Except for WBS that increased with the selection programs, the phenotypic changes in other reproductive traits observed were not statistically significant (P > 0.05). It is concluded that the selection program for red tilapia has resulted in very little changes in reproductive performance of the animals after three generations. However, periodic monitoring of genetic changes in fecundity and fitness related traits such as NDF or MFH should be made in selective breeding programs for red tilapia.
1. Introduction Body weight has been the sole selection criterion in majority of genetic improvement programs for farmed aquaculture species because it is the primary determinant of animal performance and market price (Nguyen, 2016). However, reproductive performance and fitness related traits are also important to hatcheries to maximize revenue and return per unit of production. It is, therefore, crucial to understand the consequences of selecting for increased body weight on other traits of commercial importance, including reproductive performance.
⁎
Corresponding author. E-mail address: [email protected] (N.P. Thoa).
http://dx.doi.org/10.1016/j.anireprosci.2017.07.003 Received 15 March 2017; Received in revised form 26 May 2017; Accepted 4 July 2017 0378-4320/ © 2017 Published by Elsevier B.V.
Please cite this article as: Thoa, N.P., Animal Reproduction Science (2017), http://dx.doi.org/10.1016/j.anireprosci.2017.07.003
Animal Reproduction Science xxx (xxxx) xxx–xxx
N.P. Thoa et al.
Several studies in model species or farmed animals have shown that selection for high growth rate might not have negative effects on reproductive traits (Bünger et al., 2005), whereas breeding objectives placing emphasis on high efficiency of meat production might result in a reduction in fertility and litter size of the animals as well as increase the incidence of genetic defects (Rauw et al., 1998). In aquatic animal species, the genetic relationships between body traits and other economically important traits have been examined in unselected populations (Gall and Neira, 2004; Gjerde, 1986; Kristjánsson and Arnason, 2014; Su et al., 2002). Whereas the heritability and correlations have enabled the prediction of the consequences of selection for increased body weight on traits of economic importance, there have been limited selection experiments to measure the realised response in reproductive traits to selection for improved growth rate. To date, only one study was conducted to examine correlated responses in reproductive traits to the long term selection program for genetically improved farmed tilapia (GIFT) strain (Hamzah et al., 2016). A selective breeding program for a red tilapia population was conducted since 2010 to improve production performance, survival and body colour in Malaysia. Substantial direct response was achieved for body weight (the main selection criterion) after three generations (years) of selection (one generation per year). Although selection was practised solely on growth, correlated responses in other dimensions of body measurements, body colour and survival were also achieved (Nguyen, unpublished results). Nevertheless, genetic changes in reproductive traits as an indirect consequence of short term selection for enhanced growth are still unknown. In this present paper, we hypothesise that there have been no significant impacts on reproductive performance traits of a red tilapia population which has underwent three generations of selection from 2008 to 2012. In addition, we examined quantitative genetic variation in reproductive characters and their relationships with growth performance. The results obtained from the present study are expected to assist with the review and monitoring of genetic progress achieved in the genetic programs. If unfavourable changes in reproductive traits occurred, the breeding objectives for red tilapia would need to be refined to sustain long term response to selection in a preferred direction. 2. Materials and methods 2.1. Origin of the red tilapia strain Data were collected from three generations of selection for improved growth in a population of red tilapia (Oreochromis spp.) at Aquaculture Extension Centre of Department of Fisheries, Malaysia. A full detailed account of the population, family production, selection procedures and management of the animals are given in Nguyen et al. (2017). In brief, the base population (G0) was established in 2009 from a full 3 × 3 diallele cross involving three strains which originated from hatcheries in Malaysia, Taiwan and Thailand. The selected line was formed by within- family sampling from 103 families (four to 10 females and males per family). They came from 103 families produced from the full diallele cross in 2008. The line was selected for high breeding values for body weight at harvest. A combined between and within family selection was practised in the selected line. The average proportion of selected animals was about 3% in females and 2.3% in males. Note, however, that selection was on breeding values but not by truncation (due to the inability of some selected breeders to reproduce or due to mortality we had to resort to selecting lower ranking fish; also, the number of selected individuals contributed by each family was restricted to avoid later inbreeding). In generations G0 (year 2009), G1 (year 2010), G2 (year 2011) and G3 (year 2012) they were the progenies of 32–50 dams and 28–51 sires for the selected line. Matings were made among genetically unrelated broodstock based on their estimated breeding values (EBVs) and their relationship to other animals in the pedigree to produce full-sibs and (paternal) half-sibs families. Over three generations, a total of 328 breeding females (offspring of 111 sires and 116 dams) successfully produced progenies used for the analyses in this study (Table 1). The breeding scheme is given in Nguyen (2016). 2.2. Family production, rearing to harvest Family (full- and half-sibs) production was normally completed within 30–45 days. Mating was conducted in hapas, following the mating design prepared from annual routine genetic evaluation and mate allocation analyses. After 7 days of mating, only fertilized eggs at eyed stage were collected from the mouth of the female and immediately transferred to hatching jars for artificial incubation. Fry often hatched after about 5–7 days. Soon after yolk-sac absorption, the hatched fry of each family were transferred from the incubators to the nursery hapas (1 m3 with 2 mm mesh size), stocked at a density of 200 fry per m3. At least three nursery hapa replicates for each family were maintained in the same pond to reduce environmental differences between families. When the fingerlings reached an average weight of 5–10 g, about 100–150 individuals per family were randomly sampled and physically Table 1 Number of breeding female, sire and dam in the population. Spawning years
Population
Breeding female
Sire
Dam
2009 2010 2011 2012 Total
Base Selection Selection Selection
110 78 65 75 328
51 28 32 111
50 32 34 116
2
Animal Reproduction Science xxx (xxxx) xxx–xxx
N.P. Thoa et al.
Table 2 Basic statistics of reproduction and growth characteristics of female red tilapia. Category
Traits
Unit
N
Mean
SD
CV
Min
Max
Reproduction
WBS NFH TFW NDF
g no. kg no.
258 318 318 308
237.4 658.5 0.9 14.8
60.2 384.7 1.6 48.0
25.4 58.4 176.8 324.9
101.3 65 0.1 0
397.8 2120 13 748
Growth at harvest
WT LG DP WD
g cm cm cm
216 216 216 216
272.2 18.5 8.0 3.5
113.0 2.5 1.3 0.5
41.5 13.6 16.8 14.7
96.7 12.4 5.2 2.2
616 25.4 11.5 4.8
WBS = Weight before spawning, NFH = Number of fry at hatching, TFW = Total fry weight, NDF = Number of dead fry. WT = Weight, LG = Length, DP = Depth and WD = Width.
identified, using PIT (passive integrated transponder) tags. After tagging, the fingerlings were pooled and conditioned for about three days in fibreglass tanks without feeding. Then representatives of each family were sent to communal grow-out testing in freshwater earthen pond (1000 m2) and cages (3 m × 3 m × 3 m). The initial stocking density in cage was 5 fish per square meter of surface water while the density in each pond was 2–3 fish per square meter of surface water. In both testing environments, the same diet (32% protein content), feeding rate (3–5% of total body mass), culture and management practices were applied. Water quality parameters (temperature, pH, dissolved oxygen and total ammonia) were monitored once a week. The same management procedures were practised in all generations during the course of the selection program for red tilapia from stocking to harvest. At harvest, fish were measured of body traits (live weight, standard length, body width and depth). Body weight data collected at harvest were then used in the computation of estimated breeding values (EBVs) for all individuals in the pedigree. The best animals (highest EBV for body weight) were selected to become parents of the succeeding generations. In each generation, reproductive performance parameters of individual breeding females were recorded. The same breeding schemes regarding the family production, genetic evaluation and selection process were repeated in subsequent generations 2009–2012. 2.3. Measurements of reproductive traits Table 2 shows data on the reproductive traits measured in this study. Measurements of the traits were carried out during fry production. Females’ body weights before spawning (WBS) were recorded as live weight of a ready to spawn female (by visual assessment of its genital papillae condition) prior to mating. The total number of live fry at hatching (NFH) per spawning per female was recorded by counting of fry collected from each hatching jar assigned for each female breeder. The total fry weight (TFW) after hatching was measured by using a digital balance. The traits under analyses were body weight of female breeders before spawning (WBS), number of fry alive at hatching per spawning per female (NFH), total fry weight per spawning per female (TFW) and number of dead fry at hatching (NDF). The number of dead fry (NDF) at hatching included the number of eggs that were not hatched during artificial incubation. NDF was also expressed as a ratio trait (called as mortality of fry at hatching, MFH) to examine its relationship with female weight. However, this ratio trait (MFH) was not included in genetic analysis. NFH and TFW were recorded in all spawning seasons (2008–2011), whereas WBS, WAS and NDF were from three recent generations (2009–2011). 2.4. Statistical analyses 2.4.1. Genetic parameters The original scale was used to analyse data on WBS, TFW and NDF traits whereas the NFH was square root transformed in all analyses to improve the distribution of residuals. Variance (genetic and environmental) components for all the traits were estimated by restricted maximum likelihood method, applied to a standard animal model using ASReml (Gilmour et al., 2009). In matrix notation, the mixed model is written as follows: y = Xβ + Zu + e
(1)
where y is the vector of observations for reproductive traits, β is the vector of the fixed effect that included spawning season or generation (1–3), and the linear covariate of female body weight prior to spawning within line for NFH, TFW and NDF, and u is the vector of the random animal additive genetic effects ∼ (0, Aσa2 ) where A is the additive genetic (numerator) relationship matrix among the animals, and e is the vector of residual effects ∼ (0, Iσe2 ). X and Z are incidence matrices relating observations to the fixed effects and the additive genetic effect of the individual animal, respectively. The remaining effects are assumed to be distributed as var(e) = R = Iσe2 , where I is an identity matrix. The expectations of all random effects are zero, cov (u,e) = 0 and thus var (y) = ZGZ’ + WIW’ + R. In the preliminary analyses, the likelihood ratio test showed that the random effects of common effects to full-sibs and mating hapas nested within spawning season (or generation) were not significant for NFH, TFW and NDF, and they were removed from the final model for these traits. These effects on WBS and body traits at harvest were also not significant (P > 0.05). Hence, model 1 was used to analyse all traits included in this study. 3
Animal Reproduction Science xxx (xxxx) xxx–xxx
N.P. Thoa et al.
Heritabilities (h2) for all traits were estimated from a single trait model [Eq. (1)] and calculated as: h2 = additive genetic variance and the residual variance analyses, using model 1.
(σe2 ).
σˆa2 σˆa2 + σe2
where σa2 is the
Phenotypic and genetic correlations were estimated from a series of bivariate
2.4.2. Correlated selection responses The same models as described above (Eq. (1)) were used to estimate correlated responses in reproduction traits (WBS, NFH, TFW and NDF) to the selection program for improved growth in red tilapia. Furthermore, reproductive performance traits (i.e. NFH, TFW and NDF) were also analysed with a model that female body weight prior to spawning (WBS) within line was not fitted as a covariate. The correlated changes in all reproduction traits were measured as differences between least square means between generations of selection or spawning season. 2.5. Ethics statement The study and the protocol were reviewed and approved by the animal ethics committee of the Department of Fisheries, Malaysia (the approval number ANE070312). 3. Results 3.1. Basic statistics The study included a total of 328 breeding females (offspring of 116 sires and 118 dams) that had reproductive performance records collected from 2008 to 2012 (Table 1). Basic statistics including means, standard deviations, coefficients of variation, and minimum and maximum values for the reproductive traits in all generations are shown in Table 2. In the present study, fecundity is defined as the total number of fry collected at hatching from a female breeder in each spawning. When fecundity was expressed as the number of fry produced per gram of female body weight, it ranged from 2.5 to 2.8. For NDF, the coefficient of variation was large due to the limited data records for this trait and due to the high mortality that was observed in some families during incubation. 3.2. Prediction of fecundity based on female body weight prior to spawning The relationship between NFH and female body weight prior to spawning (WBS) are essentially linear (TNF = 461.1 × WBS + 0.895). However, the relationship between total fry weight (TFW) and WBS was not significant (P > 0.05). The linear regression coefficients indicated a 0.90 (s.e. 0.02) g increase in the numer of fry at hatching (NFH) and a 0.63 (s.e. 0.14) kg increase in total fry weight (TFW) for every gram increase in WBS (Fig. 1). Conversely, with increased NFH, individual fry weight (TFW divided by TNF) decreased by −1.78 × 10−6 (5.17 × 10−7) kg (results not presented). 3.3. Relationships between fecundity and total fry weight and fry mortality There was a proportional reduction of 0.0025% in mortality rate at hatching for every fry increase per spawn until the number of fry (NFH) equalled around 500 (Fig. 2A), after which there was a proportionate decrease in mortality of 0–0.1% per unit increase in
Fig. 1. Relationship between number of fry at hatching (NFH) and total fry weight (TFW, g), TFW = 0.63 × NFH + 0.006.
4
Animal Reproduction Science xxx (xxxx) xxx–xxx
N.P. Thoa et al.
Fig. 2. A) Relationship between morality of fry at hatching (MFH, %) and total number of fry (NFH), MFH = 0.413 × NFH – 2.533 × 10−5, and B) Relationship between morality of fry at hatching (MFH) and total fry weight (TFW, g), MFH = 0.043 × TFW – 0.004.
NFH. The mortality at hatching decreased rapidly with increasing total fry weight, with mortality of about 0.02% for fry with greater weight than 10 g (Fig. 2B).
3.4. Heritabilities for traits Except for the heritability (h2) for average body weight of females prior to spawning (WBS) that was fairly high (h2 = 0.80), these estimates for other reproductive traits recorded in the present study were low and not significantly different from zero (Table 3). This was due to a large proportion of variance explained by the environmental effects. The large standard errors associated with the h2 estimates for reproductive traits were likely a result of the limited sample size of breeding females recorded, although the data were collected over four generations including the base population from 2008 to 2012.
3.5. Correlations among reproductive traits Genetic and phenotypic correlations among reproductive traits are both favourable and unfavourable (Table 4). The genetic correlations between WBS and fecundity related traits (NDF and TFW) were moderate and positive (0.13–0.30), and the estimates were associated with high standard errors. On the other hand, WBS was genetically correlated negatively with fitness related traits of the fry (NFH), although magnitude of the estimates was low and not different from zero. Overall, phenotypic correlations were generally low and not significant, ranging from −0.02 to 0.16. Table 3 Variance components, heritability (h2) and common environmental effects (c2) for body traits. Category
h2
Traits
σa2
σe2
Reproduction
WBS NFH TFW NDF
2619.1 0.84 0.05 67.6
638.9 50.9 2.60 2119.6
0.80 0.02 0.02 0.03
± ± ± ±
0.16 0.11 0.12 0.12
Growth
WT LG DP WD
854.4 0.96 0.11 0.092
2679.5 1.74 0.59 0.061
0.24 0.36 0.16 0.60
± ± ± ±
0.14 0.15 0.13 0.16
σa2 = Additive genetic variance and σe2 = Residual variance. WBS = Weight before spawning, NFH = Number of fry at hatching, TFW = Total fry weight, NDF = Number of dead fry. WT = Weight, LG = Length, DP = Depth and WD = Width.
5
Animal Reproduction Science xxx (xxxx) xxx–xxx
N.P. Thoa et al.
Table 4 Phenotypic (above) and genetic (below) correlations among traits. Traits
WBS
WBS NFH TFW NDF
n.e. 0.30 ± 1.10 0.13 ± 0.84
NFH
TFW
NDF
0.16 ± 0.06
−0.03 ± 0.06
n.e. n.e.
−0.02 ± 0.06 −0.01 ± 0.06 −0.02 ± 0.06
n.e.
WBS = Weight before spawning, NFH = Number of fry at hatching, TFW = Total fry weight, NDF = Number of dead fry. WT = Weight, LG = Length, DP = Depth and WD = Width. n.e. = not estimable.
3.6. Correlations between harvest weight and reproduction traits Phenotypic (rp) and genetic correlations (rg) of body traits with reproduction traits are presented in Table 5. The genetic correlations between body traits of grow-out stocks at harvest and WBS of breeding females were close to one (0.7–0.94). There were also strong negative genetic association between body traits of the grow-out stocks and total fry weight (0.43–0.77). The genetic correlations of body traits with fecundity (NFH) were moderate and positive (0.12–0.72, respectively). The phenotypic correlations between body traits of grow-out stocks and those of females (WBS) and females’ reproductive performance (NFH and TFW) were of similar size and magnitude to those of the genetic correlations. However, both the phenotypic and genetic correlation estimates between body traits and reproductive performance (except WBS) had high standard errors and not significant different from zero. 3.7. Correlated selection responses The correlated responses in reproduction traits estimated as the changes in least squares means (LSMs) for each generation is given in Table 6. Significant differences between successive generations were observed for body weight of females prior to spawning (P < 0.01). The differences between generations or spawning seasons were not statistically significant for other reproductive traits (NFH, TFW and NDF), except that an increase in NDF in the latest generation (2012) was due to mass mortality of some families. 4. Discussions Overall, reproductive performance of the red tilapia line used in this study is comparable to other stocks reported in the literature (Eguia, 1996; Suresh and Lin, 1992; Watanabe et al., 1989). The average number of fry per gram of female body weight of the present red tilapia stock was, however, lower than those reported for Nile tilapia (2.8 vs. 3–8 fry per gram of female) (Bhujel et al., 2007; Campos-Mendoza et al., 2004; El-Sayed et al., 2003; Osure and Phelps, 2006; Siddiqui et al., 1998). Comparing fecundity among studies is relative because different strains of tilapia, size of broodstocks and breeding animals are used. Furthermore, reproductive performance of fish is largely influenced by environmental factors such as rations, testing environments and management practices. Interestingly, our results showed that the selection program for improved growth in red tilapia increased WBS. However, there were no significant changes in reproductive performance (NFH and TFW) of the animal. This is consistent with the non-significant genetic association between body traits at harvest/or female body weight at mating and fecundity (total number of fry at hatching, fry weight). The weak genetic correlations between body weight and fecundity are also reported in other aquaculture species (Arcos et al., 2004; Gall and Neira, 2004; Gjerde et al., 1994; Su et al., 1997). Additionally, it could have been because the current selection program is still at the early stage to cause any causal correlated changes in fitness related traits. We support this hypothesis on the basis of the genetic changes in the same traits measured from a long term selection program in the GIFT strain. After 6 generations of the GIFT strain in Philippines and 12 generations in Malaysia, there was also non-significant change per unit of female body weight across six measures of the reproductive traits (Hamzah et al., 2016). Selection for improved growth in farmed animals resulted in correlated improvement in birth weight in pigs and cattle (Burrow and Prayaga, 2004). However, breeding objective emphasising high efficiency of feed utilisation may impair reproductive performance of the animals (Kerr and Cameron, 1995). Except for the realised response in reproductive traits estimated in the GIFT strains, the literature is not available in other aquatic animal species to make a relative comparison with the results obtained from our present study. Table 5 Phenotypic (rp) and genetic (rg) correlations between reproductive traits and growth. Traits
WBS NFH TFW NDF
WT
LG
DP
WD
rg
rp
rg
rp
rg
rp
rg
rp
0.80 ± 0.14 0.72 ± 1.34 −0.43 ± 1.58 n.e.
0.75 ± 0.04 0.10 ± 0.07 −0.07 ± 0.08 n.e.
0.81 ± 0.14 0.74 ± 2.26 −0.72 ± 3.25 n.e.
0.69 ± 0.04 0.06 ± 0.07 −0.09 ± 0.08 n.e.
0.94 ± 0.10 0.59 ± 1.68 −0.77 ± 11.5 −0.64 ± 1.31
0.81 ± 0.03 0.12 ± 0.07 0.14 ± 0.08 −0.06 ± 0.06
0.77 ± 0.15 0.12 ± 1.13 −0.54 ± 3.77 0.07 ± 0.79
0.59 0.08 0.05 0.05
± ± ± ±
0.06 0.07 0.08 0.06
WBS = Weight before spawning, NFH = Number of fry at hatching, TFW = Total fry weight, NDF = Number of dead fry. WT = Weight, LG = Length, DP = Depth and WD = Width.
6
Animal Reproduction Science xxx (xxxx) xxx–xxx
N.P. Thoa et al.
Table 6 Least square means of reproductive traits over generations. Traits
Generation 1
WBS NFH TFW NDF
2
3
a
215.0 ± 6.38 555.2 ± 41.94 0.71 ± 0.18 a 6.61 ± 5.67 a
a
600.6 ± 49.49 0.92 ± 0.21 a 9.71 ± 6.24 a
a
271.0 ± 6.73 b 588.8 ± 43.06 a 0.94 ± 0.19 a 35.6 ± 5.43 b
WBS = Weight before spawning, NFH = Number of fry at hatching, TFW = Total fry weight, NDF = Number of dead fry. WT = Weight, LG = Length, DP = Depth and WD = Width. Means with different superscript letters in the same line are significant different (P < 0.05).
The low and non-significant heritabilities estimated for a range of reproductive traits here show that making genetic progress by direct selection for the reproductive traits, albeit possible, may be slow. With traits that are of low heritability, advanced statistical and genetic evaluation methods such as Best Linear Unbiased Prediction (BLUP) should be applied to utilise all the information available in the pedigree. Further, a large amount of data records and in-depth pedigree data should be routinely collected to obtain reliable accuracy of breeding value estimation. Also note that a zero heritability for a trait of interest does not mean that there is no genetic variation to allow scope for selection because heritability is a population parameter and as expected the reproductive traits are largely influenced by environmental factors. The low heritability for reproduction and fitness related traits have been reported in different farmed aquaculture species ranging from fish (Gima et al., 2014; Gjerde, 1986; Trọng et al., 2013a, 2013b) to crustaceans (Caballero-Zamora et al., 2015; Macbeth et al., 2007). The heritability estimates for the same traits as recorded in the present study were from 0.13 to 0.17 in the GIFT strain (Hamzah et al., 2016) or 0.03–0.25 for early survival and sexual maturity in the salinity tolerance line of Nile tilapia (Thoa et al., 2015; Thoa et al., 2016). A synthesised review from literature results together with those obtained in our present study suggest that fecundity of red tilapia can be improved through direct or indirect selection for increased growth. This is also demonstrated in the GIFT strain by Hamzah et al. (2016). However, the improvement in fecundity depends upon specific breeding objectives of genetic improvement programs to produce stocks for hatcheries or commercial production. Commercial hatcheries may prefer having highly fecund females as their main interest is selling fry to production. On the contrary, grow-out producers prefer animals that from each unit of food intake, waste least in reproduction and retain most to favour the efficient conversion of metabolic energy into growth/or protein deposition. An alternative is to develop specialised genetic lines such as a maternal line for hatcheries and a paternal line for production. This approach has been successfully implemented in well-structured animal industry and by multi-national companies (Smith, 1964). However, maintenance of multiple lines would require investments and recourses that are generally not generous in major aquaculture producing countries. There is no single simple answer in this regard. The genetic improvement of reproductive traits should be considered on the basis of its merits for individual cases. In summary, the finding of our study indicates that selection for improved growth increased female body weight before spawning. However, there were no significant differences in fecundity related traits (NFH and TFW) between generations of selection (Table 6). Although our results are generally consistent with those reported in the GIFT strain and other fishes, periodically monitoring of the selective breeding program will warrant a rigorous conclusion regarding genetic changes in reproductive characteristics to selection for high growth in this red tilapia stock. 5. Conclusions Fecundity and fitness related traits of red tilapia are lowly heritable. The genetic association for body traits in grow-out animals were retained in reproductive breeding stage in broodstock as indicated by the high and positive genetic correlations between body traits and WBS. Genetic correlations between body traits and reproductive performance, albeit positive, were associated with large standard errors and hence, the estimates were not significant. Short term selection for improved growth in the present study did not cause significant changes in reproduction productivity (NFH and TFW). Routine collection of reproductive characters should be practised to accumulate further data to obtain more reliable genetic parameter estimates as well to monitor possible genetic changes in these traits as a result of long term selection for high production performance. Conflict of interest The authors declare that they have no competing interests. Acknowledgements This study was funded by Department of Fisheries, Malaysia. We would like to thank Hoong Yip Yee and Mohd A Aziz for their kindly assistance in the management of the animals. The Aquaculture Extension Center of the Malaysian Department of Fisheries in Jitra, Kedah provided facilities and additional manpower for the study. 7
Animal Reproduction Science xxx (xxxx) xxx–xxx
N.P. Thoa et al.
References Arcos, F.G., Racotta, I.S., Ibarra, A.M., 2004. Genetic parameter estimates for reproductive traits and egg composition in Pacific white shrimp (Penaeus (Litopenaeus) vannamei). Aquaculture 236, 151–165. Bünger, L., et al., 2005. Relationships between quantitative and reproductive fitness traits in animals. Phil. Trans. R. Soc. B : Biol. Sci. 360, 1489–1502. Bhujel, R.C., Little, D.C., Hossain, A., 2007. Reproductive performance and the growth of pre-stunted and normal Nile tilapia (Oreochromis niloticus) broodfish at varying feeding rates. Aquaculture 273, 71–79. Burrow, H., Prayaga, K., 2004. Correlated responses in productive and adaptive traits and temperament following selection for growth and heat resistance in tropical beef cattle. Livest. Prod. Sci. 86, 143–161. Caballero-Zamora, A., et al., 2015. Genetic parameters for spawning and growth traits in the Pacific white shrimp (Penaeus (Litopenaeus) vannamei). Aquacult. Res. 46, 833–839. Campos-Mendoza, A., McAndrew, B., Coward, K., Bromage, N., 2004. Reproductive response of Nile tilapia (Oreochromis niloticus) to photoperiodic manipulation; effects on spawning periodicity, fecundity and egg size. Aquaculture 231, 299–314. Eguia, M.R.R., 1996. Reproductive performance of four red tilapia strains in different seed production systems. Israeli J. Aquacult. Bamidgeh 48, 10–18. El-Sayed, A.-F.M., Mansour, C.R., Ezzat, A.A., 2003. Effects of dietary protein level on spawning performance of Nile tilapia (Oreochromis niloticus) broodstock reared at different water salinities. Aquaculture 220, 619–632. Gall, G.A., Neira, R., 2004. Genetic analysis of female reproduction traits of farmed coho salmon (Oncorhyncus kisutch). Aquaculture 234, 143–154. Gima, M.E., et al., 2014. Realized heritability and response to selection for fecundity, hatching rate and fry/Kg for channel catfish females (Ictalurus punctatus) induced to ovulate and fertilized with blue catfish (Ictalurus punctatus) males for the production of hybrid catfish embryos. Aquaculture 40–421, S36–S41. Gjerde, B., Simianer, H., Refstie, T., 1994. Estimates of genetic and phenotypic parameters for body weight, growth rate and sexual maturity in Atlantic salmon. Livest. Prod. Sci. 38, 133–143. Gjerde, B., 1986. Growth and reproduction in fish and shellfish. Aquaculture 57, 37–55. Hamzah, A., et al., 2016. Genetic parameters and correlated responses in female reproductive traits in the GIFT strain. Aquacult. Res. 47, 1488–1498. Kerr, J., Cameron, N., 1995. Reproductive performance of pigs selected for components of efficient lean growth. Anim. Sci. 60, 281–290. Kristjánsson, T., Arnason, T., 2014. Strong phenotypic and genetic correlation between size and first maturity in Atlantic cod Gadus morhua L. reared in commercial conditions. Aquacult. Res. 46, 2185–2193. Macbeth, M., et al., 2007. Heritability of reproductive traits and genetic correlations with growth in the black tiger prawn Penaeus monodon reared in tanks. Aquaculture 270, 51–56. Nguyen, N.H., 2016. Genetic improvement for important farmed aquaculture species with a reference to carp, tilapia and prawns in Asia: achievements, lessons and challenges. Fish Fish. 17, 483–506. Nguyen, N.H., Hamzah, A., Ngo, P.T., 2017. Effect of genotype by environment interaction on genetic gain and genetic parameter estimates in red tilapia (Oreochromis spp.). Frontiers in Genetics 8. Osure, G.O., Phelps, R.P., 2006. Evaluation of reproductive performance and early growth of four strains of Nile tilapia (Oreochromis niloticus, L) with different histories of domestication. Aquaculture 253, 485–494. Rauw, W., Kanis, E., Noordhuizen-Stassen, E., Grommers, F., 1998. Undesirable side effects of selection for high production efficiency in farm animals: a review. Livestock Production Science 56, 15–33. Siddiqui, A., Al‐Hafedh, Y., Ali, S., 1998. Effect of dietary protein level on the reproductive performance of Nile tilapia, Oreochromis niloticus (L.). Aquaculture research 29, 349–358. Smith, C., 1964. The use of specialised sire and dam lines in selection for meat production. Anim. Prod. 6, 7–344. Su, G.-S., Liljedahl, L.-E., Gall, G.A., 1997. Genetic and environmental variation of female reproductive traits in rainbow trout (Oncorhynchus mykiss). Aquaculture 154, 115–124. Su, G.-S., Liljedahl, L.-E., Gall, G.A., 2002. Genetic correlations between body weight at different ages and with reproductive traits in rainbow trout. Aquaculture 213, 85–94. Suresh, A.V., Lin, C.K., 1992. Tilapia culture in saline waters: a review. Aquaculture 106, 201–226. Thoa, N.P., et al., 2015. Genetic variation in survival of tilapia (Oreochromis niloticus, Linnaeus, 1758) fry during the early phase of rearing in brackish water environment (5–10ppt). Aquaculture 442, 112–118. Thoa, N.P., Ninh, N.H., Knibb, W., Nguyen, N.H., 2016. Does selection in a challenging environment produce Nile tilapia genotypes that can thrive in a range of production systems? Sci. Rep. 6. Trọng, T.Q., van Arendonk, J.A., Komen, H., 2013a. Genetic parameters for reproductive traits in female Nile tilapia (Oreochromis niloticus): I. Spawning success and time to spawn. Aquaculture 416, 57–64. Trọng, T.Q., van Arendonk, J.A.M., Komen, H., 2013b. Genetic parameters for reproductive traits in female Nile tilapia (Oreochromis niloticus): II. Fecundity and fertility. Aquaculture 416–417, 72–77. Watanabe, W.O., Burnett, K.M., Olla, B.L., Wicklund, R.I., 1989. The effects of salinity on reproductive performance of Florida red tilapia. J. World Aquacult. Soc. 20, 223–229.
8