- Email: [email protected]
Response to Divergent Selection for Early Growth of Chickens Fed a Diet Deficient in Selenium
Response to Divergent Selection for Early Growth of Chickens Fed a Diet Deficient in Selenium D. L. CUNNINGHAM, G. F. COMBS, JR., J. A. SAROKA, and M. W. LaVORGNA Department of Poultry and Avian Sciences, Cornell University, Ithaca, New York 14853 (Received for publication June 9, 1986)
1987 Poultry Science 66:209-214 INTRODUCTION
Previous studies have indicated that growth responses of young chicks to severe uncomplicated nutritional deficiency of selenium (Se) may involve an hereditary component. In 1981, Bunk and Combs observed that Leghorn chicks fed an amino acid-based diet containing an exceedingly low amount (less than .10 ppm) of Se, but adequate with respect to all other known nutrients, showed considerable variation in effects on growth and survival. They found that although one-third of the population showed severely depressed growth with associated pancreatic exocrine dysfunction, an equal proportion was able to grow apparently normally. This observation was also made by LaVorgna and Combs (1983), who tested the hypothesis that variance in the growth response to severe Se deficiency is due to a hereditary factor. Their results snowed the feasibility of developing, through selective breeding, lines of Single Comb White Leghorn chickens that differed in sensitivity to dietary Se-deficiency as measured by impairment in the growth of young chicks. Further,
they showed that such line-related differences in growth were associated with analogous differences in methionine-methyl group oxidation rate, which suggests that a lesion in the metabolism of the sulfur-containing amino acids may be the site of hereditary involvement in the metabolic need for Se. Subsequent studies by Halpin and Baker (1984) found similar evidence of aberrant sulfur-amino acid metabolism in one meat-type strain of chicken but no effects in a Leghorn strain or a crossbred strain. The extent to which the consequences of nutritional Se deficiency may differ among genotypes is of fundamental importance to understanding the roles of Se in normal metabolism and of practical significance to poultry feeding particularly in parts of the world with endemic Se deficiency. The purpose of this study, therefore, was to test the hypothesis that response of the chick to severe Se-deficiency is heritable. MATERIALS AND METHODS
Animals and Diets. The Athens-Canadian Randombred (AC) population of meat-type
209
Downloaded from http://ps.oxfordjournals.org/ at University of Georgia on June 23, 2015
ABSTRACT Three generations of divergent selection for 21-day growth response to a diet deficient in selenium (-Se) were bred using a meat-type chicken. The Athens-Canadian Randombred (AC) population of chickens served as the base population for this study. Mass selection was used to establish a -Se refractory line (SDR) and a -Se susceptible line (SDS). A genetic control line was maintained during the selection process to facilitate evaluation of the responses of the selected lines. The SDR males and females had an average of 17% increase in weight gain at 21 days of age when fed the -Se diet compared with the control line males and females fed the same diet for the three generations of selection. The SDS line had an average reduction in weight gain of 27% during the same period of selection. A difference of 25 g was observed between the mean body weights of SDR males and females and between the SDS males and females after one generation of selection. By the third generation of selection, the difference between SDR and SDS males had increased to 31 g, whereas the SDR and SDS females differed by 41 g. Early response to selection was asymmetrical because the response was greater in the SDS than in the SDR line. Response to selection, however, generally declined after the first generation. Realized heritability estimates for individual generations for this trait were variable (ranging from -.26 to .95), but cumulative estimates (.15 to .39) for growth through three generations were similar to those reported for chickens and quail. Females had larger estimates of heritability than males, which suggests the possiblity of sex-linked effects for this trait. This study establishes the response of the chick to nutritional Se deficiency as an heritable trait. (Key words: selection, selenium, growth response, heritability)
210
CUNNINGHAM ET AL.
An unselected control line was established by using 25 males and 100 females from the group of chicks set aside to form the control. Initial selection was random, except males and females were selected to give the broadest possible representation of the original 60 sire families of the AC line. The control line was perpetuated by forming 25 pedigreed sire families with matings arranged such that full and half-sib relationships were avoided. All matings of controls were made by artificial insemination. For each generation of the control, approximately 400 chicks were assigned randomly to either of two dietary treatment groups of equal size. One group received the Se-deficient basal diet, whereas the second group received the same basal diet supplemented with . 15 ppm Se as Na 2 Se0 3 . The two control groups were used for comparisons with the selected lines. For each generation, means of 21-day-weight gain for each selected line were adjusted accordTABLE 1. Composition of the Se-deficient basal diet1 Ingredient Glucose monohydrate Amino acid mix 2 Corn oil, refined Monoolein Linoleic acid Sodium taurocholate Vitamin mix 3 Mineral mix 4 Choline chloride, 70% solution Total 1
Percentage 65.31 23.53 4.00 .50 .50 .10 .20 5.56 .20 100.00
Contained (by analysis) .010 ppm Se.
'Amino acid mix supplied (per kilogram diet): (g) L-arginine HCl, 11.6; L-glutamic acid, 122.7; glycine 12; L-histidine, 4.7; L-isoleucine, 9.5; L-leucine, 16; L-lysine HCl, 12; DL-methionine, 8.2; L-phenylalanine, 8.2; L-proline, 3; L-threonine, 8.2; L-tryptophan, 2.2; L-tyrosine, 7.0; L-valine, 10.0. 3 Vitamin mix supplied (per kilogram diet): vitamin A (all-trcms-retinyl palmitate), 16,250 IU; vitamin D 3 , 800 IU; all-rac-alpha-tocopheryl acetate, 100 IU; menadione sodium bisulfite, 1.2 mg; biotin, .2 mg; vitamin B, 2 , .014 mg; d-calcium pantothenate, 30 mg; folic acid, 4 mg; niacin, 50 mg; pyridoxine, 10 mg; riboflavin, 10 mg; thiamin HCl, 20 mg; ethoxyquin, 125 mg.
"Mineral mix supplied (per kilogram diet): CaHP0 4 , 19.8 g; CaC0 3 , 7.9 g; NaHC0 3 , 11 g; KHC0 3 , 11 g; MnS0 4 , .2; MgC0 3 , 5 g; FeS0 4 7 H 2 0 , .45 g; K10 3 , 10 g; N a M o 0 4 - 2 H 2 0 , 8 g; C r 2 K 2 ( S 0 4 ) 2 - 2 4 H 2 0 , 1 mg.
Downloaded from http://ps.oxfordjournals.org/ at University of Georgia on June 23, 2015
chickens was used as the base population for this study (Hess, 1962). For the initial generation of this study, approximately 500 pedigreed AC eggs, obtained from Athens, GA, were incubated. Upon hatching, the chicks were sexed and assigned randomly to either of two populations of approximately equal size. Each group of chicks comprised offspring representing the original 60 sire families of the AC population. One group formed the basis for a genetic control line; the second group was used to establish Se-deficient susceptible (SDS) and Se-deficient refractory (SDR) lines. Chicks were reared in thermostatically controlled, wire floored battery brooders in a room with a 15-hr day. A Se-deficient basal diet (Bunk and Combs, 1981) was fed from 1 to 21 days of age (Table 1). This diet contained DLmethionine as the only source of sulfur-containing amino acids; it was supplemented with 100 IU of vitamin E. Fluorometric analysis of the diet by the method of Olson et al. (1975) as all-rac-alpha-tocopheryl acetate indicated that it contained less than .10 ppm total Se. Feed and water were provided ad libitum during the study for all lines. Selective Breeding. The procedure used to separate the lines was mass selection for body weight gain to 21 days of age using the Se-deficient basal diet. The 60 female and 20 male chicks representing both extremes in growth formed the first selected generation. The 30 females and 10 males that phenotypically displayed a refractory response (i.e., showed best growth) to Se deficiency and the 30 females and 10 males most susceptible to the Se deficiency (i.e., showed poorest growth) were used to establish the SDR and SDS lines, respectively. For each selected line, 10 pedigreed sire families were created with each sire mated to three dams. Matings within lines were at random, except that full and half-sib matings were avoided to minimize inbreeding. All matings were made by artificial insemination. For subsequent generations, approximately 60 male and 150 female offspring from each of the selected lines were hatched and evaluated for 21-day growth response using the Se-deficient diet as described for the initial generation. Individuals selected for propagation of lines were returned to the same practical-type growing and maintenance diets, which were supplemented with .10 ppm Se as Na 2 Se0 3 at 21 days of age. Males and females were maintained in separate holding pens until reproductive age.
211
DIVERGENT SELECTION FOR SE-DEFICIENCY
RESULTS AND DISCUSSION
Use of a control population in a selection experiment enables the separation of genetic and nongenetic sources of variation that may occur over time. If, however, time trends are present in a control population, adjustments of selected lines to control levels can produce errors in evaluation of selection response (Hill, 1972). Therefore it is important to evaluate control population stability over generations. Body weight gains for control lines fluctuated from generation to generation (Table 2); how-
TABLE 2. Twenty-one-day body weight gain by genetic line, sex, and generation Generation
Genetic line
Sex
Parental weight
1
Control (+Se) Control (+Se)
M F
128 115
121 ± 3.0* 103 ± 4.0 b
SDR1 SDR
M F
89 93
Control (-Se) Control (-Se)
M F
SDS2 SDS
M F
2
3
122 ± 2.1* 118 + 2.1*
137 ± 1.9* 123 ± 1.8b
70 ± 3.0C 76 ± 2.0 c d
106 ± 2.6 b 106 ± 3.4 b
76 ± 2.8 d 88 ± 3.1 c
89 93
60 + 2.0 d e 63 ± 2.0 e
91 ± 2 . 1 d 97 + 2 . 1 c
67 ± 1.5 e 70 ± 1.5 de
89 93
45 ± 2.0 f 51 ± 2.0 f
64 ± 3.3 d 74 ± 3.0 d
45 ± 1.9f 47 ± 2.0 f
(g)
a—f Mean ± SE with different superscripts within columns are significantly different (P<.05). 1
SDR = Selenium-deficient refractory line.
2
SDS = Selenium-deficient susceptible line.
Downloaded from http://ps.oxfordjournals.org/ at University of Georgia on June 23, 2015
ever, regression of weight gain means on generations produced nonsignificant (P>.05) regression coefficients of .08 and .09 for males and females, respectively, for the control population fed the Se-supplemented diet and -.02 for either males or females of the control population fed the Se-deficient diet. These nonsignificant regression coefficients indicate that phenotypic trends over time were not present for the control. Therefore, selected line means were adjusted to control levels for each generation. Weight gains for the selected lines were significantly different (P<.05) from the control after one generation of selection (Table 2). Selection for high growth response with the Sedeficient diet (SDR) resulted in moderate improvements in 21-day-weight gains when compared with the control. Weight gains for the SDR line, averaged over the three generations of selection, were approximately 17% greater than control gains. Selection for reduced weight gain (SDS) with the Se-deficient diet, however, produced a more pronounced separation from the control than was observed with the SDR line (Figure 1). Weight gains for the SDS line averaged 27% less than the Se-deficient control for the three selected generations. Differences in 21-day-weight gain between the selected lines averaged 38% during the selection process. Males tended to be more sensitive to Se-deficiency but differences among the sexes for weight gain were not significant (P>.05) except for the third generation of selection, where females of the SDR line gained 12 grams more
ing to growth of the control line offspring fed the Se-deficient diet. Within each generation, the standardized selection differential (S) and the standardized response (R) to selection were computed. Selection differentials were weighted for numbers of progeny produced by each parent to produce a realized selection differential. Realized heritabilities (h2) were calculated from the R:S ratio for each line each generation. For Generation 3, growth rates of selected lines after the 21-day testing period were assessed by measuring body weights at 12 and 16 weeks. Hatchability for all eggs set and liveability during the 21 days of testing with the Se-deficient diet also were evaluated. Data for each generation were subjected to analysis of variance; when significant (P<.05) effects were detected, means were separated by a 5% Duncan's multiple range test (Barr et al., 1976).
212
CUNNINGHAM ET AL.
SDR(F)..S
SDRIMT 8
I<1^-^ GAIN (g)
Control
^ ^ \ ^ ^ "''"t3
SDS(MT~ Y
1
GENERATION
than males. LaVorgna and Combs (1983) also observed sex differences for weight gains of chicks fed Se-deficient diets. Although the present selection process resulted in improved growth response for SDR chicks exposed to the Se-deficient diets compared with control chicks fed the Se-deficient diet, the SDR line did not achieve body weight gains comparable to control chicks fed Se-supplemented diets (Table 2). Expected selection differentials, effective selection differentials, and response to selection for 21-day-growth rate are presented in Table 3. Selection differentials were weighted for numbers of progeny produced for each parent (Falconer, 1981). Unweighted and weighted selection differentials were in close agreement except for Generation 3 SDR males and SDS females. This difference occurred because some of the SDR line males and SDS line females were inadvertently discarded during rearing. This effectively reduced the population size of
TABLE 3. Expected selection differentials, effective selection differentials, and response to selection in standard deviation units Expected selection differentials SDR1
Effective selection differentials
SDS 2
SDR
SDS
Response SDR
SDS
Generation
Cumulative
1.38 1.22 1.20
.68 1.15 1.10
3.80
2.93
1.00 1.28 1.18 3.46
.98 1.00 .97 2.95
1
SDR = Selenium-deficient refractory line.
5
SDS = Selenium-deficient susceptible line.
1.38 1.29 .95 3.62
.63 1.10 1.00
1.02 1.31 1.08
.93 1.17 .60
2.73
3.41
2.70
.52 .20 .19 .53
.60 .19 .30 .71
.67 .60 -.28
.63 .42 .0
.99
1.05
Downloaded from http://ps.oxfordjournals.org/ at University of Georgia on June 23, 2015
FIG. 1. Response to selection for 21-day body weight gain of chicks fed a Se-deficient purified diet: Se-deficiency refractory (SDR) and Se-deficiency susceptible (SDS) males (M) and females (F).
these lines and, as a result, the effective selection differentials were also reduced. Nevertheless, expected and effective selection differentials were similar, which suggests minor influences of natural selection. Selection differentials for males were consistently larger than for females because of the greater intensity of selection possible with males with the mating system employed (Table 3). Response to selection generally declined after the first generation and was negligible for Generation 3 (Table 3). Lack of response to selection observed for Generation 3 may have resulted from a rapid reduction in additive genetic variance associated with the selection process for this trait. These results are similar to those of Hess et al., 1962, who reported that response to selection for chicks on a methionine-deficient diet was rapid for the first two generations but negligible thereafter. There appeared to be asymmetry of response with the SDS line showing the greater response to selection (Fig. 1). Because this selection program employed only a few generations, it is difficult to speculate on asymmetry of response. Nevertheless, it is useful to note that inequality in response to selection is not unusual and has a number of possible causes, as discussed by Falconer (1981). Natural selection does not appear to have been a major factor during the development of these lines to this point. Thus, a more plausible explanation may be that longterm natural selection may have favored genes that facilitate growth of chicks fed Se-deficient diets. In that case, artificial selection would be expected to produce a more rapid response in the opposite direction, as observed in the present study.
DIVERGENT SELECTION FOR SE-DEFICIENCY
+i +i +i +i *JL
213
ACKNOWLEDGMENTS
Appreciation is expressed to Henry L. Marks, Southeast Poultry Research Laboratory, Athens, GA, for supplying the base population of Athens-Canadian Ramdombred chickens for this study. The expert technical assistance of Lynne
Downloaded from http://ps.oxfordjournals.org/ at University of Georgia on June 23, 2015
As a result of the variable response to selection, cumulative values for realized estimates of heritability were also variable, ranging from .15 for the SDR males to .39 for the SDS females. Estimates of heritability for this trait (i.e., 21day gain in body weight of chicks fed a Se-deficient diet) have not previously been reported in the literature; however, the values obtained here are similar to other estimates of heritability of juvenile body weight gains for chickens (Siegel, 1962) and Japanese quail (Marks and Lepore, 1968). Estimates of heritability obtained in this study demonstrate the presence of additive genetic variability for this trait. As estimates of heritability were generally different for males and females, it is possible that sexlinked effects are operating for this trait as well. Reports by Brunson et al, (1956) and Thomas et al. (1958) suggest that sex-linked genes may be of considerable importance in the inheritance of body weight traits for chickens. Table 4 presents body weights, hatchability, and liveability rates determined for the genetic lines in Generation 3. Body weights of the SDR, SDS, and control lines were significantly (P< .05) different at the end of 21 days of feeding the Se-deficient diet. By 12 weeks of age, the SDR line males and females and the SDS line males had compensated in body weight gain and were equal to or greater in body weight in comparison with control line individuals previously fed the Se-supplemented diet. The SDS females achieved body weights comparable to the control line females by 16 weeks of age. Hatchability of fertile eggs of each selected line was not different from that of the control line, except for SDR males, which demonstrated a significant reduction for this trait. Survival rate to 21 days of age was not significantly affected for any of the selected lines. These results demonstrate successful divergent selection for susceptibility of chicks to dietary Se-deficiency as measured by 21-dayweight gain. Two lines differing in this trait have been established: SDR and SDS. These lines should provide useful material for studies of the metabolic bases of the nutritional action of Se and related nutrients.
CUNNINGHAM ET AL.
214
P. Deuschle and the clerical assistance of Barbara J. Smagner in preparing this manuscript is also greatly appreciated. REFERENCES
Downloaded from http://ps.oxfordjournals.org/ at University of Georgia on June 23, 2015
Barr, A. J., J. H. Goodnight, J. P. Sail, and J. T. Helwig, 1976. A User's Guide to SAS 76. SAS Inst., Raleigh, NC. Bunk, M. J., and G. F. Combs, Jr., 1981. Relationship of selenium-dependent glutathione peroxidase activity and nutritional pancreatic atrophy in selenium-deficient diets. J. Nutr. 111:1611-1620. Brunson, C. C , G. F. Godfrey, and B. L. Goodman, 1956. Types of gene action in the inheritance of ten week body weight and breast angle of broilers. Poultry Sci. 35:524-532. Falconer, D. S., 1981. Introduction to Quantitative Genetics. Ronald Press Co., New York, NY. Halpin, K. M., andD. H. Baker, 1984. Selenium deficiency and transsulfuration in the chick. J. Nutr. 114:606612. Hess, C. W., 1962. Randombred populations of the Southern Regional Poultry Breeding Project. World's Poult.
Sci. J. 18:147-152. Hess, C. W., H. M. Edwards, Jr., and E. F. Dembnicki, 1962. Growth-rate selection on a methionine deficient diet. Poultry Sci. 41:1042-1047. Hill, W. G., 1972. Estimation of realized heritabilities from selection experiments. II. Selection in one direction. Biometrics 28:767-780. LaVorgna, M. W., and G. F. Combs, Jr., 1983. Evidence of a hereditary factor affecting the chick's response to uncomplicated selenium deficiency. Poultry Sci. 62:164-168. Marks, H. L., and P. D. Lepore, 1968. Growth rate inheritance in Japanese quail. 2. Early responses to selection under different nutritional environments. Poultry Sci. 47:1540-1546. Olson, O. E., I. S. Palmer, and E. E. Cary, 1975. Modification of the official method for selenium implants. J. Assoc. Offic. Agric. Chem. 58:117-121. Siegel, P. B., 1962. Selection for body weight at eight weeks of age. 1. Short term response and heritabilities. Poultry Sci. 41:954-962. Thomas, C. H., W. L. Blow, C. C. Cockerham, and E. W. Glazener, 1958. The heritability of body weight gain, feed consumption, and feed conversion in broilers. Poultry Sci. 37:862-869.
1987 Poultry Science 66:209-214 INTRODUCTION
Previous studies have indicated that growth responses of young chicks to severe uncomplicated nutritional deficiency of selenium (Se) may involve an hereditary component. In 1981, Bunk and Combs observed that Leghorn chicks fed an amino acid-based diet containing an exceedingly low amount (less than .10 ppm) of Se, but adequate with respect to all other known nutrients, showed considerable variation in effects on growth and survival. They found that although one-third of the population showed severely depressed growth with associated pancreatic exocrine dysfunction, an equal proportion was able to grow apparently normally. This observation was also made by LaVorgna and Combs (1983), who tested the hypothesis that variance in the growth response to severe Se deficiency is due to a hereditary factor. Their results snowed the feasibility of developing, through selective breeding, lines of Single Comb White Leghorn chickens that differed in sensitivity to dietary Se-deficiency as measured by impairment in the growth of young chicks. Further,
they showed that such line-related differences in growth were associated with analogous differences in methionine-methyl group oxidation rate, which suggests that a lesion in the metabolism of the sulfur-containing amino acids may be the site of hereditary involvement in the metabolic need for Se. Subsequent studies by Halpin and Baker (1984) found similar evidence of aberrant sulfur-amino acid metabolism in one meat-type strain of chicken but no effects in a Leghorn strain or a crossbred strain. The extent to which the consequences of nutritional Se deficiency may differ among genotypes is of fundamental importance to understanding the roles of Se in normal metabolism and of practical significance to poultry feeding particularly in parts of the world with endemic Se deficiency. The purpose of this study, therefore, was to test the hypothesis that response of the chick to severe Se-deficiency is heritable. MATERIALS AND METHODS
Animals and Diets. The Athens-Canadian Randombred (AC) population of meat-type
209
Downloaded from http://ps.oxfordjournals.org/ at University of Georgia on June 23, 2015
ABSTRACT Three generations of divergent selection for 21-day growth response to a diet deficient in selenium (-Se) were bred using a meat-type chicken. The Athens-Canadian Randombred (AC) population of chickens served as the base population for this study. Mass selection was used to establish a -Se refractory line (SDR) and a -Se susceptible line (SDS). A genetic control line was maintained during the selection process to facilitate evaluation of the responses of the selected lines. The SDR males and females had an average of 17% increase in weight gain at 21 days of age when fed the -Se diet compared with the control line males and females fed the same diet for the three generations of selection. The SDS line had an average reduction in weight gain of 27% during the same period of selection. A difference of 25 g was observed between the mean body weights of SDR males and females and between the SDS males and females after one generation of selection. By the third generation of selection, the difference between SDR and SDS males had increased to 31 g, whereas the SDR and SDS females differed by 41 g. Early response to selection was asymmetrical because the response was greater in the SDS than in the SDR line. Response to selection, however, generally declined after the first generation. Realized heritability estimates for individual generations for this trait were variable (ranging from -.26 to .95), but cumulative estimates (.15 to .39) for growth through three generations were similar to those reported for chickens and quail. Females had larger estimates of heritability than males, which suggests the possiblity of sex-linked effects for this trait. This study establishes the response of the chick to nutritional Se deficiency as an heritable trait. (Key words: selection, selenium, growth response, heritability)
210
CUNNINGHAM ET AL.
An unselected control line was established by using 25 males and 100 females from the group of chicks set aside to form the control. Initial selection was random, except males and females were selected to give the broadest possible representation of the original 60 sire families of the AC line. The control line was perpetuated by forming 25 pedigreed sire families with matings arranged such that full and half-sib relationships were avoided. All matings of controls were made by artificial insemination. For each generation of the control, approximately 400 chicks were assigned randomly to either of two dietary treatment groups of equal size. One group received the Se-deficient basal diet, whereas the second group received the same basal diet supplemented with . 15 ppm Se as Na 2 Se0 3 . The two control groups were used for comparisons with the selected lines. For each generation, means of 21-day-weight gain for each selected line were adjusted accordTABLE 1. Composition of the Se-deficient basal diet1 Ingredient Glucose monohydrate Amino acid mix 2 Corn oil, refined Monoolein Linoleic acid Sodium taurocholate Vitamin mix 3 Mineral mix 4 Choline chloride, 70% solution Total 1
Percentage 65.31 23.53 4.00 .50 .50 .10 .20 5.56 .20 100.00
Contained (by analysis) .010 ppm Se.
'Amino acid mix supplied (per kilogram diet): (g) L-arginine HCl, 11.6; L-glutamic acid, 122.7; glycine 12; L-histidine, 4.7; L-isoleucine, 9.5; L-leucine, 16; L-lysine HCl, 12; DL-methionine, 8.2; L-phenylalanine, 8.2; L-proline, 3; L-threonine, 8.2; L-tryptophan, 2.2; L-tyrosine, 7.0; L-valine, 10.0. 3 Vitamin mix supplied (per kilogram diet): vitamin A (all-trcms-retinyl palmitate), 16,250 IU; vitamin D 3 , 800 IU; all-rac-alpha-tocopheryl acetate, 100 IU; menadione sodium bisulfite, 1.2 mg; biotin, .2 mg; vitamin B, 2 , .014 mg; d-calcium pantothenate, 30 mg; folic acid, 4 mg; niacin, 50 mg; pyridoxine, 10 mg; riboflavin, 10 mg; thiamin HCl, 20 mg; ethoxyquin, 125 mg.
"Mineral mix supplied (per kilogram diet): CaHP0 4 , 19.8 g; CaC0 3 , 7.9 g; NaHC0 3 , 11 g; KHC0 3 , 11 g; MnS0 4 , .2; MgC0 3 , 5 g; FeS0 4 7 H 2 0 , .45 g; K10 3 , 10 g; N a M o 0 4 - 2 H 2 0 , 8 g; C r 2 K 2 ( S 0 4 ) 2 - 2 4 H 2 0 , 1 mg.
Downloaded from http://ps.oxfordjournals.org/ at University of Georgia on June 23, 2015
chickens was used as the base population for this study (Hess, 1962). For the initial generation of this study, approximately 500 pedigreed AC eggs, obtained from Athens, GA, were incubated. Upon hatching, the chicks were sexed and assigned randomly to either of two populations of approximately equal size. Each group of chicks comprised offspring representing the original 60 sire families of the AC population. One group formed the basis for a genetic control line; the second group was used to establish Se-deficient susceptible (SDS) and Se-deficient refractory (SDR) lines. Chicks were reared in thermostatically controlled, wire floored battery brooders in a room with a 15-hr day. A Se-deficient basal diet (Bunk and Combs, 1981) was fed from 1 to 21 days of age (Table 1). This diet contained DLmethionine as the only source of sulfur-containing amino acids; it was supplemented with 100 IU of vitamin E. Fluorometric analysis of the diet by the method of Olson et al. (1975) as all-rac-alpha-tocopheryl acetate indicated that it contained less than .10 ppm total Se. Feed and water were provided ad libitum during the study for all lines. Selective Breeding. The procedure used to separate the lines was mass selection for body weight gain to 21 days of age using the Se-deficient basal diet. The 60 female and 20 male chicks representing both extremes in growth formed the first selected generation. The 30 females and 10 males that phenotypically displayed a refractory response (i.e., showed best growth) to Se deficiency and the 30 females and 10 males most susceptible to the Se deficiency (i.e., showed poorest growth) were used to establish the SDR and SDS lines, respectively. For each selected line, 10 pedigreed sire families were created with each sire mated to three dams. Matings within lines were at random, except that full and half-sib matings were avoided to minimize inbreeding. All matings were made by artificial insemination. For subsequent generations, approximately 60 male and 150 female offspring from each of the selected lines were hatched and evaluated for 21-day growth response using the Se-deficient diet as described for the initial generation. Individuals selected for propagation of lines were returned to the same practical-type growing and maintenance diets, which were supplemented with .10 ppm Se as Na 2 Se0 3 at 21 days of age. Males and females were maintained in separate holding pens until reproductive age.
211
DIVERGENT SELECTION FOR SE-DEFICIENCY
RESULTS AND DISCUSSION
Use of a control population in a selection experiment enables the separation of genetic and nongenetic sources of variation that may occur over time. If, however, time trends are present in a control population, adjustments of selected lines to control levels can produce errors in evaluation of selection response (Hill, 1972). Therefore it is important to evaluate control population stability over generations. Body weight gains for control lines fluctuated from generation to generation (Table 2); how-
TABLE 2. Twenty-one-day body weight gain by genetic line, sex, and generation Generation
Genetic line
Sex
Parental weight
1
Control (+Se) Control (+Se)
M F
128 115
121 ± 3.0* 103 ± 4.0 b
SDR1 SDR
M F
89 93
Control (-Se) Control (-Se)
M F
SDS2 SDS
M F
2
3
122 ± 2.1* 118 + 2.1*
137 ± 1.9* 123 ± 1.8b
70 ± 3.0C 76 ± 2.0 c d
106 ± 2.6 b 106 ± 3.4 b
76 ± 2.8 d 88 ± 3.1 c
89 93
60 + 2.0 d e 63 ± 2.0 e
91 ± 2 . 1 d 97 + 2 . 1 c
67 ± 1.5 e 70 ± 1.5 de
89 93
45 ± 2.0 f 51 ± 2.0 f
64 ± 3.3 d 74 ± 3.0 d
45 ± 1.9f 47 ± 2.0 f
(g)
a—f Mean ± SE with different superscripts within columns are significantly different (P<.05). 1
SDR = Selenium-deficient refractory line.
2
SDS = Selenium-deficient susceptible line.
Downloaded from http://ps.oxfordjournals.org/ at University of Georgia on June 23, 2015
ever, regression of weight gain means on generations produced nonsignificant (P>.05) regression coefficients of .08 and .09 for males and females, respectively, for the control population fed the Se-supplemented diet and -.02 for either males or females of the control population fed the Se-deficient diet. These nonsignificant regression coefficients indicate that phenotypic trends over time were not present for the control. Therefore, selected line means were adjusted to control levels for each generation. Weight gains for the selected lines were significantly different (P<.05) from the control after one generation of selection (Table 2). Selection for high growth response with the Sedeficient diet (SDR) resulted in moderate improvements in 21-day-weight gains when compared with the control. Weight gains for the SDR line, averaged over the three generations of selection, were approximately 17% greater than control gains. Selection for reduced weight gain (SDS) with the Se-deficient diet, however, produced a more pronounced separation from the control than was observed with the SDR line (Figure 1). Weight gains for the SDS line averaged 27% less than the Se-deficient control for the three selected generations. Differences in 21-day-weight gain between the selected lines averaged 38% during the selection process. Males tended to be more sensitive to Se-deficiency but differences among the sexes for weight gain were not significant (P>.05) except for the third generation of selection, where females of the SDR line gained 12 grams more
ing to growth of the control line offspring fed the Se-deficient diet. Within each generation, the standardized selection differential (S) and the standardized response (R) to selection were computed. Selection differentials were weighted for numbers of progeny produced by each parent to produce a realized selection differential. Realized heritabilities (h2) were calculated from the R:S ratio for each line each generation. For Generation 3, growth rates of selected lines after the 21-day testing period were assessed by measuring body weights at 12 and 16 weeks. Hatchability for all eggs set and liveability during the 21 days of testing with the Se-deficient diet also were evaluated. Data for each generation were subjected to analysis of variance; when significant (P<.05) effects were detected, means were separated by a 5% Duncan's multiple range test (Barr et al., 1976).
212
CUNNINGHAM ET AL.
SDR(F)..S
SDRIMT 8
I<1^-^ GAIN (g)
Control
^ ^ \ ^ ^ "''"t3
SDS(MT~ Y
1
GENERATION
than males. LaVorgna and Combs (1983) also observed sex differences for weight gains of chicks fed Se-deficient diets. Although the present selection process resulted in improved growth response for SDR chicks exposed to the Se-deficient diets compared with control chicks fed the Se-deficient diet, the SDR line did not achieve body weight gains comparable to control chicks fed Se-supplemented diets (Table 2). Expected selection differentials, effective selection differentials, and response to selection for 21-day-growth rate are presented in Table 3. Selection differentials were weighted for numbers of progeny produced for each parent (Falconer, 1981). Unweighted and weighted selection differentials were in close agreement except for Generation 3 SDR males and SDS females. This difference occurred because some of the SDR line males and SDS line females were inadvertently discarded during rearing. This effectively reduced the population size of
TABLE 3. Expected selection differentials, effective selection differentials, and response to selection in standard deviation units Expected selection differentials SDR1
Effective selection differentials
SDS 2
SDR
SDS
Response SDR
SDS
Generation
Cumulative
1.38 1.22 1.20
.68 1.15 1.10
3.80
2.93
1.00 1.28 1.18 3.46
.98 1.00 .97 2.95
1
SDR = Selenium-deficient refractory line.
5
SDS = Selenium-deficient susceptible line.
1.38 1.29 .95 3.62
.63 1.10 1.00
1.02 1.31 1.08
.93 1.17 .60
2.73
3.41
2.70
.52 .20 .19 .53
.60 .19 .30 .71
.67 .60 -.28
.63 .42 .0
.99
1.05
Downloaded from http://ps.oxfordjournals.org/ at University of Georgia on June 23, 2015
FIG. 1. Response to selection for 21-day body weight gain of chicks fed a Se-deficient purified diet: Se-deficiency refractory (SDR) and Se-deficiency susceptible (SDS) males (M) and females (F).
these lines and, as a result, the effective selection differentials were also reduced. Nevertheless, expected and effective selection differentials were similar, which suggests minor influences of natural selection. Selection differentials for males were consistently larger than for females because of the greater intensity of selection possible with males with the mating system employed (Table 3). Response to selection generally declined after the first generation and was negligible for Generation 3 (Table 3). Lack of response to selection observed for Generation 3 may have resulted from a rapid reduction in additive genetic variance associated with the selection process for this trait. These results are similar to those of Hess et al., 1962, who reported that response to selection for chicks on a methionine-deficient diet was rapid for the first two generations but negligible thereafter. There appeared to be asymmetry of response with the SDS line showing the greater response to selection (Fig. 1). Because this selection program employed only a few generations, it is difficult to speculate on asymmetry of response. Nevertheless, it is useful to note that inequality in response to selection is not unusual and has a number of possible causes, as discussed by Falconer (1981). Natural selection does not appear to have been a major factor during the development of these lines to this point. Thus, a more plausible explanation may be that longterm natural selection may have favored genes that facilitate growth of chicks fed Se-deficient diets. In that case, artificial selection would be expected to produce a more rapid response in the opposite direction, as observed in the present study.
DIVERGENT SELECTION FOR SE-DEFICIENCY
+i +i +i +i *JL
213
ACKNOWLEDGMENTS
Appreciation is expressed to Henry L. Marks, Southeast Poultry Research Laboratory, Athens, GA, for supplying the base population of Athens-Canadian Ramdombred chickens for this study. The expert technical assistance of Lynne
Downloaded from http://ps.oxfordjournals.org/ at University of Georgia on June 23, 2015
As a result of the variable response to selection, cumulative values for realized estimates of heritability were also variable, ranging from .15 for the SDR males to .39 for the SDS females. Estimates of heritability for this trait (i.e., 21day gain in body weight of chicks fed a Se-deficient diet) have not previously been reported in the literature; however, the values obtained here are similar to other estimates of heritability of juvenile body weight gains for chickens (Siegel, 1962) and Japanese quail (Marks and Lepore, 1968). Estimates of heritability obtained in this study demonstrate the presence of additive genetic variability for this trait. As estimates of heritability were generally different for males and females, it is possible that sexlinked effects are operating for this trait as well. Reports by Brunson et al, (1956) and Thomas et al. (1958) suggest that sex-linked genes may be of considerable importance in the inheritance of body weight traits for chickens. Table 4 presents body weights, hatchability, and liveability rates determined for the genetic lines in Generation 3. Body weights of the SDR, SDS, and control lines were significantly (P< .05) different at the end of 21 days of feeding the Se-deficient diet. By 12 weeks of age, the SDR line males and females and the SDS line males had compensated in body weight gain and were equal to or greater in body weight in comparison with control line individuals previously fed the Se-supplemented diet. The SDS females achieved body weights comparable to the control line females by 16 weeks of age. Hatchability of fertile eggs of each selected line was not different from that of the control line, except for SDR males, which demonstrated a significant reduction for this trait. Survival rate to 21 days of age was not significantly affected for any of the selected lines. These results demonstrate successful divergent selection for susceptibility of chicks to dietary Se-deficiency as measured by 21-dayweight gain. Two lines differing in this trait have been established: SDR and SDS. These lines should provide useful material for studies of the metabolic bases of the nutritional action of Se and related nutrients.
CUNNINGHAM ET AL.
214
P. Deuschle and the clerical assistance of Barbara J. Smagner in preparing this manuscript is also greatly appreciated. REFERENCES
Downloaded from http://ps.oxfordjournals.org/ at University of Georgia on June 23, 2015
Barr, A. J., J. H. Goodnight, J. P. Sail, and J. T. Helwig, 1976. A User's Guide to SAS 76. SAS Inst., Raleigh, NC. Bunk, M. J., and G. F. Combs, Jr., 1981. Relationship of selenium-dependent glutathione peroxidase activity and nutritional pancreatic atrophy in selenium-deficient diets. J. Nutr. 111:1611-1620. Brunson, C. C , G. F. Godfrey, and B. L. Goodman, 1956. Types of gene action in the inheritance of ten week body weight and breast angle of broilers. Poultry Sci. 35:524-532. Falconer, D. S., 1981. Introduction to Quantitative Genetics. Ronald Press Co., New York, NY. Halpin, K. M., andD. H. Baker, 1984. Selenium deficiency and transsulfuration in the chick. J. Nutr. 114:606612. Hess, C. W., 1962. Randombred populations of the Southern Regional Poultry Breeding Project. World's Poult.
Sci. J. 18:147-152. Hess, C. W., H. M. Edwards, Jr., and E. F. Dembnicki, 1962. Growth-rate selection on a methionine deficient diet. Poultry Sci. 41:1042-1047. Hill, W. G., 1972. Estimation of realized heritabilities from selection experiments. II. Selection in one direction. Biometrics 28:767-780. LaVorgna, M. W., and G. F. Combs, Jr., 1983. Evidence of a hereditary factor affecting the chick's response to uncomplicated selenium deficiency. Poultry Sci. 62:164-168. Marks, H. L., and P. D. Lepore, 1968. Growth rate inheritance in Japanese quail. 2. Early responses to selection under different nutritional environments. Poultry Sci. 47:1540-1546. Olson, O. E., I. S. Palmer, and E. E. Cary, 1975. Modification of the official method for selenium implants. J. Assoc. Offic. Agric. Chem. 58:117-121. Siegel, P. B., 1962. Selection for body weight at eight weeks of age. 1. Short term response and heritabilities. Poultry Sci. 41:954-962. Thomas, C. H., W. L. Blow, C. C. Cockerham, and E. W. Glazener, 1958. The heritability of body weight gain, feed consumption, and feed conversion in broilers. Poultry Sci. 37:862-869.