- Email: [email protected]
Response to ten generations of divergent selection for tibial dyschondroplasia in broiler chickens: growth, egg production, and hatchability
BREEDING AND GENETICS Responses to Ten Generations of Divergent Selection for Tibial Dyschondroplasia in Broiler Chickens: Growth, Egg Production, and Hatchability X. ZHANG,*,1 G. R. MCDANIEL,* D. A. ROLAND,* and D. L. KUHLERS† Departments of *Poultry Science and †Animal and Dairy Sciences, Alabama Agricultural Experiment Station, Auburn University, Alabama 36849-5416 later (6 to 10) generations, in contrast to nonsignificant responses for both durations in the LTD line. The 4-wk BW of the HTD line was slightly heavier than or similar to that of the LTD line within generations. The HTD line birds tended to decrease 7-wk BW with advancing generations. The trend of changes in BW was not as clear in the LTD line as in the HTD line. The variability of 7-wk BW had an increased trend with advancing generations in the HTD line, accompanied by a decreased additive genetic variability of TD due to continued selection. The average EP in the LTD hens was 7.6 percentage points higher than in the HTD from Generations 1 through 10. Mean hatchability in the LTD line did not differ from that in the HTD line within generation. Responses of EP and hatchability, components associated with fitness, appeared slower towards increased fitness than towards decreased fitness.
(Key words: body weight, broiler, egg production, hatchability, tibial dyschondroplasia) 1998 Poultry Science 77:1065–1072
Early selection experiments of tibial dyschondroplasia (TD) in broiler chickens found that: 1) individual selection was effective to increase or decrease TD incidence of progeny (Leach and Nesheim, 1965, 1972; Sheridan, 1974; Riddell, 1976; Sheridan et al., 1976); 2) males and females differed in TD incidence (Leach and Nesheim, 1965; Sheridan, 1974; Sheridan et al., 1976); 3) the estimate of heritability (h2) for TD incidence changed from 0.2 to 0.3 (Sheridan et al., 1978; Burton et al., 1981); 4) the genetic correlation between BW and TD incidence was negative (Sheridan et al., 1976; Burton et al., 1981) or positive (Riddell, 1975; Poulos et al., 1978; Kiiskinen and Andersson, 1982). A short-term selection experiment was initiated to study the genetic susceptibility of broiler chickens to TD
and relationships of TD incidence with growth and reproduction traits. Divergent selection for TD resulted in two lines: high (HTD) and low (LTD) incidence of TD. A randombred control line was maintained to estimate the trend of environmental fluctuations between generations. Direct (Wong-Valle et al., 1993a) and correlated (Wong-Valle et al., 1993b) responses to selection have been reported for the first four generations. No difference in average BW was found between HTD and LTD lines, although TD incidence in these two lines differed by 20-fold. Egg production (EP) and semen concentration were greater in the LTD line than in the HTD line. The present study reports on 1) direct response of TD incidence to 10 generations of selection for increased and decreased TD; 2) correlated responses of BW at 4 and 7 wk of age, egg production, and hatchability; and 3)
Received for publication September 10, 1997. Accepted for publication March 31, 1998. 1To whom correspondence should be [email protected]
Abbreviation Key: CSDA = adjusted cumulative selection differential; CSDU = unadjusted cumulative selection differential; EP = egg production; H = high incidence; h2 = heritability; L = low incidence; TD = tibial dyschondroplasia.
INTRODUCTION
addressed:
1065
Downloaded from http://ps.oxfordjournals.org/ at Serials Section, Dixson Library on June 25, 2015
ABSTRACT Continued genetic selection for improved BW gain has met an obstacle of skeletal disorders in broiler chickens. Two broiler chicken lines (HTD and LTD) were developed by 10 generations of divergent selection for tibial dyschondroplasia (TD) incidence originating from commercial primary breeders. The reference population was a randombred control line maintained along with the selected lines. Relationships of TD incidence with BW, egg production (EP), and hatchability were assessed using these lines. The response of TD to selection was asymmetric, favoring an increased TD incidence. Mean TD incidence increased 7.6 percentage points per generation during Generation 1 through 10 in males and 9.1 percentage points in females of the HTD line but did not change significantly in the LTD line at 4 wk of age. Responses of the HTD line in early (1 to 4) generations were greater than in
1066
ZHANG ET AL.
comparison of early (Generation 1 to 4) vs later (6 to 10) selection response.
MATERIALS AND METHODS
Maintenance of Lines
Measurements and Analyses of Traits Incidence of TD was scored as either 0 for normal or 1 for abnormal by examining the left and right proximal tibiotarsus of birds at 4 (TD4) and 7 (TD7) wk of age using a lixiscope (Bartels et al., 1989). Body weight was also obtained at 4 (BW4) and 7 (BW7) wk of age. The TD4 was measured consecutively for 10 and BW4 for eight generations, whereas TD7 and BW7 were recorded for seven generations. Before selection of breeders in the lines, 60 hens were selected randomly within each line to test for EP. Daily egg production per hen was recorded from 24 to 36 wk of age in Generations 1 through 7 and 24 to 30 wk in Generations 8 through 10. Abnormal eggs (broken, poorshelled, shell-less, and miniature) were also recorded. Starting from 32 wk of age, selected breeder hens were inseminated artificially twice a week. Eggs were collected from the 3rd to 18th d once commencing insemination. Hatchability of eggs set was recorded from selected breeders in all generations.
Ingredients
Starter
Grower
Pullet developer
Corn Soybean (48% protein) Poultry oil Alfalfa meal (17% protein) Dibasic calcium phosphate Calcium carbonate Salt Vitamin premix1 Mineral premix2 DL-methionine Biocox Calculated analysis Protein Calcium Available phosphorus ME, kcal/kg
50.15 38.50 7.00 . . . 1.90 1.20 0.35 0.25 0.25 0.30 0.10
(%) 60.30 30.55 5.25 . . . 1.65 1.10 0.35 0.25 0.25 0.20 0.10
68.56 15.30 . . . 13.54 1.10 0.50 0.35 0.25 0.25 0.15 . . .
22.85 1.07 0.47 3,203
19.91 0.96 0.42 3,214
15.54 0.79 0.30 2,895
1Vitamin premix provided per kilogram of finished feed: vitamin A (retinyl palmitate), 7,350 IU; cholecalciferol, 2,200 IU; vitamin E (dl-atocopheryl acetate), 8 IU; riboflavin, 5.5 mg; d-pantothenic acid, 13.0 mg; niacin, 36 mg; choline, 500 mg; vitamin B12, 0.02 mg; menadione, 2 mg; folic acid, 0.5 mg; thiamine mononitrate, 1.0 mg; pyridoxine, 2.2 mg; dbiotin, 0.05 mg. 2Mineral premix provided in milligrams per kilogram of finished feed: Cu, 6.0; Fe, 54.8; I, 1.0; Mn, 65.3; Se, 0.3; Zn, 55.0.
Linear regression of responses on generations was used to evaluate selection response trend across generations. Regression coefficients were tested for significance from zero by t test. Realized h2 of TD4 and TD7 were estimated by: 1) regressing mean responses of the population on cumulative selection differentials unadjusted for the number of progeny and 2) regressing mean responses of offspring on values of sires and dams within generations to assess changes of genetic variation with the progress of selection. Incidence of TD in dam and sire families was regressed on Generation 1 to 4 and 6 to 10 to compare responses of families to selection of earlier vs later generations. The cumulative selection differential of TD incidence adjusted (CSDA) and unadjusted (CSDU) for the number of progeny were regressed separately on generations. The CSDA measures the joint effects of natural and artificial selection, whereas CSDU measures the effect of artificial selection (Falconer, 1981). The comparison between these two regression coefficients might reflect the influence of natural selection on TD incidence. The ANOVA for BW was performed by generation using the General Linear Models (GLM) procedure (SAS Institute, 1988) with the following model: Yijklm = u + Li + Sij + Dijk + Gl + (LG)il + eijklm [1]
where Yijklm = individual observation for a trait; u = overall mean for the trait; Li = fixed effect of the ith line; Sij = random effect of the jth sire within the ith line; Dijk = random effect of the kth dam mated to the jth sire within
Downloaded from http://ps.oxfordjournals.org/ at Serials Section, Dixson Library on June 25, 2015
Details of the early generations of selection in the HTD and LTD lines were given by Wong-Valle et al. (1993a). Briefly, this divergent selection experiment commenced from a base population consisting of 200 males and 200 females of primary broiler breeders. From the base population, 15 sires and 58 dams were divergently selected to develop the first generation of the HTD and LTD lines. Ten sires and 50 dams were randomly selected to reproduce the first generation of a randombred control line from the same base population. From Generation 2 to 10, 15 sires and 60 dams were selected from each of the three lines (about 400 birds each line) to produce the next generation of each line. In each generation of the HTD and LTD lines, sires and dams were selected based on half-sib family performance. In each generation of the control line, sires and dams were randomly selected. A selected sire was randomly mated to four dams within each line, avoiding matings of pairs with a common parent or grandparent. All chicks were wing-banded and vaccinated against Marek’s disease at hatch, and reared with lines intermingled in three floor pens with side-windowed ventilation. Chicks were given 24 h of light during the first 4 d, after which birds in all generations except Generation 5 (8 h) received 22 h of light up to 7 wk of age. Birds consumed feed and water ad libitum up to 7 wk of age and then feed was restricted according to Ross breeder’s recommendations. A starter broiler diet was fed from 0 to 3 wk, a grower from 3 to 7 wk, and a pullet developer from 7 to 24 wk of age (Table 1).
TABLE 1. Composition of the starter, grower, and pullet developer diets
1067
TIBIAL DYSCHONDROPLASIA, GROWTH, AND REPRODUCTION TABLE 2. Linear regression coefficients ± SE of means for tibial dyschondroplasia (TD) and BW on generations in the randombred control line Generations of selection Trait
Sex
TD, %
Female Male
BW, g
Female Male
Age
Overall1
(wk) 4 7 4 7 4 7 4 7
2.4 1.6 3.6 1.3 7.7 –43.8 8.6 –54.8
± ± ± ± ± ± ± ±
Earlier2 1.2 0.9 1.6 1.8 20.5 34.6 21.8 44.6
0.5 0.9 0.1 –0.9 –36.2 –105.0 –29.5 –132.8
± ± ± ± ± ± ± ±
Later3 1.4 2.3 2.3 4.5 65.9 46.3 64.7 65.1
–4.3 . . . –5.4 . . . 87.8 . . . 111.6 . . .
± 2.0 ± 2.7 ± 44.6 ± 38.6
1Overall:
1 to 10 for TD and 1 to 8 for BW at 4 wk of age; 1 to 7 for TD and BW at 7 wk of age. 1 to 4 for TD and BW. 3Later: 6 to 10 for TD; 6 to 8 for BW at 4 wk of age. 2Earlier:
Yijklm = u + Li + Sij + Dijk + eijklm
[2]
Components in the model were the same as in Equation 1 except for sex and line by sex interaction terms. Regressions of measurements for these traits on generations were used to evaluate correlated responses to selection.
RESULTS AND DISCUSSION
Response of Tibial Dyschondroplasia Incidence to Divergent Selection Fluctuation of Environments. The overall environmental fluctuation trend was not significant for incidence of TD at 4 and 7 wk of age, as estimated by linear regression of mean TD on generations in the control line (Table 2). A nonsignificant generation effect on TD incidence may demonstrate no directional environmental influence of this selection study. Considering that there was a decrease in TD incidence due to shortened photoperiod (8 h light) in Generation 5, segmented regressions of TD on generations showed environmental similarity of early (1 to 4) and later generations (6 to 10). Selected Lines. Mean TD incidence increased 7.6 percentage points per generation during Generation 1 through 10 in males and 9.1 percentage points in females of the HTD line but did not change significantly in the
LTD line at 4 wk of age, as expressed by deviations from corresponding values in the control line (Figure 1A). At 7 wk of age, TD incidence increased 10.8 percentage points per generation during Generations 1 through 7 in males and 12.7 percentage points in females of the HTD line and decreased 3.5 percentage points in males and 1.7 percentage points in females of the LTD line (Figure 1B). Response to selection in the HTD line was positive during Generation 1 through 4 but was negative or zero during Generation 6 through 10 at 4 wk of age (Table 3). No response was noted for both durations in the LTD line. At 7 wk of age, TD incidence had an increase about 9 (females) and 10 percentage points (males) per generation in full-sib families of the HTD line, in comparison to a decrease of 3 (females) and 8 (males) percentage points in the LTD line during Generations 1 to 4 (Table 3). Similar responses were observed in sire families (data not shown). Responses to divergent selection were asymmetric. Selection favored the HTD line and resulted in a ratio of mean TD incidence (HTD: LTD) greater than 5 starting from Generation 3 (data not shown). Males had a slightly higher incidence of TD than females. There was a higher incidence of TD at 7 than at 4 wk of age. The LTD line varied around 6% incidence of TD in males and 3% in females at 4 wk of age across generations. Later generations (after 5) had a poor stability of TD incidence, especially in the HTD line. Selection Differentials. Linear regression coefficients for cumulative selection differentials of TD incidence adjusted for the number of progeny on generation did not differ from zero (Table 4), indicating that selection differentials remained constant from Generations 0 through 10. However, cumulative selection differentials unadjusted for the number of progeny decreased more than 11 percentage points per generation in the HTD line and remained unchanged in the LTD line. The linear regression coefficients for adjusted selection differentials differed from those for unadjusted ones in the HTD line (Table 4). A comparison between adjusted and unadjusted selection differentials suggested that natural selection
Downloaded from http://ps.oxfordjournals.org/ at Serials Section, Dixson Library on June 25, 2015
the ith line; Gl = fixed effect of the lth sex; (LG)il = effect of line by sex interaction; and eijklm = random error. The mean squares of sires within lines were used as the error term to test line effects. The effect of sires within lines was tested against dams within sires and lines. Mean BW of birds with or without TD was regressed on generations within line and sex to estimate relationships between BW and TD. The ANOVA for EP and hatchability was performed by generation using the General Linear Models (GLM) procedure with model:
1068
ZHANG ET AL. TABLE 3. Linear regression coefficient ± SE of tibial dyschondroplasia (TD) incidence in full-sib families on generations Generations of selection Line1
Overall2
Sex
1 to 4
6 to 10
(%) 4 wk of age HTD Female Male LTD Female Male 7 wk of age HTD Female Male LTD Female Male
13.0 11.9 –0.1 –0.6
± ± ± ±
0.8** 0.9** 0.2 0.4
5.7 11.9 –0.4 –1.6
± ± ± ±
1.7** 2.1** 0.6 1.1
–2.4 0.5 –0.9 –1.1
13.2 11.1 –0.8 –2.9
± ± ± ±
0.8** 0.8** 0.3** 0.5**
8.9 10.4 –2.6 –7.8
± ± ± ±
2.0** 2.3** 0.7** 1.5**
. . . .
. . . .
± ± ± ±
1.1* 0.9 0.5 0.8
. . . .
FIGURE 1. Changes in tibial dyschondroplasia (TD) incidence at 4 (TD4, A) and 7 wk of age (TD7, B); HTD = line with high incidence of TD; LTD = line with low incidence of TD (LTD). Values are represented as deviations from the randombred control line.
exert some sort of influence upon artificial selection in the HTD line and little impact in the LTD line. This implies that the HTD line, once selection was relaxed, would regress toward the level of the base population from which selection had been initiated, whereas the LTD line might move slowly to the original level of the base population. Realized Heritability. Realized h2 for TD4, estimated by regressing mean responses of the population on CSDU were greater than 0.5 in the HTD line and less than 0.5 in the LTD line (Table 5). Corresponding estimates for TD7 were uncomfortably larger than 0.8 in the HTD line and close to 0.1 in the LTD line (Table 5). The estimate for males was higher than that for females (Table 5). Realized h2 for the HTD line were greater in this report than those in previous studies (Sheridan et al., 1978; Burton et al., 1981; Wong-Valle et al., 1993a; Zhang et al., 1995; and Kuhlers and McDaniel, 1996). Because CSD were decreased (Table 4) and mean responses increased
with advancing generations in the HTD line, current estimates based on 10 generations of selection are assumably greater than based on the first few generations of selection in previous papers. Divergent selection produced different estimates of realized h2 between the HTD and LTD lines. They were the consequence of asymmetric responses. The incidence of TD, as a component of characters associated with natural fitness, responded to selection much more slowly towards increased fitness than towards decreased fitness. Divergent selection operated in favor of the HTD line, whereas the LTD line fluctuated around 5%
TABLE 4. Linear regression selection differentials for incidence adjusted and of progeny
coefficient ± SE of cumulative tibial dyschondroplasia (TD) unadjusted for the number on generations Cumulative selection differential
Line1
Sex
Adjusted
Unadjusted (%)
4 wk of age2 HTD LTD 7 wk of age HTD LTD
Female Male Female Male
–0.5 –0.6 0.0 –0.1
± ± ± ±
0.4 0.3 0.2 0.1
–13.2 –12.0 0.1 0.7
± ± ± ±
4.3* 3.2* 0.3 0.9
Female Male Female Male
–0.6 0.2 –0.1 0.2
± ± ± ±
0.3 0.5 0.2 0.2
–13.6 –11.3 0.8 3.0
± ± ± ±
1.9** 2.2** 0.7 1.7
1HTD = line with high incidence of TD; LTD = line with low incidence of TD. 2Generations 0 to 10 at 4 wk of age and 0 to 7 at 7 wk of age. *P < 0.05. **P < 0.01.
Downloaded from http://ps.oxfordjournals.org/ at Serials Section, Dixson Library on June 25, 2015
1HTD = line with high incidence of TD; LTD = line with low incidence of TD. 2Generations 1 to 10 at 4 wk of age and 1 to 7 at 7 wk of age. *P < 0.05. **P < 0.01.
TIBIAL DYSCHONDROPLASIA, GROWTH, AND REPRODUCTION TABLE 5. Realized heritabilities (h2) of tibial dyschondroplasia (TD) incidence estimated by regressing mean responses of the population on cumulative selection differentials unadjusted for the number of progeny h2 for TD at Line1
Sex
Week
HTD
Female Male Female Male
0.65 0.52 0.50 0.40
LTD
± ± ± ±
42
Week 7
0.53 0.47 0.32 0.22
1.06 0.86 0.10 0.05
± ± ± ±
0.25 0.33 0.14 0.13
1HTD = line with high incidence of TD; LTD = line with low incidence of TD. 2Generations 0 to 10 at 4 wk of age and 0 to 7 at 7 wk of age.
comparison of positive regression coefficients of birds afflicted with TD at 4 wk of age and negative values at 7 wk of age (Table 6) and 2) the fact that BW4 in the HTD line birds was slightly heavier than or similar to that in the LTD line within generation (data not shown). Afflicted birds at early ages were reluctant to move and reduced ability of competing for feed and water, causing a decrease of BW7 in the HTD line. Therefore, the relationship of TD incidence with BW appeared to be age-related. The incidence of TD exhibited negative genetic correlations with BW estimated from sire’s and dam’s components of variance and covariance of the seventh generation of divergently selected lines (Zhang et al., 1995). Sheridan et al. (1976) and Burton et al. (1981) found a negative correlation between BW7 and TD7, whereas a positive association between BW and TD was observed prior to 7
Correlated Responses to Divergent Selection Body Weight. The general environmental fluctuation trend was not clear for BW due to a broad range of variation, as shown in the control line (Table 2). The HTD line birds tended to decrease BW7 during Generations 1 through 7 (Figure 2) but this was uncertain due to great fluctuation for BW4 (Table 6). Although mean BW of the LTD birds demonstrated a positive change at 4 wk of age and negative at 7 wk of age with advancing generations of selection (Table 6), the trend was not significantly evident as in the HTD line. The TD incidence appeared, to some extent, to be associated with an enhanced BW prior to 4 wk of age, but not strongly. This association was determined by: 1) a
FIGURE 2. Changes in 7-wk body weight (BW7) of birds with advancing generations of divergent selection for tibial dyschondroplasia (TD); HTD = line with high incidence of TD; A) with TD and B) with no TD.
Downloaded from http://ps.oxfordjournals.org/ at Serials Section, Dixson Library on June 25, 2015
incidence of TD at 4 wk of age across generations. A main cause for asymmetrical responses could be that TD incidence in the base population was not midway between the two theoretical limits of the HTD and LTD lines (100 and 0%, respectively). Therefore, responses to divergent selection had further to go in the HTD line than in the LTD. This genetic asymmetry was caused by disequilibrium gene frequency of the base population. Sheridan (1974) presumed a major dominant sex-linked gene controlling TD. Later, a significant relationship between the sex-linked gene for rapid feathering and the incidence of TD in the first three generations of selection was observed by Sheridan et al. (1976). Additive genetic effect, rather than dominant or maternal, could be important for TD, as demonstrated by a study of diallel crosses between HTD and LTD lines developed by divergent selection of seven generations (Yalcin et al., 1995). They found that progeny from the cross of the HTD × HTD had a higher incidence of TD at 4 and 7 wk of age than from either of the reciprocal crosses, which, in turn, were higher than from the cross of the LTD × LTD. The reciprocal crosses were similar in TD incidence at both ages. A poor stability of TD incidence in higher generations, especially in the HTD line, was the consequence of the loss or decrease of additive genetic variability accompanying selection.
1069
1070
ZHANG ET AL. TABLE 6. Linear regression coefficient ± SE of 4- and 7-wk mean BW on generations1 Mean BW at
Line2
Sex
TD3
Week 4
HTD
Female
HTD
Male
LTD
Female
LTD
Male
0 1 0 1 0 1 0 1
–3.7 3.8 –15.6 0.8 11.1 3.5 7.4 12.2
Week 7 (g)
1Generations
± ± ± ± ± ± ± ±
17.7 17.4 21.3 18.3 20.4 15.6 21.8 23.5
–61.9 –47.2 –87.7 –73.4 –43.0 –49.8 –59.1 –9.7
± ± ± ± ± ± ± ±
28.1† 25.4 36.7 34.2† 40.8 45.7 48.8 61.5
1 to 8 at 4 wk of age and 1 to 7 at 7 wk of age.
2HTD = line with high incidence of tibial dyschondroplasia (TD); LTD
wk of age (Riddell, 1975; Poulos et al., 1978; Kiiskinen and Andersson, 1982). An age-related correlation between TD and BW was reported also in turkeys (Rath et al., 1994). There was no correlation between BW and TD of turkeys at early ages, whereas turkeys with severe TD had heavier BW relative to those without or with mild lesions during Weeks 14 and 15. Variability of BW7 had an increasing trend with advancing generations of selection in the HTD line (Figure 3A,B) and also of BW7 of males with no TD in the LTD line (Figure 3C). The increased variability of BW7 in the HTD line was accompanied by a decreasing additive genetic variability of TD due to continued selection. Does this presuppose that pleiotropic genes might be excluded to interpret the relationship between BW and TD? Of course, the increase of variability for BW7 could be the consequence of no selection for BW or environmental fluctuation. Egg Production. Egg production of the control line was influenced little from Generation 1 to 10. Egg production increased from Generation 1 to 5, in both the HTD and LTD lines. A reverse trend occurred from Generation 7 to 10 (Figure 4). The LTD line hens had a greater (P < 0.05) EP than the HTD hens within every generation (Figure 4). The average EP in the LTD line hens was 7.6 percentage points higher than in the HTD line from Generation 1 to 10, ranging from 4 to 14 percentage points. The increase (P < 0.05) of EP in both HTD and LTD lines from Generation 1 to 5 was because of an environmental influence instead of correlated responses to selection. This result can be seen from the fact that the randombred control line had a similar increase (P < 0.05) in EP from Generation 1 to 5. However, the decrease of EP from Generation 6 to 10 in both HTD (P < 0.01) and LTD (P = 0.06) lines could be due to correlated responses to selection, because the control line (P = 0.15) did not show the same trend in EP during that period.
FIGURE 3. Changes in CV for 7-wk body weight (BW7) with advancing generations of divergent selection for tibial dyschondroplasia (TD): A) birds with TD and B) with no TD in the line of high incidence of TD (HTD); and C) with no TD in the line of low incidence of TD. The CV of 7-wk BW for HTD females with TD, LTD males and females with TD, and LTD females without TD was not different from generation to generation (P > 0.1) and not showed.
Paired differences in TD incidence between the LTD and HTD lines became greater with advancing generations of selection, to which paired differences in mean EP between lines were not proportional. A plot of paired differences in mean EP between the LTD and HTD lines against corresponding differences in TD incidence was nearly flat across generations (data not shown). This effect weakened the conclusion that TD incidence decreases EP in broiler hens. The flat plot occurred possibly because selection of hens for testing for EP was not completely random. Hens with TD in the HTD line and those with no TD in the LTD line were more likely to be selected as testers.
Downloaded from http://ps.oxfordjournals.org/ at Serials Section, Dixson Library on June 25, 2015
= line with low incidence of TD. 30 = no TD, 1 = TD. †P < 0.1.
TIBIAL DYSCHONDROPLASIA, GROWTH, AND REPRODUCTION
1071
REFERENCES FIGURE 4. Changes in mean egg production (EP, per hen-day) of 60 hens in selected and randombred control lines with advancing generations of divergent selection for tibial dyschondroplasia (TD); HTD = line with high incidence of TD; LTD = line with low incidence of TD. Values were recorded for a period of 12 wk (24 to 36 wk old) in Generations 1 through 7 and only for 6 wk (24 to 30 wk old) in Generations 8 through 10.
FIGURE 5. Changes in hatchability of selected and randombred control lines with advancing generations (6 to 10) of divergent selection for tibial dyschondroplasia (TD); HTD = line with high incidence of TD; LTD = line with low incidence of TD. The data of the first 5 generations was not accessible due to the leave of a person involved in this study during that period.
Bartels, J. E., G. R. McDaniel, and F. J. Hoerr, 1989. Radiographic diagnosis of tibial dyschondroplasia in broilers: a field selection technique. Avian Dis. 33:254–257. Burton, R. W., A. K. Sheridan, and C. R. Howlett, 1981. The incidence and importance of tibial dyschondroplasia to the commercial broiler industry in Australia. Br. Poult. Sci. 22: 153–160. Falconer, D. S., 1981. Introduction to Quantitative Genetics. 2nd ed. Longman, New York, NY. Kiiskinen, T., and P. Andersson, 1982. The incidence of tibial dyschondroplasia in two broiler strains and their performance on different diets. Ann. Agric. Fenn. 21:169–176. Kuhlers, D. L., and G. R. McDaniel, 1996. Estimates of heritabilities and genetic correlations between tibial dyschondroplasia expression and body weight at two ages in broilers. Poultry Sci. 75:959–961. Leach, R. M., Jr., and M. C. Nesheim, 1965. Nutritional, genetic and morphological studies of an abnormal cartilage formation in young chicks. J. Nutr. 86:236–244. Leach, R. M., Jr., and M. C. Nesheim, 1972. Further studies on tibial dyschondroplasia (cartilage abnormality) in young chicks. J. Nutr. 102:1673–1680. Poulos, P. W., S. Reiland, K. Elwinger, and S. E. Olsson, 1978. Skeletal lesions in the broiler, with special reference to dyschondroplasia (osteochondrosis). Pathology, frequency and clinical significance in two strains of birds on high and low energy feed. Acta-Radiologica. Suppl. 358:229–275. Rath, N. C., G. R. Bayyari, J. N. Beasley, W. E. Huff, and J. M. Balog, 1994. Age-related changes in the incidence of tibial dyschondroplasia in turkeys. Poultry Sci. 73:1254–1259. Riddell, C., 1975. Studies on the pathogenesis of tibial dyschondroplasia in chickens. III. Effect of body weight. Avian Dis. 19:497–505. Riddell, C., 1976. Selection of broiler chickens for high and low incidence of tibial dyschondroplasia with observations on spondylolisthesis and twisted legs (Perosis). Poultry Sci. 55: 145–151. SAS Institute, 1988. SAS/STAT User’s Guide. 6th ed. SAS Institute Inc., Cary, NC.
Downloaded from http://ps.oxfordjournals.org/ at Serials Section, Dixson Library on June 25, 2015
Hatchability. Mean hatchability of the control line was not affected from Generation 6 through 10 (Table 2) and data for Generation 1 through 5 was missing due to lost communication with a person involved in this project during that period of time. The trend of changes in hatchability with generations was not clear in the selected lines. The LTD line did not differ (P > 0.1) from the HTD line within every generation (Figure 5). Similarly, the plot of paired differences in mean hatchability between the LTD and HTD lines against corresponding differences in TD incidence was flat across generations (data not shown). The flat plot demonstrated that the association of hatchability with TD incidence was not as tight as expected. The lack of close relationships between hatchability and TD incidence from Generation 6 to 10 did not contradict the report of no differences in sexual maturity and semen volume and concentration between the HTD and LTD lines in early generations (Wong-Valle et al., 1993b).
1072
ZHANG ET AL.
Sheridan, A. K., 1974. Genetic studies of tibial dyschondroplasia in the broiler chicken. Page: 86 in: Proceedings Australian Poultry Science Convention, World’s Poultry Science Association, Hobart, Tasmania, Australia. Sheridan, A. K., C. R. Howlett, and R. W. Burton, 1978. The inheritance of tibial dyschondroplasia in broilers. Br. Poult. Sci. 19:491–499. Sheridan, A. K., C. R. Howlett, and E. B. Dettmann, 1976. Selection for tibial dyschondroplasia in Australian meat chickens. Pages: 176–181 in: Proceedings First Australasian Poultry Stock Feed Convention. Vol. 2. Melbourne, Australia. Wong-Valle, J., G. R. McDaniel, D. L. Kuhlers, and J. E. Bartels, 1993a. Divergent genetic selection for incidence of tibial
dyschondroplasia in broilers at seven weeks of age. Poultry Sci. 72:421–428. Wong-Valle, J., G. R. McDaniel, D. L. Kuhlers, and J. E. Bartels, 1993b. Correlated responses to selection for high or low incidence of tibial dyschondroplasia in broilers. Poultry Sci. 72:1621–1629. Yalcin, S., X. Zhang, G. R. McDaniel, and D. L. Kuhlers, 1995. Effect of selection for high or low incidence of tibial dyschondroplasia for seven generations on live performance. Poultry Sci. 74:1411–1417. Zhang, X., G. R. McDaniel, Z. S. Yalcin, and D. L. Kuhlers, 1995. Genetic correlations of tibial dyschondroplasia incidence with carcass traits in broilers. Poultry Sci. 74:910–915.
Downloaded from http://ps.oxfordjournals.org/ at Serials Section, Dixson Library on June 25, 2015
(Key words: body weight, broiler, egg production, hatchability, tibial dyschondroplasia) 1998 Poultry Science 77:1065–1072
Early selection experiments of tibial dyschondroplasia (TD) in broiler chickens found that: 1) individual selection was effective to increase or decrease TD incidence of progeny (Leach and Nesheim, 1965, 1972; Sheridan, 1974; Riddell, 1976; Sheridan et al., 1976); 2) males and females differed in TD incidence (Leach and Nesheim, 1965; Sheridan, 1974; Sheridan et al., 1976); 3) the estimate of heritability (h2) for TD incidence changed from 0.2 to 0.3 (Sheridan et al., 1978; Burton et al., 1981); 4) the genetic correlation between BW and TD incidence was negative (Sheridan et al., 1976; Burton et al., 1981) or positive (Riddell, 1975; Poulos et al., 1978; Kiiskinen and Andersson, 1982). A short-term selection experiment was initiated to study the genetic susceptibility of broiler chickens to TD
and relationships of TD incidence with growth and reproduction traits. Divergent selection for TD resulted in two lines: high (HTD) and low (LTD) incidence of TD. A randombred control line was maintained to estimate the trend of environmental fluctuations between generations. Direct (Wong-Valle et al., 1993a) and correlated (Wong-Valle et al., 1993b) responses to selection have been reported for the first four generations. No difference in average BW was found between HTD and LTD lines, although TD incidence in these two lines differed by 20-fold. Egg production (EP) and semen concentration were greater in the LTD line than in the HTD line. The present study reports on 1) direct response of TD incidence to 10 generations of selection for increased and decreased TD; 2) correlated responses of BW at 4 and 7 wk of age, egg production, and hatchability; and 3)
Received for publication September 10, 1997. Accepted for publication March 31, 1998. 1To whom correspondence should be [email protected]
Abbreviation Key: CSDA = adjusted cumulative selection differential; CSDU = unadjusted cumulative selection differential; EP = egg production; H = high incidence; h2 = heritability; L = low incidence; TD = tibial dyschondroplasia.
INTRODUCTION
addressed:
1065
Downloaded from http://ps.oxfordjournals.org/ at Serials Section, Dixson Library on June 25, 2015
ABSTRACT Continued genetic selection for improved BW gain has met an obstacle of skeletal disorders in broiler chickens. Two broiler chicken lines (HTD and LTD) were developed by 10 generations of divergent selection for tibial dyschondroplasia (TD) incidence originating from commercial primary breeders. The reference population was a randombred control line maintained along with the selected lines. Relationships of TD incidence with BW, egg production (EP), and hatchability were assessed using these lines. The response of TD to selection was asymmetric, favoring an increased TD incidence. Mean TD incidence increased 7.6 percentage points per generation during Generation 1 through 10 in males and 9.1 percentage points in females of the HTD line but did not change significantly in the LTD line at 4 wk of age. Responses of the HTD line in early (1 to 4) generations were greater than in
1066
ZHANG ET AL.
comparison of early (Generation 1 to 4) vs later (6 to 10) selection response.
MATERIALS AND METHODS
Maintenance of Lines
Measurements and Analyses of Traits Incidence of TD was scored as either 0 for normal or 1 for abnormal by examining the left and right proximal tibiotarsus of birds at 4 (TD4) and 7 (TD7) wk of age using a lixiscope (Bartels et al., 1989). Body weight was also obtained at 4 (BW4) and 7 (BW7) wk of age. The TD4 was measured consecutively for 10 and BW4 for eight generations, whereas TD7 and BW7 were recorded for seven generations. Before selection of breeders in the lines, 60 hens were selected randomly within each line to test for EP. Daily egg production per hen was recorded from 24 to 36 wk of age in Generations 1 through 7 and 24 to 30 wk in Generations 8 through 10. Abnormal eggs (broken, poorshelled, shell-less, and miniature) were also recorded. Starting from 32 wk of age, selected breeder hens were inseminated artificially twice a week. Eggs were collected from the 3rd to 18th d once commencing insemination. Hatchability of eggs set was recorded from selected breeders in all generations.
Ingredients
Starter
Grower
Pullet developer
Corn Soybean (48% protein) Poultry oil Alfalfa meal (17% protein) Dibasic calcium phosphate Calcium carbonate Salt Vitamin premix1 Mineral premix2 DL-methionine Biocox Calculated analysis Protein Calcium Available phosphorus ME, kcal/kg
50.15 38.50 7.00 . . . 1.90 1.20 0.35 0.25 0.25 0.30 0.10
(%) 60.30 30.55 5.25 . . . 1.65 1.10 0.35 0.25 0.25 0.20 0.10
68.56 15.30 . . . 13.54 1.10 0.50 0.35 0.25 0.25 0.15 . . .
22.85 1.07 0.47 3,203
19.91 0.96 0.42 3,214
15.54 0.79 0.30 2,895
1Vitamin premix provided per kilogram of finished feed: vitamin A (retinyl palmitate), 7,350 IU; cholecalciferol, 2,200 IU; vitamin E (dl-atocopheryl acetate), 8 IU; riboflavin, 5.5 mg; d-pantothenic acid, 13.0 mg; niacin, 36 mg; choline, 500 mg; vitamin B12, 0.02 mg; menadione, 2 mg; folic acid, 0.5 mg; thiamine mononitrate, 1.0 mg; pyridoxine, 2.2 mg; dbiotin, 0.05 mg. 2Mineral premix provided in milligrams per kilogram of finished feed: Cu, 6.0; Fe, 54.8; I, 1.0; Mn, 65.3; Se, 0.3; Zn, 55.0.
Linear regression of responses on generations was used to evaluate selection response trend across generations. Regression coefficients were tested for significance from zero by t test. Realized h2 of TD4 and TD7 were estimated by: 1) regressing mean responses of the population on cumulative selection differentials unadjusted for the number of progeny and 2) regressing mean responses of offspring on values of sires and dams within generations to assess changes of genetic variation with the progress of selection. Incidence of TD in dam and sire families was regressed on Generation 1 to 4 and 6 to 10 to compare responses of families to selection of earlier vs later generations. The cumulative selection differential of TD incidence adjusted (CSDA) and unadjusted (CSDU) for the number of progeny were regressed separately on generations. The CSDA measures the joint effects of natural and artificial selection, whereas CSDU measures the effect of artificial selection (Falconer, 1981). The comparison between these two regression coefficients might reflect the influence of natural selection on TD incidence. The ANOVA for BW was performed by generation using the General Linear Models (GLM) procedure (SAS Institute, 1988) with the following model: Yijklm = u + Li + Sij + Dijk + Gl + (LG)il + eijklm [1]
where Yijklm = individual observation for a trait; u = overall mean for the trait; Li = fixed effect of the ith line; Sij = random effect of the jth sire within the ith line; Dijk = random effect of the kth dam mated to the jth sire within
Downloaded from http://ps.oxfordjournals.org/ at Serials Section, Dixson Library on June 25, 2015
Details of the early generations of selection in the HTD and LTD lines were given by Wong-Valle et al. (1993a). Briefly, this divergent selection experiment commenced from a base population consisting of 200 males and 200 females of primary broiler breeders. From the base population, 15 sires and 58 dams were divergently selected to develop the first generation of the HTD and LTD lines. Ten sires and 50 dams were randomly selected to reproduce the first generation of a randombred control line from the same base population. From Generation 2 to 10, 15 sires and 60 dams were selected from each of the three lines (about 400 birds each line) to produce the next generation of each line. In each generation of the HTD and LTD lines, sires and dams were selected based on half-sib family performance. In each generation of the control line, sires and dams were randomly selected. A selected sire was randomly mated to four dams within each line, avoiding matings of pairs with a common parent or grandparent. All chicks were wing-banded and vaccinated against Marek’s disease at hatch, and reared with lines intermingled in three floor pens with side-windowed ventilation. Chicks were given 24 h of light during the first 4 d, after which birds in all generations except Generation 5 (8 h) received 22 h of light up to 7 wk of age. Birds consumed feed and water ad libitum up to 7 wk of age and then feed was restricted according to Ross breeder’s recommendations. A starter broiler diet was fed from 0 to 3 wk, a grower from 3 to 7 wk, and a pullet developer from 7 to 24 wk of age (Table 1).
TABLE 1. Composition of the starter, grower, and pullet developer diets
1067
TIBIAL DYSCHONDROPLASIA, GROWTH, AND REPRODUCTION TABLE 2. Linear regression coefficients ± SE of means for tibial dyschondroplasia (TD) and BW on generations in the randombred control line Generations of selection Trait
Sex
TD, %
Female Male
BW, g
Female Male
Age
Overall1
(wk) 4 7 4 7 4 7 4 7
2.4 1.6 3.6 1.3 7.7 –43.8 8.6 –54.8
± ± ± ± ± ± ± ±
Earlier2 1.2 0.9 1.6 1.8 20.5 34.6 21.8 44.6
0.5 0.9 0.1 –0.9 –36.2 –105.0 –29.5 –132.8
± ± ± ± ± ± ± ±
Later3 1.4 2.3 2.3 4.5 65.9 46.3 64.7 65.1
–4.3 . . . –5.4 . . . 87.8 . . . 111.6 . . .
± 2.0 ± 2.7 ± 44.6 ± 38.6
1Overall:
1 to 10 for TD and 1 to 8 for BW at 4 wk of age; 1 to 7 for TD and BW at 7 wk of age. 1 to 4 for TD and BW. 3Later: 6 to 10 for TD; 6 to 8 for BW at 4 wk of age. 2Earlier:
Yijklm = u + Li + Sij + Dijk + eijklm
[2]
Components in the model were the same as in Equation 1 except for sex and line by sex interaction terms. Regressions of measurements for these traits on generations were used to evaluate correlated responses to selection.
RESULTS AND DISCUSSION
Response of Tibial Dyschondroplasia Incidence to Divergent Selection Fluctuation of Environments. The overall environmental fluctuation trend was not significant for incidence of TD at 4 and 7 wk of age, as estimated by linear regression of mean TD on generations in the control line (Table 2). A nonsignificant generation effect on TD incidence may demonstrate no directional environmental influence of this selection study. Considering that there was a decrease in TD incidence due to shortened photoperiod (8 h light) in Generation 5, segmented regressions of TD on generations showed environmental similarity of early (1 to 4) and later generations (6 to 10). Selected Lines. Mean TD incidence increased 7.6 percentage points per generation during Generation 1 through 10 in males and 9.1 percentage points in females of the HTD line but did not change significantly in the
LTD line at 4 wk of age, as expressed by deviations from corresponding values in the control line (Figure 1A). At 7 wk of age, TD incidence increased 10.8 percentage points per generation during Generations 1 through 7 in males and 12.7 percentage points in females of the HTD line and decreased 3.5 percentage points in males and 1.7 percentage points in females of the LTD line (Figure 1B). Response to selection in the HTD line was positive during Generation 1 through 4 but was negative or zero during Generation 6 through 10 at 4 wk of age (Table 3). No response was noted for both durations in the LTD line. At 7 wk of age, TD incidence had an increase about 9 (females) and 10 percentage points (males) per generation in full-sib families of the HTD line, in comparison to a decrease of 3 (females) and 8 (males) percentage points in the LTD line during Generations 1 to 4 (Table 3). Similar responses were observed in sire families (data not shown). Responses to divergent selection were asymmetric. Selection favored the HTD line and resulted in a ratio of mean TD incidence (HTD: LTD) greater than 5 starting from Generation 3 (data not shown). Males had a slightly higher incidence of TD than females. There was a higher incidence of TD at 7 than at 4 wk of age. The LTD line varied around 6% incidence of TD in males and 3% in females at 4 wk of age across generations. Later generations (after 5) had a poor stability of TD incidence, especially in the HTD line. Selection Differentials. Linear regression coefficients for cumulative selection differentials of TD incidence adjusted for the number of progeny on generation did not differ from zero (Table 4), indicating that selection differentials remained constant from Generations 0 through 10. However, cumulative selection differentials unadjusted for the number of progeny decreased more than 11 percentage points per generation in the HTD line and remained unchanged in the LTD line. The linear regression coefficients for adjusted selection differentials differed from those for unadjusted ones in the HTD line (Table 4). A comparison between adjusted and unadjusted selection differentials suggested that natural selection
Downloaded from http://ps.oxfordjournals.org/ at Serials Section, Dixson Library on June 25, 2015
the ith line; Gl = fixed effect of the lth sex; (LG)il = effect of line by sex interaction; and eijklm = random error. The mean squares of sires within lines were used as the error term to test line effects. The effect of sires within lines was tested against dams within sires and lines. Mean BW of birds with or without TD was regressed on generations within line and sex to estimate relationships between BW and TD. The ANOVA for EP and hatchability was performed by generation using the General Linear Models (GLM) procedure with model:
1068
ZHANG ET AL. TABLE 3. Linear regression coefficient ± SE of tibial dyschondroplasia (TD) incidence in full-sib families on generations Generations of selection Line1
Overall2
Sex
1 to 4
6 to 10
(%) 4 wk of age HTD Female Male LTD Female Male 7 wk of age HTD Female Male LTD Female Male
13.0 11.9 –0.1 –0.6
± ± ± ±
0.8** 0.9** 0.2 0.4
5.7 11.9 –0.4 –1.6
± ± ± ±
1.7** 2.1** 0.6 1.1
–2.4 0.5 –0.9 –1.1
13.2 11.1 –0.8 –2.9
± ± ± ±
0.8** 0.8** 0.3** 0.5**
8.9 10.4 –2.6 –7.8
± ± ± ±
2.0** 2.3** 0.7** 1.5**
. . . .
. . . .
± ± ± ±
1.1* 0.9 0.5 0.8
. . . .
FIGURE 1. Changes in tibial dyschondroplasia (TD) incidence at 4 (TD4, A) and 7 wk of age (TD7, B); HTD = line with high incidence of TD; LTD = line with low incidence of TD (LTD). Values are represented as deviations from the randombred control line.
exert some sort of influence upon artificial selection in the HTD line and little impact in the LTD line. This implies that the HTD line, once selection was relaxed, would regress toward the level of the base population from which selection had been initiated, whereas the LTD line might move slowly to the original level of the base population. Realized Heritability. Realized h2 for TD4, estimated by regressing mean responses of the population on CSDU were greater than 0.5 in the HTD line and less than 0.5 in the LTD line (Table 5). Corresponding estimates for TD7 were uncomfortably larger than 0.8 in the HTD line and close to 0.1 in the LTD line (Table 5). The estimate for males was higher than that for females (Table 5). Realized h2 for the HTD line were greater in this report than those in previous studies (Sheridan et al., 1978; Burton et al., 1981; Wong-Valle et al., 1993a; Zhang et al., 1995; and Kuhlers and McDaniel, 1996). Because CSD were decreased (Table 4) and mean responses increased
with advancing generations in the HTD line, current estimates based on 10 generations of selection are assumably greater than based on the first few generations of selection in previous papers. Divergent selection produced different estimates of realized h2 between the HTD and LTD lines. They were the consequence of asymmetric responses. The incidence of TD, as a component of characters associated with natural fitness, responded to selection much more slowly towards increased fitness than towards decreased fitness. Divergent selection operated in favor of the HTD line, whereas the LTD line fluctuated around 5%
TABLE 4. Linear regression selection differentials for incidence adjusted and of progeny
coefficient ± SE of cumulative tibial dyschondroplasia (TD) unadjusted for the number on generations Cumulative selection differential
Line1
Sex
Adjusted
Unadjusted (%)
4 wk of age2 HTD LTD 7 wk of age HTD LTD
Female Male Female Male
–0.5 –0.6 0.0 –0.1
± ± ± ±
0.4 0.3 0.2 0.1
–13.2 –12.0 0.1 0.7
± ± ± ±
4.3* 3.2* 0.3 0.9
Female Male Female Male
–0.6 0.2 –0.1 0.2
± ± ± ±
0.3 0.5 0.2 0.2
–13.6 –11.3 0.8 3.0
± ± ± ±
1.9** 2.2** 0.7 1.7
1HTD = line with high incidence of TD; LTD = line with low incidence of TD. 2Generations 0 to 10 at 4 wk of age and 0 to 7 at 7 wk of age. *P < 0.05. **P < 0.01.
Downloaded from http://ps.oxfordjournals.org/ at Serials Section, Dixson Library on June 25, 2015
1HTD = line with high incidence of TD; LTD = line with low incidence of TD. 2Generations 1 to 10 at 4 wk of age and 1 to 7 at 7 wk of age. *P < 0.05. **P < 0.01.
TIBIAL DYSCHONDROPLASIA, GROWTH, AND REPRODUCTION TABLE 5. Realized heritabilities (h2) of tibial dyschondroplasia (TD) incidence estimated by regressing mean responses of the population on cumulative selection differentials unadjusted for the number of progeny h2 for TD at Line1
Sex
Week
HTD
Female Male Female Male
0.65 0.52 0.50 0.40
LTD
± ± ± ±
42
Week 7
0.53 0.47 0.32 0.22
1.06 0.86 0.10 0.05
± ± ± ±
0.25 0.33 0.14 0.13
1HTD = line with high incidence of TD; LTD = line with low incidence of TD. 2Generations 0 to 10 at 4 wk of age and 0 to 7 at 7 wk of age.
comparison of positive regression coefficients of birds afflicted with TD at 4 wk of age and negative values at 7 wk of age (Table 6) and 2) the fact that BW4 in the HTD line birds was slightly heavier than or similar to that in the LTD line within generation (data not shown). Afflicted birds at early ages were reluctant to move and reduced ability of competing for feed and water, causing a decrease of BW7 in the HTD line. Therefore, the relationship of TD incidence with BW appeared to be age-related. The incidence of TD exhibited negative genetic correlations with BW estimated from sire’s and dam’s components of variance and covariance of the seventh generation of divergently selected lines (Zhang et al., 1995). Sheridan et al. (1976) and Burton et al. (1981) found a negative correlation between BW7 and TD7, whereas a positive association between BW and TD was observed prior to 7
Correlated Responses to Divergent Selection Body Weight. The general environmental fluctuation trend was not clear for BW due to a broad range of variation, as shown in the control line (Table 2). The HTD line birds tended to decrease BW7 during Generations 1 through 7 (Figure 2) but this was uncertain due to great fluctuation for BW4 (Table 6). Although mean BW of the LTD birds demonstrated a positive change at 4 wk of age and negative at 7 wk of age with advancing generations of selection (Table 6), the trend was not significantly evident as in the HTD line. The TD incidence appeared, to some extent, to be associated with an enhanced BW prior to 4 wk of age, but not strongly. This association was determined by: 1) a
FIGURE 2. Changes in 7-wk body weight (BW7) of birds with advancing generations of divergent selection for tibial dyschondroplasia (TD); HTD = line with high incidence of TD; A) with TD and B) with no TD.
Downloaded from http://ps.oxfordjournals.org/ at Serials Section, Dixson Library on June 25, 2015
incidence of TD at 4 wk of age across generations. A main cause for asymmetrical responses could be that TD incidence in the base population was not midway between the two theoretical limits of the HTD and LTD lines (100 and 0%, respectively). Therefore, responses to divergent selection had further to go in the HTD line than in the LTD. This genetic asymmetry was caused by disequilibrium gene frequency of the base population. Sheridan (1974) presumed a major dominant sex-linked gene controlling TD. Later, a significant relationship between the sex-linked gene for rapid feathering and the incidence of TD in the first three generations of selection was observed by Sheridan et al. (1976). Additive genetic effect, rather than dominant or maternal, could be important for TD, as demonstrated by a study of diallel crosses between HTD and LTD lines developed by divergent selection of seven generations (Yalcin et al., 1995). They found that progeny from the cross of the HTD × HTD had a higher incidence of TD at 4 and 7 wk of age than from either of the reciprocal crosses, which, in turn, were higher than from the cross of the LTD × LTD. The reciprocal crosses were similar in TD incidence at both ages. A poor stability of TD incidence in higher generations, especially in the HTD line, was the consequence of the loss or decrease of additive genetic variability accompanying selection.
1069
1070
ZHANG ET AL. TABLE 6. Linear regression coefficient ± SE of 4- and 7-wk mean BW on generations1 Mean BW at
Line2
Sex
TD3
Week 4
HTD
Female
HTD
Male
LTD
Female
LTD
Male
0 1 0 1 0 1 0 1
–3.7 3.8 –15.6 0.8 11.1 3.5 7.4 12.2
Week 7 (g)
1Generations
± ± ± ± ± ± ± ±
17.7 17.4 21.3 18.3 20.4 15.6 21.8 23.5
–61.9 –47.2 –87.7 –73.4 –43.0 –49.8 –59.1 –9.7
± ± ± ± ± ± ± ±
28.1† 25.4 36.7 34.2† 40.8 45.7 48.8 61.5
1 to 8 at 4 wk of age and 1 to 7 at 7 wk of age.
2HTD = line with high incidence of tibial dyschondroplasia (TD); LTD
wk of age (Riddell, 1975; Poulos et al., 1978; Kiiskinen and Andersson, 1982). An age-related correlation between TD and BW was reported also in turkeys (Rath et al., 1994). There was no correlation between BW and TD of turkeys at early ages, whereas turkeys with severe TD had heavier BW relative to those without or with mild lesions during Weeks 14 and 15. Variability of BW7 had an increasing trend with advancing generations of selection in the HTD line (Figure 3A,B) and also of BW7 of males with no TD in the LTD line (Figure 3C). The increased variability of BW7 in the HTD line was accompanied by a decreasing additive genetic variability of TD due to continued selection. Does this presuppose that pleiotropic genes might be excluded to interpret the relationship between BW and TD? Of course, the increase of variability for BW7 could be the consequence of no selection for BW or environmental fluctuation. Egg Production. Egg production of the control line was influenced little from Generation 1 to 10. Egg production increased from Generation 1 to 5, in both the HTD and LTD lines. A reverse trend occurred from Generation 7 to 10 (Figure 4). The LTD line hens had a greater (P < 0.05) EP than the HTD hens within every generation (Figure 4). The average EP in the LTD line hens was 7.6 percentage points higher than in the HTD line from Generation 1 to 10, ranging from 4 to 14 percentage points. The increase (P < 0.05) of EP in both HTD and LTD lines from Generation 1 to 5 was because of an environmental influence instead of correlated responses to selection. This result can be seen from the fact that the randombred control line had a similar increase (P < 0.05) in EP from Generation 1 to 5. However, the decrease of EP from Generation 6 to 10 in both HTD (P < 0.01) and LTD (P = 0.06) lines could be due to correlated responses to selection, because the control line (P = 0.15) did not show the same trend in EP during that period.
FIGURE 3. Changes in CV for 7-wk body weight (BW7) with advancing generations of divergent selection for tibial dyschondroplasia (TD): A) birds with TD and B) with no TD in the line of high incidence of TD (HTD); and C) with no TD in the line of low incidence of TD. The CV of 7-wk BW for HTD females with TD, LTD males and females with TD, and LTD females without TD was not different from generation to generation (P > 0.1) and not showed.
Paired differences in TD incidence between the LTD and HTD lines became greater with advancing generations of selection, to which paired differences in mean EP between lines were not proportional. A plot of paired differences in mean EP between the LTD and HTD lines against corresponding differences in TD incidence was nearly flat across generations (data not shown). This effect weakened the conclusion that TD incidence decreases EP in broiler hens. The flat plot occurred possibly because selection of hens for testing for EP was not completely random. Hens with TD in the HTD line and those with no TD in the LTD line were more likely to be selected as testers.
Downloaded from http://ps.oxfordjournals.org/ at Serials Section, Dixson Library on June 25, 2015
= line with low incidence of TD. 30 = no TD, 1 = TD. †P < 0.1.
TIBIAL DYSCHONDROPLASIA, GROWTH, AND REPRODUCTION
1071
REFERENCES FIGURE 4. Changes in mean egg production (EP, per hen-day) of 60 hens in selected and randombred control lines with advancing generations of divergent selection for tibial dyschondroplasia (TD); HTD = line with high incidence of TD; LTD = line with low incidence of TD. Values were recorded for a period of 12 wk (24 to 36 wk old) in Generations 1 through 7 and only for 6 wk (24 to 30 wk old) in Generations 8 through 10.
FIGURE 5. Changes in hatchability of selected and randombred control lines with advancing generations (6 to 10) of divergent selection for tibial dyschondroplasia (TD); HTD = line with high incidence of TD; LTD = line with low incidence of TD. The data of the first 5 generations was not accessible due to the leave of a person involved in this study during that period.
Bartels, J. E., G. R. McDaniel, and F. J. Hoerr, 1989. Radiographic diagnosis of tibial dyschondroplasia in broilers: a field selection technique. Avian Dis. 33:254–257. Burton, R. W., A. K. Sheridan, and C. R. Howlett, 1981. The incidence and importance of tibial dyschondroplasia to the commercial broiler industry in Australia. Br. Poult. Sci. 22: 153–160. Falconer, D. S., 1981. Introduction to Quantitative Genetics. 2nd ed. Longman, New York, NY. Kiiskinen, T., and P. Andersson, 1982. The incidence of tibial dyschondroplasia in two broiler strains and their performance on different diets. Ann. Agric. Fenn. 21:169–176. Kuhlers, D. L., and G. R. McDaniel, 1996. Estimates of heritabilities and genetic correlations between tibial dyschondroplasia expression and body weight at two ages in broilers. Poultry Sci. 75:959–961. Leach, R. M., Jr., and M. C. Nesheim, 1965. Nutritional, genetic and morphological studies of an abnormal cartilage formation in young chicks. J. Nutr. 86:236–244. Leach, R. M., Jr., and M. C. Nesheim, 1972. Further studies on tibial dyschondroplasia (cartilage abnormality) in young chicks. J. Nutr. 102:1673–1680. Poulos, P. W., S. Reiland, K. Elwinger, and S. E. Olsson, 1978. Skeletal lesions in the broiler, with special reference to dyschondroplasia (osteochondrosis). Pathology, frequency and clinical significance in two strains of birds on high and low energy feed. Acta-Radiologica. Suppl. 358:229–275. Rath, N. C., G. R. Bayyari, J. N. Beasley, W. E. Huff, and J. M. Balog, 1994. Age-related changes in the incidence of tibial dyschondroplasia in turkeys. Poultry Sci. 73:1254–1259. Riddell, C., 1975. Studies on the pathogenesis of tibial dyschondroplasia in chickens. III. Effect of body weight. Avian Dis. 19:497–505. Riddell, C., 1976. Selection of broiler chickens for high and low incidence of tibial dyschondroplasia with observations on spondylolisthesis and twisted legs (Perosis). Poultry Sci. 55: 145–151. SAS Institute, 1988. SAS/STAT User’s Guide. 6th ed. SAS Institute Inc., Cary, NC.
Downloaded from http://ps.oxfordjournals.org/ at Serials Section, Dixson Library on June 25, 2015
Hatchability. Mean hatchability of the control line was not affected from Generation 6 through 10 (Table 2) and data for Generation 1 through 5 was missing due to lost communication with a person involved in this project during that period of time. The trend of changes in hatchability with generations was not clear in the selected lines. The LTD line did not differ (P > 0.1) from the HTD line within every generation (Figure 5). Similarly, the plot of paired differences in mean hatchability between the LTD and HTD lines against corresponding differences in TD incidence was flat across generations (data not shown). The flat plot demonstrated that the association of hatchability with TD incidence was not as tight as expected. The lack of close relationships between hatchability and TD incidence from Generation 6 to 10 did not contradict the report of no differences in sexual maturity and semen volume and concentration between the HTD and LTD lines in early generations (Wong-Valle et al., 1993b).
1072
ZHANG ET AL.
Sheridan, A. K., 1974. Genetic studies of tibial dyschondroplasia in the broiler chicken. Page: 86 in: Proceedings Australian Poultry Science Convention, World’s Poultry Science Association, Hobart, Tasmania, Australia. Sheridan, A. K., C. R. Howlett, and R. W. Burton, 1978. The inheritance of tibial dyschondroplasia in broilers. Br. Poult. Sci. 19:491–499. Sheridan, A. K., C. R. Howlett, and E. B. Dettmann, 1976. Selection for tibial dyschondroplasia in Australian meat chickens. Pages: 176–181 in: Proceedings First Australasian Poultry Stock Feed Convention. Vol. 2. Melbourne, Australia. Wong-Valle, J., G. R. McDaniel, D. L. Kuhlers, and J. E. Bartels, 1993a. Divergent genetic selection for incidence of tibial
dyschondroplasia in broilers at seven weeks of age. Poultry Sci. 72:421–428. Wong-Valle, J., G. R. McDaniel, D. L. Kuhlers, and J. E. Bartels, 1993b. Correlated responses to selection for high or low incidence of tibial dyschondroplasia in broilers. Poultry Sci. 72:1621–1629. Yalcin, S., X. Zhang, G. R. McDaniel, and D. L. Kuhlers, 1995. Effect of selection for high or low incidence of tibial dyschondroplasia for seven generations on live performance. Poultry Sci. 74:1411–1417. Zhang, X., G. R. McDaniel, Z. S. Yalcin, and D. L. Kuhlers, 1995. Genetic correlations of tibial dyschondroplasia incidence with carcass traits in broilers. Poultry Sci. 74:910–915.
Downloaded from http://ps.oxfordjournals.org/ at Serials Section, Dixson Library on June 25, 2015