Estimates of Genetic Parameters in Turkeys.

Estimates of Genetic Parameters in Turkeys. 1. Body Weight and Skeletal Characteristics 1 G. B. HAVENSTEIN, K. E. NESTOR, V. D. TOELLE, and W. L. BACON

(Received for publication January 19, 1988) ABSTRACT Heritabilities (h2) of and genetic (rG) correlations among body weight and a number of skeletal characteristics were estimated from data on 1,088 pedigree turkeys (504 females and 584 males) of a randombred control line. All measurements were made at 16 wk of age. The h 2 estimates (sire component) obtained from females (F) and males (M), respectively, were: BW .23, .60; shank width (SW) .54, .47; shank length (SL) .43, .54; drum length .66, .60; rough-cleaned weights of the thigh (nonestimatable, .60), drum .09, .57, and shank .28, .69; cleaned weights of the drum ,37, .71, and shank .30, .40; fat-free weights of the drum .44, .93, and shank, .45, .82; bone density measured at 40 and 60% of the length from the proximal end of the drum .68, .80 and .34, .92, respectively, and shank .28, .69 and .31, .55. Genetic correlations (rG) among the various bone weights were all above .66, with most above .85. The rG between BW and SW were .33 from M and .47 from F. These correlations suggest there is a relatiyejy weak relationship, indicating that selection for BW alone might not cause a large enough increase in SW to support the increase brought about in BW. Shank width also had relatively low correlations with bone weight measurements, ranging from .27 to .53 from M and .17 to .56 from F. The h2 of walking ability score (WA) wasj^06 and the rcjrfWA with BW was -.73, which indicated thatjhe low body weight families tended_to have- pnnr WA (i.e., higher scores). This may be a spurious rGj_tec^use_poorwalking ability scores tended to be grouped in families. Birds with poor~walking~abllity would have difficulty eating, and^ould^therefore, tend to have low-BW, The rr, between SW:WA (-.09) and"SL?WA~(703) inaTcajeainHI£riJaHonsi^^ The WA and the bone density readings had high negative rn(range -.75 to -LjiT), indicating^ that femilies with poor WA ratings also had low bone density scores. (Key words: turkeys, heritability, genetic correlation, body weight, skeletal measurements) 1988 Poultry Science 67:1378-1387 INTRODUCTION

Breeders of broiler chickens and turkeys have for many years placed their primary selection emphasis on body weight at market age and subjective body conformation scores for improving breast muscle width. A secondary emphasis has been placed on livability, feed efficiency, and freedom from leg problems and other defects. The assumption has been made that selection for growth rate would result in proportional increases in all body parts. It appears, however, that selection for body weight and breast conformation has resulted in greater increases in breast than in leg muscle and skeletal mass. Furthermore, such disproportionate change in body parts appears to have contri-

' Salaries and research support provided by state and federal funds appropriated to the Ohio Agricultural Research and Development Center, The Ohio State University. Journal Article Number 2-88.

buted to leg and skeletal problems in fast-growing poultry meat lines. A genetic basis for skeletal problems in meattype poultry has been clearly documented by studies showing that: 1) leg problems in broilers are moderately to highly heritable (Sheridan et al., 1974, 1978; Burton et al., 1981; Mercer and Hill, 1984); 2) different broiler and turkey strains differ considerably in their levels of leg problems (Haye and Simons, 1978; Veltman and Jensen, 1981; Nestor, 1984; Nestor et al., 1985, 1987); 3) broilers respond to selection for a decreased incidence of leg abnormalities (Serfontein and Payne, 1934; Leach and Nesheim, 1965, 1972; Riddell, 1976); 4) selection for increased 16-wk body weight increased the number of leg problems in turkeys (Nestor, 1984); and 5) selection for increased shank width reduced the number of leg abnormalities and improved walking ability in turkeys (Nestor et al., 1985, 1987). Nestor et al. (1985) hypothesized that direct selection for increased

1378

Downloaded from http://ps.oxfordjournals.org/ at Ernst Mayr Library of the Museum Comp Zoology, Harvard University on July 1, 2015

Department of Poultry Science, The Ohio State University, Ohio Agricultural Research and Development Center, Wooster, Ohio 44691

GENETIC PARAMETERS OF SKELETAL TRAITS IN TURKEYS

MATERIALS AND METHODS

Data were collected at 16 wk of age on 1,088 fully pedigreed individuals (504 females and 584 males) of a randombred line (RBC2) of turkeys maintained at the Ohio Agricultural Research and Development Center, Wooster, OH. The RBC2 line was formed in 1966 from reciprocal crosses of two commercial strains (Nestor et al., 1969), and has been maintained with a paired mating system since that time using 36 randomly chosen parental pairs (Nestor, 1977). Until 1984, 72 individuals, 36 males and 36 females, were chosen at random from the previous year's paired matings (one male and one female from each to be mated at random with the exception of the avoidance of brother by sister matings) for reproduction of the line. In the nineteenth generation, 36 males and 180 females were chosen at random (five females from each of the previous generation's matings), with the intention of mating each sire with five nonsister females. This was done to produce the family structure needed for a genetic analysis of body weight, walking ability (WA), and body composition measurements to be collected on their offspring. Two of the selected males and 12 selected females failed to produce an adequate sample of progeny, so 34 males and 168 females actually produced the progeny used for this study. Two hatches placed 2 wk apart were used to produce birds used in this study. Offspring were hatched and pedigree wingbanded; both sexes were grown intermingled in an enclosed shed to 16 wk of age. Continuous lighting was provided for the 1st 4 wk of age, and 12 h light/day

were provided thereafter. A four-diet feeding system with declining levels of protein was provided for both sexes (Naber and Touchburn, 1970), based on the schedule provided for males. During the 16th wk of age, all birds were sacrificed over a 5-day period. The day of kill was recorded for each bird. During the killing period, approximately 250 to 300 birds were chosen at random each day to be fasted overnight (a procedure which, according to Salmon, 1979, does not affect eviscerated carcass yield). Prior to the fast, all individuals were weighed and measured for shank width at the point of the dew claw (SW). Males were rated by one individual for their WA at this time. Because of the very low incidence of leg problems in females (F), WA was not estimated for them in this study. The WA of each male (described previously by Nestor, 1984) was rated from 1 to 5, with 1 indicating no apparent leg defects or walking problems, 5 indicating a male whose legs showed extreme lateral deviation or who had great difficulty in walking. Ratings of 2, 3, or 4 indicated intermediate and increased problems with walking. Following the overnight fast, birds were killed by severing the jugular vein. Birds were then scalded at 68 C for about 30 s, plucked by a mechanical picker, and then eviscerated. Abdominal fat (AF), i.e., the leaf fat (LF) plus the fat from the gizzard, was removed and weighed. The LF was also weighed separately. Carcasses were then sawed in half longitudinally with a band saw, and placed in cold water for 3 h for chilling, before one of the halves was bagged and frozen for later dissection. The other half of the carcass, along with onehalf of the leaf fat, was placed in a Hobart VCM 30 vertical cutter (Hobart Corporation, Troy, OH) and rough ground. The rough-ground carcass was regrouped twice using a Hobart A-200 mixer-grinder using a 2.75-mm mesh plate. Three subsamples were then taken (approximately 40 g each) from each carcass for freezedrying to obtain percentage moisture. Percentage of moisture was then recorded as the average percentage loss from freeze-drying for the three samples. The three freeze-dried subsamples were then blended into one sample by mixing in a Waring blender. Percentage of ash was determined on duplicate 1 to 2-g samples by ashing them at 600 C for 4 h, and was recorded as the average of the two samples. Percentage of fat was measured on duplicate 3-g samples by the

Downloaded from http://ps.oxfordjournals.org/ at Ernst Mayr Library of the Museum Comp Zoology, Harvard University on July 1, 2015

amount of breast muscles as well as for greater total body weight had caused total body weight and breast muscles to increase at a faster rate than the muscles and bones of the legs, and that this disproportionate change had caused an inherent weakness in the bird, which results in leg problems. The objective of this study was to examine the inheritance of and the genetic relationships among a number of muscle, fat, and skeletal measurements, as well as the genetic and phenotypic relationships of these factors with 16-wk body weight in turkeys. The genetic and phenotypic relationships of these traits were also examined in relation to a visual rating score for walking ability of the males taken at 16 wk of age.

1379

1380

HAVENSTEIN ET AL.

Yijklm = |x + Si + d/Sy + Hk + SX, + b(XijkIm - x) + e;jklni where: |x = the overall mean for trait Y; Sj = the random effect of the ith sire; d/Sy = the random effect of the j t h dam mated to the i,h sire; Hk = the fixed effect of the kth hatch; SX,

= the fixed effect of the 1 sex; b = the regression of Yijklm on day of kill; Xjjklm = day of kill for the individual; x = average day of kill; and ejjklm = the random error associated with the measurement of each individual which is assumed to be randomly and independently distributed, with x = 0 and a variance of cr2. The data were also analyzed separately by sex using the same analysis as shown above, except that sex was deleted from the model. Sex was deleted for two reasons: namely, that WA was only measured in the males, and because turkeys have an extreme degree of sexual dimorphism for skeletal and other body measures, which results in large differences between the means and variances for the two sexes. Genetic and phenotypic parameters are often reported separately by sex. Phenotypic variances from the within-sex analyses were tested for heterogeneity by the method of Bartlett (1937) as shown by Neter and Wasserman (1974), and were found to be heterogeneous for most of the traits studied. Therefore, heritabilities (h2) and genetic (rG) and phenotypic (rP) correlations were estimated within sex. Standard errors of h2 and rG estimates were estimated using modifications of the methods of Tallis (1959) and Swiger et al. (1964), as described by Harvey (1986). In spite of the heterogeneity of variances between the two sexes, because ANOVA is such a robust technique, the data from the two sexes were combined and analyzed together with sex in the model to compare the estimates obtained with those obtained from the within-sex analyses. RESULTS AND DISCUSSION

The results reported herein cover only part of the data collected in the overall study described in the Materials and Methods section. Body weight and the skeletal characteristics are reported in this paper, and body weight and its relationship to the muscle and fat characteristics are reported in a companion paper (Havenstein etal., 1988). Means and Phenotypic Standard Deviations. Overall x and phenotypic SD from both withinsex and combined-data set analyses for BW and skeletal characteristics are presented in Table 1. Males had significantly larger mean values and significantly larger variances for all of the traits reported herein. Heritability Estimates. The h2 for BW and for all of the skeletal characteristics measured,

Downloaded from http://ps.oxfordjournals.org/ at Ernst Mayr Library of the Museum Comp Zoology, Harvard University on July 1, 2015

method of Folch et al. (1957), and was recorded as the average of the two samples. Nitrogen was determined on duplicate 1-g samples by the Kjeldahl procedure. The percentage of crude protein was estimated by percentage N X 6.25, and recorded as the average of the two samples. As time permitted, half-carcasses were thawed, and legs were removed from the carcass and separated into drumsticks and thighs by a cut through the femuro-tibial and patellar joint. Drum and thigh muscles (DM and TM) were carefully dissected from the bone and weighed. Breast muscles (BM) were removed from the breast bone, and were also weighed. Thigh (femur), drum (tibiotarsus), and shank (tarsometatarsus) bones were weighed immediately after separation from the muscles; the drum and shank were then refrozen. No cleaning was done on the shank bone before freezing. The thigh bone was discarded. The bone weights will be referred to as the rough-cleaned weights of the thigh (RCT), drum (RCD), and shank (RCS). Again, as time permitted, drum and shank bones were thawed and placed into boiling water for 15 min, and any adhering tissues were removed. The cleaned drum (CD) and shank (CS) were then reweighed. Drum and shank bones were fat extracted using a methanolchloroform mixture (1:2, vol/vol), and were then dried at 100 C for 24 h. The fat-free drum (FFD) and shank (FFS) bones were then reweighed, and shank length (SL) and drum length (DL) were recorded. Density of the fat-extracted drum and shank bones was then measured at 40 and 60% of the length from the proximal end of each bone using photon absorption with an I 125 source (Cantor et al., 1980). Density measurements will be referred to as D40, D60, S40, and S60 for the two drum and shank measurements, respectively. Data were analyzed and variance and covariance components were estimated using the least squares and maximum likelihood procedures outlined by Harvey (1986). The data set containing both sexes was analyzed using the following model:

GENETIC PARAMETERS OF SKELETAL TRAITS IN TURKEYS

1381

TABLE 1. Overall means and phenotypic standard deviations (op) for traits studied as calculated 'within sex and with sexes combined Females

Males Trait

1

X

7.34

P .56

Sexes combined <7p

X

5.25

.42

Op

X

6.37

.50

SW, mm SL, cm DL, cm

13.3 15.6 20.0

.5 .5 .6

11.8 12.8 17.8

.5 .5 .6

12.6 14.3 19.0

.5 .5 .6

RCT.g RCD,g RCS, g

63.3 100.3 78.0

5.6 9.1 6.2

36.8 58.8 46.2

4.6 6.4 5.2

51.0 81.1 63.3

5.1 8.0 5.8

CD,g CS, g

53.8 29.3

4.7 2.7

32.4 16.1

3.3 1.7

43.9 23.2

4.1 2.3

FFD, g FFS, g D40 D60 S40 S60

26.4 15.1

2.5 1.2

17.1 9.1

1.6 .8

22.1 12.3

2.1 1.1

51.3 56.3 37.1 40.1

5.4 5.9 3.3 3.1

39.1 42.0 28.6 32.4

3.5 3.8 2.5 2.5

45.6 49.6 33.2 36.6

4.6 5.0 3.0 2.9

WA

1.67

.66

1 SW = Shank width; SL = shank length; DL = drum length; RCT, RCD, and RCS = rough-cleaned weights of the thigh, drum and shank, respectively; CD and CS = cleaned weights of the drum and shank, respectively; FFD and FFS = fat-free weights of the drum and shank, respectively; D40, D60, S40, and S60 = density measurements (Cantor et al., 1980) of the drum and shank taken at 40 and 60% of the length from the proximal end, respectively ; WA = walking ability score (1 = normal through 5 = extreme difficulty in walking).

as estimated from the sire (cr2) and dam (o^) components of variance are summarized in Table 2 for both the within-sex and the pooled analyses. Estimates from the pooled analysis are, as expected, generally intermediate in value to the corresponding values obtained from the within-sex analyses, but because of the larger amount of data involved in the combined-data set, the estimates have smaller standard errors. As reported previously by Krueger et al, (1972), Arthur and Abplanalp (1975), Nestor (1984), and Delabrosse (1986), h2 for most turkey traits studied have generally been found to be high. The only exception to this finding in these data was the h2 for WA (.06 ± .08). This result is not surprising, as WA is a subjective score, and the incidence of leg problems in the RBC2 line is low in relation to the incidence in modern growth-selected lines of turkeys. Using a binomial scoring system for leg problems in broiler chickens, Sheridan et al. (1978) and Mercer and Hill (1984) reported somewhat higher h2 estimates for leg disorders than was found from this population of male turkeys. The paternal half-sib h 2 estimates from males (M) were generally higher than comparable es-

timates from F (Table 2). This finding is in agreement with estimates reported by Abplanalp and Kosin (1952) and by Krueger et al. (1972) from data on broad-breasted bronze turkeys. However, the maternal half-sib h 2 estimates from F were generally higher than those from M. This is not what one would expect if sex-linkage were an important contributor to h 2 estimates. The expectation is that a 2 from the F data (heterogametic sex) would contain one-half of the sex-linkage variance, whereas a 2 from the M data (homogametic sex) would contain only one-fourth of the sex-linkage variance (Bohidar, 1964; James, 1973). If one assumes a completely additive model, then the sex linkage variance of the homogametic sex is twice the sex-linkage variance of the heterogametic sex. Thus, the expectation is that the sex-linkage variance contained in the sire component estimated from the two sexes would be equal in absolute value. However, if one assumed that the sex-linkage variance is the same in the two sexes, as suggested by Thomas et al. (1958), then the h2 estimate from the F data should be larger than the h 2 from the M data. What was actually observed fits neither of

Downloaded from http://ps.oxfordjournals.org/ at Ernst Mayr Library of the Museum Comp Zoology, Harvard University on July 1, 2015

BW, kg

a

HAVENSTEIN ET AL.

1382

TABLE 2. Heritabilities (h2) and their SE as estimated from sire (oj) and dam (o%) components of variance when measured within sex and with sexes combined Females

Males

oi )

Sexes combined

("V

(o%)

(oD

(oh)

(oh)

Trait 1

h2

SE

h2

SE

h2

SE

h2

SE

h2

SE

h2

SE

.60

.20

.63

.17

.23

.14

1.22

.20

.45

.13

.82

.12

.47 .54 .60

.17 .19 .20

.48 .58 .57

.17 .17 .17

.54 .43 .66

.20 .18 .22

.68 .05 .09

.20 .18 .18

.46 .51 .62

.14 .14 .16

.33 .27 .34

.11 .10 .10

RCT RCD RCS

.60 .57 .69

.20 .19 .21

.59 .21 .54

.17 .16 .17

NE2 .09 .28

.10 .15

1.67 .37 NE

.19 .19

.36 .34 .40

.11 .11 .12

.60 .25 .38

.11 .09 .10

CD CS

.76 .45

.22 .17

.48 .23

.17 .16

.37 .30

.17 .15

.38 .20

.19 .19

.57 .37

.16 .12

.36 .25

.10 .09

FFD FFS

.93 .82

.25 .23

.43 .46

.17 .17

.44 .45

.18 .18

.60 .43

.20 .19

.66 .61

.17 .16

.48 .41

.11 .10

S40 S60 D40 D60

.69 .55 .80 .92

.21 .19 .23 .25

.39 .24 .58 .46

.17 .16 .17 .17

.28 .31 .68 .34

.15 .15 .22 .16

.78 .76 .49 .79

.20 .20 .19 .20

.46 .38 .65 .62

.14 .12 .17 .16

.37 .35 .50 .42

.10 .10 .11 .10

WA

.06

.08

.23

.16

1 SW = Shank width; SL = shank length; DL = drum length; RCT, RCD, and RCS = rough-cleaned weights of the thigh, drum and shank, respectively; CD and CS = cleaned weights of the drum and shank, respectively; FFD and FFS = fat-free weights of the drum and shank, respectively: D40, D60, S40, and S60 = density measurements (Cantor et al., 1980) of the drum and shank taken at 40 and 60% of the length from the proximal end, respectively; WA = walking ability score (1 = normal through 5 = extreme difficulty in walking). 2

NE = Nonestimatable, due to negative variance component.

these assumptions, and the difference in h2 values can probably be interpreted as being due to the sex-influenced expression of genotype. Studies by Mullen and Swatland (1979) and Walser et al. (1982) support the hypothesis that physiological differences exist between the two sexes in skeletal growth. Heritability estimates from maternal half-sibs (MHS) were approximately equal to those from paternal half-sibs (PHS) for BW, for roughcleaned bone weights, and for SW, SL, and DL from M (Table 2). The MHS h2 estimates from M were consistently lower than PHS estimates from M for cleaned and fat-free bone weights and for bone density measurements. No such pattern is present for MHS vs. PHS h 2 estimates from F. When the sexes were combined into a single analysis with sex as a fixed effect, PHS and MHS h2 estimates were nearly equal for most of the traits, and were intermediate in value to those from the within-sex analyses. The only exceptions to this were the combined MHS h 2 estimates for BW and RCT, which were higher

than PHS h2, presumably due to the high estimate for a2; from the F data. The h2 estimate for WA was low, suggesting that it would be very difficult to improve WA through selection using such a subjective score. The h2 estimates from fat-free bone weights were consistently higher than those from roughcleaned or cleaned bone weights (Table 2). However, due to the relatively high h2 for the rough-cleaned bone weights, it is doubtful that the improvement in selection accuracy that one might achieve by using fat-free bone weights over rough-cleaned weights would be worth the cost and effort required to obtain the data. Genetic and Phenotypic Correlation Estimates. The rG and rP estimates from the sire components of variance and covariance from the F data, the M data, and from the combineddata set are summarized in Tables 3,4, and 5, respectively. The rG estimates from the combined analysis (Table 5) are in good agreement with those from the within-sex analyses in Tables 3 and 4. Thus, because of the lower standard

Downloaded from http://ps.oxfordjournals.org/ at Ernst Mayr Library of the Museum Comp Zoology, Harvard University on July 1, 2015

BW SW SL DL

GENETIC PARAMETERS OF SKELETAL TRAITS IN TURKEYS

more moderate degree with shank and drum density scores (rG = .23 to .79). These relationships were consistently higher from M than they were from F, and the combined estimates were intermediate to those from the individual sexes. The rG of SW with BW was relatively low: M = .33, F = .47, and combined = .34 (Tables 3, 4, and 5). This indicates that selection for BW alone might not increase SW enough to support the changes brought about in BW. The

TABLE 3. Genetic (YQ) and phenotypic (rp) correlation estimates from sire components of variance and covariance calculated from data on females' Trait 2

BW

BW SE3

FFS

D40

D60

S40

S60

.97 .09

.76 .17

.48 .24

.23 .31

.51 .27

.25 .32

.22 .27

.27 .24

.25 .24

.55 .18

.16 .27

.36 .25

.49 .22

.08

.90 .10

.99 .08

.87 .09

.63 .17

.91 .14

.61 .22

.61 .21

.97 .07

.89 .12

.97 .06

.85 .10

.60 .16

.80 .15

.57 .21

.64 .19

SW

SL

DL

RCT

RCD

RCS

CD

CS

FFD

.47 .24

.84 .19

1.10

NE4

1.72

1.20

1.03

.62

.15

.10

.90 .17

.10 .26

.18 .23

NE

.56 .35

.55 .21

.17 .26

1.01

NE

1.15

.64 .19

1.03

.90 .13

.12

SW SE

.51

SL SE

.33

DL SE

.46

RCT SE

.41

.32

.42

.34

RCD SE

.51

.41

.55

.45

RCS SE

.47

.42

.54

.43

.52

.66

CD SE

.60

.38

.54

.62

.50

.67

CS SE

.46

.38

.63

.46

.52

.65

.64

.79

FFD SE

.66

.41

.52

.65

.50

.64

.59

.82

.66

FFS SE

.54

.41

.66

.52

.56

.70

.72

.73

.82

.81

D40 SE

.40

.34

.40

.34

.33

.45

.41

.52

.47

.65

.60

D60 SE

.49

.30

.36

.34

.36

.46

.39

.59

.46

.72

.59

.84

S40 SE

.43

.30

.34

.32

.35

.38

.46

.42

.46

.56

.68

.64

.64

S60

.43

.35

.34

.33

.32

.39

.45

.45

.47

.58

.67

.69

.70

.19

.32

.05 .18

NE

.66

1.50 .49

1

NE .64

NE

NE

NE

NE

NE

NE

NE

NE

NE

1.20

1.05 .25

.83 .25

1.02

.28

.26

.72 .26

.77 .32

.59 .35

.62 .38

.29 .44

.84 .13

.85 .13

.88 .12

.75 .14

.61 .19

.49 .25

.75 .19

.65 .21

.83 .10

1.00 .03

.74 .13

.69 .15

.87 .12

.61 .22

.52 .23

.92 .09

.94 .05

.61 .18

.72 .19

.81 .17

.71 .19

.87 .07

.80 .10

.90 .08

.77 .15

.60 .19

.73 .13

.77 .13

.98 .07

.85 .11

.92 .05

.74 .14

.72 .14

.61 .20

.47 .23

.60

.87 .09 .82

The TQ are above and rp are below the diagonal.

2

SW = Shank width; SL = shank length; DL = drum length; RCT, RCD, and RCS = rough-cleaned weights of the thigh, drum and shank, respectively; CD and CS = cleaned weights of the drum and shank, respectively; FFD and FFS = fat-free weights of the drum and shank, respectively; D40, D60, S40, and S60 = density measurements (Cantor et al., 1980) of the drum and shank taken at 40 and 60% of the length from the proximal end, respectively. 3

SE = Standard error of the genetic correlation estimate.

4

NE = Nonestimatable, due to negative variance component.

Downloaded from http://ps.oxfordjournals.org/ at Ernst Mayr Library of the Museum Comp Zoology, Harvard University on July 1, 2015

errors associated with them, it appears that the combined estimates would be the ones of choice even though the variances from the two sexes were heterogeneous for most traits. The rG between BW and the various bone weight measurements studied were extremely high (>.75), and the rough-cleaned bone weights were as highly correlated with BW as were the cleaned and fat-free bone weights (Tables 3 , 4 , and 5). The BW was correlated to a

1383

HAVENSTEIN ET AL.

1384

turkeys would be expected to maintain relative bone length better than bone width. However, such selection could be expected to result in a proportionate decrease in both length and width of the leg bones in relation to BW. Relative bone density would also be expected to decline under selection for BW alone. All bone weight measurements were highly correlated (r G >.66; Tables 3, 4, and 5), indicat-

TABLE 4. Genetic (YQ) and phenotypic (rp) correlation estimates from sire components of variance and covariance calculated from the data on males1 Trait 2

CS

FFD

S40

WA

BW SW

SL

DL

RCT

RCD

RCS

CD

.33 .21

.53 .18

.55 .17

.80 .10

.90 .08

.79 .10

.94 .05

.82 .12

.86 .07

.81 .09

.59 .16

.79 .10

.66 .14

.66 .15

-.73

.32 .23

.18 .24

.27 .22

.34 .22

.45 .19

.39 .20

.53 .19

.40 .20

.40 .20

.57 .17

.23 .22

.41 .20

.48 .19

-.09

.91 .05

.75 .12

.73 .13

.86 .08

.63 .14

.90 .09

.70 .12

.74 .11

.62 .15

.45 .18

.47 .19

.41 .20

.03 .44

.76 .12

.80 .11

.91 .07

.58 .15

.78 .13

.64 .13

.67 .13

.54 .17

.36 .19

.39 .20

.32 .21

.52 .45

.97 .04

.84 .08

.91 .06

.94 .07

.75 .10

.66 .13

.51 .17

.70 .12

.41 .19

.39 .20

-.08

.96 .04

.99 .06

.90 .06

.85 .08

.71 .13

.85 .08

.68 .14

.64 .15

-.31

.93 .05

1.00

.87 .06

.72 .12

.67 .13

.67 .13

.58 .16

-.36

.05

.89 .06

.99 .03

.94 .04

.83 .08

.70 .12

.90 .06

.63 .14

.68 .13

-.42

.98 .05

.91 .06

.91 .09

.90 .08

.72 .13

.79 .12

-.60

.98 .02

.88 .06

.95 .03

.88 .07

.87 .07

-.80

.92 .05

.87 .07

.94 .04

.91 .05

-.83

.77 .09

.91 .05

.91 .06

-.75

.80 .09

.81 .09

-.87

BW SE3

FFS

D40

D60

SW SE

.39

SL SE

.32

.09

DL SE

.36

.09

.79

RCT SE

.55

.27

.45

RCD SE

.52

.31

.47

.49

.73

RCS SE

.55

.44

.55

.50

.64

.69

CD SE

.57

.33

.52

.52

.63

.65

.58

CS SE

.38

.34

.49

.37

.51

.51

.61

FFD SE

.61

.33

.53

.55

.59

.64

.62

.76

.56

FFS SE

.53

.38

.61

.50

.54

.57

.74

.67

.71

.79

D40 SE

.30

.35

.39

.30

.35

.41

.42

.48

.40

.68

.60

D60 SE

.43

.30

.41

.30

.47

.51

.45

.63

.46

.75

.61

.77

S40 SE

.41

.35

.36

.32

.32

.36

.51

.44

.46

.58

.74

.63

.60

S60 SE

.43

.36

.37

.30

.34

.37

.49

.46

.49

.60

.74

.64

.62

.87

WA

.07

.03

.03

.02

.08

.07

-.01

.08

.02

-.01

-.13

-.08

-.05

-.18

.45

.96 .04

.75

S60

.45 .45

.43 .43 .43 .42 .47 .46 .50 .48 .48

.95 -1.47 .03 .69

-1.26 .63

1

-.15

The TQ are above and rp are below the diagonal.

2

SW = Shank width; SL = shank length; DL = drum length; RCT, RCD, and RCS = rough-cleaned weights of the thigh, drum and shank, respectively; CD and CS = cleaned weights of the drum and shank, respectively; FFD and FFS = fat-free weights of the drum and shank, respectively; D40, D60, S40, and S60 = density measurements (Cantor et al., 1980) of the drum and shank taken at 40 and 60% of the length from the proximal end, respectively ; WA = walking ability score (1 = normal through 5 = extreme difficulty in walking). 3

SE = Standard error of the genetic correlation estimate.

Downloaded from http://ps.oxfordjournals.org/ at Ernst Mayr Library of the Museum Comp Zoology, Harvard University on July 1, 2015

BW was more highly correlated with SL and DL than it was with SW: .53 and .84 for SL from M and F; and .55 and 1.10 for DL from M and F. The SW had low correlations with SL and DL: .32 and . 10 for SL from M and F; and .18 and .18 for DL from M and F. The SW measurements were moderately correlated with drum and shank density measurements (range .16 to .57). Thus, selection for BW alone in

GENETIC PARAMETERS OF SKELETAL TRAITS IN TURKEYS

(r G >.47, with most above .75). The WA scores showed a high negative rG (-.73; Table 4) with BW, indicating that those families with the lowest BW tended to have the poorest WA (i.e., the highest scores). This result is contrary to what one would expect, and may be a spurious correlation. It is suspected that those individuals with poor WA would tend not to be able to eat and drink satisfactorily, would tend to drop in social order, and would therefore tend to have low BW. Sheridan et al. (1978)

TABLE 5. Genetic (YQ) and phenotypic (rp) correlation estimates from sire components of variance and covariance calculated from the analyses of both sexes combined1 Trait 2

BW

BW SE3

sw

sw

SL

DL

RCT

RCD

RCS

CD

CS

FFD

FFS

D40

D60

S40

S60

.34 .19

.60 .15

.72 .11

.82 .09

.94 .06

.87 .07

.91 .05

.78 .11

.86 .06

.75 .10

.55 .15

.71 .11

.60 .14

.58 .15

.18 .21

.17 .20

.26 .21

.36 .20

.38 .19

.28 .20

.38 .19

.29 .19

.28 .19

.54 .15

.18 .20

.31 .20

.43 .18

.97 .03

.80 .10

.78 .10

.84 .08

.73 .10

.84 .08

.74 .10

.75 .09

.60 .14

.55 .15

.50 .16

.48 .17

.83 .09

.83 .09

.90 .06

.72 .11

.78 .10

.71 .11

.71 .11

.57 .14

.50 .16

.44 .17

.49 .17

.99 .03

.87 .07

.94 .05

.92 .06

.77 .10

.66 .13

.55 .16

.74 .11

.44 .18

.37 .20

.96 .04

.96 .04

.94 .06

.87 .06

.80 .09

.65 .13

.80 .09

.67 .14

.64 .15

.95 .04

.96 .04

.90 .05

.85 .07

.70 .12

.70 .12

.69 .12

.68 .13

.94 .04

.95 .03

.80 .08

.66 .12

.87 .06

.64 .13

.61 .14

.95 .04

.91 .05

.80 .10

.86 .08

.76 .11

.74 .11

.95 .03

.83 .07

.95 .03

.85 .07

.80 .09

.86 .06

.87 .06

.96 .03

.91 .05

.78 .08

.86 .07

.85 .07

.81 .08

.74 .10

.44

SE SL SE

.33

.13

DL SE

.39

.11

.72

RCT SE

.51

.29

.45

.41

RCD SE

.52

.35

.50

.45

.70

RCS SE

.53

.44

.55

.45

.60

.68

CD SE

.58

.34

.52

.54

.58

.66

.58

CS SE

.40

.36

.53

.38

.51

.55

.62

.76

FFD SE

.63

.36

.52

.55

.56

.64

.61

.78

.59

FFS SE

.54

.40

.62

.47

.55

.61

.74

.68

.74

.79

D40 SE

.33

.35

.38

.28

.34

.42

.42

.49

.42

.67

.60

D60 SE

.44

.29

.38

.30

.43

.49

.42

.62

.46

.74

.60

.79

S40 SE

.42

.34

.35

.29

.33

.37

.50

.43

.47

.58

.72

.63

.60

S60

.43

.37

.36

.29

.33

.38

.48

.45

.48

.59

.71

.66

.64

1

.96 .02 .85

The XQ are above and rp are below the diagonal.

2

SW = Shank width; SL = shank length; DL = drum length; RCT, RCD, and RCS = rough-cleaned weights of the thigh, drum and shank, respectively; CD and CS = cleaned weights of the drum and shank, respectively; FFD and FFS = fat-free weights of the drum and shank, respectively; D40, D60, S40, and S60 = density measurements (Cantor et al., 1980) of the drum and shank taken at 40 and 60% of the length from the proximal end, respectively. 3

SE = Standard error of the genetic correlation estimate.

Downloaded from http://ps.oxfordjournals.org/ at Ernst Mayr Library of the Museum Comp Zoology, Harvard University on July 1, 2015

ing that rough-cleaned measurements could serve as satisfactory selection criteria to improve the other measures. The weight of the RCS, which is the easiest bone to measure, was highly correlated with other bone weights (r G >.74, with most above .83), and would appear to be the measurement of choice for a commercial selection program. The RCS was also moderately correlated with bone density measurements (r G >.49). Shank and drum density measurements were highly correlated with each other

1385

HAVENSTEIN ET AL.

1386

REFERENCES Abplanalp, H., and I. L. Kosin, 1952. Heritability of body measurements in turkeys. Poultry Sci. 31:781-791. Arthur, J. A., and H. Abplanalp, 1975. Linear estimates of heritability and genetic correlation for egg production, body weight, conformation, and egg weight of turkeys. Poultry Sci. 54:11-23. Bartlett, M. S., 1937. Properties of sufficiency and statistical tests. Proc. R. Soc. Lond., Series A 160:268-282. Bohidar, N. R., 1964. Derivation and estimation of variance and covariance components associated with covariance between relatives under sex-linked transmission. Biometrics 20:505-521. 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. Cantor, A. H., M. A. Musser, W. L. Bacon, and A. B. Hellewell, 1980. The use of bone mineral mass as an

Nestor, K. E., W. L. Bacon, P. D. Moorhead, Y. M. Saif, G. B. Havenstein, and P. A. Renner, 1987. Comparison of bone and muscle growth in turkey lines selected for increased body weight and increased shank width. Poultry Sci. 66:1421-1428. Nestor, K. E., W. L. Bacon, Y. M. Saif, and P. A. Renner, 1985. The influence of genetic increases in shank width on body weight, walking ability, and reproduction of turkeys. Poultry Sci. 64:2248-2255. Nestor, K. E., M. G. McCartney, and N. Bachev, 1969. Relative contributions of genetics and environment to turkey improvement. Poultry Sci. 48:1944-1949. Neter, J., and W. Wasserman, 1974. Applied Linear Statistical Models. RichardD. Irwin, Inc. Homewood, IL. Riddell, C , 1976. Selection of broiler chickens for a high and low incidence of tibial dyschondroplasia with observations on spondylolisthesis and twisted legs (Perosis). Poultry Sci. 55:145-151. Salmon, R. E., 1979. Slaughter losses and carcass composition of the medium white turkeys. Br. Poult. Sci. 20:247-302. Serfontein, P. J., and L. F. Payne, 1934. Inheritance of

Downloaded from http://ps.oxfordjournals.org/ at Ernst Mayr Library of the Museum Comp Zoology, Harvard University on July 1, 2015

indicator of vitamin D status in turkeys. Poultry Sci. reported a similar negative rG between BW and 59:563-568. leg disorders in broiler chickens. The WA scores V., M. Dounaire, and J. Mallard, 1986. Les (Table 4) also had high negative correlations Delabrosse, parametres genetiques de la composition corporelle with bone density measurements (>-.75) and chez la dinde. Pages 171-175 in: Proc. 7th European with bone weights (range = -.08 to -.83). Poultry Conf., Vol. 2, Paris, France, October24-28. These correlations may also be inflated if WA Folch, J., M. Lees, and G. H. Sloane-Stanley, 1957. A simple method for the isolation and purification of affected feed consumption and therefore mineral total lipids from animal tissues. J. Biol. Chem. intake and metabolism. 226:497-509. In conclusion, the low to moderate rG esti- Harvey, W. R., 1986. User's guide for LSMLMW. The mates between BW and several skeletal characOhio State Univ., Columbus, Ohio. teristics such as bone weights, bone lengths, Havenstein, G. B., V. D. Toelle, K. E. Nestor, and W. L. Bacon, 1988. Estimates of genetic parameters in turand bone density scores, indicate that selection keys: 2. Body weight and carcass characteristics. Poulfor BW alone would result in disproportionately try Sci. 67:1388-1399. low increases in skeletal characteristics in rela- Haye, U., and P.CM. Simons, 1978. Twisted legs in broilers. Br. Poult. Sci. 19:549-557. tion to the gains that would be brought about in BW. Therefore, selection for BW or BW and James, J. W., 1973. Covariance between relatives due to sex-linked genes. Biometrics 29:584-588. some measure of breast conformation may lead Krueger, W. F., R. L. Atkinson, J. H. Quisenberry, and to less than optimal changes in the skeletal supJ. W. Bradley, 1972. Heritability of body weight and port system, thereby leading to genetically inconformation traits and their genetic association in turkeys. Poultry Sci. 51:1276-1282. duced leg problems. Genetic correlations of less than unity with breast and leg muscle weights Leach, R. M., Jr., and M. C. Nesheim, 1965. Nutritional, genetic and morphological studies of an abnormal car(Havenstein et al., 1988) may also contribute tilage formation in young chicks. J. Nutr. 86:236-244. to genetically induced leg problems. Nestor et Leach, R. M., Jr., and M. C. Nesheim, 1972. Further al. (1987) recently reported comparisons of studies on tibial dyschondroplasia (cartilage abnormality) in young chicks. J. Nutr. 102:1673-1680. growth-selected and shank width-selected lines of turkeys with the RBC2 line used herein, Mercer, J. T., and W. G. Hill, 1984. Estimation of genetic "^ parameters for skeletal defects in broiler chickens. which support this conclusion. Heredity 53:193-203. These and other data from the Ohio State Mullen, K., and H. J. Swatland, 1979. Linear skeletal growth in male and female turkeys. Growth 43:151University indicate that breeders of domestic 3?3 V 159. turkeys should use a balanced selection approach E. C , and S. P. Touchburn, 1970. Ohio Poultry that includes some measure of skeletal growth. ^.? Naber,Rations. Ohio Coop. Ext. Serv. Bull. 343, Ohio State It appears that SW and SL and possibly the RCS^C.$ Univ., Columbus, OH. c could provide the measures of skeletal develop- ° Nestor, K. E., 1977. The use of a paired mating system for the maintenance of experimental populations of turment necessary for bringing about desirable keys. Poultry Sci. 56:60-65. changes in the skeletal support system in a breedNestor, K. E., 1984. Genetics of growth and reproduction ing program that is primarily aimed at the imin the turkey. 9. Long-term selection for increased provement of growth rate. 16-wk body weight. Poultry Sci. 63:2114-2122.

GENETIC PARAMETERS OF SKELETAL TRAITS IN TURKEYS

Tallis, G. M., 1959. Sampling errors of genetic correlation coefficients calculated from analyses of variance and covariance. Aust. J. Stat. 1:39-43. 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. Veltmann, J. R., Jr., and L. S. Jensen, 1981. Tibial dyschondroplasia in broilers: comparison of dietary additives and strains. Poultry Sci. 60:1473-1478. Walser, M. M., F. L. Cherms, and H. E. Dziuk, 1982. Osseus development and tibial dyschondroplasia in five lines of turkeys. Avian Dis. 26:265-271.

Downloaded from http://ps.oxfordjournals.org/ at Ernst Mayr Library of the Museum Comp Zoology, Harvard University on July 1, 2015

abnormal anatomical conditions in the tibial metatarsal joints. Poultry Sci. 13:61-63. Sheridan, A. K., C. R. Howlett, and J. A. Bruyn, 1974. Genetic factors influencing tibial dyschondroplasia in Australian meat chickens. Pages 34-35 in: Proc. XV World's Poult. Cong., New Orleans, LA. Sheridan, A. K., C. R. Howlett, and R. W. Burton, 1978. The inheritance of tibial dyschondroplasia in broilers. Br. Poult. Sci. 19:491-499. Swiger, L. A., W. R. Harvey, D. O. Everson, and K. E. Gregory, 1964. The variance of intra-class correlation involving groups with one observation. Biometrics 20:818-826.

1387