Genetic and Phenotypic Relationships among Type Traits Scored Linearly in Holsteins1

Genetic and Phenotypic Relationships Among Type Traits Scored Linearly in Holsteins 1 G. B. S C H A E F F E R , W. E. V I N S O N , R. E. PEARSON, and R. G. LONG 2

Department of Dairy Science Virginia Polytechnic Institute and State University Blacksburg 24061 ABSTRACT

descriptions may vary both among and within organizations. Objectives of this study were to evaluate a linear type appraisal system for Holstein cows operated by one AI organization and to compare results with those from the official breed association program (5) and from two other AI organizations (2, 4).

Phenotypic and genetic parameters were estimated for 12 type traits of Holstein cows scored linearly by personnel of an artificial insemination organization. Estimates were compared with those from programs offered by the breed association and by two other artificial insemination organizations. All programs were based on scoring guidelines of the National Association of Animal Breeders. Heritabilities of linear type traits ranged from .40 for stature to .14 for fore udder attachment with heritabilities for udder traits smaller than for nonudder traits. Largest genetic and phenotypic correlations were among udder traits with genetic correlations consistently exceeding phenotypic correlations. Similarities in relative magnitudes of phenotypic and genetic parameters suggest that linear type appraisal programs do not differ greatly among organizations using scoring guidelines of the National Association of Animal Breeders.

MATERIALS AND METHODS

INTRODUCTION

Linear evaluations for individual type traits are assigned both by breed associations and artificial insemination (AI) organizations, frequently yielding dual evaluations for individual sires. Systems of appraisal used by different organizations seemingly are very similar in that they are based on guidelines of the National Association of Animal Breeders (NAAB). However, point ranges for scoring differ among organizations, and interpretations of scoring

Received April 8, 1985. Partially supported by a grant from Select Sires, Inc. 2Select Sires, Inc., Plain City, OH 43064. 1985 J Dairy Sci 68:2984-2988

Complete, valid, first available linear evaluations for 12 type traits (Table 1) were from 56,642 Holstein cows ("all" records), of which 19,020 included sire identity ("sire identified" records). Traits were scored by trained evaluators of Select Sires, Inc. using a nine-point (1 to 9) scoring range. Records were from 1,854 herds in 41 states. All records were used to determine effects on linear type scores of parity (1, 2, and 3 or greater), year (1979 to 1981), season (January to June and July to December), evaluator, as well as interactions of parity with season and evaluator. All effects except random error were assumed fixed and estimated after absorbing herds. Multiplicative factors for adjusting linear scores to a first parity basis were obtained by fitting least squares constants for parity after absorbing herd-year-month of evaluation. Adjustment factors were determined within four regions defined as combinations of the geographical regions used in computing Dairy Herd Improvement (DHI) age adjustment factors for milk and fat yields. Combining of DHI regions was to provide sufficient numbers of records for accurate determination of adjustment factors. Phenotypic correlations between linear type traits were product-moment correlations between linear scores adjusted for parity, using all records. Heritabilities of, and genetic correlations among linear type traits were estimated from a subset of sire identified records including

2984

PARAMETERS FOR LINEAR TYPE

17,280 daughters of 235 sires having a m i n i m u m of 15 daughters. The m o d e l used was: Yijkl =/a + H i + Pj + S k + eijkl

2985

p r o c e d u r e o f the Statistical Analysis System (3) with h e r d - y e a r - m o n t h subclasses absorbed. A p p r o x i m a t e standard errors of heritabilities and genetic correlations were as given by Dickerson (1).

where: /~ is an effect c o m m o n to all records, H i is an effect c o m m o n to all records in the ith herd-year-month, Pj is an effect c o m m o n to all records of the jth parity, Sk is an effect c o m m o n to all daughters o f the k th sire, and eijkl is an effect peculiar to the i th daughter of the kth sire in the jth parity and i th herd-year-month. Sire and error effects were assumed indep e n d e n t l y distributed r a n d o m variables with 2 2 means zero and variances as and Oe, respectively. Solution was b y the general linear models

RESULTS A N D DISCUSSION

Means and standard deviations for linear t y p e traits (Table 2) were similar for all and sire identified records. Means exceeded the average of integers (5 points) used for scoring as r e p o r t e d for o t h e r evaluation programs (2, 4, 5). Also similar to o t h e r progams, teat p l a c e m e n t and fore udder a t t a c h m e n t were m o s t variable whereas legs, side view, and pelvic angle were least variable. Standard deviations were 12 to 15% of the scoring range for all traits in all programs. Mean squares f r o m analyses o f variance for linear t y p e traits are in "Fable 3. Parity, year, season, evaluator, and parity-by-evaluator sig-

TABLE 1. Linear type traits and interpretation of high scores. Trait

High scores

Trait

High scores

Stature Dairyness Strength Pelvic width Pelvic angle Legs, side view

Tall Angular Strong Wide Slope Sickled

Heel depth Fore udder attachment Rear udder height Rear udder width Udder support Teat placement

Deep Tight High Wide Strong Close

TABLE 2. Means and standard deviations for linear type traits. Trait

Stature Dairyness Strength Pelvic width Pelvic angle Heel depth Legs, side view Fore udder attachment Rear udder height Rear udder width Udder support Teat placement

All records

Sire identified records

.X

SD

.X

SD

5.69 6.49 6.16 5.79 5.77 5.52 6.20 5.76 5.69 5.83 6.92

1.26 1.31 1.22 1.20 1.21 1.35 1.12 1.46 1.24 1.19 1.23 1.51

5.68 6.62 6.15 5.70 5.86 5.62 6.14 5.84 5.76 5.98 7.15 6.13

1.28 1.28 1.19 1.22 1.07 1.32 1.11 1.43 1.22 1.24 !.17 1.51

6.03

Journal of Dairy Science Vol. 68, No. 11, 1985

2986

SCHAEFFER ET AL.

season were significant ( P < . 0 5 ) for 4 o f t h e 12 traits, b u t m e a n squares generally were small. Heritabilities f o r linear t y p e t r a i t s (Table 4) r a n g e d f r o m .40 for s t a t u r e to .14 for fore u d d e r a t t a c h m e n t a n d generally were smaller for u d d e r traits t h a n for n o n u d d e r traits. Heritabilities in t h e c u r r e n t s t u d y were somew h a t larger t h a n t h o s e r e p o r t e d f r o m b r e e d a s s o c i a t i o n data (5) a n d d a t a f r o m o n e AI o r g a n i z a t i o n (2) b u t smaller t h a n r e p o r t e d in (4). However, r a n k o f h e r i t a b i l i t i e s was similar in all f o u r programs, i n d i c a t i n g stature, s t r e n g t h , pelvic w i d t h , a n d pelvic angle were a m o n g t h e m o s t h e r i t a b l e , a n d legs, side view, heel d e p t h , a n d u d d e r s u p p o r t t o be a m o n g t h e least h e r i t a b l e traits. Largest d i f f e r e n c e s a m o n g p r o g r a m s in relative h e r i t a b i l i t i e s were l o w e r

n i f i c a n t l y ( P < . 0 5 ) a f f e c t e d all traits. Mean squares for m a i n effects generally were largest for year a n d season. Largest effects o f p a r i t y were for stature, pelvic w i d t h , fore u d d e r a t t a c h m e n t , a n d s t r e n g t h . Results were similar to t h o s e for o t h e r p r o g r a m s (2, 4, 5) e x c e p t t h a t in o t h e r p r o g r a m s a larger effect o f p a r i t y o n scores for dairyness was n o t e d . E v a l u a t o r s d i f f e r e d m o s t in assigning scores for u d d e r s u p p o r t , s t r e n g t h , rear u d d e r height, a n d dairyness a n d were m o s t c o n s i s t e n t for rear u d d e r w i d t h a n d fore u d d e r a t t a c h m e n t . Interactions between parity and evaluator suggested e v a l u a t o r s a d j u s t e d d i f f e r e n t l y for effects o f parity, p a r t i c u l a r l y for rear u d d e r h e i g h t a n d heel d e p t h . Similar i n t e r a c t i o n s were r e p o r t e d in (5). I n t e r a c t i o n s b e t w e e n p a r i t y a n d

TABLE 3. Means squares for effects of environmental factors on individual type traits. Source of variation ~,2 Y (2)

S (1)

E

P × S

P × E

Trait

P (2)

(25)

(2)

(45)

Residual (54710)

Stature Dairyness Strength Pelvic width Pelvic angle tteel depth Legs, side view Fore udder attachment Rear udder height Rear udder width Udder support Teat placement

38.4 16.3 19.1 37.8 8.9 14.0 7.2 29.0 6.5 14.1 10.4 17.6

23.3 110.4 40.9 70.6 59.2 130.5 98.7 189.8 96.9 19.5 38.5 44.3

48.4 93.0 48.2 116.8 46.4 176.1 7.3 33.0 93.1 4.2 41.5 36.8

I1.0 23.0 32.1 17.4 11.5 20.1 15.1 7.3 27.1 5.8 36.2 14.9

1.9 ~ 1.83 2.83 .93 13.0 5.5 5.9 1.53 4.7 1.23 .83 .373

8.8 13.5 13.3 6.0 7.8 33.9 3.3 14.9 47.7 15.9 3.3 6.4

1.4 1.4 1.2 1.1 1.1 1.5 1.1 1.9 1.2 1.2 1.1 2.0

Sources of variation with herds absorbed were p = parity, Y = year, S = season, E = evaluator. 2 Degrees of freedom in parentheses. 3P>.05, all others P<.05.

TABLE 4. Heritabilities (h 2 ) and approximate standard errors for linear type traits. Trait

h2

SE

Trait

h2

SE

Stature Dairyness Strength Pelvic width Pelvic angle Heel depth

.40 .31 .29 .29 .25 .18

.039 .032 .030 .030 .027 .021

Legs, side view Fore udder attachment Rear udder height Rear udder width Udder support Teat placement

.16 .14 .20 .24 .15 .22

.020 .018 .023 .026 .019 .025

Journal of Dairy Science Vol. 68, No. 11, 1985

PARAMETERS

FOR LINEAR

relative estimates from one AI organization for strength and dairyness (2) and higher estimates from another for fore udder attachment (4). Comparisons of heritabilities from different programs did not show any apparent tendency for 50-point scoring systems to yield higher heritabilities than 9-point systems. This may be partly from classifier inexperience in applying 50-point scoring systems. Genetic and phenotypic correlations among linear type traits are in Table 5. Largest phenotypic correlations were among udder traits, ranging from .56 between udder support and teat placement to .18 between rear udder width and teat placement. Phenotypic correlations among udder traits computed in this study were very similar to those from other AI organizations (2, 4) but generally less than correlations from breed association data (5). Phenotypic correlations among nonudder traits and among udder and nonudder traits were generally in the range of -+ .2 and were very similar to correlations from other AI organizations (2, 4). Except for the correlation between stature and strength (near .50 from breed association data and .25 from AI organization data), agreement also was good with correlations from the breed association program (5). Largest genetic correlations were among udder traits, ranging from .75 between udder support and teat placement and .72 between rear udder height and width to .27 between rear udder height and teat placement. Genetic correlations among udder traits generally were similar in AI organization programs and slightly higher for breed association traits. Genetic correlations among nonudder traits were largest for strength with pelvic width (.65), dairyness (-.56), and heel depth (.55), and for heel depth with legs, side view (-.61). Those among udder and nonudder traits were largest for pelvic width and rear udder width (.45), dairyness, and udder support (.41), and for pelvic angle with rear udder height (-.34). In general, agreement among the programs examined was less for genetic than for phenotypic correlations. This might be expected from the relative sizes of sampling errors of the two types of correlations. There was a tendency for differences among estimates of genetic correlations to involve the traits, stature and strength, and the results from one AI organization (4). However, estimates from the current

2987

TYPE

I

I

.-Z

v ~ q o O O ° q ~ I I

.

.

.

£ t~

E O

O

I

i

III

I

8 ~O~OO I

. . . . . . I II

I

r~

< I

I

Ill q

£ I

I

O

i

E O

O it~

I

I

e~

I

I

I

I O

0~

II

I

I

8

.u

-7 &

I

£ < O v v

,,4

E X £ e~ e~

J o u r n a l o f D a i r y S c i e n c e V o l . 6 8 , N o . 11, 1 9 8 5

2988

SCHAEFFER ET AL.

study differed substantially from previous estimates for genetic correlations between strength and pelvic width (.65) and between pelvic width and fore udder attachment (.01).

association or AI organization evaluation. However, variation from these differences is likely to be small relative to variation from differing samples of daughters scored by AI organizations and breed association programs.

CONCLUSIONS

REFERENCES

Similarities in relative magnitudes of genetic and phenotypic parameters suggests linear type appraisal programs do not differ greatly among organizations using NAAB scoring guidelines. The generally good agreement among phenotypic correlations suggests that variation in estimates of genetic correlations for some traits was likely from sampling variation or from differences in populations of bulls examined rather than from substantive differences in interpretation and application of the NAAB scoring guidelines. Such differences as do exist are likely sufficient to expect variation in sire rankings for linear type traits based on breed

1 Dickerson, G. E. 1969. Techniques for research in quantitative animal genetics. Page 36 in Tech. Proc. Anita. Sci. Res. 2 Edlund, D. P., J. M. White, and W. E. Vinson. 1979. Genetic parameters of a linearized type appraisal program. J. Dairy Sci. 62(Suppl. 1):144. (Abstr.) 3 Statistical Analyses System Institute Inc. 1982. SAS User's guide: statistics. 1982 ed. Cary, NC. 4 Thompson, J. R., A. E. Freeman, D. J. Wilson, Ca. A. Chapin, and P. J. Berger. 1981. Evaluation of a linear type program in Holsteins. J. Dairy Sci. 64:1610. 5 Thompson, J. R., K. L. Lee, A. E. Freeman, and L. P. Johnson. 1983. Evaluation of a linearized type appraisal system for Holstein cattle. J. Dairy Sci. 66:325.

Journal of Dairy Science Vol. 68, No. 11, 1985