- Email: [email protected]
Relation of Registration with Yield and Termination of First Lactation in Dairy Goats1
Relation of Registration with Yield and Termination o f First L a c t a t i o n in D a i r y G o a t s 1 T. A. GIPSON and M. GROSSMAN ~ Department of Animal Sciences
University of Illinois at Urbana-Champaign Urbana 61801
G. R. WIGGANS Animal Improvement Programs Laboratory Agricultural Research Service United States Department of Agriculture
Beltsville, MD 20706
ABSTRACT
production than grades within herd for all breeds except Brown Swiss; however, yield differences were small. Dentine et al. (2) found no significant difference between registered and grade Holsteins in first lactation Modified Contemporary Deviation for milk. Powell and Norman (10) reported that average milk yield generally was higher in 2-yr-old registered Holsteins than in grades within herd and overall. United States Department of Agriculture studies (9, 10) found that culling within herd on yield was more severe for grades than for registered Holsteins. Studies by Dentine et al. (2, 3) indicated that first lactation grade Holsteins were culled more intensely than purebreds. They also found that grade cows were culled more often than registered cows because of low yield and recommended that registration of the daughter should be considered when evaluating sires for culling rate or productive lifespan (2). Several registries are availabile for dairy goats in the United States. Two major registries INTRODUCTION are purebred and American programs of the An important measure of production in the dairy industry is how well a dairy animal per- American Dairy Goat Association (ADGA). forms in her first lactation. Differences in regis- Purebred dairy goats are registered in breed herdbooks and are defined by ADGA (1) as tration of first lactation animals are of interest dairy goats and their linear descendants "imto dairy breeders because of possible relationported from the Eastern Hemisphere that were ships between registration and production. Norman and Powell (9) reported that 2-yr- duly registered in the countries of their origin before their importation; and those goats duly old registered cows (1980 calvings) had greater and correctly registered by ADGA in conformity with its registry rules." These rules specify Received December 29, 1986. that sire and dam must be registered purebreds Accepted April 9, 1987. aContribution of Project Number 35-306 of the of the same breed, except for LaManchas. LaAgricultural Experiment Station, University of Illinois Manchas with three generations of American at Urbana-Champaign, as a part of the work of NC-2, LaMancha ancestry, born after January 1, and supported in part by the University of Illinois 1980, qualify for registration in the purebred Research Board. LaMancha herdbook. Nonpurebred dairy goats 2Reprint requests. First lactation records of 9941 Alpine, 2383 LaMancha, 11,203 Nubian, 3714 Saanen, and 4649 Toggenburg dairy goat does were obtained from the USDA to evaluate relation of registration (purebred, American, or grade) with lactation yield traits (mature equivalent milk and fat yields) and with reason for termination of lactation (normal, sold for dairy, or culled). Does were grouped by age at first kidding (11 through 16 mo or 21 ~hrough 26 mo) and by calendar year (1976 through 1978, 1979 through 1981, or 1982 through 1984). Data were analyzed within breed and year group for yield traits and within breed for reason of termination. Registration and age at first kidding accounted for little variation in mature equivalent milk and fat yields or in reason for termination.
1987 J Dairy Sci 70:1710-1718
1710
REGISTRATION AND MILK YIELD IN GOATS
1711
(code 8) nor records not used in genetic evaluations (code 9). Data were from 9941 Alpine, 2383 LaMancha, 11,203 Nubian, 3714 Saanen, and 4649 Toggenburg does. Of these 31,890 does, 16,185 (51%) were registered as purebred and 5551 (17%) as American; 10,154 (32%) were recorded as grade (Table 1). Records were grouped by age at first kidding (11 thorugh 16 mo or 21 through 26 mo) because o f a reported significant effect of age within parity on lactation yields (4, 8). Due to seasonality of breeding, goats do not normally kid at ages 17 through 20 too, and those that did were excluded as being unrepresentative of does kidding as yearlings or 2-yr-olds. F o r each breed and registration, about 75% of does were 11 through 16 mo of age at first kidding, except Nubian, which had a percentage of 70% for this age group. Records also were grouped into 3-yr intervals (1976 through 1978, 1979 through 1981, or 1982 through 1984) to account for changes in distribution of registration. The model for the least squares analysis of variance used to evaluate relation o f registration with mature equivalent (ME) milk and fat yields was:
that have least 7/8 purebred lineage are included in A D G A American herdbooks (1). Other animals are eligible to be recorded in A D G A grade programs (1). Whether registration is related to lactation yield or to reason for termination of lactation is not known for dairy goats. Grossman and Wiggans (7) examined distribution of lactation records based on reason for termination by breed, but registration and parity were not considered. Gipson et aI. (6) found no significant relationship between registration and buck PD for yield traits. The objective of this study was to evaluate relation of registration with first lactation yield traits and with reason for termination of first lactation for dairy goat does. MATERIALS AND METHODS
First lactation records o f dairy goat does enrolled in the National Cooperative Dairy Herd Improvement Program ( N C D H I P ) f r o m 1 9 7 6 through 1984 were obtained from the Animal Improvement Programs Laboratory, Agricultural Research Service, USDA. Only first lactation records were included because later lactations might be affected by selection. A t least two records per herd-year were required. Reasons for termination of lactation were based on codes reported in records (12): normal lactation terminated by dry date or 305 d in milk if doe is still in the herd (code 0); sold for dairy purposes (code 2); or sold because of low yield or other problems (codes 3 through 7). Termination codes 3 through 7 were combined because some dairy records processing centers do not use all codes. We did not include records terminated because o f sampling interval more than 75 d (code 1) or because of abortion
y = X/3+e
where y is a vector of ME milk and fat yields; X is the known incidence matrix;/3 is a vector of fixed effects that includes overall mean, age as a covariate, herd-year, and registration-age group subclasses; and e is a vector of random errors. Effect of herd-year was absorbed.
TABLE 1. Distribution of does by breed and registration.
Breed Alpine LaMancha Nubian Saanen Toggenburg Overall
Purebred
Registration American
4633 846 6815 1479 2412
2032 762 1049 942
16,185
Grade
Total
766
3276 775 3339 1293 1471
9941 2383 11,203 3714 4649
5551
10,154
31,890
Journal of Dairy Science Vol. 70, No. 8, 1987
1712
GIPSON ET AL.
Linear combinations of subdass solutions were constructed to test for effects of registration and age group using the CONTRAST and ESTIMATE statements of PROC GLM in SAS [p. 433 in (11)]. Age was included in the analysis because adjustment factors for calculating ME milk and fat yields might not fully remove effect of age. Each breed-year group combination was analyzed separately because confounding of year groups with herd-year made comparisons across groups impossible. A categorical data analysis, PROC CATMOD in SAS [p. 171 in (11)], was used to evaluate relation of registration with reason for termination of lactation. The model was similar to that for yields except that y was the vector of reason for termination category and/3 included overall mean, age as a covariate, and registration-age group-year group subclasses. Analysis was within breed, and effect of herd-year was not included because of computational difficulties associated with the large number of herd-years (2950 in Alpine, 875 in LaMancha,
3890 in Nubian, 1339 in Saanen, and 1576 in Toggenburg).
RESU LTS A N D D ISCUSSION
Within breed, the largest proportion of production tested does (Table 1) was purebred (36 to 61%), followed by grade (30 to 35%), and American (9 to 33%). Overall, purebreds were 51%, grades 32%, and Americans 17%. From 1976 through 1984, number of does registered as purbred or American or recorded as grade increased for each breed (Table 2). Percentage of purebred and American does increased for each breed, whereas percentage of grade does decreased. Overall, purebred registrations had l~rger increases (from 14.2 to 62.8%) than did American registrations (from 11.4 to 21.6%). From 1976 through 1978, almost 75% of first lactation records were from grade does, but from 1982 through 1984 less than 16% of records were from grades. From 1976 to present, there has been improvement in the
TABLE 2. Distribution of registration by breed and years. Registration American
Grade
Total number
Breed
Years
Purebred
Alpine
1976--1978 1979--1981 1982--1984
14.31 56.5 56.3
14.3 17.0 26.5
71.4 26.5 17.2
2309 3376 4256
LaMancha
1976-- 1978 1979--1981 1982--1984
5.1 33.1 50.8
9.0 45.6 33.3
85.9 21.3 15.9
509 748 1126
Nubian
1976--1978 1979--1981 1982-1984
17.2 70.5 73.9
5.3 8.0 12.1
77.5 21.5 14.0
2370 3540 5293
Saanen
1976-1978 1979-1981 1982-1984
11.7 54.6 45.1
8.8 18.4 36.3
69.5 27.0 18.6
963 1323 1428
1976-1978
1979-1981 1982-1984
13.8 61.8 67.1
13.1 15.9 19.0
73.1 22.3 13.9
1169 1571 1909
1976-1978 1979-1981 1982-1984
14.2 60.1 62.8
11.4 16.0 2!.6
74.4 23.9 15.6
7320 10,558 14,012
(%)
Toggenburg
Overall
i Percentage of breed-years group. Journal of Dairy Science Vol. 70, No. 8, 1987
TABLE 3. Means and standard deviations for mature equivalent milk yield (kg) by breed, years, and registration. Registration Breed
Years
Purebred
Alpine
1976--1978 1979-1981 1982-1984
1023.3 973.3 981.9
321.1 329.3 337.8
1040.9 1018.8 1035.6
LaMancha
1976--1978 1979--1981 1982-1984
821.2 796.5 897.9
374.8 303.9 292.7
Nubian
1976--1978 1979--1981 1982-1984
890.3 804.3 801.3
Saanen
1976-1978 1979-1981 1982-1984 1976--1978 1979--1981 1982--1984
Grade
American
SD
SD
SD
329.1 339.4 358.0
965.9 883.0 941.8
333.8 342.3 322.2
918.4 857.3 884.0
334.2 337,0 329.1
793,2 763.9 910.9
294.0 311.8 303.9
294.5 297,1 290.9
878.2 827.5 804.1
301.9 302.0 293.0
788.2 750.1 806.0
285.7 288.0 299.9
1047.6 983.1 1008.5
345.8 369.6 363~2
964.O 1086.9 1055.5
333.6 401.3 383.2
963.2 929.9 1011.1
357.2 355.7 346.3
1083.8 983.1 992,3
328,7 323.8 317.7
1019.7 1002.7 997.8
337.2 331.7 335,8
982.0 913.7 969.1
334.0 333.9 325.7
,-]
Z > ~z
e~
F Toggenburg
<
Z
O
,g
0 >
e"
t O
~a
TABLE 4. Means and standard deviations for mature equivalent fat yield (kg) by" breed, years, and registration. < O
Registration Z
Breed
Years
Purebred
R
SD
Alpine
1976-1978 1979-1981 1982--1984
35.6 34.2 34.6
11.7 12.7 12.8
36.3 36.2 35.9
11.9 13.0 13.6
33.7 31.3 33.6
12.1 12.8 13.3
LaMancha
1976-1978 1979-1981 1982--1984
30.3 30.1 34.6
14.0 11.4 11.5
33.4 32.4 33.8
11.1 12.7 12.6
30.3 29.3 33.3
11.4 12.5 11.0
Nubian
1976-1978 1979-1981 1982-1984
40.8 37.1 37.5
14.0 14.3 14.4
39.7 37.0 36.6
13.8 14.O 14.1
36.3 34.1 36.1
14.1 13.3 14.3
Saanen
1976--1981 1979--1981 1982--1984
35.7 33.5 33.8
14.4 13.0 12.6
34.0 36.3 35.6
12.4 14.6 13.9
33.7 32.1 35.2
12.9 12.6 12.6
Toggenburg
1976--1978 1979--1981 1982--1984
35.1 31.7 33.0
11.1 11.5 11,7
33.6 32.3 32.5
11.1 11.3 11.8
32.6 29.9 32.7
11.8 11.9 13.7
American
Grade
O
.q
SD
SD
ct~
O Z >
.r-
REGISTRATION AND MILK YIELD IN GOATS
1715
TABLE 5. Regression coefficients (fl, kg/mo) and standard errors for age on mature equivalent (ME) milk and fat yields by breed and year group. Breed
Years
ME Milk
Alpine
1976--1978 1979--1981 1982--1984
LaMancha
1976--1978 1979--1981 1982--1984
--3.9 --16.7 --11.0
Nubian
1976--1978 1979--1981 1982--1984
Saanen
Toggenburg
ME Fat SE
B
SE
5.7 4.6 4.0
--.001 --.020 --.067
.204 .172 .156
10.4 9.4 7.9
--.264 --.306 --.362
.408 .375 .302
--8.7 --10.7"* --9.3**
4.8 4.0 3.2
--.476* --.265 --.378*
.228 .188 .152
1976--1978 1979--1981 1982--1984
--30.8** --4.8 .3
10.0 8.7 8.3
--1.074"* --.196 .175
.362 .299 .293
1976-1978 1979--1981 1982-1984
--10.8 -10.5 4.5
8.0 5.8 6.0
-.344 -.041 .026
.277 .212 .219
--2.1 --5.2 --8.3*
*P<.O5. **P<.01.
TABLE 6. Estimated differences (D) t and standard errors for mature equivalent milk yield by breed, registration, and year group. Breed
Registration
1976-- 1978
Alpine
American Grade
--84.4 --69.6
74.9 68.2
LaMancha
American Grade
205.5 286.1
197.9 175.9
Nubian
American Grade
--89.2 -143.1"*
80.1 55.3
Saanen
American Grade
-191.7 -237.8
144.8 136.1
Toggenburg
American Grade
79.2 -87.6
104.8 96.5
D
SE
1979-1982
1982-1984
D
SE
D
SE
34.9 -78.4*
40.5 36.5
14.8 -12.4
33.2 38.8
131.2 120.0
79.0 88.8
114.0" 47.8
57.8 99.1
48.7 34.8
18.8 34.7
31.1 30.7
23.2 --15.5
86.1 75.1
44.3 140.0
63.5 79.4
31.7 -37.8
51.0 50.0
40.0 60.2
54.2 70.7
95.8* -4.0
t Expressed as deviation from purebreds (kg). *P<.05. **P<.O1. Journal of Dairy Science Vol. 70, No. 8, 1987
1716
GIPSON ET AL.
TABLE 7. Estimated differences (D) ~ and standard errors for mature equivalent fat yield by breed, registration, and year group. Breed
Registration
Alpine
American Grade
LaMancha
1976--1978 D --2.97 -2.86
SE 2.68 2.45
American Grade
7.87 9.56
7.81 6.94
Nubian
American Grade
--9.09* -7.18"
Saanen
American Grade
Toggenburg
American Grade
1979-1982 D 2.28 - 1.34
1982-1984
SE 1.51 1.36
D .28 1.02
SE 1.29 1.51
7.80* 5.49
3.14 3.53
4.58* 1.38
2.22 3.82
3.79 2.62
1.42 -.39
2.29 1.63
.44 .36
1.51 1.47
--5.54 -7.26
5.23 4.91
1.02 .76
2.96 2.59
3.20 7.20*
2.22 2.80
4.53 - 1.42
3.63 3.34
4.27* 1.84
1.82 1.78
1.78 3.87
2.58
2.00
xExpressed as deviation from purebreds (kg). *P<.05.
incorporation of registration information in NCDHIP, this might account for some of the shift in registration of production tested does. Means for ME yield by breed, year, and registration are in Table 3 for milk and in Table 4 for fat. Generally, American does had higher ME milk yield than purebred does, which had higher yields than grades. For ME fat yield, Americans and purebreds had comparable yields but both were higher than grades. Regression of ME milk yield on age (Table 5) was significant and negative for Saanens from 1976 through 1978, for Nubians from 1979 through 1981, and for Alpines and Nubians from 1982 through 1984. Regression of ME fat yield on age (Table 5) also was significant and negative for Nubians and Saanens from 1976 through 1978 and for Nubians from 1982 through 1984. These negative slopes indicate an overadjustment for age for milk and fat yields. Presently used adjustment factors for ME milk and fat yields were calculated using purebred and American does and were from 1976 to 1981. Heterogeneity of slopes (5) for the two age groups was tested but was not significant (P> .05). Linear combinations of registration-age group solutions to estimate differences due to Journal of Dairy Science Vol. 70. No. 8, 1987
registration for ME milk and fat yield are in Tables 6 and 7. Generally, these results agree with means presented in Tables 3 and 4. American does had higher ME milk yield than purebreds, which had higher yield than grades. Ranking differs slightly for ME fat; American does had higher ME fat yield than grades, which had higher than purebreds. Except for ME milk and fat yield for Toggenburgs from 1976 through 1978, age group had no effect (P>.05) on ME milk and fat yields. The interaction of registration with age group was not significant (P>.05) for any breed or time period. There were a large number of contrasts made in this study (five breeds, three year groups, two dependent variables, and four contrasts per analysis). Ten contrasts yielded significant effects due to registration and two to age group. An experiment error rate of .05 would yield six significant results for this same number of contrasts, if none was significant. Distribution of first lactation records by breed, reason for termination, and registration is in Table 8. Normally terminated lactations (code 0) were 68% of total records; sold for dairy purposes (code 2), 23%; and sold because of low yield or other problems (codes 3
REGISTRATION AND MILK YIELD IN GOATS
1717
TABLE 8. Distribution of first lactation records by breed, reason for termination, and registration. Termination code t Breed
Registration
0
2
3--7
(%) Alpine
Purebred American Grade
67.8 u 70.0 65.9
24.2 22.5 22.1
8.0 7.5 12.0
LaMancha
Purebred American Grade
69.7 69.7
22.2 23.8 19.1
8.1 6.5
Purebred American Grade
69.5
63.5 63.1
22.1 26.7 26.4
8.4 9.8 10.5
Saanen
Purebred American Grade
68.7 68.5 67.6
23.8 22.5 21.9
7.5 9.0 10.5
Toggenburg
Purebred American Grade
68.8 65.0 68.5
24.1 26.0 22.2
7.1 9.0 9.3
Overall
Purebred American Grade
68.8 67.8 66.0
23.2 24.0 23.3
8.0 8.3 10.8
Nubian
71.0
9.9
0 = Normal lactation terminated by dry date or 305 days in milk if doe is still in the herd ; 2 = sold for dairy purposes; and 3 - 7 = sold because of low yield or other problems. 2Percentage of breed-registration group.
through 7), 9%. These percentages agree with those of G r o s s m a n and Wiggans (7). F o r each breed, a higher percentage of grade does were sold because of low yield or other problems (codes 3 through 7) than o f purebred or American does. Effect of registration on reason for t e r m i n a t i o n was significant (P<.05) o n l y for Nubians.
In general, however, registration or age group a c c o u n t e d for little variation in first lactation ME milk and fat yields or in reason for t e r m i n a t i o n of first lactation. Differences a m o n g purebred, A m e r i c a n , and grade populations probably are due to m a n a g e m e n t , yearly fluctuations, or trends of increased purebred and A m e r i c a n registrations.
CONCLUSIONS
A C K N O W L E D G M ENTS
Current status of the dairy goat doe population is represented best by data f r o m 1982 through 1984. F o r this period, A m e r i c a n and grade does had higher e x p e c t e d ME milk and fat yields than purebred does. Registration was related significantly to ME milk yield for LaManchas and ME fat yield for LaManchas and Saanens. First lactation records of Nubians and Alpines m a y be overadjusted for age for milk yield; fat yield of Nubians also m a y be overadjusted.
M. R. Dentine, University o f Wisconsin, and S. M. Hubbard, US D e p a r t m e n t o f Agriculture made valuable suggestions for this study. REFERENCES
1 American Dairy Goat Association. 1986. By-laws, article VII: rules for registration and recordation. Page 21 in American Dairy Goat Association guidebook. Spindale, NC. 2 Dentine, M. R., B. T. McDaniel, and H. D. Norman. 1985. Comparison of culling rates, reasons for disJournal of Dairy Science Vol. 70, No. 8, 1987
1718
3
4
5
6
7
GIPSON ET AL.
posal, and yields for registered and grade cows. J. Dairy Sci. 68(Suppl. 1):224. (Abstr.) Dentine, M. R., B. T. McDaniel, and H. D. Norman. 1985. Evaluation of sires for traits associated with longevity of grade and registered daughters. J. Dairy Sci. 68 (Suppl. 1):209. (Abstr.) Finley, C. M., J. R. Thompson, and G. E. Bradford. 1984. Age-parity-season adjustment factors for milk and fat yields of dairy goats. J. Dairy Sci. 67 : 1868. Freund, R. J., and R. C. Littet. 1981. Page 200 in SAS series in statistical applications: SAS for linear models. A guide to the ANOVA and GLM procedures. SAS Inst., Inc., Cary, NC. Gipson, Ti A., M. Grossman, and R. L. Fernando. 1986. Relationships between Predicted Difference for milk production traits and buck designation in dairy goats. J. Dairy Sci. 69:591. Grossman, M., and G. R. Wiggans. 1980. Dairy goat
Journal of Dairy Science Vol. 70, No. 8, 1987
8
9
10
11 12
lactation records and potential for buck evaluation. J. Dairy Sci. 63:1925. Kennedy, B. W., C. M. Finley, E. J, Pollak, and G. E. Bradford. 1981. Joint effects of parity, age, and season of kidding on milk and fat yields in dairy goats. J. Dairy Sci. 64:1707. Norman, H. D., and R. L. Powell. 1983. Use of grade animals in United States Department of Agriculture programs. J. Dairy Sci. 66:1567. Powell, R. L., and H. D. Norman. 1986. Genetic and environmental differences between registered and grade Holstein cows. J. Dairy Sci. 69:2897. SAS Institute, Inc. 1985. SAS User's guide: statistics, version 5 ed. Cary, NC. Wiggans, G. R. 1986. Codes for transfer of NCDHIP data. Natl. Coop. Dairy Herd Improvement Prog. Handbook, Fact Sheet H-11. J. L. Majeskie, ed. Ext. Serv., US Dep. Agric., Washington, DC.
University of Illinois at Urbana-Champaign Urbana 61801
G. R. WIGGANS Animal Improvement Programs Laboratory Agricultural Research Service United States Department of Agriculture
Beltsville, MD 20706
ABSTRACT
production than grades within herd for all breeds except Brown Swiss; however, yield differences were small. Dentine et al. (2) found no significant difference between registered and grade Holsteins in first lactation Modified Contemporary Deviation for milk. Powell and Norman (10) reported that average milk yield generally was higher in 2-yr-old registered Holsteins than in grades within herd and overall. United States Department of Agriculture studies (9, 10) found that culling within herd on yield was more severe for grades than for registered Holsteins. Studies by Dentine et al. (2, 3) indicated that first lactation grade Holsteins were culled more intensely than purebreds. They also found that grade cows were culled more often than registered cows because of low yield and recommended that registration of the daughter should be considered when evaluating sires for culling rate or productive lifespan (2). Several registries are availabile for dairy goats in the United States. Two major registries INTRODUCTION are purebred and American programs of the An important measure of production in the dairy industry is how well a dairy animal per- American Dairy Goat Association (ADGA). forms in her first lactation. Differences in regis- Purebred dairy goats are registered in breed herdbooks and are defined by ADGA (1) as tration of first lactation animals are of interest dairy goats and their linear descendants "imto dairy breeders because of possible relationported from the Eastern Hemisphere that were ships between registration and production. Norman and Powell (9) reported that 2-yr- duly registered in the countries of their origin before their importation; and those goats duly old registered cows (1980 calvings) had greater and correctly registered by ADGA in conformity with its registry rules." These rules specify Received December 29, 1986. that sire and dam must be registered purebreds Accepted April 9, 1987. aContribution of Project Number 35-306 of the of the same breed, except for LaManchas. LaAgricultural Experiment Station, University of Illinois Manchas with three generations of American at Urbana-Champaign, as a part of the work of NC-2, LaMancha ancestry, born after January 1, and supported in part by the University of Illinois 1980, qualify for registration in the purebred Research Board. LaMancha herdbook. Nonpurebred dairy goats 2Reprint requests. First lactation records of 9941 Alpine, 2383 LaMancha, 11,203 Nubian, 3714 Saanen, and 4649 Toggenburg dairy goat does were obtained from the USDA to evaluate relation of registration (purebred, American, or grade) with lactation yield traits (mature equivalent milk and fat yields) and with reason for termination of lactation (normal, sold for dairy, or culled). Does were grouped by age at first kidding (11 through 16 mo or 21 ~hrough 26 mo) and by calendar year (1976 through 1978, 1979 through 1981, or 1982 through 1984). Data were analyzed within breed and year group for yield traits and within breed for reason of termination. Registration and age at first kidding accounted for little variation in mature equivalent milk and fat yields or in reason for termination.
1987 J Dairy Sci 70:1710-1718
1710
REGISTRATION AND MILK YIELD IN GOATS
1711
(code 8) nor records not used in genetic evaluations (code 9). Data were from 9941 Alpine, 2383 LaMancha, 11,203 Nubian, 3714 Saanen, and 4649 Toggenburg does. Of these 31,890 does, 16,185 (51%) were registered as purebred and 5551 (17%) as American; 10,154 (32%) were recorded as grade (Table 1). Records were grouped by age at first kidding (11 thorugh 16 mo or 21 through 26 mo) because o f a reported significant effect of age within parity on lactation yields (4, 8). Due to seasonality of breeding, goats do not normally kid at ages 17 through 20 too, and those that did were excluded as being unrepresentative of does kidding as yearlings or 2-yr-olds. F o r each breed and registration, about 75% of does were 11 through 16 mo of age at first kidding, except Nubian, which had a percentage of 70% for this age group. Records also were grouped into 3-yr intervals (1976 through 1978, 1979 through 1981, or 1982 through 1984) to account for changes in distribution of registration. The model for the least squares analysis of variance used to evaluate relation o f registration with mature equivalent (ME) milk and fat yields was:
that have least 7/8 purebred lineage are included in A D G A American herdbooks (1). Other animals are eligible to be recorded in A D G A grade programs (1). Whether registration is related to lactation yield or to reason for termination of lactation is not known for dairy goats. Grossman and Wiggans (7) examined distribution of lactation records based on reason for termination by breed, but registration and parity were not considered. Gipson et aI. (6) found no significant relationship between registration and buck PD for yield traits. The objective of this study was to evaluate relation of registration with first lactation yield traits and with reason for termination of first lactation for dairy goat does. MATERIALS AND METHODS
First lactation records o f dairy goat does enrolled in the National Cooperative Dairy Herd Improvement Program ( N C D H I P ) f r o m 1 9 7 6 through 1984 were obtained from the Animal Improvement Programs Laboratory, Agricultural Research Service, USDA. Only first lactation records were included because later lactations might be affected by selection. A t least two records per herd-year were required. Reasons for termination of lactation were based on codes reported in records (12): normal lactation terminated by dry date or 305 d in milk if doe is still in the herd (code 0); sold for dairy purposes (code 2); or sold because of low yield or other problems (codes 3 through 7). Termination codes 3 through 7 were combined because some dairy records processing centers do not use all codes. We did not include records terminated because o f sampling interval more than 75 d (code 1) or because of abortion
y = X/3+e
where y is a vector of ME milk and fat yields; X is the known incidence matrix;/3 is a vector of fixed effects that includes overall mean, age as a covariate, herd-year, and registration-age group subclasses; and e is a vector of random errors. Effect of herd-year was absorbed.
TABLE 1. Distribution of does by breed and registration.
Breed Alpine LaMancha Nubian Saanen Toggenburg Overall
Purebred
Registration American
4633 846 6815 1479 2412
2032 762 1049 942
16,185
Grade
Total
766
3276 775 3339 1293 1471
9941 2383 11,203 3714 4649
5551
10,154
31,890
Journal of Dairy Science Vol. 70, No. 8, 1987
1712
GIPSON ET AL.
Linear combinations of subdass solutions were constructed to test for effects of registration and age group using the CONTRAST and ESTIMATE statements of PROC GLM in SAS [p. 433 in (11)]. Age was included in the analysis because adjustment factors for calculating ME milk and fat yields might not fully remove effect of age. Each breed-year group combination was analyzed separately because confounding of year groups with herd-year made comparisons across groups impossible. A categorical data analysis, PROC CATMOD in SAS [p. 171 in (11)], was used to evaluate relation of registration with reason for termination of lactation. The model was similar to that for yields except that y was the vector of reason for termination category and/3 included overall mean, age as a covariate, and registration-age group-year group subclasses. Analysis was within breed, and effect of herd-year was not included because of computational difficulties associated with the large number of herd-years (2950 in Alpine, 875 in LaMancha,
3890 in Nubian, 1339 in Saanen, and 1576 in Toggenburg).
RESU LTS A N D D ISCUSSION
Within breed, the largest proportion of production tested does (Table 1) was purebred (36 to 61%), followed by grade (30 to 35%), and American (9 to 33%). Overall, purebreds were 51%, grades 32%, and Americans 17%. From 1976 through 1984, number of does registered as purbred or American or recorded as grade increased for each breed (Table 2). Percentage of purebred and American does increased for each breed, whereas percentage of grade does decreased. Overall, purebred registrations had l~rger increases (from 14.2 to 62.8%) than did American registrations (from 11.4 to 21.6%). From 1976 through 1978, almost 75% of first lactation records were from grade does, but from 1982 through 1984 less than 16% of records were from grades. From 1976 to present, there has been improvement in the
TABLE 2. Distribution of registration by breed and years. Registration American
Grade
Total number
Breed
Years
Purebred
Alpine
1976--1978 1979--1981 1982--1984
14.31 56.5 56.3
14.3 17.0 26.5
71.4 26.5 17.2
2309 3376 4256
LaMancha
1976-- 1978 1979--1981 1982--1984
5.1 33.1 50.8
9.0 45.6 33.3
85.9 21.3 15.9
509 748 1126
Nubian
1976--1978 1979--1981 1982-1984
17.2 70.5 73.9
5.3 8.0 12.1
77.5 21.5 14.0
2370 3540 5293
Saanen
1976-1978 1979-1981 1982-1984
11.7 54.6 45.1
8.8 18.4 36.3
69.5 27.0 18.6
963 1323 1428
1976-1978
1979-1981 1982-1984
13.8 61.8 67.1
13.1 15.9 19.0
73.1 22.3 13.9
1169 1571 1909
1976-1978 1979-1981 1982-1984
14.2 60.1 62.8
11.4 16.0 2!.6
74.4 23.9 15.6
7320 10,558 14,012
(%)
Toggenburg
Overall
i Percentage of breed-years group. Journal of Dairy Science Vol. 70, No. 8, 1987
TABLE 3. Means and standard deviations for mature equivalent milk yield (kg) by breed, years, and registration. Registration Breed
Years
Purebred
Alpine
1976--1978 1979-1981 1982-1984
1023.3 973.3 981.9
321.1 329.3 337.8
1040.9 1018.8 1035.6
LaMancha
1976--1978 1979--1981 1982-1984
821.2 796.5 897.9
374.8 303.9 292.7
Nubian
1976--1978 1979--1981 1982-1984
890.3 804.3 801.3
Saanen
1976-1978 1979-1981 1982-1984 1976--1978 1979--1981 1982--1984
Grade
American
SD
SD
SD
329.1 339.4 358.0
965.9 883.0 941.8
333.8 342.3 322.2
918.4 857.3 884.0
334.2 337,0 329.1
793,2 763.9 910.9
294.0 311.8 303.9
294.5 297,1 290.9
878.2 827.5 804.1
301.9 302.0 293.0
788.2 750.1 806.0
285.7 288.0 299.9
1047.6 983.1 1008.5
345.8 369.6 363~2
964.O 1086.9 1055.5
333.6 401.3 383.2
963.2 929.9 1011.1
357.2 355.7 346.3
1083.8 983.1 992,3
328,7 323.8 317.7
1019.7 1002.7 997.8
337.2 331.7 335,8
982.0 913.7 969.1
334.0 333.9 325.7
,-]
Z > ~z
e~
F Toggenburg
<
Z
O
,g
0 >
e"
t O
~a
TABLE 4. Means and standard deviations for mature equivalent fat yield (kg) by" breed, years, and registration. < O
Registration Z
Breed
Years
Purebred
R
SD
Alpine
1976-1978 1979-1981 1982--1984
35.6 34.2 34.6
11.7 12.7 12.8
36.3 36.2 35.9
11.9 13.0 13.6
33.7 31.3 33.6
12.1 12.8 13.3
LaMancha
1976-1978 1979-1981 1982--1984
30.3 30.1 34.6
14.0 11.4 11.5
33.4 32.4 33.8
11.1 12.7 12.6
30.3 29.3 33.3
11.4 12.5 11.0
Nubian
1976-1978 1979-1981 1982-1984
40.8 37.1 37.5
14.0 14.3 14.4
39.7 37.0 36.6
13.8 14.O 14.1
36.3 34.1 36.1
14.1 13.3 14.3
Saanen
1976--1981 1979--1981 1982--1984
35.7 33.5 33.8
14.4 13.0 12.6
34.0 36.3 35.6
12.4 14.6 13.9
33.7 32.1 35.2
12.9 12.6 12.6
Toggenburg
1976--1978 1979--1981 1982--1984
35.1 31.7 33.0
11.1 11.5 11,7
33.6 32.3 32.5
11.1 11.3 11.8
32.6 29.9 32.7
11.8 11.9 13.7
American
Grade
O
.q
SD
SD
ct~
O Z >
.r-
REGISTRATION AND MILK YIELD IN GOATS
1715
TABLE 5. Regression coefficients (fl, kg/mo) and standard errors for age on mature equivalent (ME) milk and fat yields by breed and year group. Breed
Years
ME Milk
Alpine
1976--1978 1979--1981 1982--1984
LaMancha
1976--1978 1979--1981 1982--1984
--3.9 --16.7 --11.0
Nubian
1976--1978 1979--1981 1982--1984
Saanen
Toggenburg
ME Fat SE
B
SE
5.7 4.6 4.0
--.001 --.020 --.067
.204 .172 .156
10.4 9.4 7.9
--.264 --.306 --.362
.408 .375 .302
--8.7 --10.7"* --9.3**
4.8 4.0 3.2
--.476* --.265 --.378*
.228 .188 .152
1976--1978 1979--1981 1982--1984
--30.8** --4.8 .3
10.0 8.7 8.3
--1.074"* --.196 .175
.362 .299 .293
1976-1978 1979--1981 1982-1984
--10.8 -10.5 4.5
8.0 5.8 6.0
-.344 -.041 .026
.277 .212 .219
--2.1 --5.2 --8.3*
*P<.O5. **P<.01.
TABLE 6. Estimated differences (D) t and standard errors for mature equivalent milk yield by breed, registration, and year group. Breed
Registration
1976-- 1978
Alpine
American Grade
--84.4 --69.6
74.9 68.2
LaMancha
American Grade
205.5 286.1
197.9 175.9
Nubian
American Grade
--89.2 -143.1"*
80.1 55.3
Saanen
American Grade
-191.7 -237.8
144.8 136.1
Toggenburg
American Grade
79.2 -87.6
104.8 96.5
D
SE
1979-1982
1982-1984
D
SE
D
SE
34.9 -78.4*
40.5 36.5
14.8 -12.4
33.2 38.8
131.2 120.0
79.0 88.8
114.0" 47.8
57.8 99.1
48.7 34.8
18.8 34.7
31.1 30.7
23.2 --15.5
86.1 75.1
44.3 140.0
63.5 79.4
31.7 -37.8
51.0 50.0
40.0 60.2
54.2 70.7
95.8* -4.0
t Expressed as deviation from purebreds (kg). *P<.05. **P<.O1. Journal of Dairy Science Vol. 70, No. 8, 1987
1716
GIPSON ET AL.
TABLE 7. Estimated differences (D) ~ and standard errors for mature equivalent fat yield by breed, registration, and year group. Breed
Registration
Alpine
American Grade
LaMancha
1976--1978 D --2.97 -2.86
SE 2.68 2.45
American Grade
7.87 9.56
7.81 6.94
Nubian
American Grade
--9.09* -7.18"
Saanen
American Grade
Toggenburg
American Grade
1979-1982 D 2.28 - 1.34
1982-1984
SE 1.51 1.36
D .28 1.02
SE 1.29 1.51
7.80* 5.49
3.14 3.53
4.58* 1.38
2.22 3.82
3.79 2.62
1.42 -.39
2.29 1.63
.44 .36
1.51 1.47
--5.54 -7.26
5.23 4.91
1.02 .76
2.96 2.59
3.20 7.20*
2.22 2.80
4.53 - 1.42
3.63 3.34
4.27* 1.84
1.82 1.78
1.78 3.87
2.58
2.00
xExpressed as deviation from purebreds (kg). *P<.05.
incorporation of registration information in NCDHIP, this might account for some of the shift in registration of production tested does. Means for ME yield by breed, year, and registration are in Table 3 for milk and in Table 4 for fat. Generally, American does had higher ME milk yield than purebred does, which had higher yields than grades. For ME fat yield, Americans and purebreds had comparable yields but both were higher than grades. Regression of ME milk yield on age (Table 5) was significant and negative for Saanens from 1976 through 1978, for Nubians from 1979 through 1981, and for Alpines and Nubians from 1982 through 1984. Regression of ME fat yield on age (Table 5) also was significant and negative for Nubians and Saanens from 1976 through 1978 and for Nubians from 1982 through 1984. These negative slopes indicate an overadjustment for age for milk and fat yields. Presently used adjustment factors for ME milk and fat yields were calculated using purebred and American does and were from 1976 to 1981. Heterogeneity of slopes (5) for the two age groups was tested but was not significant (P> .05). Linear combinations of registration-age group solutions to estimate differences due to Journal of Dairy Science Vol. 70. No. 8, 1987
registration for ME milk and fat yield are in Tables 6 and 7. Generally, these results agree with means presented in Tables 3 and 4. American does had higher ME milk yield than purebreds, which had higher yield than grades. Ranking differs slightly for ME fat; American does had higher ME fat yield than grades, which had higher than purebreds. Except for ME milk and fat yield for Toggenburgs from 1976 through 1978, age group had no effect (P>.05) on ME milk and fat yields. The interaction of registration with age group was not significant (P>.05) for any breed or time period. There were a large number of contrasts made in this study (five breeds, three year groups, two dependent variables, and four contrasts per analysis). Ten contrasts yielded significant effects due to registration and two to age group. An experiment error rate of .05 would yield six significant results for this same number of contrasts, if none was significant. Distribution of first lactation records by breed, reason for termination, and registration is in Table 8. Normally terminated lactations (code 0) were 68% of total records; sold for dairy purposes (code 2), 23%; and sold because of low yield or other problems (codes 3
REGISTRATION AND MILK YIELD IN GOATS
1717
TABLE 8. Distribution of first lactation records by breed, reason for termination, and registration. Termination code t Breed
Registration
0
2
3--7
(%) Alpine
Purebred American Grade
67.8 u 70.0 65.9
24.2 22.5 22.1
8.0 7.5 12.0
LaMancha
Purebred American Grade
69.7 69.7
22.2 23.8 19.1
8.1 6.5
Purebred American Grade
69.5
63.5 63.1
22.1 26.7 26.4
8.4 9.8 10.5
Saanen
Purebred American Grade
68.7 68.5 67.6
23.8 22.5 21.9
7.5 9.0 10.5
Toggenburg
Purebred American Grade
68.8 65.0 68.5
24.1 26.0 22.2
7.1 9.0 9.3
Overall
Purebred American Grade
68.8 67.8 66.0
23.2 24.0 23.3
8.0 8.3 10.8
Nubian
71.0
9.9
0 = Normal lactation terminated by dry date or 305 days in milk if doe is still in the herd ; 2 = sold for dairy purposes; and 3 - 7 = sold because of low yield or other problems. 2Percentage of breed-registration group.
through 7), 9%. These percentages agree with those of G r o s s m a n and Wiggans (7). F o r each breed, a higher percentage of grade does were sold because of low yield or other problems (codes 3 through 7) than o f purebred or American does. Effect of registration on reason for t e r m i n a t i o n was significant (P<.05) o n l y for Nubians.
In general, however, registration or age group a c c o u n t e d for little variation in first lactation ME milk and fat yields or in reason for t e r m i n a t i o n of first lactation. Differences a m o n g purebred, A m e r i c a n , and grade populations probably are due to m a n a g e m e n t , yearly fluctuations, or trends of increased purebred and A m e r i c a n registrations.
CONCLUSIONS
A C K N O W L E D G M ENTS
Current status of the dairy goat doe population is represented best by data f r o m 1982 through 1984. F o r this period, A m e r i c a n and grade does had higher e x p e c t e d ME milk and fat yields than purebred does. Registration was related significantly to ME milk yield for LaManchas and ME fat yield for LaManchas and Saanens. First lactation records of Nubians and Alpines m a y be overadjusted for age for milk yield; fat yield of Nubians also m a y be overadjusted.
M. R. Dentine, University o f Wisconsin, and S. M. Hubbard, US D e p a r t m e n t o f Agriculture made valuable suggestions for this study. REFERENCES
1 American Dairy Goat Association. 1986. By-laws, article VII: rules for registration and recordation. Page 21 in American Dairy Goat Association guidebook. Spindale, NC. 2 Dentine, M. R., B. T. McDaniel, and H. D. Norman. 1985. Comparison of culling rates, reasons for disJournal of Dairy Science Vol. 70, No. 8, 1987
1718
3
4
5
6
7
GIPSON ET AL.
posal, and yields for registered and grade cows. J. Dairy Sci. 68(Suppl. 1):224. (Abstr.) Dentine, M. R., B. T. McDaniel, and H. D. Norman. 1985. Evaluation of sires for traits associated with longevity of grade and registered daughters. J. Dairy Sci. 68 (Suppl. 1):209. (Abstr.) Finley, C. M., J. R. Thompson, and G. E. Bradford. 1984. Age-parity-season adjustment factors for milk and fat yields of dairy goats. J. Dairy Sci. 67 : 1868. Freund, R. J., and R. C. Littet. 1981. Page 200 in SAS series in statistical applications: SAS for linear models. A guide to the ANOVA and GLM procedures. SAS Inst., Inc., Cary, NC. Gipson, Ti A., M. Grossman, and R. L. Fernando. 1986. Relationships between Predicted Difference for milk production traits and buck designation in dairy goats. J. Dairy Sci. 69:591. Grossman, M., and G. R. Wiggans. 1980. Dairy goat
Journal of Dairy Science Vol. 70, No. 8, 1987
8
9
10
11 12
lactation records and potential for buck evaluation. J. Dairy Sci. 63:1925. Kennedy, B. W., C. M. Finley, E. J, Pollak, and G. E. Bradford. 1981. Joint effects of parity, age, and season of kidding on milk and fat yields in dairy goats. J. Dairy Sci. 64:1707. Norman, H. D., and R. L. Powell. 1983. Use of grade animals in United States Department of Agriculture programs. J. Dairy Sci. 66:1567. Powell, R. L., and H. D. Norman. 1986. Genetic and environmental differences between registered and grade Holstein cows. J. Dairy Sci. 69:2897. SAS Institute, Inc. 1985. SAS User's guide: statistics, version 5 ed. Cary, NC. Wiggans, G. R. 1986. Codes for transfer of NCDHIP data. Natl. Coop. Dairy Herd Improvement Prog. Handbook, Fact Sheet H-11. J. L. Majeskie, ed. Ext. Serv., US Dep. Agric., Washington, DC.