C.P. Van Tassell 1, * G.R. Wiggans 1, J.C. Philpot 1, and I. Misztal Animal Improvement Programs Laboratory Agricultural Research Service, USDA, Beltsville, MD 2 2 University of Georgia, Athens 2002 Improvements in Dystocia National Genetic Evaluation System and Data Processing
ADSA 2002 (2) C.P. Van Tassell 2002 History 1977 – NAAB-sponsored evaluations within AI organizations 1980 – NAAB-sponsored evaluations across AI organizations 1988 – Ordered categorical threshold model 1990 – Semiannual evaluations 1999 – Processing change from Iowa State University to USDA (November) 2002 – Sire-maternal grandsire (MGS) model
ADSA 2002 (3) C.P. Van Tassell 2002 Improving Calving Ease Evaluation Data quality Increase MGS identification Improve detection of duplicate records Genetic evaluation system Add MGS effect to threshold model Estimate parameters
ADSA 2002 (4) C.P. Van Tassell 2002 Calving Ease Scores 1 = No problem during calving or calving not observed 2 = Slight problem during calving 3 = Moderate difficulty during calving 4 = Considerable force used during calving 5 = Extreme difficulty during calving
ADSA 2002 (5) C.P. Van Tassell 2002 Frequency of Calving Difficulty Lactation Calving Ease Score 1 st 2 nd 3 rd 163%80%81% 214%10%9% 314%7% 45%2% 53%1%
ADSA 2002 (6) C.P. Van Tassell 2002 Data 10 million calving reports since 1980 Pedigree data From production evaluation system Increased MGS ID from 57% to 73% Additional 2.5 million pedigree records for cows without unique national identification Allowing inclusion of data that does not meet requirements for production
ADSA 2002 (7) C.P. Van Tassell 2002 Data Extraction Requirements Calving Data Single birth Dam and sire of Holstein or Red & White breed Sire age >18 mo and <15 yr Birth date of Jan. 1, 1980, or later
ADSA 2002 (8) C.P. Van Tassell 2002 Sire – MGS Model MGS effect included Evaluations of bulls as service sires improved by partially accounting for mates Maternal effect on calving difficulty provided
ADSA 2002 (9) C.P. Van Tassell 2002 Threshold model Scores assumed to be observations on a continuous underlying scale Thresholds estimated to relate observed scores to underlying scale Procedure allows for differences in amount of change between consecutive scores Observed scores Underlying scale
ADSA 2002 (10) C.P. Van Tassell 2002 Calving Ease Evaluation Statistics (Sire-MGS Model) Range Effect LevelsMinimumMaximum Herd-year 204, Year-season Parity-gender Sire birth year MGS birth year Sire 120, MGS 120,
ADSA 2002 (11) C.P. Van Tassell 2002 Presentation of Evaluations Evaluations converted to expected percentage of births that would be scored 4 or 5 (difficult) for heifers (%DBH)
ADSA 2002 (12) C.P. Van Tassell 2002 %DBH Distributions
ADSA 2002 (13) C.P. Van Tassell 2002 Changes in %DBH
ADSA 2002 (14) C.P. Van Tassell 2002 Sire and MGS %DBH Distributions
ADSA 2002 (15) C.P. Van Tassell 2002 Reliability Calculated as inverse of diagonal Overestimated due to assumption of perfect accuracy of parents
ADSA 2002 (16) C.P. Van Tassell 2002 Variance Component Estimation Required sire and MGS be among 2600 most common Required 20 records/herd-year Created 5 distinct random sub-samples of ~200,000 records based on herd code
ADSA 2002 (17) C.P. Van Tassell 2002 Variance Component Estimation Min Max Mean Selected
ADSA 2002 (18) C.P. Van Tassell 2002 Status of Evaluation System Test evaluations distributed to AI organizations Web queries to display calving ease data available First release of sire-MGS evaluations for calving ease in August 2002