|
|
||||||||
Animal Husbandry Research Division, ARS, USDA, Beltsville, Maryland 20705
ABSTRACT
A total of 214,400 different estimates of lactation milk yield was obtained by the Test Interval Method from 536 actual lactations of Holstein cows by simulating five sampling plans, five test intervals, and 16 time spans from calving to first test per lactation. The five sampling plans were Official Dairy Herd Improvement, alternate am-pm, alternate pm-am, all-am, and all-pm. The test intervals were 20, 30, 40, 50, and 60 days. Time spans from calving to first test varied from three to 83 days. Daily milking intervals in this herd were 11 hours (am to pm) and 13 hours (pm to am). Important differences in bias and random variation in deviations of estimated lactation milk yield from true yield (obtained from summing individual milking weights) were caused by sampling plan, test interval, time span from calving to first test, and daily milking interval.
It is apparent that milk yields estimated from alternate am-pm sampling plans will contain bias and random error greater than that found in the Official Dairy Herd Improvement plan, unless appropriate controls and adjustments can be incorporated into the alternate sampling plans to decrease errors. Further investigation into the causes of these increased biases and random errors and development of suitable controls and adjustments are needed.
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |