HERD TESTING.

Taranaki Daily News, 20 July 1933, Page 14

HERD TESTING.

(To the Editor.) Sir,—ln your issue this morning appears a letter over the non de plume of “Puzzled” in which the writer , seeks information regarding comparison of factory and herd-testing returns. The letter being dated from Hawera, you might permit me to amplify your reply as published with the letter. Herd testing was originally designed to point out the comparative value of one cow with another in the same herd. It was never intended as a check on factory figures, or vice versa. The herd testing sample is taken direct from the cow before it has been affected by any treatment whatever; the sample is also taken for an assumed 24 hour period, and on this basis is calculated for 30 days. The result is therefore based on average, whereas the factory sample is taken daily and such sample accurately represents the milk, or cream, actually supplied to the factory after the various treatments it has received on the farm. It is during the period between the milking of the individual cows and the delivery of the milk, or cream, to the factory that the greatest difference between the two sets of figures is brought about. Some of the causes for difference are as follows: —(a) Losses through spilling; (b) calf'and house milk, and cream; (c) separator losses; (d) “beastings” included in’ association returns but not in factory; (e) loss of the cream “skin” forming on night’s milk; (f) incorrect calving dates given to association, or none at all, calling for approximations, similarly with drying off dates; (g) drinking by cats and dogs; (h) addition of water either as rain or as flushings. Most of these occur on every farm, and there are other causes which occur (often unknown to the farmer) which will greatly increase the difference between the two sets of figures. Incorrect herd management has, for instance, a very great bearing on the subject. It will be readily understood that it is possible for association and factory figures to show a great variation, but such fact does not affect ■ the value of the association returns for the individual cows which still, show—for culling and breeding purposes—the best and worst cows in accurate comparison, as all have been tested under the same conditions. Notwithstanding the number of avenues for discrepancy it is the exception rather than the rule to find variations of over 10 per cent, after the known causes for difference, such as calf and house milk, have been allowed. In most cases where the testing farmer has treated testing as a matter of routine and made no efforts to “boost” or otherwise vary normal procedure the variations will be 5 per cent., or even lower. In this connection it might be as well to mention that herd testing associations are careful to watch' that individual records are as ( correct as possible—checking association totals with factory returns —when issuing production certificates or accepting records for calf marking or bull registration.—l am, etc., E. M. BARDSLEY, (Sec. S. Taranaki Herd Testing Assn.) Hawera, July 18.