Errores de predicción de la producción de leche por lactancia a partir de pesajes a distintos intervalos de tiempo en vacas de doble propósito

摘要

Therecordsusedforthisstudyreferto902lactationsof615cows,mainly(96%)BostaurusxBosindicuscrossbreds,keptonthreedualpurposefarmslocatedinGuáricoState,Venezuela.ManagementwastypicalofthedualpurposesystemintheAmericantropics:milkingbyhandoncedailywithcalfsuckling,andgrazingwithlittleornosupplementation..Dailymilkrecordswereusedtoestimatetotalproduction,simulatingdifferentsimple(oneday/interval)orcombined(twoormoreconsecutivedays/interval)samplingfrequencies.Theabsolutevalue(withoutsign)ofthedifferencesbetweentheestimatedandactualproductionlevelswerecalculatedtodeterminetherealsizeoftheerror.Resultswereexpressedinunits(kg)andas%ofthemeanlevelofmilkyield.Theerrors(withsigns)werethensubjectedtoananalysisofvariancetodeterminetheeffectsofherd(1,4,5),breedgroup(?,?and?Europeancrossbreds),andlactationnumber(1,2-4,5-10).Unadjustedmeanyieldwas904±427kg,withameandurationof235±79days.Simpleweekly(W),two-weekly(TW)andmonthly(M)samplesgeneratedmeanerrorsof2.3%,3.9%and6.9%,respectively.However,therangeofextremeerrorswaswide,reaching21.9%(W),50.6%(TW)and53.8%(M).Combinedsamplingimprovedtheprecisionforeachinterval,generatingmeanerrorsof1.5%,1.2%;3.2%,2.9%;6.3%,5.9%and5.6%for2and3days/week(2W,3W);2and3days/twoweeks(2TW,3TW);2,3and4days/month(2M,3M,4M),respectively.Therangesofextremeerrorsforthesesamplingswere12.5%(2W),9.3%(3W),32.7%(2TW),33.0%(3TW),52.2%(2M),46.6%(3M)and40.1%(4M).Ingeneral,nocombinedsamplingforagivenintervalexceededorevenreachedtheprecisionobtainedfromsamplingcarriedoutatashorterinterval.Variationintheerrorswasobservedduetoherd(W,2M,3M,4M),breedgroup(M)andlactationnumber(TW,2TW,M).Howevertheeffectswerenotsufficientlyimportanttochangethedecisionastowhichintervalshouldbeused.ItwasconcludedthatasimpleTWintervalisthemaximumacceptableforsamplingintraditionallymanageddualpurposeherds.However,ifitisrequiredtoreducethewidevariationintheerrors,thecombinedTWandWorsimpleWsamplingsarenecessary,althoughthisincreasesrecordingcosts.Furtherstudiesarerecommendedtoexplorepracticaloptionsformilkrecordingundertheseconditions.