| Lactation curves and mathematical models of dairy cattle are one of effective methods accurately to describe rules of milk yield against time. Research in lactation curves and mathematical models is of great importance in genetics and breeding,feeding and management etc.of dairy cattle.The fitting effect of Wood model that was tabled by Wood in 1967 was studied by the numbers used different grouping and sampling data of Chinese Holstein Cattle,.Heritaility, genetic correlation, phenotypic correlation and influencing factors of parameters of Wood model were estimated and analyzed, and the effect of 305-day milk yield and monthly milk yield were forecast according to different Wood models. The main results of the present study were as following:1.There has different fitting effect of Wood models using different lactation data. The analysis of fitting lactation curves using entirely lactation data (1～10 months milk yield data) indicate that, it is not good fitting effect for individual, the percent of lactations of R2≥0.90 are just 5.82%(test-day models) and 42.80% (month mean models), and there have several atypical lactation curves;It is better fitting effect for herds,R2 of test-da models and month mean models are 96.13% and 96.75% respectively; There also have better fitting effect for grouping models,R2 of test-day models and month mean models are 0.9068～0.9764 and 0.9237～0.98 respectively in grouped by bull families, parities, years of freshing seasons of freshing, houses, 305-day milk yield and peak milk yield. but there are no significant difference in parameters of models. The analysis of fitting effect of lactation curves using part lactation data (1～4 months milk yield data) indicate that the grouping month mean models (R2: 0.8876～0.9950) are more better fitting effect than test-day models(R2:0.5195～0.9957), and R2 of test-da models are very instability and there are significant difference in parameters of models.2. There are significant affect in bull families, parities, yeare of freshing, seasons of freshing,houses to parametere A, b, c of test-da tnodels besides mp of bohing tO A, and in all bo tO tOparameters A, b, c of month mean modeis besideS parities, years of heshing, seasons of fhahiDg toA and M of freshing tO c. The the of bo boons to modls parambo are gewtlySInailer sighfiCan de are jUSt in ~esx bull families interactions tO b, c, Paritiesx seasons ofbohing, bul families X hoUSeS, years Of ffe~x seasons of ffeshing, howt X seasons of eqx years of beshing, Parities x buIl bolies x seasons of ffeshing x years of wtng intendons tob in teSttw models, and in Parihesxbull families, parities Xseasons of ffeshing, bull familiesxhouses, interaCtions tO b, parities X yearS of bohing interaCtions to A in month mean models. TheWt deN of M models are more than month mean models.3. The heritability of parameters of teSt~ and month mean models in complete herds aregenetally lowef, the heritaility of A, b, c, PT P, PM of test-day models are 0.l l25, 0.l094, 0.l718,0.0976, 0.l589, 0.0596 reSPeCtively, jUSt c is signifiCan (P<0.05). The heritability of pmbo ofmonth mean modls are 0.l403, 0.075l, 0.l573, 0.l581, 0.l525, 0.0775 res~vely, bot there hasno swifiCan in ParameteI'S. The twis of poeters of genetic cormlation and PhenOtyPiccormlallon indibe that, for hav sighfiCan genedc cormlation foreen A and b, P; b and c; c andv, rr PM; pp tal P; anu ail pnenotwic cormlation boiaes A anu b; rv nd PM in test'ar mwtls.Thot are all signifiCan cormlation besides pboic correlation between A and b; PT and PM, theposthe and wt Symbol of the coefficient of genetic the phenOtyPic corrlallon are basicallysame. The p-- of mOnth mean models seem more heritability than testday models.4. The 305wi milk yields Of herds have been foreCaS using testtw and month mean modlsof COmlete and pat ha The relboly MPE are 8.00O/ and 4.04% of Part bo and 0.038O/o and0.03l% of Wlere data, the M Of models of comlete h...
|
PDF Full Text Request