Feed intake and production in dairy breeds dependent on the ration

Selection applied to populations of dairy cattle has produced a genetic increase in milk production. This will be increased further in the Netherlands by the introduction of Holstein Friesians. In general the high yielding cow is not capable of taking in enough nutrients to meet the requirements for maintenance and milk production. However the knowledge of the variation in feed intake between animals is limited. It requires detailed observations on each cow.The variation in feed intake and production characteristics has been studied mostly on feeding regimes with concentrates fed according to milk production. Both characteristics were confounded in that situation. However, feeding systems with concentrates independent of the milk yield, e.g., a fixed concentrate level for all the individuals, and ad libitum roughage shows a variation in milk yield dependenton the variation in roughage intake, mobilization or deposition of body reserves and/or utilization of nutrients.At present decisions were taken in selection programs m the desirable characteristics of the dairy cow over 10 or 12 years. The present selection is based on performance on a high concentrate level. Enviromnental circumstances such as nutrient supply (roughage/concentrates) may be changed and the mechanisms for regulating the feed intake may vary according to its digestibility. In addition, the import of semen of Holstein Friesians is increasing in the Netherlands. The Dutch Friesians and the Holstein Friesians show a genetic difference for milk production and the subpopulations were selected in different enviromnental circumstances. Reports of the importance of a genotype-ration interaction with different dairy breeds in temperate zones were not found in the literature.This study describes the variation in feed intake (energy and roughage) and production characteristics (milk production and composition, and live weight) for two subpopulations dependent on the ration. The importance of the genotyperation interaction for these characteristics was also tested. These objectives were studied in an experiment over two successive lactations (experiments I and II) and in the dry period between these two experiments.The two subpopulations were characterized as Dutch Friesian (DF) and the crosses between Holstein- and Dutch Friesians (HF). In the first experiment the cows were in their second or later lactation. The analysis of the preceding lactation and the lactation as heifers (concentrates fed according to milk yield) resulted in a contrast between subpopulations of -974 (15.2%) and -813 kg (14.0%) respectively.The rations contained ad libitum roughage (experiment I - hay; dry period and experiment II - grass silage) and the amount of concentrates was independent of the milk production and restricted to the treatment. A low (Roughage group) and a high concentrate level (Concentrate group) were used in the two experiments. The concentrates were allocated over the lactation in three fixed steps (figure 3.1) and the total concentrate intake per lactation for the roughage and concentrate groups was approximately 570 and 2310 kg respectively. During the dry period all cows were offered ad libitum roughage and, in the last 6 weeks before calving, 1 kg concentrates per day. The carry-over effects from the first to the second experiment were studied on a ration change for all the individuals (roughage to concentrate group and the reverse).Full lactation data were analysed from 91 cows in the first experiment and these were allocated over the four genotype-ration groups: DF- Roughage 23, DF-Concentrate 22, HF-Roughage 23 and HF-Concentrate 23. The total number in the second experiment and the dry period was 64 (DF- Roughage 16, DF-Concentrate 18, HF-Roughage 17 and HF-Concentrate 13).Energy and roughage intake. The energy intake was calculated on the roughage and concentrate intake. The roughage dry matter intake was recorded for individual cows on four successive days every 3 or 4 weeks during the lactation and every 2 weeks during the dry period. The roughage offered was analysed weekly for in-vitro digestibility and composition. The concentrate intake was determined daily and the estimated feeding value was 940 VEM kg -1(1 VEM = 6.904 k net energy).The maximum average energy intake for the roughage and concentrate groups was 14631 and 18988 VEM d -1. The energy intake of the roughage groups in the first and second experiments was approximately 3372 (21%) and 4269 VEM d -1(27%) lower than that of the concentrate groups. The carry-over effect was shown in the second experiment but the ration effect of the first experiment was not significant during the dry period. The roughage groups had a longer time and level of underfeeding during the first experiment in comparison with the concentrate groups. A feeding regime with the same concentrate level for all the individuals during the dry period did not indicate compensation by a higher roughage intake. The contrast between rations (Roughage-Concentrate) for roughage dry matter intake in the first and second experiments was 2.9 (22%) (overall mean: 11.7) and 2.1 kg d -1(17%) (overall mean: 11.5) respectively. These results showed that the rations produced different nutritional environments.The genotype effect was significant (P ≤ 0.05) only at the end of both experiments and the contrasts between DF and HF groups for energy intake in the first and second experiments were -408 (3%) and -516 VEM d -1(3%) respectively. The contrasts for the roughage dry matter intake were -0.5 (4%) and -0.6 kg d -1(5%) respectively.The interaction between genotype and ration was not significant (P>0.05) during the two experiments and the dry period. The coefficient of variation (adjusted for number of lactation, season and days open) for roughage dry matter intake during the two experiments was approximately 7% with the lowest values in the mid stage of lactation. This stage had also the highest repeatabilities between the experiments (+0.49 - +0.74 for the four genotype-ration groups).Milk production and components. The milk production and composition for each cow were determined once a week. The adjusted coefficient of variation for total milk yield (app. 10%) was lower in the first experiment than in the preceding lactation (app. 14%) (table 5.1). The contrasts between the two subpopulations were also smaller. In the first experiment this contrast was 594 kg (10.1%) (overall mean: 5596) and in the second 407 kg (7.2%) (overall mean: 5416) in favour of the HF group. The HF group showed a higher persistency in production during the experiments. However the smaller peak yield contrast combined with greater persistency was still not sufficient to reach the differences expected between genotypes in total yield. The allocation of the concentrates over the cows (dependent or independent of milk production) in feeding systems with ad libitum roughage has an influence on the coefficient of variation in total milk yield. The expression of the differences between subpopulations in total milk yield and the phenotypic variation was smaller in a system of a fixed concentrate level for individuals independent of the milk yield.The genotype contrasts (DF-HF) (10 g kg -1d -1) in the two experiments were, for milk fat, +0.21 and +0.32 respectively and for milk protein, +0.08 and +0.05. These results caused the contrasts for the milk energy equivalence (FPCM = (0.349 + 0.107 Milk Fat percentage + 0.067 Milk Protein percentage)x Milk Yield) to be smaller than for milk yield (7.7% and 3.7%).The ration had a clear effect on the milk yield and fat protein corrected milk yield (FPCM) in the two experiments (app. 20% - 1250 kg). As a result of the difference in milk fat to milk protein ratio between the rations during the lactation the contrasts for FPCM were smaller in the early stage of the lactation compared with the milk production, but at the end the position was reversed. The average components over the total lactation were not significantly (P>0.05) affected by the ration. A carry-over effect for milk yield was shown at the beginning of the second experiment but it was compensated at the end (table 5.2).The genotype-ration interaction was not significant (P>0.05) for the total lactation milk production characteristics and generally also during the lactation.Live weight change. The live weight of the individuals during the first experiment was accurately fitted by a model with the following parameters: level, time of minimum live weight during the lactation, maximum live weight losses and a pregnancy parameter. The maximum live weight losses (including variation per group) was clearly influenced by the ration (DF- Roughage -71 kg, DF-Concentrate -41 kg, HF-Roughage -64 kg, HF-Concentrate -50 kg). The cows on the low concentrate level had a greater live weight loss and for a longer time of the lactation than the cows on the high concentrate level. The live weight level and the maximum losses were more highly correlated for the roughage groups than for the concentrate groups (Roughage: -0.79 and -0.57; Concentrate: -0.22 and -0.38).A carry-over effect was clearly shown in the second experiment. The live weight change in weeks 1-12 in the two experiments were, for the roughage groups, -66 and -71 kg respectively and for the concentrate groups -37 and -16 kg. Differences between ration groups at the end of the first experiment were not compensated in the dry period. The effect of the genotype and the interaction between genotype and ration were small and not significant (P>0.05) during the experiments.Difference between energy intake and requirement . This characteristic was based on the energy intake and the requirements for maintenance and milk production. A deficiency in energy existed at the beginning of the lactation. The energy equilibrium in the first experiment was reached between weeks 6 and 9 and weeks 9 and 12 for the concentrate and roughage rations respectively. The second experiment showed a carry-over effect for the time of reaching energy equilibrium. The energy deficiency resulted in a high ratio of FPCM yield to energy input at the beginning of the lactation and the lack had to be compensated by the mobilization of body reserves. No close relationship between live weight losses and the difference between energy intake and requirement was observed. This was in accordance with the literature.Relationships between production characteristics. The analysis of the relations between characteristics was confined to energy supply and demand processes namely fat protein corrected milk, live weight, live weight change and energy intake. The coefficients of correlation and a multiple regression analysis with (FPCM as dependent variable were calculated within a genotype-ration group.The general relation between live weight, (FPCM and energy intake during the lactation in the first experiment per genotype-ration group was presented in figure 4.19. These curves were based on a model with the parameters mentioned in the section on live weight change.The roughage groups showed a closer relation between FPCM yield and energy intake than the concentrate groups (correlation coefficients total lactation: +0.78 and +0.58). The relation between live weight change and FPCM also showed a. difference between the rations but the values of the concentrate groups were higher than those of the roughage groups. In summary the relations were more dependent on the ration than on the genotype. The concentrate level and the allocation of concentrates to individuals have an influence on the relationships. Knowledge of the genetic parameters is only available for systems of feeding concentrates according to milk yield. Parameter estimates on different feeding systems and the relationships between characteristics of the performance test of the young bull and his lactating daughers are necessary for predicting the response to selection in different environments.In addition to these analyses the coefficients of correlation between these charachteristics and the ratio FPCM to energy intake (efficiency) were presented in the discussion. The efficiency was highly correlated with FPCM (a tendency for a ration difference) but the values for efficiency and energy intake were low. However, the economic importance of the energy intake is in the exchange between roughage and concentrates. This was illustrated by two price ratios per net energy from roughage and concentrates. The same prices for both sources showed a difference between milk production returns and feed costs per cow in favour of the high concentrate ration. A price ratio of 1 : 2 (roughage : concentrates) resulted in the reverse ranking of the two rations