Associations of breed and feeding management with milk production curves at herd level using a random regression test-day model

Earlier studies identified large between-herd variation in estimated lactation curve parameters from test-day milk yield and milk composition records collected in Ragusa province, Italy. The objective of this study was to identify sources of variation able to explain these between-herd differences in milk production curves, by estimating associations of animal breed (Holstein Friesian vs. Brown Swiss), feeding system [separate feeding (SF) vs. total mixed ration (TMR)], and TMR chemical composition on milk and milk components herd curves. Data recorded from 1992 through 2007 for test-day (TD) milk, fat, and protein yields from 1,287,019 records of 148,951 lactations of 51,489 cows in 427 herds were processed using a random regression TD model. Random herd curves (HCUR) for milk, fat, and protein yields were estimated from the model per herd, year, and parity (1, 2, and 3+) using 4-order Legendre polynomials. From March 2006 through December 2007, samples of TMR were collected every 3 mo from 37 farms in Ragusa province. Samples were analyzed for dry matter, ash, crude protein, soluble nitrogen, acid detergent lignin, neutral detergent fiber, acid detergent fiber, and starch. Traits used to describe milk production curves were peak, days in milk at peak, persistency, and mean. Association of feeding system and animal breed with HCUR traits was investigated using a general mixed model procedure. Association of TMR chemical composition with HCUR traits was investigated using multivariate analysis with regression and stepwise model selection. Results were consistent for all traits and parities. Feeding system was significantly associated with HCUR peak and mean, with higher values for TMR. Animal breed was significantly associated with HCUR persistency, with higher values for Brown Swiss herds. Furthermore, animal breed influenced HCUR peak and mean, with higher values for Holstein Friesian herds. Crude protein had the largest effect on HCUR peak and mean, whereas the interaction between crude protein and dry matter mainly affected persistency. When provided by a national evaluation system, HCUR can be used as an indicator of herd feeding management.     

Variance components for test-day milk, fat, and protein yield, and somatic cell score for analyzing management information

Test-day (TD) models are used in most countries to perform national genetic evaluations for dairy cattle. The TD models estimate lactation curves and their changes as well as variation in populations. Although potentially useful, little attention has been given to the application of TD models for management purposes. The potential of the TD model for management use depends on its ability to describe within- or between-herd variation that can be linked to specific management practices. The aim of this study was to estimate variance components for milk yield, milk component yields, and somatic cell score (SCS) of dairy cows in the Ragusa and Vicenza areas of Italy, such that the most relevant sources of variation can be identified for the development of management parameters. The available data set contained 1,080,637 TD records of 42,817 cows in 471 herds. Variance components were estimated with a multilactation, random-regression, TD animal model by using the software adopted by NRS for the Dutch national genetic evaluation. The model comprised 5 fixed effects [region × parity × days in milk (DIM), parity × year of calving × season of calving × DIM, parity × age at calving × year of calving, parity × calving interval × stage of pregnancy, and year of test × calendar week of test] and random herd × test date, regressions for herd lactation curve (HCUR), the animal additive genetic effect, and the permanent environmental effect by using fourth-order Legendre polynomials. The HCUR variances for milk and protein yields were highest around the time of peak yield (DIM 50 to 150), whereas for fat yield the HCUR variance was relatively constant throughout first lactation and decreased following the peak around 40 to 90 DIM for lactations 2 and 3. For SCS, the HCUR variances were relatively small compared with the genetic, permanent environmental, and residual variances. For all the traits except SCS, the variance explained by random herd × test date was much smaller than the HCUR variance, which indicates that the development of management parameters should focus on between-herd parameters during peak lactation for milk and milk components. For SCS, the within-herd variance was greater than the between-herd variance, suggesting that the focus should be on management parameters explaining variances at the cow level. The present study showed clear evidence for the benefits of using a random regression TD model for management decisions.     

Use of herd solutions from a random regression test-day model for diagnostic dairy herd management

In a random regression test-day model, environmental effects in addition to individual animal factors can be included and analyzed. Moreover, instead of herd-year classification of the management groups, the herd-test-day classification within the model better accounts for month-to-month short-term environmental variation in production and somatic cell count (SCC) traits. The herd management levels of milk yield (milk deviation from whole-country mean, kilograms/day), protein and fat concentration (protein and fat deviation, %), and SCC (SCC deviation, 1,000 cells/mL) are used in the dairy herd management Web application “Maitoisa” (in English, “Milky”). This management tool helps to recognize several management problems. For recognition of systematic patterns and single unusual test-days, a monthly time-trend analysis was developed to smooth the random fluctuations and display the yearly production pattern. In addition to analyzing single test-day deviations from the mean, modeled herd solutions assist users in identifying repeated phenomena and enable the forecasting of the management pattern for the subsequent year. The solutions are displayed either as tables or graphs plotted by calendar months. In addition to management effects of the farmer's own herd, he or she can request country or region percentiles to be displayed in the graphs. The Web service has been offered to farmers and dairy advisors since 2001, and it has proved to be a powerful tool for herd monitoring and planning.     

Prediction of daily milk, fat, and protein production by a random regression test-day model

Test-day genetic evaluation models have many advantages compared with those based on 305-d lactations; however, the possible use of test-day model (TDM) results for herd management purposes has not been emphasized. The aim of this paper was to study the ability of a TDM to predict production for the next test day and for the entire lactation. Predictions of future production and detection of outliers are important factors for herd management (e.g., detection of health and management problems and compliance with quota). Because it is not possible to predict the herd-test-day (HTD) effect per se, the fixed HTD effect was split into 3 new effects: a fixed herd-test month-period effect, a fixed herd-year effect, and a random HTD effect. These new effects allow the prediction of future production for improvement of herd management. Predicted test-day yields were compared with observed yields, and the mean prediction error computed across herds was found to be close to zero. Predictions of performance records at the herd level were even more precise. Discarding herds enrolled in milk recording for <1 yr and animals with very few tests in the evaluation file improved correlations between predicted and observed yields at the next test day (correlation of 0.864 for milk in first-lactation cows as compared with a correlation of 0.821 with no records eliminated). Correlations with the observed 305-d production ranged from 0.575 to 1 for predictions based on 0 to 10 test-day records, respectively. Similar results were found for second and third lactation records for milk and milk components. These findings demonstrate the predictive ability of a TDM.     

Genetic analysis of milk production traits of polish black and white cattle using large-scale random regression test-day models

Genetic parameters for milk, fat, and protein yield and persistency in the first 3 lactations of Polish Black and White cattle were estimated. A multiple-lactation model was applied with random herd-test-day effect, fixed regressions for herd-year and age-season of calving, and random regressions for the additive genetic and permanent environmental effects. Three data sets with slightly different edits on minimal number of days in milk and the size of herd-year class were used. Each subset included more than 0.5 million test-day records and more than 58,000 cows. Estimates of covariance components and genetic parameters for each trait were obtained by Bayesian methods using the Gibbs sampler. Due to the large size and a good structure of the data, no differences in estimates were found when additional criteria for record selection were applied. More than 95% of the genetic variance for all traits and lactations was explained by the first 2 principal components, which were associated with the mean yield and lactation persistency. Heritabilities of 305-d milk yield in the first 3 lactations (0.18, 0.16, 0.17) were lower than those for fat (0.12, 0.11, 0.12) and protein (0.13, 0.14, 0.15). Estimates of daily heritabilities increased in general with days in milk for all traits and lactations, with no apparent abnormalities at the beginning or end of lactation. Genetic correlations between yields in different lactations ranged from 0.74 (fat yield in lactations 1 and 3) to 0.89 (milk yield in lactations 2 and 3). Persistency of lactation was defined as the linear regression coefficient of the lactation curve. Heritability of persistency increased with lactation number for all traits and genetic correlations between persistency in different lactations were smaller than those for 305d yield. Persistency was not genetically correlated with the total yield in lactation.     

Genetic parameter estimates of portuguese dairy cows for milk, fat, and protein using a spline test-day model

A spline animal model was fitted to 152,103 test-day milk, fat, and protein yield records from 14,423 first-lactation cows. The models included age at calving and the herd-test-month as fixed effects. Model fitting was carried out using Restricted Maximum Likelihood with ASREML. For milk yield, the heritability at 18 d in milk was 0.19, which increased to the maximum estimated value of 0.23 at midlactation and then decreased. On average, milk, fat, and protein yield heritabilities were 0.22, 0.14, and 0.19, respectively.For milk yield, all correlations were positive and ranged from 0.54 to 0.99 for the genetic component and from 0.32 to 0.78 for the phenotypic component. Genetic correlations were higher than phenotypic ones. For fat and protein yields, all genetic correlations were positive, ranging from 0.43 to 0.99. The phenotypic correlations for fat yield had the lowest correlations of the 3 traits.Curves of estimated breeding values for milk, fat, and protein over lactation had positive deviations from mean curves for sires with high genetic merit, but there was considerable variability in the shapes of the curves for different sires. More research is needed to compare the spline function with other mathematical functions used as submodels of lactation curve.     

Comparison of random regression test-day models for Polish Black and White cattle

Test-day milk yields of first-lactation Black and White cows were used to select the model for routine genetic evaluation of dairy cattle in Poland. The population of Polish Black and White cows is characterized by small herd size, low level of production, and relatively early peak of lactation. Several random regression models for first-lactation milk yield were initially compared using the “percentage of squared bias” criterion and the correlations between true and predicted breeding values. Models with random herd-test-date effects, fixed age-season and herd-year curves, and random additive genetic and permanent environmental curves (Legendre polynomials of different orders were used for all regressions) were chosen for further studies. Additional comparisons included analyses of the residuals and shapes of variance curves in days in milk. The low production level and early peak of lactation of the breed required the use of Legendre polynomials of order 5 to describe age-season lactation curves. For the other curves, Legendre polynomials of order 3 satisfactorily described daily milk yield variation. Fitting third-order polynomials for the permanent environmental effect made it possible to adequately account for heterogeneous residual variance at different stages of lactation.     

Genetic parameters for tunisian holsteins using a test-day random regression model

Genetic parameters of milk, fat, and protein yields were estimated in the first 3 lactations for registered Tunisian Holsteins. Data included 140,187; 97,404; and 62,221 test-day production records collected on 22,538; 15,257; and 9,722 first-, second-, and third-parity cows, respectively. Records were of cows calving from 1992 to 2004 in 96 herds. (Co)variance components were estimated by Bayesian methods and a 3-trait-3-lactation random regression model. Gibbs sampling was used to obtain posterior distributions. The model included herd × test date, age × season of calving × stage of lactation [classes of 25 days in milk (DIM)], production sector × stage of lactation (classes of 5 DIM) as fixed effects, and random regression coefficients for additive genetic, permanent environmental, and herd-year of calving effects, which were defined as modified constant, linear, and quadratic Legendre coefficients. Heritability estimates for 305-d milk, fat and protein yields were moderate (0.12 to 0.18) and in the same range of parameters estimated in management systems with low to medium production levels. Heritabilities of test-day milk and protein yields for selected DIM were higher in the middle than at the beginning or the end of lactation. Inversely, heritabilities of fat yield were high at the peripheries of lactation. Genetic correlations among 305-d yield traits ranged from 0.50 to 0.86. The largest genetic correlation was observed between the first and second lactation, potentially due to the limited expression of genetic potential of superior cows in later lactations. Results suggested a lack of adaptation under the local management and climatic conditions. Results should be useful to implement a BLUP evaluation for the Tunisian cow population; however, results also indicated that further research focused on data quality might be needed.     

An autoregressive repeatability animal model for test-day records in multiple lactations

Test-day (TD) models are becoming a standard for genetic evaluation of production traits in dairy cattle. Various approaches to model covariances between TD records include random regression, autoregressive repeatability, orthogonal polynomials, and models based on character processing. The applicability of these models is mainly associated with the number of parameters to estimate, incorporation of multiple lactations, and the accuracy of correlations generated by the cow's repeated expression of milking performance (TD yields) within and across lactations. We define and evaluate a multiple-lactation, autoregressive-repeatability model that disentangles environmental effects due to cow within and between lactations. Simulated records either included or ignored a long-term environmental effect between lactations. Our autoregressive TD animal model correctly detected presence and the absence of this effect and accurately recovered the assumed variance components and correlations underlying the data (10 parameters for three lactations). Estimates of variance components and autocorrelation coefficients were obtained using DFREML-simplex methodology. Given the value of this approach to reduce the size of residual variance components, autoregressive animal models are a preferable alternative to classical methods based on cumulative lactation yield to improve milk production in dairy cattle.     

Autoregressive single-step test-day model for genomic evaluations of Portuguese Holstein cattle

The multiple-lactation autoregressive test-day (AR) model is the adopted model for the national genetic evaluation of dairy cattle in Portugal. Under this model, animals' permanent environment effects are assumed to follow a first-order autoregressive process over the long (auto-correlations between parities) and short (auto-correlations between test-days within lactation) terms. Given the relevance of genomic prediction in dairy cattle, it is essential to include marker information in national genetic evaluations. In this context, we aimed to evaluate the feasibility of applying the single-step genomic (G)BLUP to analyze milk yield using the AR model in Portuguese Holstein cattle. In total, 11,434,294 test-day records from the first 3 lactations collected between 1994 and 2017 and 1,071 genotyped bulls were used in this study. Rank correlations and differences in reliability among bulls were used to compare the performance of the traditional (A-AR) and single-step (H-AR) models. These 2 modeling approaches were also applied to reduced data sets with records truncated after 2012 (deleting daughters of tested bulls) to evaluate the predictive ability of the H-AR. Validation scenarios were proposed, taking into account young and proven bulls. Average EBV reliabilities, empirical reliabilities, and genetic trends predicted from the complete and reduced data sets were used to validate the genomic evaluation. Average EBV reliabilities for H-AR (A-AR) using the complete data set were 0.52 (0.16) and 0.72 (0.62) for genotyped bulls with no daughters and bulls with 1 to 9 daughters, respectively. These results showed an increase in EBV reliabilities of 0.10 to 0.36 when genomic information was included, corresponding to a reduction of up to 43% in prediction error variance. Considering the 3 validation scenarios, the inclusion of genomic information improved the average EBV reliability in the reduced data set, which ranged, on average, from 0.16 to 0.26, indicating an increase in the predictive ability. Similarly, empirical reliability increased by up to 0.08 between validation tests. The H-AR outperformed A-AR in terms of genetic trends when unproven genotyped bulls were included. The results suggest that the single-step GBLUP AR model is feasible and may be applied to national Portuguese genetic evaluations for milk yield.     

Genotype by environment interaction for milk production traits between organic and conventional dairy cattle production in The Netherlands

Estimates of genetic parameters for organic dairy farming have not been published previously, and neither is information available on the magnitude of genotype by environment interaction (G×E) between organic and conventional farming. However, organic farming is growing worldwide and basic information about genetic parameters is needed for future breeding strategies for organic dairy farming. The goal of this study was to estimate heritabilities of milk production traits under organic farming conditions and to estimate the magnitude of G×E between organic and conventional dairy farming. For this purpose, production records of first-parity Holstein heifers were used. Heritabilities of milk, fat and protein yield, and somatic cell score (SCS) were higher under organic farming conditions. For percentages of fat and protein, heritabilities of organic and conventional production were very similar. Genetic correlations between preorganic and organic, and organic and conventional milk production were 0.79 and 0.80, respectively. For fat yield, these correlations were 0.86 and 0.88, and for protein yield, these were 0.78 and 0.71, respectively. Our findings indicate that moderate G×E was present for yield traits. For percentage of fat and protein and SCS, genetic correlations between organic and conventional and preorganic production were close to unity, indicating that there was no G×E for these traits.     

The accuracy of seven mathematical functions in modeling dairy cattle lactation curves based on test-day records from varying sample schemes

Daily milk yield over the course of the lactation follows a curvilinear pattern, so a suitable function is required to model this curve. In this study, 7 functions (Wood, Wilmink, Ali and Schaeffer, cubic splines, and 3 Legendre polynomials) were used to model the lactation curve at the phenotypic level, using both daily observations and data from commonly used recording schemes. The number of observations per lactation varied from 4 to 11. Several criteria based on the analysis of the real error were used to compare models. The performance of models showed few discrepancies in the comparison criteria when daily or 4-weekly (with first test at days in milk 8) data by lactation were used. The performance of the Wood, Wilmink, and Ali and Schaeffer models were highly affected by the reduction of the sample dimension. The results of this work support the idea that the performance of these models depends on the sample properties but also shows considerable variation within the sampling groups.     

A stochastic model simulating pathogen-specific mastitis control in a dairy herd

The objective of this study was to develop a model simulating mastitis control in dairy herds and to investigate how sensitive the model is when varying the effect parameters according to the uncertainty. The model simulates 9 pathogen-specific mastitis types, each of which can be subclinical or clinical. The clinical cases can be 1 of 4 severities defined according to the effect of the mastitis case: mild, moderate, severe, and permanent effect. The risk factors include lactation stage, parity, yield level, previous diseases, season, and contagious spread of the infection from herd mates. Occurrence of mastitis is modeled to have direct effects on feed intake, body weight, milk yield, somatic cell count in the milk, subsequent mastitis cases within the cow and in herd mates, voluntary and involuntary culling, mortality, and milk withdrawal. Thirty-five scenarios were simulated to study model behavior and model sensitivity. The consequences per cow/yr of mastitis in the default simulated herd included 0.42 clinical mastitis occurrences, 0.56 subclinical mastitis occurrences, loss of 385-kg milk yield, a 1.3% reduced feed intake, 61-kg milk withdrawal and €146 in reduced economic net return. Based on scenarios demonstrating model behavior and sensitivity analysis, the model appears to produce valid consequences of mastitis control strategies. Representation of the effect of subclinical mastitis and of variation in mastitis severity was concluded in this study to be important when modeling mastitis economics in a dairy herd. The model offers the opportunity to study the long-term herd specific effects of a wide range of control strategies against mastitis.     

Relationship between herd size and measures of animal welfare on dairy cattle farms with freestall housing in Germany

The objective of this study was to examine the association of herd size with animal welfare in dairy cattle herds. Therefore, 80 conventional dairy cattle farms were classified by the number of cows into 4 herd size classes, C1 (<100 cows), C2 (100–299 cows), C3 (300–499 cows), and C4 (≥500 cows), and assessed using multiple animal-based measures of the Welfare Quality Assessment protocol for dairy cattle. Data were recorded from April 2014 to September 2016 by an experienced single assessor in northern Germany. Each farm was visited 2 times at an interval of 6 mo (summer period and winter period) to avoid seasonal effects on the outcome. The average herd size was 383 ± 356 Holstein-Friesian cows (range 45 to 1,629). Only farms with freestall (cubicle) housing and a maximum of 6 h access to pasture per day were included in the study. Data were statistically analyzed using a generalized linear mixed model. None of the farms reached the highest overall rating of “excellent.” The majority of the farms were classified as “enhanced” (30%) or “acceptable” (66%), and at 6 assessments the farms were rated as “not classified” (4%). Regarding single indicators, mean trough length per cow, percentage of cows with nasal discharge, and vulvar discharge increased with increasing herd size, whereas it was vice versa for displacements of cows. Percentage of lean cows, percentage of dirty lower legs, and duration of the process of lying down showed a curvilinear relationship with the number of cows per farm. Herd size was not associated with any other measures of the Welfare Quality protocol. In conclusion, herd size effects were small, and consequently herd size cannot be considered as a feasible indicator of the on-farm animal welfare level. Housing conditions and management practices seem to have a greater effect on animal welfare than the number of dairy cows per farm.     

Genetic analysis of herd life in Canadian dairy cattle on a lactation basis using a Weibull proportional hazards model

The objectives of this study were to identify the most important factors that influence functional survival and to estimate the genetic parameters of functional survival for Canadian dairy cattle. Data were obtained from lactation records extracted for the May 2002 genetic evaluation of Holstein, Jersey, and Ayrshire breeds that calved between July 1, 1985 and April 5, 2002. Analysis was performed using a Weibull proportional hazard model, and the baseline hazard function was defined on a lactation basis instead of the traditional analysis of the whole length of life. The statistical model included the effects of stage of lactation; season of production; the annual change in herd size; type of milk recording supervision; age at first calving; effects of milk, fat, and protein yields calculated within herd-year-parity deviations; and the random effects of herd-year-season of calving and sire. All effects fitted in the model had a significant effect on functional survival of cows in all breeds. Milk yield was by far the most important factor influencing survival, and the hazard increased as the milk production of the cows decreased. The hazard also increased as the fat content increased compared with the average group. Heifers that were older at calving were at higher risk of being culled, and expanding herds were at lower risk of being culled compared with stable herds. More culling was found in unsupervised herds than in supervised herds. The heritability values obtained were 0.14, 0.10, and 0.09 for Holstein, Jersey, and Ayrshire, respectively. Rank correlation between estimated breeding values (EBV) obtained from the current national genetic evaluation of direct herd life and the survival kit used in this study ranged from 0.65 to 0.87, depending on the number of daughters per sire. Estimated genetic trend obtained using the survival kit was overestimated.     

Diversity of dairy goat lactation curves in France

A high level of production at the peak of lactation may be associated with animal health disorders, high feeding costs, and reduced milk supply throughout the year. The objective of this study was to typologize the lactation curves in French dairy goats and analyze the influence of environmental and genetic factors on these curves. The data set consisted of 2,231,720 monthly test-day records of 213,534 French Saanen and Alpine goats recorded between September 2008 and June 2012. First, principal component analysis classified the shape of the lactation curves into 3 principal components: the first component accounted for milk yield level throughout lactation, the second component accounted for lactation persistency, and the third component accounted for milk yield in mid-lactation. Then, from the principal component scores, the lactations were clustered into 5 different groups. Most lactations had a similar shape to the mean curve, except 30% of the lactations that fell into 3 clusters that had a high production level at the peak and then a different persistency according to cluster. Estimated breeding value for milk yield and home region of breeding were the factors most related to lactation production level. Month of kidding, breed, and gestation stage had the biggest effect on persistency. Month of kidding was the factor most strongly linked to mid-lactation production. A herd effect was observed on all 3 principal components.     

Negative influence of high maternal milk production before and after conception on offspring survival and milk production in dairy cattle

There is a paucity of studies on the effect of intrauterine conditions on subsequent progeny performance in dairy cows. Using a large national data set on Irish Holstein-Friesian dairy cows, the objective of this study was to determine if intrauterine conditions, quantified by a maternal genetic variance component, significantly affected milk production, age at first calving, calving interval, somatic cell score (natural logarithm of somatic cell count) and survival in first-, second-, and third-parity female offspring. Maternal genetic variance for each trait in each parity was estimated in a linear mixed model which included, other than fixed effects, direct additive genetic, maternal genetic, cytoplasmic and permanent environmental effect of the dam, and residual component. A covariance was also estimated between the direct additive and maternal genetic components where possible. Because calves in Irish dairy herds are separated from dams at birth, a significant maternal genetic variance (with all other random effects in the model) indicates a prepartum maternal effect. A significant maternal genetic variance was estimated for 305-d milk yield in first and third lactation, somatic cell score in first lactation, and survival to second lactation from 188,144 lactations on 80,881 animals. Where estimated, a negative correlation existed between the direct additive and maternal genetic components. Regression of maternal mixed model solutions on dam milk production at different stages relative to conception revealed that greater milk yield preconception and during gestation was associated with reduced survival and milk yield and greater somatic cell count in the progeny. This study suggests that offspring survival and performance are affected by prepartum conditions that offspring experience as an oocyte, embryo, or fetus, one of which is mediated through milk production (or factors related to milk production) of the dam.     

The effect of diets on milk production and composition, and on lactation curves in pastured dairy goats

A 2-yr study investigated effects of different levels of concentrate supplementation on milk production, composition, and lactation curves in pastured dairy goats. For both years, 44 Alpine goats (Capra hircus; 55 ± 11 kg body weight) were randomly allocated to 4 groups. Animals were supplemented with 0.66 (treatments A and B), 0.33 (treatment C), or 0 kg of concentrate (treatment D) per kg of milk over 1.5 kg/d. Mixed vegetative forages were rotationally grazed by the goats (treatments B, C, and D), except that treatment A was confined and fed alfalfa hay. Individual milk production was recorded daily, and milk samples were collected once every 2 wk for the 7-mo period (March to September) and analyzed for fat, protein, lactose, urea-N, nonesterified fatty acids, and allantoin (second year only). Milk yield and composition varied among dietary treatments, with some measures affected by year. Average daily milk yield was lowest for treatment D. The increased level of concentrate supplementation in treatment A led to 22% greater milk yield compared with treatment D. Milk production increased by 1.7 and 0.9 kg for each additional kilogram of concentrate fed per day during the first and second years, respectively. Average peak yield, time of peak yield, and persistency were lower for treatment D than for other treatments. The percentage of milk fat was lower for treatment D than for other treatments. Concentration of milk protein was greater for treatments A and B during the first year, and was higher for treatment C than for other treatments during the second year. Average milk lactose concentration was higher for treatments B and C than for other treatments. However, milk urea-N concentration in treatment A was higher than other treatments. Milk allantoin, used to estimate microbial proteins synthesis, was 20 to 25% greater for treatment A than for other treatments. Averaged across year, plasma urea-N and nonesterified fatty acids concentration were lowest for treatment B. Average organic matter intake was similar among treatments during both years. Ratios of acetate and propionate concentrations for treatment A were lowest among treatments. In conclusion, milk production and composition were affected by the feeding treatment and year. Increased level of nutrition lead to an increase in daily milk yield, peak yield, time of peak yield, and persistency compared with treatment D. Alpine dairy goats grazing on fresh forages without concentrate supplementation can produce milk inexpensively, and response to concentrate supplementation is greater for low quality pasture.