The objectives of the first and second study were to evaluate the effect of crude protein (CP) supplementation on the metabolic characteristics, performance, muscle, and mammary gland development and expression of genes involved in the urea cycle, and muscle tissue development of Holstein × Gyr crossbreed heifers grazing Brachiaria decumbens throughout the year. Thirty-eight heifers were randomly assigned to four treatments: three protein supplements (SUP) fed at 5g/kg of body weight, plus a control group. The supplement CP levels were 12, 24, and 36%. The experimental period was divided into four seasons: rainy, dry, rainy-dry transition, and dry-rainy transition. The data were analyzed using PROC GLIMMIXED of the SAS with repeated measures. SUP animals had a greater intake of dry matter, metabolizable energy, and metabolizable protein. Furthermore, SUP animals had a greater average daily gain, rib eye area and fat thickness than non-supplemented animals. Among SUP animals, we observed a quadratic response to ADG, with the highest level in S24. No supplementation effects were detected on mammary gland development. In muscle, we observed greater expression of AMPK in non-supplemented animals than SUP animals. No differences were observed for mTOR. We observed greater urea excretion and retention coefficient in SUP animals than non-supplemented animals. In this sense, we also observed greater gene expression of CPS, ASL, and ARG in SUP animals than non-supplemented, and among SUP animals, the supplement CP linearly affected CPS expression. We observed a positive linear effect of urea excretion, nitrogen intake, nitrogen retention, and retention coefficient among SUP animals. In conclusion, SUP animals had greater intake, performance than non- supplemented animals, with S24% demonstrating the best results. The third study aimed to describe the FA profile, as predicted using Fourier transform mid-infrared (FTIR) spectroscopy, of bulk tank milk from automated milking system (AMS) farms and to assess the association of management and housing factors with the bulk tank milk composition and FA profile of those AMS farms. The data used were collected from 124 commercial Canadian Holstein dairy farms. Information regarding individual cow milk yield (kg/d), days in milk (DIM), parity, and the number of milking cows were automatically collected by the AMS units on each farm. Multivariable regression models were used to associate herd-level housing factors and management practices with milk production, composition, and FA profile. Milk yield was positively associated with using a robot feed pusher (+2.1 kg/d) and the use of deep bedding (+2.6 kg/d). The use of a robot feed pusher, deep bedding, and greater stall raking frequency were positively associated with greater yield (kg/d) of de novo FA, mixed FA, preformed FA, and de novo + mixed FA. Greater frequency of PMR delivery (>2x/d vs. 1 and 2 x/d) was positively associated with a greater proportion (g/100 g of FA) of de novo, mixed, and de novo + mixed FA and negatively associated with the proportion of preformed FA. Overall, these associations indicate that bulk tank FA profile can be used to monitor and adjust management and housing in AMS farms. Keywords: Nitrogen metabolism. Pasture. Robotic milking. Season.