Predição de emissão de metano entérico de vacas leiteiras

Abstract

Enteric methane (CH4) in ruminants is a product of the enteric fermentation process, mainly ruminal, and is largely eliminated by eructation. A meta-analysis was performed to elucidate the factors affect enteric CH4 intensity and enteric CH4 production in dairy cows, to develop models to predict the production of enteric CH4 in dairy cows, using only characteristics associated with animals as predictive variables and to compare with the models of predictor variables with diets; and to evaluate the prediction quality of the proposed models with 43 existing models to predict enteric CH4 production using an independent data set. The data set consisted of 115 articles (419 average treatments of 125 experiments) published in scientific journals. The data set was randomly divided into two sub-sets. The first subset was used to construct the models to predict predict enteric CH4 production (80 articles, 90 experiments, 301 means). And the second sub-set was used to evaluate the precision and accuracy of the prediction of prosposed and existing models (35 articles, 35 experiments, 118 treatment means). The data set originated in 18 countries, showed a wide variation in the characteristics of the animals, diet and methane emission. Chamber, SF6 technique and GreenFeed® were the methods used to measure enteric CH4 production, intensity of enteric CH4 production and yield of enteric CH4 production. We did not observe differences in the means of enteral CH4 yield (19.9 ± 1.0 g methane / kg DMI, P = 0.13) and intensity of enteric CH4 (19.9 ± 1.0 g methane / kg DMI; = 0.13) observed in studies using methods to quantify enteric CH4 production. he variables related to the animal had a greater influence on the production of enteric CH4 than the variables associated with the diet (90.18% versus 3.30% of the total change in enteric CH4 production per cow / day). Feed conversion had a greater effect on the intensity of enteric CH4 production than the yield of enteric CH4 production (59.0% versus 39.6% of the total change in intensity of enteric CH4 production). In the Animal I models (without dry matter intake (DMI)) and II (with DMI) milk production and metabolic weight were used as predictor variables. The Diet I and II models were adjusted with ethereal extract contents and digestible organic matter. The Animal + Diet I and II models were adjusted with all variables of the Diet and Animal models. The animal model I presented the lowest correlation coefficient (CCC) and highest (P <0.05) square root of the mean square of the prediction error (RMSEP) in relation to the Diet II, Animal + Diet I and II . The Animal II model, although presented a CCC similar to the Diet II, Animal + Diet I and II models, the RMSEP was higher (P <0.05) than these models. Thus, the prediction models of enteric CH4 production using only animal variables did not present similar accuracy to models with variables associated to diet and animal. Among the six models proposed, the Animal + Diet I, Animal + Diet II and Diet II models presented the best predictive qualities. The models II and III of Nielsen et al. (2013) and Model III of Storlien et al. (2014) were the most similar models to the Animal + Diet I model. We recommend using the Animal + Diet I or Animal + Diet II or Diet II model to predict enteric CH4 production of dairy cows.This research is part of a project conducted by the UFMT-Sinop Dairy Cattle Research Laboratory to develop a new nutritional system applied to dairy cattle, the Nutrition System to Dairy Cattle (NS-Dairy Cattle).