Assessing the carbon footprint in dairy cattle farms in the northern temperate region of Spain

The availability of models constitutes a key factor when selecting a decision support tool aimed at improving the production and environmental aspects of farms. There is a need for robust models that are user-friendly, facilitating the estimation of farm emissions and the analysis of their temporal fluctuations. The objectives of this study were i) to calculate both the partial (PCF) and total carbon (TCF) footprints of 212 dairy farms, distinguishing those with and without maize cultivation; ii) to identify critical variables related to feed, nutrition, productivity and environmental efficiency; and iii) to formulate and validate prediction equations based on available data from dairy farms. The database encompasses information from 212 dairy cattle farms situated in the temperate-humid zone of northern Spain, spanning the period from 2014 to 2018. Farm classification was based on the presence (CcMz) or absence (ScCMz) of maize cultivation for silage production, resulting in 96 farms in the CcMz category and 116 farms in the ScCMz category.

Among the variables considered, the variable herd N-use efficiency (NUE_CR) for (PCF) showed the lowest root mean square error of prediction at 0.39% and the correspondingly lowest root men. The root mean squared percentage error (RMSPE): standard deviation ratio (RSR) at 0.52. In the case of total carbon footprint (TCF), herd N-use efficiency (NUE_CR) again showed the lowest root mean square error of prediction at 0.52%. Regarding TCF, herd feed efficiency (EA_CR) was the variable with the lowest both RMSPE and RSR, with 0.65 and 0.64, respectively. Consequently, the estimation of the PCF and TCF of 1 ​kg of milk from the temperate-humid zone of northern Spain at the farm gate can be feasibly accomplished utilizing NUE_CR and EA_CR, respectively.