Spatial variability and exploratory inference of abiotic factors in barn compost confinement for cattle in the semiarid.

The assessment of environmental conditions in the creation of dairy cattle in feedlots must be carried out to identify possible causes of stress and assist producers in the decision-making process. The objective was to characterize the abiotic factors and bed temperature, in the comfort conditions of dairy cows, through geostatistics and exploratory data inference, in a Compost Barn System in the Brazilian semiarid region. The data were obtained in a milk production unit, located in the municipality of Lajedo, Pernambuco, Brazil. The variables air temperature (Tair), relative humidity (RH), wind speed (WS), illuminance (Lux), skin temperature (Ts), bed temperature (Tb) were recorded and the temperature index was determined and humidity (THI). Data were recorded at 9:00 a.m., 12:00 p.m., and 03:00 p.m., over 5 days in the summer season. For geostatistical analysis, the classic semivariances were determined. The principal component analysis was performed to establish an index that characterized the condition of animal comfort. The variables Tair, RH, Tb, and THI showed a low coefficient of variation for all times. The best fit to the models of the semivariograms was the Gaussian at 9:00 a.m. and 03:00 p.m., and the spherical at 12:00 p.m. The Tb spatial variability was low for all studied hours. Tair showed a strong correlation with Tb, due to the process of heat transfer by convection from the floor to the environment. Geostatistics and exploratory data analysis allowed the establishment of a comfort index for Compost Barn production systems in the Brazilian semiarid region (R[2] = 0.996; p < 0.0001).