BIOCLIMATIC ZONING FOR CATTLE IN THE STATE OF BAHIA AND STATISTICAL MODELING

##Título: Bioclimatic Zoning for Cattle in the State of Bahia and Statistical Modeling ##Authors Talia Silva Ribeiroa, Dian Júnior Bomfim Borges, Luanna Chácara Pires,Gerson dos Santos Lisboa, Rafael Henrique de Freitas Noronha #Overview This dataset was collected to perform bioclimatic and thermal comfort zoning for beef and dairy cattle in the intermediate regions of Bahia. The data were used to estimate the Temperature-Humidity Index (THI) and the decline in milk production (DMP). Climatological data from the Daily Meteorological Data Grid for Brazil (BR-DWGD, 2025), with a spatial resolution of 0.1° × 0.1°, were used for the period from 1961 to 2024. The BR-DWGD is composed of time series from 1,226 pluviometric stations and 87 meteorological stations (Xavier et al., 2022). The following climatic variables were considered: mean accumulated precipitation (mm), maximum temperature (Tmax, °C), minimum temperature (Tmin, °C), mean temperature (Tmean, °C), and mean relative humidity (RH, %) for the period from 1993 to 2023, totaling 4,601 grids across the state of Bahia. ##Temperature-Humidity Index (THI) The THI was calculated for the two seasonal periods, dry and rainy, using the following equation (Buffington, Collier, & Canton, 1982): THI = 0.8 × Ta + [UR × (Ta − 14.3)] / 100 + 46.3 where Ta is the mean air temperature (°C) and UR is the mean relative humidity (%). ##Decline in Milk Production (DMP) The decline in milk production (DMP, kg/day) for lactating Holstein cows was calculated according to Berry et al. (1964) and Hahn (1993) as follows: DMP = −1.075 − 1.736 × PN + 0.02474 × PN × THI Where: PN is the normal milk production level (kg/day) and THI is the Temperature-Humidity Index (dimensionless). Three normal production levels (PN) were considered: 10, 20, and 30 kg/day, representing different productive capacities of the animals under good thermal conditions. The analyses focus on: · The effect of intermediate regions and season (10 × 2) on THI and DMP10, DMP20, and DMP30, the data were analyzed at a 5% significance level (P < 0.05). · The statistical model included the fixed effects of region, season, and their interaction. When significant differences among means were observed, the Kruskal–Wallis test was applied, followed by Dunn’s post hoc test with significance correction (P < 0.05) for multiple comparisons. · To estimate the probability of thermal discomfort occurrence (THI > 70) throughout the months in each region, the data were modeled using logistic regression with binomial distribution, considering month as the explanatory factor. In cases of a significant time effect (P < 0.05), pairwise comparisons between months were performed using the nonparametric Bonferroni test. The database was provided as CSV files and analyzed using R scripts to ensure full reproducibility of the results. ##Content This repository includes the following files: Script_R_Dados.R: R script that analyzes the effect of intermediate regions and season (10 × 2) on THI and DMP10, DMP20, and DMP30, considering the fixed effects of region, season, and their interaction. The script also estimates the probability of thermal discomfort occurrence (THI > 70) throughout the months in each region, using data from dataset_point.csv. dataset_point.csv: CSV file containing the raw climatic data, as well as the calculations of THI and DMP. #Data Description ##dataset_point.csv This file contains the following columns: ano: year of observation. mes: month of observation. periodo: dry season or rainy season regiao: intermediate region of Bahia (Barreiras, Feira de Santana, Guanambi, Ilhéus–Itabuna, Irecê, Juazeiro, Paulo Afonso, Salvador, Santo Antônio de Jesus, and Vitória da Conquista) Tmed: monthly mean air temperature (°C). UR: monthly mean relative humidity (%). Tmax: monthly maximum temperature (°C). Tmin: monthly minimum temperature (°C). ITU: calculated Temperature-Humidity Index. DPL10: Decline in Milk Production for a normal production level of 10 kg/day. DPL20: Decline in Milk Production for a normal production level of 20 kg/day. DPL30: Decline in Milk Production for a normal production level of 30 kg/day. To analyze the dataset, open and run the corresponding R scripts, which contain all the code necessary to reproduce the analyses described in the Overview.