Clinical mastitis situation in Brazilian dairy farms: incidence rate and geospatial distribution

• Ano de defesa: 2024
• Autor(a) principal: Mitsunaga, Thatiane Mendes
• Tipo de documento: Tese
• Tipo de acesso: Acesso aberto
• Idioma: eng
• Instituição de defesa: Biblioteca Digitais de Teses e Dissertações da USP
• Palavras-chave em Português: Análise geoespacial; Bacteria; Bactéria; Clusters; Cows; Geospatial analysis; Glândula mamária; Mammary gland; Vacas
• Link de acesso: https://www.teses.usp.br/teses/disponiveis/11/11139/tde-06022025-111538/

Mastitis is the primary disease that affects the mammary gland of cows in dairy herds, responsible for negative impacts on productivity and farm profitability. Its rapid detection is the main method of control and prevention. Therefore, there is a clear need for rapid diagnostic methods as essential tools for controlling this disease. This thesis was developed in four parts. The first part consists of a literature review focusing on the methods for the identification and diagnosis of mastitis, considering new technologies. The second part involves a bibliometric review seeking the main applications of artificial intelligence (AI) in the context of bovine mastitis, identifying the main AI tools for diagnosing and predicting mastitis. A total of 62 articles from the Scopus database, published between 2011 and 2021, were used, based on keyword searches with terms related to AI models. The results identified machine learning and mastitis as the most cited terms, with a significant increase between 2018 and 2021. The most cited model was artificial neural networks. It was concluded that the use of AI in bovine mastitis is related to mastitis detection as a vital tool for preventing this disease. Moving on to the third chapter, the study was developed to determine the incidence rate of clinical mastitis (CM) in dairy herds in Brazil and its association with risk factors. A total of 117,296 records from 2019 to 2023 of CM cases in dairy farms across the country were used, including data on date, location, CM severity, herd size, housing system, and pathogen diagnosis for calculating the clinical mastitis incidence rate (IRCM). The results showed that the average IRCM was 8.05 cases per month. The most isolated pathogens were Non-aureus Staphylococcus (NAS), Streptococcus agalactiae/dysgalactiae, Escherichia coli, Streptococcus uberis, and Staphylococcus aureus, and incidence rates were calculated for each pathogen. There was an association between IRCM by pathogen and the respective variables: region, housing system, climate, and herd size. Finally, the fourth chapter of the thesis was developed as a continuation of the IRCM analysis, but with a geospatial approach. The data from the third chapter were filtered to include only those with information on location, herd size, and microbiological results, resulting in 148,191 observations, which were then merged with a database containing geographic coordinates available from IBGE. The data were analyzed to identify the relationship between IRCM and the distribution of pathogens according to immediate geographic regions (IGR). The average IRCM was higher in regions located in the South and Southeast of the country, with clusters forming among regions within the same state. Clusters of the main pathogens were mostly found in the region of Minas Gerais. However, there are no studies indicating the specific occurrence of pathogens in this region, suggesting that possible local factors and movements between herds within the same region may influence these results.