The aims of the present study were to compare centrality and other measures of bovine movement networks in the state of Mato Grosso, Brazil, in 2007, by examination of the positive and negative herds identified in the brucellosis survey conducted in 2003. In addition, we aimed to examine the association between herd size (measured by the number of females older than 24 months, denoted FEM24+) and animal trade between herds. We found a statistical association between brucellosis occurrence and both the total degree (number of traded animals) and out-degree (for the number of animals sold and the number of neighboring herds). This finding suggests that positive herds in the 2003 survey were trading (particularly selling) more frequently in 2007 than negative herds, thereby presumably increasing the risk of disease spread. Statistical differences observed (p-values) in the network of movements for reproduction purposes were more significant than those observed in other networks for average herd size; degrees of animals, batches, and neighbors; outdegree of neighbors; and betweenness. We found positive associations (p < 0.001) between the following variables: number of traded animals and the FEM24+ herd size; FEM24+ herd size and FEM24+ herd size of neighbors within the network of cattle movement; FEM24+ herd size of neighbors and number of traded animals; and weighted degree (animals) of neighbors and degree (animals) of the herd of origin. A comparison of positive and negative herds stratified by herd size (≤ 10 FEM24+; and > 50 FEM24+), within the network of movements for reproduction purposes, revealed significant differences (p < 0.05) in the > 50 FEM24+ category only, for average herd size, total degree (animals, batches, and neighbors), outdegree (animals, batches, and neighbors), and outcloseness. Logistic regression analysis, in which average herd size and degree (animals) were both factors, showed that a tenfold increase in herd size and the number of animals traded would lead to an increased chance of a herd being positive by 41% (OR = 1.41 [1.01; 1.99]) and 39% (OR = 1.39 [1.05; 1.85]), respectively. We concluded that the presence of bovine brucellosis is associated with a larger herd size, which is consistent with other reports. We also found an association between the increased trade of bovines and the presence of bovine brucellosis. Furthermore, an association was noted between herd size and animal trade, both of which could contribute to the spread of brucellosis.
ResumoO objetivo do presente estudo foi comparar medidas de centralidade e outras medidas da rede de trânsito bovino no estado de Mato Grosso, Brasil, em 2007, pela análise de rebanhos positivos e negativos identificados no levantamento de brucelose realizado em 2003. Além disso, examinamos a associação entre tamanho de rebanho (medido pelo número de fêmeas acima de 24 meses, denotado FEM24+) e o comércio animal entre rebanhos. Foi observada uma associação estatística entre brucelose e tanto o grau total (número de animais comerci...