BIASES IN THE KNOWLEDGE OF THE GEOGRAPHICAL DISTRIBUTION OF CAVES IN BRAZIL

Abstract

Many caves have yet to be discovered and/or mapped in Brazil. Various socio-economic and accessibility factors may be responsible for biasing the knowledge of Brazilian caves geographical distribution. In this study, using Bayesian modeling techniques, it was predicted that knowledge about the Brazilian caves geographic distribution is strongly biased towards locations close to urban centers and mining activities. More caves are known near mining environments and/or more accessible locations than far from them for instance. The models were built by associating the known occurrence data of digitally accessible caves in the Speleological Information National Register (Cadastro Nacional de Informações Espeleológicas (CANIE) with different accessibility and infrastructure variables. The findings suggest that the regions with the highest occurrence of known caves are mainly concentrated within the states of Minas Gerais, São Paulo, Goiás, Pará and Piauí. Meanwhile some Brazilian states, such as Amazonas, Roraima, Mato Grosso, Mato Grosso do Sul, Paraná and Maranhão are still under-explored. They have large gaps and high potential for the discovery of new caves. In conclusion, the economic exploitation has generated biased knowledge about the distribution of caves. This may be linked to the historical clashes between exploitation and conservation legislation, making it essential that new policies on the conservation of Brazil's speleological heritage consider the entire speleological landscape, its biases (better studied sites), Racovitzan shortfalls (less studied or
possibly neglected sites), and the socio-economic processes that imply these.