Extending beyond individual caves: a graph theory approach broadening conservation priorities in Amazon iron ore caves.

The Amazon is renowned worldwide for its biological significance, but it also harbors substantial mineral reserves. Among these, the ferruginous geosystems of the region are critical for iron ore extraction, accounting for 10% of Brazil's export revenue. Additionally, this region holds a significant speleological heritage with more than 1,000 caves. However, cave conservation efforts are often in conflict with land use, necessitating mediation through environmental regulations. While conservation decisions typically consider only the caves' characteristics, such an approach fails to account for the interactions among cave communities and their surrounding landscape. This poses a challenge to reserve design for cave conservation purposes. To address this issue, we assessed the predictors that influence the similarity among cave communities, suggesting the use of this parameter as a proxy for subterranean connectivity. Applying graph theory, we proposed a tool to aid in the selection of priority caves for conservation purposes. Our study involved the sampling of invertebrates in 69 iron ore caves and analyzing 28 environmental variables related to these subterranean habitats and adjacent landscape. Our analysis revealed that landscape and habitat characteristics are more important than geographical distance in determining patterns of similarity among caves. Our graph approach highlighted densely interconnected clusters based on similarity. However, specific caves stood out for harboring exclusive fauna and/or exhibiting habitat specificity, making them unique in the study area. Thus, we recommend prioritizing cave clusters for conservation, assembling both singular caves and others that influence them. It is crucial to note that protocols for the protection of subterranean biodiversity must consider measures that encompass both the caves and the surrounding landscape. Our methodology provides insights into the connectivity among caves, identifies existing groups, highlights singular (or unique) cavities that require preservation, and recognizes those influencing these unique habitats. This methodological advancement is crucial for the development of better conservation policies for the speleological heritage in areas under constant economic pressure.