ESTIMATING CAVE ENTRANCE DENSITY USING GIS ANALYSIS: QUINTANA ROO, MEXICO

Caves and their entrance locations are of interest to scientists, engineers, and urban planners for reasons including sinkhole hazard assessment and cave roof collapse. Obtaining unbiased entrance location data is complicated by hindrance to cave accessibility by factors such as terrain considerations, proximity to human development, limited areas of exploration, access to databases, and landowner property access restrictions. Any work that requires an understanding of cave entrance density over a large area must manage detection uncertainty due to these and other biases. One approach is to evaluate cave development in a small area and then test the hypothesis that similar densities exist across similar geologic formations.

A cost map analysis in ArcGIS was used to assess the relationship between entrance density and the proximity to roads along the eastern shoreline of Quintana Roo, Mexico. Quintana Roo is widely known for its extensive cave systems at the existing water table, as well as a level of cave development occurring near the surface, which is prone to collapse. As urban development expands it is increasingly important to assess risk of sinkhole formation, groundwater contamination, and impacts on habitat. Variables used in this analysis include slope, land use, vegetation type, and distance from roads. Spatial Analyst tools were used to evaluate entrance density across geologic units. Entrance locations were overlain by buffers of varying distances from roads in order to evaluate if observed density is a function of accessibility. Cave density per square kilometer decreased geometrically with increased distance from roads. While it is difficult to remove access bias from a data set, an alternative approach is to use similarly biased data for purposes of comparison.

Entrance densities in the 1 km buffer, but from adjacent geologic units (Miocene limestone ≥10km from the coast, or Pleistocene limestone), were compared to determine whether an observed density increase toward the coast reflects geologic controls or sampling bias. Miocene limestone has a much lower entrance density of 0.47 per km² compared to Pleistocene limestone, which is 1.7 per km². These results suggest that significant differences exist in the likelihood of entrance development between the two geologic units.