Caves are relatively protected from the main external erosional factors. Therefore, they constitute potentially reliable places for long-term conservation of continental history. Moreover their secondary carbonated deposits, the speleothems, can be dated precisely on the 0-500 ka time-scale using U-series isotopes measured by thermo-ionisation mass spectrometry (TIMS). Tectonic events (tectonic displacements and earthquakes) may change cave morphology and induce speleothem breaking or displacement as has been shown by previous studies performed mainly in Italy [Forti et Postpischl, 1984; Postpischl et al., 1991 for example]. Nevertheless, collapses of speleothems observed today in caves are difficult to interpret as their origin may be linked to several other natural processes.
We studied the Aven de la Salamandre cave located in southeastern France (Gard), an area between the Cevennes fault and the Nîmes fault, where evidence of Quaternary vertical movements was previously described. However, this region is not considered to be a seismic active zone.
The Aven de la Salamandre cave is characterized by numerous broken speleothems. Some of them are very large and a lot of them are covered by growth of new calcite or new speleothem generation.
We report here 13 TIMS U/Th analysis performed on two broken speleothems recovered by second generation calcite growth. Dating results are discussed through time corrections due to detrital content of calcite. In the first sequence, a 7 ± 0.35 ka fracture event is identified. In the second sequence, the age of the breakdown is between 1.1 ± 0.1 and 6.3 ± 2 ka. These events could thus be contemporary. Hypotheses for the origin of this fracture event are presented and discussed: (i) karstic catastrophic event due to intense climatic changes or to cavity collapse (break down of hanging wall, gravity, landslide…); (ii) co-seismic ruptures.
The first conclusion of this study is that these collapse episodes in the Aven de la Salamandre cave cannot be due to the direct effect of an earthquake or climatic event. Only indirect effects of flooding (by mobilization of the argillaceous components of the floor and consecutive destabilization of the speleothems growing upon it) or earthquake effects (more likely by local effects than by wave front passage) are privileged. By comparing our dating with regional destructive known events (in other karsts, in cliffs and scarps), dated by relative chronology, we are lead to propose a regional generalized event precisely dated here at 7 ka. This second conclusion doesn’t contradict the presence of unbroken speleothems older than 100 ka found in other caves in the neighborhood as local effects is one of the predominant factors relative to speleothem stability.
As a final conclusion, this paper promotes the use of speleothems as reliable datable tools for assessing regional stability problems (sensitivity to seismic hazards, to destructive intense climatic events…) as is done for paleoclimatic reconstruction.