Maximum Ages of the Coa Valley (Portugal) Engravings Measured with Chlorine-36

Article excerpt

Panel faces in the Coa valley, Portugal, were available for engraving during the Upper Palaeolithic, according to 36Cl exposure ages of 16,000 to 136,000 years.

Introduction

The Coa valley petroglyphs in Portugal have a style indicative of a Palaeolithic age (Bahn 1995; Clottes et al. 1995; Zilhao 1995; Zuchner 1995). These engravings have become controversial because they were in danger of being flooded by a proposed dam, and because radiocarbon (Watchman 1995; 1996) and microerosion (Bednarik 1995a; 1995b; 1995c; 1995d) dating results suggest that they are Holocene in age.

There are also claims that the geomorphic system in the Coa valley is too unstable to support Palaeolithic art on the grounds of 4000-6000 b.p. luminescence ages for inset river sediments; close proximity to the river channel and flooding; and slope instability (Bednarik 1995c; 1995d; Watchman 1996). Bednarik (1995b: 98) writes, 'in summary, there is not one iota of evidence supporting a Pleistocene antiquity of the petroglyphs, be [TABULAR DATA FOR TABLE 1 OMITTED] it quantifiable or circumstantial, deductive or inductive.'

We focused on the question of whether the rock panels were available for engraving during the Palaeolithic. We address this issue by measuring the accumulation of cosmogenic 36Cl in samples from petroglyph panels.

Overview of 36Cl analysis at Coa

One clear test of the age of the Coa petroglyphs is to determine the time at which the rock faces containing the art were exposed. The age of the carvings cannot exceed that of the rock surfaces: a Holocene age for the exposure of panels would preclude a Palaeolithic date for the art.

The maximum age can be determined in a straightforward fashion using the accumulation of cosmogenic nuclides in minerals at the rock face. Cosmogenic nuclides are produced by reactions of cosmic-ray particles with elements in the atmosphere or rock (Cerling & Craig 1994). One to two metres of rock will block most cosmic radiation; where thick slabs of rock spall off to expose fresh faces, the accumulation of cosmogenic nuclides will be initiated at the fresh face at the time of the spall event. This condition is well met in the Coa Valley where the panels on which the art was engraved are created by slabbing along joint planes in schist spaced at approximately 1.5-m intervals.

Methods

Samples for 36Cl measurement were taken from two engraved panels at the Canada do Inferno site, one at Penascosa, and one at Ribeira de Piscos. In addition, two unengraved joint faces at Ribeira de Piscos, the top of the hill above the Penascosa site, and a medieval castle were sampled to help assess rates of erosion in the area.

One sample (FC95-BD-4) was collected from the dam excavation, 16 m below the riverbed. This sample assesses background concentrations of 36Cl. The 36Cl concentration of this sample, one to two orders of magnitude less than the surface samples, indicates a 'background' signal of no more than a few thousand years.

Samples were prepared by dissolution in HF and HN[O.sub.3]; the chlorine was extracted and purified using standard procedures (Zreda et al. 1991). A 35Cl-labelled carrier was added during dissolution; 36Cl and total chlorine were then determined simultaneously by isotope dilution during the accelerator mass spectrometric (AMS) analysis (Elmore & Phillips 1987) at PRIME Laboratory, Purdue University (TABLE 1). Exposure ages were calculated for panel surfaces, exposed by spalling events, assuming a minimum erosion rate of zero and a maximum erosion rate of 2 mm/ka (where 1 ka=[10.sup.3] years) - using 36Cl production rates (Phillips et al. 1996) and erosion corrections (Liu et al. 1994b); this is because higher rates of surface erosion would have obliterated the engravings.

Results for panel faces

The senior author intuitively anticipated that engraved joint surfaces were exposed in the Holocene. …