Page 4 - Shell_growth_2008

312 Geo-Mar Lett (2008) 28:309–325

Work was undertaken between November 2002 and was cut along the outer surface of each shell to expose the
October 2004 at Cala Grande and Monte Cofano, and from junction of the prismatic and nacreous layers (Fig. 3). To
July 2004 to October 2005 at Mazzaforno. The areas of maximise the resolution from each sample, the groove is
shore selected were in the intertidal zone, with no isolated on the periphery of the whorl, where larger growth
rock pools in which warming and evaporation could have increments are laid down. Consecutive holes are then
affected the oxygen isotopes in the ambient water or in the drilled into this groove to obtain samples of carbonate
shells of O. turbinatus. Our observations at numerous powder (~200 μg) for the isotope analyses. In accordance
shores across the western Mediterranean basin suggest that with our procedure for sampling archaeological shells
O. turbinatus can occur in rock pools low on the shore (Mannino et al. 2007), samples are drilled at 0.7–0.9 mm
which are regularly flushed out by tides but rarely (if ever) intervals with a 0.6 mm drill bit, starting from the aperture
in rock pools higher on the shore. These latter pools are edge of the shell backwards to encompass one or more
often inhabited by Osilinus articulatus Lamarck, which is years of growth (Fig. 3).
adapted to eurythermal and euryhaline environments For oxygen and carbon isotope analyses, the powdered
(Menzies et al. 1992). O. articulatus was absent from the shell samples were weighed to yield sub-samples of around
areas selected for the surveys, although it was present in 70±10 μg. These were analysed, along with a laboratory
other parts of the shores. standard calcite (KCM), using an Optima mass spectrom-
For the first research aim (above), two or three speci- eter with an automated carbonate device (IsoCarb). Isotope
18
16
12
mens of O. turbinatus were harvested each month at each ratios ( O/ O and 13 C/ C) are expressed in delta units,
13
18
locality. A single sample of shell carbonate powder δ O and δ C(‰), relative to the international Vienna Pee
(~200 μg) for isotope analysis was taken from the growing Dee Belemnite (VPDB) standard. The error of the analyses
edge of the shell showing the most recent growth. A total of was generally <0.09‰ (one standard deviation for the
24 shells were therefore sampled monthly at Cala Grande laboratory standard for both isotopes).
and at Monte Cofano. At Mazzaforno, single shells were At each monthly visit to the localities, seawater
sampled in July 2004 and in August to October 2005 but samples were collected and five readings of seawater
two shells, a larger one and a smaller one, were sampled temperature were taken, the mean value giving the surface
each month from August 2004 to July 2005. seawater temperature at the time of sampling. The
For the second aim, a shell from Cala Grande, marked seawater samples were analysed by the equilibration
with a notch at its aperture on 18 September 2003, was method for oxygen (Epstein and Mayeda 1953), using a
harvested on 22 October 2004. This shell had a conspic- VG Isoprep 18 and Sira 10 mass spectrometer, with an
18
uous growth line just behind the aperture when notched and analytical precision of ±0.05‰ for δ O. Isotope ratios are
another one behind the aperture when harvested. A series of defined in relation to the international standard, Vienna
24 shell carbonate samples was taken from the shell edge Standard Mean Ocean Water (VSMOW). The salinities of
back to just beyond the growth line of September 2003 seawater samples were measured relative to the interna-
(Fig. 3). tional Practical Salinity Scale using a conductivity meter,
Scanning electron microscopy and X-ray diffraction the resulting dimensionless values being expressed in
(XRD) analyses of modern shells of O. turbinatus practical salinity units (psu).
confirmed these to be entirely aragonitic, with an outer Other datasets used here are the mean monthly surface
prismatic layer and an inner nacreous layer. The protocol seawater temperatures from marine recording stations
for obtaining a series of samples of shell powder for geographically closest to the survey locations (Fig. 1): in
isotope analysis was influenced by the archaeological the Gulf of Palermo (for comparison with the results from
objectives of our research. In archaeological specimens, Monte Cofano and Mazzaforno) and at Mazara del Vallo
which in our study region are all from limestone caves, any (for comparison with the results from Cala Grande). The
recrystallization or contamination with either younger data from these stations are available from the Agenzia per
(pedogenetic) or older (geological) carbonates would most la Protezione dell’Ambiente e per i Servizi Tecnici (http://
likely occur in the outer shell layer (we have detected small www.idromare.com/).
quantities of calcite in XRD spectra of outer-layer
18
carbonate samples in some archaeological specimens). Calculation of seawater temperatures from δ O SHELL
Sampling of shell carbonates in archaeological specimens
18
is therefore restricted to the outermost surface of the inner Values of δ O (VPDB) in shell aragonite samples
18
nacreous layer and, to ensure compatibility between (δ O SHELL ) served to estimate surface seawater temper-
modern and archaeological isotope datasets, this sampling atures using the empirically derived equation of Grossman
procedure is applied to modern shells. A shallow groove and Ku (1986) for biogenic aragonite, incorporating the