Introduction

1Lead (Pb) is a toxic metal dispersed in the environment by ancient and modern human activities such as mining, metallurgy, fossil energy burning or waste incineration (Nriagu, 1998; Shotyk & Le Roux, in press). However anthropogenic Pb was already emitted since the beginning of the metallurgy (Nriagu, 1983). Regional and global atmospheric pollution by ancient mining and smelting was already highlighted by increase in Pb concentration and shift in Pb isotopes recorded in peat bogs, lakes and ice cores (Brännvall et al., 1999; Le Roux et al., 2004; Rosman et al., 1997; Weiss et al., 1999). Little research (Labonne et al., 1998; Le Roux et al., 2003) has been done on local contaminations near ancient urban coastal centres.

2In the present study, we investigate coastal sediments from the ancient harbour of Marseilles. Our objectives are to: (i) distinguish between natural and anthropogenic Pb in the sediment using Pb concentration and Pb isotopes and therefore evidence possible Pb emissions during ancient times; (ii) to estimate the general chronology of anthropogenic Pb input in the harbour and (iii) using the isotopic composition of Pb, to trace the sources of Pb in the ancient harbour of Marseilles.

1. Material & methods

1.1. Site

3Briefly, the ancient harbour of Marseilles is located in the centre of the actual French Mediterranean town of Marseilles. Marseilles was founded ca. 600 BC by the Phoceans; it was the first Greek colony in France. The site had been earlier occupied before as testified by a shelly deposit and by the scattered prehistoric discoveries around the Ancient Harbour (Morhange et al., 1996). A bracelet dated from the end of the 7^th century BC and beginning of the 6^th century BC seems to attest a continuous occupation until the arrival of Greeks (Hesnard et al., 1999).

4The Phoceans installed in naturally sheltered creek called Lacydon a significant port located between Spain and Aegean Area. It was a military and trading port as attested by ceramics and amphorae coming from Greece but also from other areas like Etrury or Phoenicians colonies (Hesnard et al., 1999).

5The presence of shipbuilding yards and metallurgical activities (including monetary) is attested around the harbour at the Hellenistic Time (Hermary et al., 1999; Hesnard et al., 1999). At the Roman Period (1^st century BC- end of the 4^th century AD), Marseilles preserved its commercial role and urbanization increased (Hesnard et al., 1999). For more details about the site and the impacts of the development of the harbour, the reader could refer to Morhange et al. (2001).

1.2. Material

6The sediments come from 2 sections F fig. 1). One profile, age dated by ¹⁴C, comes from the Bargemon excavation and covers 5000 years of sediment deposition. The other one come from J. Verne excavation and is dated to the Roman Period (1^st century AD) as attested by archaeological artefacts.

7Sub-samples of samples stored in frozen were taken after decontamination by cutting the outside edges. They were freeze-dried and hand-powdered in an agate mortar.

1.3. Methods

850-70 mg of sediments were digested at 120°C with an HNO₃-HCl-HF mixture in acid pre-cleaned Teflon Beakers. After evaporation, aliquots were dissolved in an HBr (0.5 N) solution. Lead was separated using Agl-X8 resin (50µl). Isotope ratios (²⁰⁴Pb, ²⁰⁶Pb, ²⁰⁷Pb, ²⁰⁸Pb) were determined by standard thermal ionisation mass spectrometry (Hamelin et al., 1990) on a Finnigan MAT 262 at CEREGE, Aix-en-Provence. Mass fractionation was corrected using NBS SRM 981. Pb concentrations were measured by isotopic dilution method using a ²⁰⁸Pb spike solution. Quality of the results were controlled by analysis of the soil standard RTC S 408 (certified value: 95.3 ± 12.4 µg g-1; measured value: 107 ± 1µg g-1). For background knowledge about Pb and Pb isotopes, the reader could refer to Shotyk and Le Roux (in press) and Dickin (1995). About the use of Pb isotopes in archaeology, the reader could refer, for example, to Gale and Stos-Gale (1982), Hosler and Mac Farlane (1996) and to the book of Lambert (1997).

2. Results

• Pre-Greeks sediments (before 600 BC), Bargemon Excavation (Fig. 1 bottom)Pb concentrations are under 15 µg g-1 and the isotopic composition does not vary and is around 1.2 for ²⁰⁶Pb/²⁰⁷Pb.

• Greek sediments, Bargemon excavation (Fig.1 bottom)
  There is a small increase of the Pb concentrations above -2.5m despite 2 samples with very low concentrations. This trend is not correlated with a variation of the Pb isotopic composition or with the mineralogy. These sediments are dated from the Archaic Period and unfortunately, we did not sample sediments from the Hellenistic Period.

• Roman sediments, J. Verne excavation (Fig. 1 top)

• The base of the «Roman» clays (samples D1, D2, D3) has very low Pb concentration (4- 9 µg g-1) and the isotopic ratios are similar to those found in the Greek clays (²⁰⁶Pb/²⁰⁷Pb > 1.200). One possibility is that dredging operations reworked older sediments.

• At the top (D5, D6), Pb concentrations are more than 10 times higher (100-300 µg g-1) than in the bottom clays and ²⁰⁶Pb/²⁰⁷Pb decreases (²⁰⁶Pb/²⁰⁷Pb < 1.183). The sample D4 (²⁰⁶Pb/²⁰⁷Pb=1.198) is intermediate considering only the isotopic signature.

3. Discussion

3.1. Evidences of anthropogenic Pb in the sediments of Marseilles harbour

9Pb isotopic composition for natural sediments is different from the composition of Pb-Cu-Ag ores (Fig. 2). The shift in the isotopic composition of the top sediments of the J.Verne site is a clear evidence of anthropogenic Pb for the period from 0 to 200 AD (= US 195 Jv 10). Moreover the high concentrations of Pb for the samples D5 and D6 could be interpreted as a large pollution event comparable with pollutions in modern harbours (Le Roux, 2001). Assuming that the natural contribution of Pb to the harbour was constant, more than 90% of Pb for the samples D5 and D6 was of anthropogenic origin. This is in good agreement with previous study that shows Pb pollution during the Roman Period.

10The other sediments do not show such a large increase in Pb concentration or a shift in Pb isotopic composition. The isotopic composition in the sediments is typical of natural sediments (Fig. 2) and does no alone show any evidence of Pb contamination. However it could not be excluded that there was Pb contamination in the harbour during the early Greek occupation or even earlier. Indeed there is a small increase of the Pb concentration during the Greek Period. Moreover some Pb ores in Greece have an isotopic composition close to the natural signature of the sediments and therefore it could not be excluded that there was a small contamination not distinguishable with the isotopic composition and only marked by a small increase in concentration. More analyses like Cu or Ag measurements would be required to detect any trace of metal contamination.

3.2. Tracing Pb sources of the Roman contamination

11On the Figure 2, are plotted different isotopic signatures of Pb-Cu-Ag ores, roman artefacts found in the Southern part of France and the samples. The polluted Roman samples are not corresponding to a precise isotopic field but more to a mixing line between different ores, for example from French and Spanish mines. Indeed, mixture of different ores because of recycling for example was common in the Roman Empire. Assuming a mixture of two sources and that one source of Pb has the signature of artefacts found in the Mediterranean French Area (corresponding more or less to mines from Massif Central), we can deduce that the second source is the SE Spanish mines and not the British mines. This is a reasonable hypothesis as Spain was the most important producer during the Roman Period (Nriagu, 1983). However a more complicated system with more than 2 sources could not be excluded and again multi-tracer analysis (Cu, Ag, Sb) will refine our analyses.

Conclusions

12Pb concentrations and Pb isotopes measurements were done in sediments from the Ancient Harbour of Marseilles. Whereas no large change in the Pb chemistry occurred with the foundation of the harbour at the Archaic Period (6^th-5^th century BC), sediments from the Roman Period are polluted as evidenced by higher Pb concentrations and shift in Pb isotopic composition. This pollution has not a clear source and it was probably a mixing of different Pb ores from Spain and France. Measurements of other metals used in the Antiquity like copper and silver will permit to better characterize the history of metallurgical activities.