Recent Tuffitic Sediments Around Santorini, Greece
The sediments from the Santorini Caldera are enriched in iron and manganese originating from the thermal springs of the Kameni islands. The only other slightly enriched element is Zn whereas the concentrations of Cu, Co, Cr, Ni and Pb are "normal". Examination of the interstitial waters revealed no evidence of a diagenetic change in the sediments.
In this study, the lateral and vertical distribution of iron and manganese will be discussed. Analyses of the surface sediments show a distinct lateral division of the iron from the manganese due to the fact that iron is more readily oxidized. The vertical distribution indicates that the Fe and Mn content is controlled by the "thinning" effect of the high rate of sedimentation and that the deposition of iron-rich sediments from the post-volcanic thermal springs took place during the past 550 years.
Quantitative estimates show that at least 300,000 t Fe and 19,000 t Mn were deposited in the Caldera at an average yearly rate of 550 t Fe and 35 t Mn.
INTRODUCTION
The Santorini archipelago in the southern Aegean Sea forms part of the Cycladic island arc, extending from Corinth via Aegina, Methana, Poros, Milos, Santorini, and the islands of the Dodecanese to the mainland of Asia Minor (Fig. 1). The archipelago is comprised of the two main islands Thera and Therasia, and the smaller island Aspronisi. These islands surround the Santorini Caldera which was formed around 1400 B.C. by an enormous eruption. Two volcanic islands, Nea and Palaea Kameni, which until recently were still active, are situated in the centre of the caldera (Fig. 2).
At present, volcanism is limited to fumarolic activity on Nea Kameni, where solfatars exhalating carbon dioxide and sulphur dioxide are present (Puchelt et al. 1971). Below sea level in the bays of Palaea and Nea Kameni, thermal springs deposit iron - and manganese - bearing solutions as well as carbon dioxide (Puchelt 1972).
The Santorini Caldera is an almost completely closed basin with depths of up to 400 m, extending 11 km from North to South and 7 km from East to West.
The geology of the archipelago has been the subject of geologic -volcanological research for more than a 100 years (e.g. Fouqué 1879). More recent publications are: Nicholls (1971), Pichler & Kussmaul (1972), Puchelt & Schock (1972), Pichler et al. (1972), Günther & Pichler (1973). Until now, the surface sediments of the Santorini Caldera have only been examined from a sedimentological and geochemical point of view by Butuzova (1968, 1969).
Sediments rich in iron from the bays of the Kameni islands were described by Harder (1960) and Puchelt (1972). A comprehensive account of the mineralogy, petrography and tephrochronology of the sediments of the Caldera in their vertical destribution was given by Petersen & Müller (1974). The results of granulometric investigations have been reported by Petersen (1975).
The object of this study is to trace the geochemical evolution in the sediment cores from the Santorini caldera.
SAMPLING AND METHODS OF INVESTIGATION
A piston corer was used in the collection of the samples (tube length 3 m, diameter 10 cm). The cores were collected on two expeditions in June 1971 and October 1972; 11 cores were taken within the Caldera and 13 cores in the surrounding sea area (Figs. 1 and 2). The core lengths vary between 35 and 250 cm. 83 surface samples, collected in 1969 (cf. Fig. 2), were also used for this investigation.
Geochemical analyses were performed on all surface samples, as well as a selection of cores in profile. Whole untreated samples were used in order to obtain the greatest accuracy possible, and for a better comparison of the individual values the clay fractions < 2 μ were also taken into account in the investigations.
The samples were treated with hydrochloric acid and measured with an atomic absorption spectrometer with regard to the following elements: Fe, Mn, Zn, Cu, Co, Cr, Ni, Pb, Ca, Mg, Sr (as well as K in the interstitial water). With this method, it was possible to measure the elements soluble in hydrochloric acid which are found on the surface of the particles. These are either adsorbed elements or amorphous hydroxides and, to a lesser extent, minerals. Comparative measurements performed on the same material, i.e. pumice from the island, showed that, using this method, the concentration of elements removed from the rock was very small. This concentration, which can be derived from the glass and the iron minerals of the pumice (e.g. magnetite), shall be termed the "background concentration". The total concentration less the background value is equal to the volcanogenic portion of an element which has accumulated on the surface of the sedimentary particles. In the following, the term "volcanogenic" will be used to designate those elements which are derived from the thermal springs of the Kameni volcanoes and whose origin probably lies in a leaching of the volcanic rocks (Schorin 1972).
Some samples were examined quantitatively by X-ray fluorescence analysis (XRF). The amount of organic material was determined in a few samples by combustion of the dry sample, whereby the loss of weight was measured.
RESULTS
- 1. The concentration of elements in the sediments
Table 1 gives a survey of the concentrations of the different elements.
It shows the averages and the variations (minimum and maximum values), the background values from the main sedimentary component -the pumice - as well as the concentrations in the clay fractions.
The concentration of iron in the Caldera sediments varies widely from very low concentrations to very strong enrichment. The same applies to manganese.
The values for zinc show little variation with only occasional enrichments, and copper, cobalt, chromium, nickel and lead are present only in trace amounts. Calcium and magnesium vary to a large extent, whereas strontium is only present in relatively small quantities. The results of the chemical analysis are given in Tables 2 to 6.
The considerable differences of the Fe, Mn and Ca values were due to three factors:
- grain size;
- lateral distribution;
- vertical distribution in the sediment.
Except for Ca (carbonates), lithological factors have no influence, since the Caldera sediments are extremely homogenous from the mineralogical point of view and consist of more than 90 per cent pumice (Petersen & Müller 1974).
- 2. Grain size dependency of metal concentrations
In the sediment fractions 20 - 6·3, 6·3 - 2 and < 2μ, a pronounced increase in all elements measured was observed: the smaller the grain size, the higher the concentration (Figs. 3 and 4). The clay fraction contains concentrations 5 to 10 times as high. The iron concentration increases in sample 11/1 from 1.72 % in the coarse silt fraction to 7.84 % in the < 2 μ fraction. At the same time, the manganese concentration increases from 0.15 % to 1.27 %.
The concentration in the total sample of 11/1 is about 3.5 % for iron and 0.48 % for manganese.
The high concentrations in the finest fractions are caused by the larger surface area on which the elements are adsorbed. In addition, the amorphous iron hydroxides precipitated from the water are also particularly prevalent in the small grain sizes. The distribution of the elements in the sediments indicates that the amount of iron derived from magnetite in the rocks is very small (background) since the magnetites in the sediments have a grain size of > 20 μ (Raab 1974).
- 3. Iron and manganese in the surface sediments
A comparison of the sediments within the Caldera with sediments from outside the Caldera shows that of the elements investigated, only iron and manganese are enriched within the Caldera. This is shown by the distribution map of iron in the surface sediments (Fig. 5). The highest iron concentrations are to be found between Nea and Palaea Kameni near the bays with the iron and manganese-producing springs (Puchelt 1972). Increased iron concentrations are also present in the northern as well as in the southern Caldera basin. The high iron concentrations outside the Caldera close to the coast and in the shallow water along the shelf between Aspronisi and Acrotiri can be explained by magnetite accumulations and/or rock fragments rich in iron. These sediments also have coarse grain sizes and can not be compared, therefore, with the fine basin sediments.
Similar observations are possible in the distribution of the manganese concentrations in the sediment surface. As in the case of iron, the manganese concentration is enriched within the Caldera (Fig. 6). Relatively high concentrations are also present in the northern and southern Caldera basins. Between the islands Palaea and Nea Kameni, the manganese concentrations are very low.
The origin of manganese as well as of iron can probably be attributed to the activity of the springs between the islands.
- 4. Vertical element distribution
In order to examine the element distribution in the vertical section, cores were selected from the northern and southern Caldera. The grain size composition of these various cores differed very little. Core C 11 and C 13 stem from the northern and Core C 15 from the southern Caldera. These cores show characteristic changes in the vertical distribution of the Fe, Mn, as well as Ca and Mg contents (Figs. 7 to 9). The elements Zn, Co, Pb do not correspond with this, while Ni, Cr and Cu show only slight variations. In all cores, the concentrations of iron and manganese are highest in the upper sections and clearly lower in the lower sections between a depth of 50 cm and 170 cm (longest core). There is no correlation with the oxidation zone which is limited to the uppermost 10 cm.
The high iron and manganese concentrations (hereafter designated as "higher" and "lower" maximum) in the upper 50 cm of the cores are separated by a minimum layer at a depth of 25 and 30 cm. There is a positive correlation between the iron and manganese concentrations in the different profiles with the exception that the manganese concentrations are especially high in the lower maximum, while iron exhibits about the same concentrations in the upper and lower maxima. The average conentrations of iron in the lower core sections of the three cores vary between 0.7 and 1.7 % iron in the total sample. In the upper part of the core, the iron concentrations reach maximum values of 4.1 % Fe in C 11 and 2.7 % Fe in C 13 as well as 2.5 % Fe in C 15, in each case related to the total sample. In contrast, the minimum layer contains only 1.7, 1.3 and 0.7 % Fe, respectively, which corresponds to the concentrations in the lower part of the core in the total samples of C 11, C 13 and C 15.
The maxima and minima of the manganese concentrations are even more pronounced. Where the manganese concentration in the middle section and in the minimum layer core C 15 is about 500 ppm, an increase is seen in the lower maximum to more than 8,000 ppm, and in the upper section to 4,800 ppm manganese in the total sample.
The distribution observed within the three cores is independent of the grain size distribution; the percentage of clay fraction and mean size vary insignificantly. Determinations of the clay fractions < 2 μ show the same variation, although many times exaggerated (Fig. 9).
Parallel to the iron and manganese concentrations, the Ca and Mg contents vary accordingly. This is caused by the carbonate contents, the variation of which is similar in all three cores (cf. Petersen & Müller 1974).
- 5. The sulphide layer in core C 13
In almost all cores, an upper oxidation zone with reddish-brown colouring can be observed. Below a depth of 10 cm, this colouring, changes to a shade of grey-green which continues down to the bottom of the core. The colouring is due to the different stages of oxidation of iron. Eh-measurements carried out on board ship showed values of + 20 mV in the oxidation zone and values of about -10 mV in the reducing zone. The upper or oxidation zone, contains Fe (III) oxides and hydroxides, whereas in the lower reducing zone, the bivalent Fe determines the colour of the sediment. Fe (II) in this zone is stable in the form Fe (II)-hydroxide and Fe(II)-carbonate (both X-ray amorphous). Black spots of iron sulphide can be observed in the reduced zone. Only in one core, C 13, of the northern Caldera is a black layer of about 8 cm thickness to be found.
The Fe-sulphides oxidize very rapidly when they come into contact with air and they are X -ray amorphous. The black colour of the sediment is not a measure of the iron concentration, since even very small amounts of finely distributed Fe-sulphides will colour the sediment black. The iron concentration in this layer is less than 1.9 %, compared with 2 to 2.5 % in the upper and underlying layers.
Sulphate-reducing bacteria produce the sulphur hydrogen necessary for the formation of sulphide. The organic substances essential for the growth of the bacteria are enriched in this layer by a factor of 3 (13.5 %) compared with the upper and underlying samples (4.3 %).
- 6. Chemistry of interstitial waters
The interstitial waters contain the same amounts of magnesium and potassium as the supernatant sea water, only the calcium concentration of approximately 700 ppm as compared with 400 ppm in sea water is somewhat higher. This enrichment can easily be explained by traces of carbonate in the sediment. According to Berner (1971), as little as 0.1 % CaCO3 with 50 % sea water in the sediment results in an increase of 100 % in the calcium concentration to 800 ppm.
The pH-value of the interstitial water varies between 7.25 and 8.3 and, in most of the samples, is between 7.5 and 8.2. This means that the pH-values of the interstitial water correspond with those of normal sea water. Hay (1966) recorded values from 7.5 to 8.1 for shallow marine water. No change in the water chemistry or the pH-values could be detected within the core profile.
DISCUSSION
A comparison of the minor and trace element concentrations in the clay fraction of the sediments with the average values for shale (Turekian & Wedepohl 1961) demonstrates that apart from Fe and Mn, which are considerably enriched, only Zn shows an increase by a factor of 2 to 5. The concentrations of Cu, Cr, Ni and Pb correspond to the average shale composition. Butuzova (1969) compared similar sediments from outside the Caldera with those from within the Caldera and found significantly higher values for Fe, Mn and somewhat higher concentrations for V and Cu in the caldera sediments. These values, however, although somewhat elevated, do not exceed the "average shale" values.
The origin of iron, manganese and zinc can probably be traced to volcanic solutions from the hot springs of the Kameni islands. These springs produced only Fe and Mn; Cu and Zn are only found in quantities which do not exceed the "average shale" values (Puchelt 1972). The solutions contain hydrosulphuric acid, their Fe and Mn concentration is derived from the leaching of andesites and dacites of the islands (Schorin 1972). A comparison of the sediments from the bays of Palaea Kameni with the normal basin sediments of the Caldera shows a general dilution of the Fe and Mn concentrations as a result of their wide lateral distribution within the Caldera.
According to Puchelt (1972), average concentrations of Fe and Mn within the bay sediments are 30 - 40 % (and less) and 100 ppm, respectively, whereas the caldera sediments have only 2 % Fe and 1000 - 3000 ppm (maximum 8000 ppm) Mn-concentrations.
Whereas iron was depleted in the Caldera sediments by a factor of more than 10, the manganese is somewhat enriched. The differences in the lateral distribution of the iron and manganese show that iron oxidizes more easily than the manganese. Iron is hardly soluble in sea water unless under reducing conditions.
It is precipitated immediately in the bays of Nea and Palaea Kameni. However, in coagulated form as Fe (III) hydroxide, it can still be transported and continues to be distributed in the sediments of the Caldera. Manganese, on the other hand, is only precipitated at greater distances from the source. Thus, the iron and the manganese are laterally separated and the iron is enriched in a locally limited area, namely in the sediments between Palaea and Nea Kameni. Such a separation is not complete, since iron in its colloidal state can be transported over the same distances as manganese. Most of the "volcanic" iron and manganese remains in the almost closed Caldera basin.
The vertical distribution of the Fe and Mn concentrations cannot be explained by diagenetic processes. Since the sediment is still in a non-compacted state and the interstitial waters are so alkaline as to dissolve iron and manganese, the iron and manganese enrichments in the vertical section of a core cannot be explained by ascending solutions as described by Rothe (1971) and Lynn & Bonatti (1965) in marine sediments.
Further results of our examination contradict the possibility of a diagenetic iron and manganese enrichment in the oxidation zone of the sediment as was reported by Butuzova (1969) :
a. There are two maxima which are separated by a minimum layer. The lower maximum (richer in manganese) is situated in the reducing zone.
b. The iron and manganese concentrations show a positive correlation with the carbonate contents.
c. The sulphide layer has only a low Fe concentration.
In order to explain the vertical distribution of the Fe, Mn and Ca concentrations, it is necessary to consider the different components of which the sediment is composed.
The three main sediment constituents are the terrigenous component (pumice), the biogenic (carbonate) and the volcanogenic (Fe and Mn) component.
They are present in the proportions 90 to 5 to < 5 %. A change in the proportion of one component affects the concentration of the other two. Two possibilites of interpretation can be derived from this:
- Change in the rate of sedimentation of the main sedimentary component (detrital pumice). The maxima of Fe and Mn are explained by lower rates of sedimentation while the minimum layer represents a periodically increased deposition of detrital pumice and a consequent dilution. This theory of dilution is backed up by the fact that the carbonate concentration also reaches its minimum in the minimum layer. Another explanation must be found for the fact that dilution occurs in the Northern as well as in the Southern basin and must therefore have the same cause. At the same time, the increasing trend of Fe and Mn concentrations from the lower portion to the upper portion of the core, as could be seen in all three cores, must also be explained.
- The variations in the Fe and Mn concentrations is due to changes in volcanic activity. The changing carbonate concentration can then be explained by changes in the carbonate production resulting from the volcanic influence on the ecological conditions of the organisms.
A factor analysis of the Fe, Mn and Ca concentrations revealed an extensive correlation of Fe and carbonate Ca and, to a certain extent, of Mn (Table 7). This correlation can only be observed in the upper portions of the core and not in the lower part. The profile must therefore be divided into two sections; according to the factor analysis, the limits lie between sample 8 and 10 in Core C 11 and at sample 12 in C 13 and in Core C 15 between sample 11 and 12.
The vertical distribution, therefore, must be described and interpreted as two separate units. In the upper section, the volcanic deposition of Fe and Mn is more or less continuous. The same holds true for the carbonate production.
The distribution is decisively influenced by the rate of sedimentation. The cataclysmic event indicated by the minimum layer might have been an earthquake which shook the whole of the island and led to an increased erosion of pumice.
A dating of the Ca and Fe/Mn minimum layer is possible by tephrochronological methods using the Kolomvos pumice, as this occurs only a little above the minimum layer (Petersen & Müller, 1974). The strong earthquake which shook the islands and was recorded by the sediments, occurred after the beginning of the eruption of the Kolomvos volcano, 1650 AD. These catastrophic earthquakes, together with the subsequent pumice erosion were documented by eye-witnesses (Fouqué 1879).
As for the lower core section, the second interpretation is applicable. Here, no iron or manganese enrichments are to be found; the lower iron and manganese values correspond with the background values. This means that during the time of deposition, the iron and manganese sources were not active, or at least only to a minor extent.
The beginning of the iron and manganese production due to post-volcanic activity of the Kameni islands can therefore be correlated with the beginning of the upper section. This boundary between the two units occurs at a depth of 45 to 50 cm. With the aid of calculated rates of sedimentation (cf. Petersen & Müller 1974), an age of 550 ± 50 years B.P. can be estimated. It is impossible to reconstruct in this study exactly the point in time when the post-volcanic activity of iron-bearing solutions commenced. It is evident, however, that the volcanic springs on the Kameni islands have been active for at least 500 years.
Quantitative estimation of the "Volcanic" iron and manganese in the Santorini-Caldera
In order to estimate the minimum amount, a "volcanic" metal concentration was calculated in the upper 45 cm of the sediment with average concentrations of 1.6 % Fe and 0.1 % Mn. The surface of the basin was assumed to be 25 km 2 (total Caldera approx. 61 km 2). The density of the total sediment was determined to be d= 1.7 (pumice = 2.4), assuming a maximum water content of 50 %. The minimum amounts of metals deposited by the thermal springs of the Kameni islands are thus 304.000 t Fe and 19.000 t Mn. Over a period of 550 years, this is equivalent to an average yearly rate of 553 t Fe and 35 t Mn.
According to Puchelt (1972), about 500 t Fe2O3 and a little less than 7 t of Mn have been deposited in the largest of the eight bays of the Kameni islands, known to be rich in iron sediments.
These estimates are of the extent to which the thermal springs of the volcano are able to deposit iron and manganese. It is also clear that despite direct precipitation, iron can still be transported over long distances and that a deposition of iron, together with manganese, takes place.
----------------------------------
| For figures and tables please refer to book. | |
| Figures and tables mentioned in this paper: | |
| Fig. 1: | Piston core sampling stations in the Santorini area. |
| Fig. 2: | Map of the Santorini island group with sampling staations. |
| Fig. 3: | The iron and manganese content of various grain size fractions of sample 11/10 from inside the caldera. |
| Fig. 4: | The zinc, copper and nickel contents of various grain size fractions of sample 11/10 from inside the caldera. |
| Fig. 5: | The iron content of the surface sediments. |
| Fig. 6: | The manganese content of the surface sediments. Important: all figures on the chart have to be multiplied by a factor of 10 to obtain the ppm reading!. |
| Fig. 7: | Graphic plot of iron, manganese and calcium contents of core C 11. |
| Fig. 8: | Graphic plot of iron, manganese and calcium contents of core C 13. |
| Fig. 9: | Graphic plot of iron, manganese and zinc contents of core C 15. a= total sample, b= fraction < 2μ. |
| Table 1: | A compilation of the concentrations of the various elements analysed including the results of all samples. The first of the two average values given includes all samples from inside the caldera and the second value (in brackets) from outside the caldera. |
| Table 2: | Chemical analyses (HCl treated) of cores 15 and 12. |
| Table 3: | Chemical analyses (CHl treated) of cores 13 and 14. |
| Table 4: | Chemical analyses (HCl treated) of the clay fraction of cores 11, 15 and 33. |
| Table 5: | Chemical analyses (HCl treated) of heavy metals in different cores. |
| Table 6: | Composition of interstitial water of Core C 33. |
| Table 7: | Factor analyses of Mn, Fe and Ca in samples of Cores 11, 13 and 15. |
----------------------------------
| Source: | "Thera and the Aegean World I" |
| Papers presented at the Second International Scientific Congress, Santorini, Greece, August 1978 | |
| Pages: | pp. 311 - 332 |
| Written by: | - M.D. Petersen - G. Müller |
| Institut für Sedimentforschung, Heidelberg, West Germany | |
| Book information: | |
| ©Thera and the Aegean World | |
| ISBN: | 0 9506133 0 4 |
| Published by: | Thera and the Aegean World, 105-109 Bishopsgate, London EC2M 3UQ, England |
| Editor: | C. Doumas |
| To order the book from amazon.co.uk: | http://www.amazon.co.uk/exec/obidos/ASIN/0950613304/qid=1141298899/sr=1-2/ref=sr_1_0_2/203-4397765-4475969 |