THE ORIGIN OF TERTIARY GYPSUM DEPOSITS, DOGALI FORMATION, ERITREA, NE AFRICA
Risto A. Kumpulainen(1), Peter Torssander(1) and Beraki Woldehaimanot(2)
1. Department of Geology and Geochemistry, Stockholm University, SE-10691 Stockholm, Sweden. email@example.com
2. Department of Earth Sciences, University of Asmara, P.O. Box 1220, Asmara, Eritrea
Tertiary gypsum-dominated evaporite deposits are known from the coastal areas of the Afar depression in Eritrea (Sagri et al. 1998). The evaporites occur west of Massawa (Fig. 1) as thin (< 1 cm) layers and constitute less than 2-3% of the succession. One of the layers close to Desset reaches the thickness of ca 5 m. The present work provides new evidence that the gypsum in the upper part of the Dogali Forma-tion is continental rather than marine in origin. Sulphur isotope signatures are used to deter-mine the processes and origin of sulphate and thereby the gypsum.
Fig 1. Selenite-type gypsum layers were collected (encircled) along the Dogali – Massawa road (around 15°36´32.7´´N / 39°21’26.7´E. Key: Basement rocks (+), Tertiary basalts (b); Dogali Fm: Mudstones (horiz. lining), Sandy mudstones (stippled), Evaporites etc (e); Desset Fm: Conglomerates etc. (circles). Modified from Sagri et al. (1998) and Ghebreab (1998).
The sequence of rocks west of Massawa begins with Neoproterozoic arc-related basaltic to ande-sitic volcaniclastic sediments and lavas intruded by subduction-related granitoids, metamorpho-sed and deformed by the pan-African orogeny. The Neoproterozoic units are also exposed south of Dogali (Fig. 1) and composed there of pelitic to semi-pelitic rocks with a flat-lying schistosity (Ghebreab 1998).
The Tertiary volcano-sedimentary succession rests unconformably on Neoproterozoic schists and granitoids. The Tertiary succession begins with Oligocene flood basalts and exceeds 500 m in thickness in these areas. The sedimentary succession comprises two formations: the Dogali Formation and the Desset Formation studied in the Desset area by Sagri et al. (1998).
The lower part of the Dogali Formation is com-posed of essentially volcaniclastic mudstones of lacustrine origin. The upper part of the formation contains mudstones, sandstones, conglomera-tes, muddy bioclastic limestones and evaporites (dominated by transparent selenite gypsum). The Desset Formation is dominated by coarse-clastic alluvial sediments and intercalating, marine reefal rocks, potential mud mounds(?) and minor evaporites.
The overall range in delta34S values for the selenite-type gypsum layers (Fig. 2) varies from -7.20 to 22.09 ‰ with an average delta34S = 6.39 ‰ but the delta34S values are very different between the two sections. In the 32-metre section, eleven selenite gypsum layers the delta34S varies from -7.20 to 7.60 ‰ with an average delta34S = 1.02 ‰, whereas in the overlying 100-centimetre section, ten gypsum layers display delta34S values between 6.40 and 22.09 ‰ and average delta34S = 11.12 ‰.
4. Origin of gypsum
Sulfur isotope composition in the gypsum layers suggests different sulphur sources and processes. Modern seawater have very uniform sulfur isotope composition with delta34S = 20.7 ‰ (e.g. Rees, 1978). In fact two determinations of Red Sea water showed delta34S values = 20.69 and 20.74 ‰. Sulphur isotope fractionation during gypsum precipitation is 1.3 ‰. Hence, two of the uppermost gypsum samples in the upper 1-metre section have a sole marine origin with delta34S values of 22.09 and 21.57 ‰.
The bedrock in the Massawa-Dogali area and westwards is composed of igneous rocks (domi-nantly basalts and granites), sedimentary and metamorphic rocks of volcanigenic or sedimen-tary origin. The 34S values of basalts depend upon tectonic setting (Taylor, 1986) and oxidation state but are in general close to 0 ‰ (Sakai et al, 1984). Granitoids and other felsic rocks display larger variation, whereas sedimen-tary rocks exhibit very large variation in delta34S because they are often affected by bacterial sul-phate reduction (BSR). BSR produces iso-topically very light sulphides and heavy sulphate in water and the formed rocks. Hence, oxidation of sulphides and subsequent mixing yield higher sulphate concentration and lowered delta34S values. Sulphate reduction decreases the sulphate concentration but increases delta34S in sulphate which remains in water until it precipitates as a mineral sulphate e.g. gypsum. Negative delta34S values in the gypsum layers arise from oxidation of sedimentary sulphides produced by BSR. Oxidation of sulphides would generally result in delta34S values around 0 ‰. A crustal average delta34S value has been estimated by comparing various reservoirs of sulphur to be delta34S = 7 ‰ (Nielsen, 1978). Variation within the lower and upper gypsum sequence amounts to a 15 ‰ difference.
Fig 2 Sulphur isotope profiles in gypsum; the five lowermost analyses relate to two specimens of selenite unrelated to the two sections above.
In the lowest sequence the highest 34S value is similar as the lowest delta34S in the upper sequence indicate little if any marine input in the lower section.
Thus, the lower gypsum section is dominated by sulphur arising from oxidation processes presumably occurring in both sediments and magmatic rocks. Mixing with crustal average delta34S is also indicated. Several waters west of the study area and on the coastal plain have extremely high SO4 concentration with delta34S values around 7 ‰ (Torssander et al, 2005). The upper gypsum section involves as previously stated almost pure sea water but mostly the delta34S value around the crustal estimate and mixing between the two in some occasions accentuated in the specimens with no stratigraphic control (much thicker layers). It is interesting to note that the overall average delta34S is 6.39 ‰. We suggest that the gypsum evaporites on the coastal plain are only marine in origin to very small part and are better characterised as non-marine evapo-rites.
Ghebreab, W., 1998: Tectonics of the Red Sea region reassessed. Earth Science Reviews, 45, 1-44.
Rees, C. E., Jenkins, W. J., Monster, J. 1978. The sulfur isotopic composition of ocean seawater sulfate. Geochim. Cosmochim. Acta, 42, 377-381.
Nielsen, H. 1978, Sulfur isotopes. In Handbook of geochemistry (ed. Wedepohl K.H.) pp16B1-16B40.
Sagri, M., Abbate, E., Azzaroli, A., Balestrieri, M.L., Ficcarelli, G., Marcucci, M., Papini, M., Pavia, G., Reale, V.,Rook, L., and Tecle, T.M., 1998: New data on the Jurassic and Neogene sedimentation in the Danakil Horst and Northern Afar Depression, Eritrea. Mémoires du Museum national d’histoire naturelle, 177, 193-214.
Sakai H., Des Marais D. J., Ueda A., and Moore J. G. 1984 Concentrations and isotope ratios of carbon, nitrogen and sulfur in ocean-floor basalts. Geochimica et Cosmochimica Acta 48, 2433-2441.
Taylor B.E. 1986 Magmatic volatiles. In Stable isotopes in high temperature geological processes. Vol 16. (ed J.W. Valley, H.P. Taylor and J.R. O’Neill) pp 185-225. Chelsea MI: Mineralogical Society of America.
Torssander, P., Kumpulainen, R.A. and Woldehaimanot, B., 2005: Investigation of water quality in a developing country. SIDA Research Conference Structures of Vulnerability. January 12-14, 2005. Stockholm University. Abstract