Geochemistry of quaternary fluvial sands from different tectonic regimes

Geochemistry of quaternary fluvial sands from different tectonic regimes

88 GEOCHEMISTRY OF THE EARTH'S SURFACE AND OF MINERAL FORMATION 2nd INTERNATIONAL SYMPOSIUM, July, 2-8, 1990, Aix en Provence, France. GEOCHEMISTRY ...

282KB Sizes 2 Downloads 33 Views

88

GEOCHEMISTRY OF THE EARTH'S SURFACE AND OF MINERAL FORMATION 2nd INTERNATIONAL SYMPOSIUM, July, 2-8, 1990, Aix en Provence, France.

GEOCHEMISTRY OF QUATERNARY FLUVIAL SANDS FROM DIFFERENT TECTONIC REGIMES

KROONENBERG S.B. Department of Soil Science and Geology, P.O. Box 37, 6700 AA Wageningen, the Netherlands.

Why bulk geochemistry of Quaternary fluvial sands? The life history of a fluvial system is recorded in its sediments. Accidents within the drainage basin such as uplift, volcanism, stream piracy and climatic change result in differences in the composition of the sediments. Bulk geochemical analysis for characterization of sediment composition is commonly used in consolidated Pre-Quaternary sediments (Ronov, 1980). It is also common practice in actual fine-grained sediments in order to study the dispersion of specific elements for exploration or environmental purposes (Horowitz, 1982). However, there is surprisingly little information on the chemical composition of unconsolidated Quaternary sands. Chemical analysis of an unconsolidated sand sample at f'n'st sight seems odd and superfluous, as sands can be studied much easier by mineralogical analysis. But actually it is much more difficult to obtain information on the bulk composition by mineralogical analysis than by chemical analysis. This is due to the following factors (1) mineralogical analysis is carried out as a rule on fractions < 0.5 mm because larger grains are cumbersome in grain slides and difficult to determine optically; there is little information on the composition of coarser grains; (2) often fractionated analysis is necessary (light/heavy fractions, opaque/transparent grains, fine/coarser fractions); (3) the composition of opaque fractions and rock fragments are not routinely determined, and special techniques as feldspar staining are necessary for analysis of light fractions. Geochemical analysis of bulk samples obviates much of these problems, and may provide many data

which often are complementary to mineralogical analysis sampling and laboratory analysis. The bulk geochemistry of sandy fluvial deposits records (1) source area characteristics; (2) differentiation during sedimentary transport, and (3) postdepositional processes such as weathering, soil formation and diagenesis. If the relation between tectonic history and sediment geochemistry is to be studied, the effects of sorting and post-depositional processes must be well-known. Therefore considerable sampling is necessary. Sorting can best be studied when several samples of different grain size are taken at a single spot. The study of downstream sorting requires also several sites alongstream to be sampled. The study of weathering and soil formation may require sampling at different depths within single soil profiles, or, in the case of fluvial terrace chronosequences, sampling at constant depth in terraces of different age. The increase of the number of samples is counteracted by the rapid laboratory treatment and easy statistical analysis of the data. In the laboratory secondary iron coatings, organic matter and calcium carbonate are removed, the latter only if working in areas of siliceous sands essentially. For the 136 samples reported here, the fraction < 16 lim has also been removed in order to avoid the effect of secondary accumulation or neoformation of clay. About 1 g of sediment, obtained by quartering, is smelted with lithium borate and analysed with X-ray fluorescence spectroscopy, on major and trace elements. In our case studies we have included about 25 elements, but this paper refers to major elements only.

GEOCHEMISTRY OF THE EARTH'S SURFACE AND OF MINERAL FORMATION 2nd I N T E R N A T I O N A L SYMPOSIUM, July, 2-8, 1990, Aix en Provence, France.

Relating geochemistry to mineralogy The site of the elements in the sand minerals sands can be determined in various ways : (1) by multivariate statistics, especially Principal Component Analysis. Highly correlated element groups are matched with published average chemical compositions of minerals known to occur in the sediments; (2) by statistical correlation of element analyses with mineral countings of the same samples; (3) by separating individual mineral species, analysing them chemically and multiplying them with their abundance in the sands as determined by mineral countings; (4) by recalculating the analyses to normative compositions as is usual in igneous and metamorphic petrology. Some results are given by Moura & Kroonenberg (1988) and Kroonenberg et al. (1988).

Bulk sand geochemistry and tectonic and climatic setting In spite of the large range of values found within individual drainage basins due to varying effects of sorting, weathering and temporal changes in sediment composition, it appears to be possible to cluster sediments of each basin or formation according to their tectonic and climatic setting by their major elements composkion. The triangular plot used (Fig. 1) portrays weight percentages SiO2 (divided by 20 to show separation between clusters better), Na20+K20 and TiO2+MgO+Fe203. The rationale for this choice is that in stable cratonic areas, weathering leads to residual enrichment of quartz (Si) and depletion of feldspars and micas (Na, K) and ferromagnesians (Ti, Mg, Fe), especially in humid tropical climates. Recent uplift in such areas leads to admixture of fresh rock fragments. In active orogenic belts and spreading rifts volcanism contributes much unstable ferromagnesians and feldspars. Several sedimentary formations from drainage basins in four different tectonic-climatic settings have been studied in this way (Fig. 1). (1) Stable Precambrian Shield areas in (semi)humid tropical areas : these are characterized by strong chemical weathering, leading to increasing residual enrichment of quartz (cf. Franzinelli & Potter, 1983). Tertiary sands in the Caqueta drainage basin (cluster A in Fig. 1) derived from highly weathered Guiana Shield rocks characterize this environment (Moura & Kroonenberg, 1988; Hoorn, 1988; Kroonenberg et al., in press). (2) Uplifted Caledonian and Hercynian Massifs constitute the framework of extra-Alpine Europe.

89

Though the peneplain remnants have developed rather strongly weathered cover sediments due to (sub)tropical Tertiary weathering, stress related to the Alpine orogeny caused uplift and rejuvenation of drainage basins, leading to increasing admixture of fresh bedrock in Pleistocene river terraces. Two clusters derive from this environment. (B) and (C) are sands derived from the uplifted low-grade metasedimentary Ardennes. (B) represents shale-rich Pleistocene Meuse sediments from South Limburg, the Netherlands, with little admixture of old Tertiary sediments; (C) represents Pleistocene Meuse sediments from North Limburg, the Netherlands, with a considerable admixture of Tertiary quartz-fich Rhine sediments (Moura & Kroonenberg, in press). (3) Continental rift valley in te perate region : (D), rE) and (F) are from the Allier River in the Limagne rift valley of the French Central Massif. This fiver mainly drains Caledonian-Hercynian crystalline basement and Miocene-Holocene alkalibasaltic volcanoes. Volcanism is related to the development of the rift valley as a result of Alpine stress in the EoceneOligocene. Cluster (D) represents Late Tertiary or Early Pleistocene pre-uplift sediments (Lezoux sands) with high amounts of quartz and minor Kfeldspar and muscovite. (E) sands are Pleistocene and Holocene basementrich alluvial terrace sediments with minor volcanic influence, (F) Pleistocene- sands are very rich in alkalibasaltic rock fragments (Kroonenberg et al., 1988; Veldkamp & Kroonenberg, 1989). (4) Active volcanic orogenic belts as found around the Pacific are represented by exclusively volcanic Holocene fluvial sands of andesitic composition in the Atlantic Zone in Costa Rica (cluster G). The fluvial sediments in the Atlantic zone are exclusively deposited by rivers draining the Quaternary stratovolcanoes of the Central Cordillera. Each cluster represents a specific life history of a drainage basin or sedimentary formation. The spread within each cluster is related to the effects of sorting, weathering and other local variations. The importance of sorting is shown by the extreme values found in placers (H). But there is also a general trend to be deduced. The Tertiary Amazonian sands (cluster A) can be regarded as one end member, reflecting extreme pre-sedimentary weathering in a stable cratonic area. The Holocene Costa Rican sands (G) represent the opposite endmember, of pure andesitic sands in an active subduction setting. All other clusters are mixed populations. Clusters E, F and G show a striking alignment, indicating a roughly constant

90

GEOCHEMISTRY OF THE EARTH'S SURFACE AND OF MINERAL FORMATION 2nd INTERNATIONAL SYMPOSIUM, July, 2-8, 1990, Aix en Provence, France.

SIO2120

rio2

+MgO-_~) +Fe203 o

,

"

. o.z

.

.

. o,

.

. o.6

.

o.e

\ ~- K'\'-20+Na20

Fig. 1. Plot of weight percentages of SiO2 (actual values divided by 20 for better visibility) Na20+K20 and TiO2+MgO+Fe203 (all Fe determined as Fe203). A: Tertiary quartz sands (Amazonian provenance), Caqueta River, Colombia. B: Pleistocene Meuse + Rhine sands, North Limburg, the Netherlands. C: Pleistocene Meuse sands, South Limburg, the Netherlands. D: Plio-Pleistocene quartzofeldspathic Allier sands, France. E: Pleistocene-Holocene basement-rich, basalt-poor Allier sands, France. F: Pleistocene alkalibasalt-rich, basement-poor sands, Allier River, France. G: Holocene andesitic sands, Atlantic Zone rivers, Costa Rica. H: Placers, Costa Rica Atlantic beach and Caqueta river, Colombian Amazones (almost purely ferromagnesian sands, shown for comparison).

value for S i O 2 / K 2 0 + N a 2 0 around 14. This is a normal ratio for igneous rocks, including basement granites, andesites and basalt. The Allier sands (E, F) are therefore hardly contaminated anymore by admixture o f the residual quartz-rich sands o f (D). While (E) is m a i n l y d e t e r m i n e d by granitic components generated during quaternary uplift and dissection o f the basement, (F) shows increasing volcanic influence. The Limburg sands (B and C) have higher S i O 2 / K 2 0 + N a 2 0 ratios indicating admixture of mature quartz-rich Tertiary Rhine sands (especially B), or admixture o f fresh sedimentary rock fragments which already are more mature (shale and quartzites from the Ardennes Massif in the case of cluster C). CONCLUSIONS The trend o f decreasing maturity from cluster A to G records increasing Quaternary tectonic activity, including uplift o f old weathered peneplains and

rifling, and increasing volcanic activity, both related to the most recent phase o f the worldwide Alpine diastrophic cycle. The g e o c h e m i s t r y o f these Quaternary continental sands therefore record part of the last tectonic megacycle in the same way as shown by Ronov (1980) for the whole Phanerozoic history. REFERENCES

Franzinelli E. & P.E. Potter 1983 - Petrology, chemistry and texture of modern river sands, Amazon fiver system. J. Geol., 91: 23-29. Hoorn, M.C. 1988 - Nota preliminar sobre la edad de los sedimentos terciarios de la zona de Araracuara (Amazonas) Bol. Geol. 29, 2 : 87-95. Horowitz, A.J. 1988 - A review of physical and chemical partitioning of inorganic constituents in sediments - USGS Circular 969 : 7-22. Kroonenberg, S.B., M.L. Moura & A.T.J. Jonker 1988 Geochemistry of the sands of the Allier fiver terraces, France -

GEOCHEMISTRY OF THE EARTH'S SURFACE AND OF MINERAL FORMATION 2nd I N T E R N A T I O N A L SYMPOSIUM, July, 2-8, 1990, Alx en Provence, France.

Geologie & Mijnbouw, 67: 75-89. Moura, M.L. & S.B. Kroonenberg 1988 Major and minor elements geochemistry and miheralogy of four soil profiles from Araracuara, Colombian Amazones. Catena 15,1: 81-97. Ronov, A.B. Osadocnaya obolocka zemli. Izdatel'stvo

91

"Nauka". Moskva 1980, 80 pp. Veldkamp A. & S.B. Kroonenberg 1989 - A comparison of the sand geochemistry of the Allier and Dore terraces, Limagne Rift Valley, France 2nd Int. Geom. Congr. Frankfurt, Abstracts. p. 304.