EXTENDED SUMMARY The main objectives of this thesis were: (i) the mapping of the current status (October 2010) of the surface lacustrine sediments of Lake Kastoria, (ii) the mapping of the older state of sediments, as recorded in subsurface lacustrine sediments, (iii) the mapping of the chemical composition of stream and floodplain sediments that are deposited in the lake, and (iv) the comparison of the geochemical results from Lake Kastoria with those of other Hellenic lakes. The purpose of this study is, apart from the diachronous recording of the chemical composition of lake sediments, the geochemical results of the surface lacustrine sediments will constitute the geochemical reference baseline for comparing the results of future geochemical surveys, and in the case of an unexpected extreme incident. For the implementation of the aforementioned objectives, the stream sediment geochemical data of 432 active stream sediment samples of the reconnaissance geochemical survey, performed by the Institute of Geology and Mineral Exploration (I.G.M.E.), under the framework of the project "Geochemical Survey of Hellas" (Public Investments Programme 461700), which are within the Lake Kastoria watershed were used. These results were complemented by the collection of different sample types under the frame of the project “Bathymetry and Sedimentology Studies over the Bottom Surface of Lake Kastoria”, which was assigned to I.G.M.E. by the Prefecture of Kastoria, i.e., stream sediment sampling from three sites on Xiropotamos stream for comparison with the results of the previous survey; collection of 18 floodplain sediment samples from six vertical profiles with the aim to study the chemical composition of sediments that are deposited in the lake during flood events; sampling of lake-bottom sediments from 112 sites, and sampling of 42 subsurface lake sediments from 14 cores. In this geochemical survey of Lake Kastoria sediments, only the inorganic chemical elements were studied: Ag, Al, As, B, Ba, Be, Bi, Ca, Cd, Ce, Co, Cr, Cu, Fe, Ga, Ge, Hg, K, La, Li, Mg, Mn, Mo, Na, Nb, Ni, P, Pb, Rb, Re, S, Sb, Sc, Se, Sn, Sr, Ta, Te, Th, Ti, Tl, U, V, W, Y, Zn and Zr. In addition, pH, total C and S were determined on only the lake-bottom sediment samples. Rigorous quality control is one of the cornerstones for the success of any geochemical mapping programme, and was installed during the planning stage of the October 2010 survey. Quality control consisted of: (i) randomisation of field sample numbers prior to the collection of lake-bottom sediment samples; (ii) collection of field duplicate lake-bottom sediment samples at a rate of 1 in 10 field samples, i.e., 10% duplication; (iii) splitting of the routine and field duplicate lake-bottom samples into sub-samples, and spreading them randomly within the analytical batch (iv) analysis of laboratory (ICP-5) and international (Till-3 & Till-4) standards; (v) repeat analysis of every 10th sample of the analytical batch, and (vi) analysis of blank samples. Upon receiving the analytical results, their quality was assessed by using different statistical techniques. Overall, the quality of the analytical results is of good standard and fit for the purposes of this study. However, some quality control issues do exist for some elements, as for example, poor precision, which is due to most samples having concentrations near to the detection limit of the analytical method for such elements as Ag (0.01 mg/kg), B (5.0 mg/kg), Bi (0.02 mg/kg), Cd (0.01 mg/kg), Ce (0.1 mg/kg), Ge (0.1 mg/kg), Hg (0.005 mg/kg), Hg (0.01), Na (0.01%), Se (0.2 mg/kg ), Ta (0.01 mg/kg), Te (0.02 mg/kg) and Tl (0.02 mg/kg). The analysis of variance has shown that for Te, Mn, Se, S (Leco), C (Leco), Na and Ta the geochemical variance is poor (20% at the 95th confidence level) are Ba, Ag, Hg, S, Ge, Na, Te, Ta, S (Leco), C (Leco) and Se, with the greatest proportion of the uncertainty to be ascribed to the analytical method, except for Ag and S, which is due to sampling. Further, the extractability of chemical elements, determined on Lake Kastoria samples, was estimated by utilising the analytical results of the internal (ICP-5) and international standards (Till-3 & Till-4) in relation to their recommended values. The extractability provides a measure of the total element content in each sample that is extracted by hot aqua regia. All the aforementioned quality control issues were taken into account during the interpretation of the analytical results. For each determinand, determined on the stream sediment and lake-bottom sediments, spatial distribution maps were plotted, using the method of proportional-size dots according to the concentration of each determinand. The proportional-size dots were plotted over (a) the lithology with the stream sediment sample sites, and (b) the grain-size distribution map with the lake-bottom sample sites. The analytical results of samples from the lake sediment cores and floodplain sediment vertical profiles were plotted as vertical histograms, according to their concentration and sampling depth interval, starting from their exact sampling site coordinates. In addition, maps of factor scores of the stream and lake-bottom sediment samples were plotted. For each factor two maps were plotted, i.e., the normal factor scores and the inverse factor scores, since all extracted factors were bipolar. Then, follows the description of the results of: (a) stream sediment samples (Ag, Cd, Co, Cr, Cu, Hg, Mn, Mo, Ni, Pb, Sr, V & Zn), (b) lake-bottom sediment samples (pH, Ag, Al, As, B, Ba, Be, Bi, Ctotal, Ca, Cd, Ce, Co, Cr, Cu, Fe, Ga, Ge, Hg, K, La, Li, Mg, Mn, Mo, Na, Nb, Ni, P, Pb, Rb, Re, S, Sb, Sc, Se, Sn, Sr, Ta, Te, Th, Ti, Tl, U, V, W, Y, Zn & Zr), and (c) lake core and floodplain sediment samples (for the same elements as the lake-bottom sediments, except pH and Ctotal). The median of the stream and surface lacustrine sediments geochemical results are compared with the corresponding mean value of the upper continental crust. Further, a comparison in the form of ratios of element median values of all Lake Kastoria sediment sample types is made with the corresponding median values of the Geochemical Atlas of Europe, namely, the results of the: Kastoria stream sediment samples are compared with the corresponding European and Hellenic stream sediment samples of the Geochemical Atlas of Europe, and Kastoria lake-bottom, lake-core, and floodplain sediment samples are compared with the corresponding European and Hellenic floodplain sediment results of the Geochemical Atlas of Europe. The geochemical results of Lake Kastoria are placed, therefore, in the context of the Global, European (continental scale), and Hellenic (national scale) geochemistry. These ratios of median values give an indication about the enrichment or depletion in element concentrations of the Lake Kastoria sample types with respect to upper continental crust, European (continental) and Hellenic (regional) geochemical results. As an example, the ratios of the median values of Lake Kastoria lake-bottom sediments in relation to the corresponding: European floodplain sediment samples of the Geochemical Atlas of Europe show elevated concentrations with respect to As, Ba, Be, Ca, Ce, Co, Cr, Cu, Fe, Hg, La, Li, Mg, Mn, Mo, Ni, P, Pb, S, Th, Tl, U, V and Zn. Hellenic floodplain sediment samples of the Geochemical Atlas of Europe exhibit higher values with respect to As, Ba, Ce, Co, Cr, Cu, Fe, Ga, Hg, La, Mg, Mn, Mo, Ni, P, Pb, S, Sb, Sn, Th, Tl, U, V and Zn. These differences in relation to the: (i) European floodplain sediment samples are due to the different dominant lithology of Hellas, where there is widespread occurrence of ophiolite and carbonate rocks in comparison to Europe, as well as polymetallic mineralisation, and (ii) Hellenic floodplain sediment samples are due to the dominant lithology of the Lake Kastoria drainage basin, where the granitic-gneiss is dominant, but there is also a significant occurrence of ophiolite and prasinite, as well as polymetallic mineralisation. The main conclusions of the geochemical study of Lake Kastoria, based on the results of (i) the reconnaissance stream sediment survey, (ii) lake-bottom sediment samples, (iii) core-sediment samples, (iv) floodplain sediment samples and (v) recent stream sediment samples, are as follows: (1) The reconnaissance stream sediment geochemical survey results (N=432) of Ag, Cd, Co, Cr, Cu, Hg, Mn, Mo, Ni, Pb, Sr, V & Zn have shown that: areas with high Co, Cr, Cu, Ni and V concentrations are due to the occurrence of ophiolite masses in the Lake Kastoria drainage basin (e.g., north and to the NW of Palaeo Idioktito); areas with high and anomalous Cd, Cu, Hg, Mn, Pb and Zn concentrations suggest the potential occurrence of mixed sulphide mineralisation (e.g., Koritsa peninsula and SSE of Korissos village), and the NE-SW pattern from the east of Melissotopos village to Fotini with anomalous concentrations of Ag, Cr, Cu, Mo, Ni, V, and Zn, and weak anomalies of Co, Mn, and Pb is worth investigating for potential mineralisation. Consequently, the stream sediments that are deposited in the lake, and which are derived from the weathering and erosion of the lithological formations within the drainage basin of Lake Kastoria, are expected to be comparatively enriched in Co, Cr, Cu, Hg, Mn, Ni, Pb, Sr, V, and Zn. (2) The comparison of the results of the older and current stream sediment samples from the Xiropotamos stream has shown that the Co, Cr, Cu, Mn, Ni, and Pb results are comparable, after taking into account the sampling and analytical errors, and the different analytical methods used in each case. The comparability of the analytical results of the two different sampling periods suggests that over time the sources of stream sediment are overall the same. This conclusion is supported by the rare earth element concentrations (La, Ce, Y, Sc), determined on the floodplain sediment samples, which exhibit a very small variation in their results, thus verifying indirectly the aforementioned assumption that the Xiropotamos stream sediment sources are the same over time. (3) The floodplain sediment results provide information about the chemical composition of the sediments that are deposited in Lake Kastoria during flood events. The Xiropotamos stream, which drains the largest area to the east of the lake, erodes mainly the granitic-gneiss and the floodplain sediments have, consequently, elevated concentrations in Ce, K, La, Nb, Rb, Sc, Sr, Ta, Th, Ti, Tl, U, W and Y. The streams Tichion and Metamorphosi, which drain the northern part of the lake catchment basin, erode again mainly the granitic-gneiss and the floodplain sediments have elevated concentrations in Be, Cd, Ce, La, Nb, Pb, Sn, Sr, Ta, Te, Th, and U. Whilst the streams Lakkos, Aposkepos and Fountoukli, which drain the north-western part of the lake catchment basin, erode mainly the formation of weakly metamorphosed rocks (metaconglomerate, metasandstone, meta-arkose, phyllite, prasinite, schist), as well as rock units of smaller area extend, such as limestone and dolomitic limestone with karstic bauxite, and ophiolite and, thus, the floodplain sediments have elevated concentrations in Al, As, Ba, Be, Bi, Co, Cr, Cu, Fe, Ga, Li, Mg, Mn, Mo, Ni, P, Sb, Sc, Ta, Te, V, Zn and Zr. (4) The relatively higher element concentrations in the lake-bottom sediment samples, they portray the geochemical characteristics of the natural terrestrial materials that are transported to the lake by the streams of Lake Kastoria, i.e., the comparatively elevated concentrations of: Ag, Cd, Pb and Zn on Koritsa peninsula, which have been found by the reconnaissance stream sediment survey, appear also in the neighbouring lake-bottom sediments; Li, Sb, Zn and Zr in the lake-bottom sediments to the east of Chloe originate from the sediments of Lakkos and Aposkepos streams, which drain the granitic-gneiss, ophiolite rocks, formation of weakly metamorphosed rocks (metaconglomerate, metasandstone, meta-arkose, phyllite, prasinite, schist), Holocene and Pliocene-Pleistocene sediments; B, Ba, Be, Ce, Fe, Ga, Ge, K, La, Nb, Rb, Sc, Se, Sr, Ta, Te, Th, Tl, U, Y and Zn in the lake-bottom sediments of the northern and central northern part of Kastoria Lake are derived mainly from the Tichion and Metamorphosis stream sediments, which drain the granitic gneiss and Holocene sediments. While, the elevated V concentrations should be due to the sediments of the Lakkos, Aposkepos and Fountoukli streams, since these drain areas with outcrops of ophiolite and metamafic rocks (prasinite); Ge, K, Na, Nb, Rb, Sr, Ti, Tl and Y in the lake-bottom sediments of the coastal area from Polikarpi and Mavrochorion villages and Krepeni locality are derived from the Xiropotamos stream and the other streams of the eastern catchment basin of Lake Kastoria, which drain the, granitic-gneiss, granite and Holocene sediments; Fe, Ga, Ge, Sc, Te, Tl and U in the lake-bottom sediments of the central southern part of Lake Kastoria are due to the sediments provided by the eastern catchment basins, which drain the granitic-gneiss and Holocene sediments, and Co, Cr, Mg, Mn, and Ni in the coastal lake-bottom sediments to the north-west of Dispilion are derived from the nearby ophiolite outcrops. (5) The study of chemical element associations in the horizons of the subsurface lake sediments, which were sampled from the cores, made it possible to reconstruct the "geochemical palaeogeography" of the lake. It is quite evident that there were two distinct periods of sedimentation in Lake Kastoria. The first and older period of sedimentation is characterised by sediments with relatively elevated concentrations in a small number of elements in which Ca and Sr are always present, and in many samples B and S, i.e., Ca, Sr, ± (B, Be, Hg, Mg, Mo, Ni, P, Pb, S, Se, Sn, Te & U). Hence, a sedimentation period completely different from the current, namely, dry climate, intense evaporation of lake water, and consequently, low lake water level, and small or no provision of sediments. This period, based on C14 dating has started before 1020 AD and continued to about 1770 AD (±90 years), although there appear to be local hiatuses in the sedimentation history of Lake Kastoria. The second, and younger period of sedimentation, is characterised by sediments with comparatively elevated concentrations in a large number of elements, i.e., Al, Ba, Be, Bi, Ce, Co, Cr, Cu, Fe, Ga, La, Li, Mg, Nb, Rb, Sc, Th, Ti, V, Y, Zn, ± (Ag, Al, As, Ca, Cd, Ge, Hg, K, Mn, Mo, Ni, P, Pb, S, Sb, Sn, Te, Tl, U, & Zr). This period is also characterised by the supply of large quantities of sediments from all the streams within the Lake Kastoria drainage basin, i.e., the climatic conditions were similar to the current ones, namely, a wet period with intensive rainfall events and, consequently, intense erosion and provision of large quantities of sediments to the lake. This period, based on C14 dating, started about 1770 A.D. (±90 years) until the present time, although locally it appears to start earlier. (6) Some comparatively elevated element concentrations that could be ascribed also to anthropogenic causes have been located in the: lake-bottom sediments in the small inlets to the south and north of Kastoria town centre and concerns the elements Ag, Al, As, B, Be, Bi, Cd, Cr, Cu, Hg, Mn, Mo, Na, P, Pb, S, Sb, Sc, Se, Sn, Te, Tl, W and Zn, and most likely caused by the disposal of urban wastes, and coastal bottom-lake sediments near the villages of Palaeo Idioktito (S), Mavrochorion (Sr) and Dispilio (P, S) from the agricultural activities. It is stressed that the inferences about the potential impact by human activities on the concentrations of chemical elements are inferred from the spatial distribution of elevated values of the aforementioned elements in relation to the inhabited and agricultural areas. However, for the extraction of safe conclusions about the contribution of human activities, the collection of additional information and data related to human activities is required as, for example, land use, wastes and their characterisation, fertilisers and pesticides used in agriculture, tasks that fall outside the objectives of this thesis. Further, the provision of large quantities of sediments to the lake, during the aforementioned second sedimentation period, may be due to deforestation, forest fires, and agricultural activities. (7) Two elements that should be studied in more detail are P and U. The former for the phenomenon of eutrophication observed in Lake Kastoria, and the latter for environmental purposes and good health status of lake biota, and especially fish. Phosphorus concentrations in lake-bottom sediments vary from 0.025% to 0.326%, with a median of 0.096% P. Whilst in core-sediments they vary from 0.053 to 0.159%, with a median of 0.099% P. The comparatively higher concentrations of P in lake-bottom sediments are obviously due to the agricultural activities (fertilisers). However, the lake sediments themselves have comparatively high natural concentrations of P, which are observed in even the deeper horizons of the core-sediments, which undoubtedly have not been affected by the agricultural activities, i.e., they reach up to 0,153% P at a depth of 90-164 cm. Concerning the U concentrations in lake-bottom sediment samples, they vary from 0.61 to 5.60 mg/kg, with a median of 3.33 mg/kg U. While, their variation in the core-sediment samples is from 1.84 to 5.02 mg/kg, with a median of 3.65 mg/kg U. The relatively high U concentrations in lake sediments are due to the detritus derived from the erosion of the granite and granitic-gneiss. It is noted that the concentrations of U in the floodplain sediment samples from streams draining the area covered by the granitic-gneiss and granite vary from 1.37 to 6.32 mg/kg, with a median of 2.58 mg/kg U. (8) The comparison of Lake Kastoria geochemical results from lake-bottom and core sediment samples with the corresponding data from other Hellenic lakes (Vegoritis, Volvi, Ioannina or Pamvotidha, Karla, Koronia and Koumoundourou) has shown that their differences in the levels of major and trace element concentrations can be explained by the local lithology of the drainage basin of each lake. Finally, it is again stres, CITATION: Demetriades, A., 2017. Geochemical study of lacustrine and stream sediments of Lake Kastoria drainage basin. Ph.D. thesis, University of Patras, Hellenic Republic, 654 pp. Nemertes: http://hdl.handle.net/10889/11156. National Documentation Centre: ND 42650: http://hdl.handle.net/10442/hedi/42650. Zenodo: https://doi.org/10.5281/zenodo.7443949 (in Greek with an extended English summary).