Proceedings of the 13 th International Conference on Environmental Science and Technology Athens, Greece, 5-7 September 2013 HYDRO-CHEMICAL CHARACTERIZATION OF WHITE MOUNTAIN KARSTIC WATER AT KOILIARIS RIVER CZO D. MORAETIS 1, G. GIANNAKIS 1, D. EFSTATHIOU 1, N.P. NIKOLAIDIS 1, P. VAN GAANS 2, M. VERCHEUL 2, M. ADRIANAKI 3 and S. BERNASCONI 3 1 Technical University Crete, Department of Environmental Engineering, Chania, Greece PC. 73100, 2 Deltares, Subsurface and Groundwater Systems, 3508 AL Utrecht, Netherlands, 3 Geological Institute ETH, Zurich, Switzerland e-mail: moraetis@mred.tuc.gr EXTENDED ABSTRACT The objective of the present study is the hydro-chemical characterization of White Mountain karstic water at Koiliaris River Critical Zone Observatory (CZO) and the identification of the factors influencing its quality. Sampling was designed as to identify the chemical characteristics of karstic water and the impacts of highland degraded soils by livestock on water quality. The analysis was comprised of monthly measurement of several chemical species for the last three years obtained from wells and springs at Koiliaris CZO. The chemical species analyzed included anions (NO 3-, SO 4 2-, Cl -, etc.) and cations (K +, Na +, Ca 2+, Mg 2+, Si 4+, Sr 2+, Ti 2+ etc.). Principal component analysis (PCA) was used for the categorization of water bodies according to their chemistry. The PCA analysis revealed two main groups of water samples with different chemical signatures, one within the alluvial aquifer having high content of chemical species and the second, the karstic water with low concentration for all chemical species. The high concentrations of anions like NO 3 - and Cl - in the first group (shallow wells), reflect water infiltration through degraded soils, whereas the low concentration of chemical species in the second group (karstic water) reflects the low reactivity of hard limestone. Significant temporal changes of the karstic water chemistry were observed at Stylos spring between January and March. The conductivity measurements showed in total four periods of differentiation with order of increasing values from April-June, November-December, July-October and January-February. The temporal variation was related to both micro-climatic, tectonic factors and anthropogenic impact. The April-June period corresponded to snow melt at higher elevation (>1200 m) where the tectonic relief is in maximum and vertical sinkholes in dolomites-limestones drain the area. The November-December period corresponded to rain from both higher and lower elevation in the karst (lack of snow cover in higher elevation). These periods showed lower electrical conductivity and corresponded to fresh water intrusion with low chemical load (Ca 2+, Mg 2+, etc.). The July-October period exhibited higher electrical conductivity compared to November-December period with relatively constant concentration of solutes which corresponds to deeper karstic water with low flow. The last period January-March showed a strong increase in all solutes (high electrical conductivity) at the Stylos spring which is related either to the wash out of dry period depositions in soils at higher elevation and/or the contribution of soil percolation from lower elevation. Thus, the water quality in the major karstic spring showed significant chemical changes during January-March and that is related to soil infiltration input either from lower (agricultural areas) or higher elevation (pasture lands). Keywords: Hydrochemistry, karstic water, karstic springs 1. INTRODUCTION Karstic water usually moves through a system of variable diameter sinkholes and conduits. The karstification of each part of the karstic system influences the discharge rate of the system and the water/rock/soil reactivity time which influences the chemical
composition of the water. The degree of karstification has been related to intergranular porosity within the matrix rock (kind of texture), fractures along joints, faults and bedding planes (White, 1999). The identification of the processes (hydrological and hydrochemical) within a karst system is a difficult task, which requires use of tracers, instrumentation deployment in springs and sinkholes, hydrochemistry etc. Hydrochemistry has been used as a tool for the characterization of temporal and spatial variability of the karstic water. The use of natural conservative tracers like Cl has been considered for the identification of soil infiltrates (Aquilina et al. 2006). Bulk chemical analysis of karstic water with statistical processing has been given successful interpretation of the origin of the karstic water (Bicalho et al. 2012). The present study seeks to identify the basic factors influencing the chemical signature of White Mountains Karstic water at Koiliaris River CZO using 3 years monitoring of hydrological and chemical status. Koiliaris CZO has been heavily instrumented and undergoes systematic monitoring. Detail hydrological studies of the site have been published previously (Nikolaidis et al., 2013, Kourgialas et al., 2010, 2011, Moraetis et al., 2010). The geological and geomorphological characteristics are comprised of a series of nappes which create a pile of successive limestone and dolomites, up to an elevation of 1500 m. Above the 1500 m elevation, autochthonous dolomites and limestones are outcropping. 2. SITE DESCRIPTION The Koiliaris River is situated 25 km east from the city of Chania. There are 17 communities in the catchment, 8 of which are at low altitude, 7 are at high altitude and 2 are at medium altitude. The climate is characterized by a separation of seasons. Summer usually is hot/dry and winter is cold/wet. The mean annual rainfall in the northern part of the catchment is 705 mm (447-1032mm). The highest catchment s slope is 43% at the White Mountains (Lefka Ori), while gentle slopes (4.8%) are present at the end of the valley and the estuary of Koiliaris river. The highest altitude is 2041 m. The total watershed area is 130 km 2 and the total length of the hydrographic system is 36 km. The Keramianos stream (Figure 1-No 1) is the main temporary stream which joins two other smaller streams, the Milavlakas (Figure 1-No 2) stream and the Mantamas stream (Figure 1-No 3). Koiliaris river has permanent flow due to the constant supply from Stylos spring and it is extended from the confluence of the 3 ephemeral rivers towards the sea. The geology of the Koiliaris River Basin (Figure 1) consists of Plattenkalk (authocthonous nappe) which is comprised mainly by black dolomites, marbles, limestone and recrystallized limestone with cherts; Trypali nappe which is comprised of recrystallized limestones with cherts and intercalations of marls; Tripolis nappe (allocthonous nappe) comprised of unbedded karstified limestones; Western Crete metamorphic rocks with phyllites and quartzites; Neogene formations with marls; Neogene formations with marly limestones; and quaternary alluvial deposits (Papanikolaou 2010). The results presented in this study correspond to monthly grab samples from 2010 up to February 2013. ICP- MS (ICP-MS-Agilent 7500-CX) was used for analysis of major cations and trace elements, where capillary electrophoresis (CE G1601BA, Agilent Technologies) for the anions and an N/C analyzer (multi N/C analyzer 2100S, Analytik Jena) was used for total nitrogen determination. 3. RESULTS 3.1 Aquifers characterization at Koiliaris watershed Table 1 presents the average chemical and physical parameter values for all sampling points and rainfall event (November 2012). The wells and springs in shallow aquifers like those in Kampi well, Xaritakis well, Katoxori spring and Keramianos ephemeral river have higher nutrient content (N-NO 3-, TKN, B 3+, Cl - ) on average compared to the other water
bodies originated mainly from the karstic area. The same grouping is observed in Figure 2, where all parameters have been fed into principal component analysis. The water samples from the shallow aquifers have higher load of chemical species and this is related to higher load of nutrients from agriculture activities and the higher reactivity of brittle thin limestone like those in alluvial formations. A1 7 6 5 4 3 2 1 Figure 1. Geological setting of the western Crete, Koiliaris watershed (thick black line) and extended watershed (thin black line) are shown. Sites of water samples are also shown. A1 line corresponds to the cross section in Figure 5 and the numbers 1-7 corresponds to the rock beds shown in Figure 5. Figure 2. The first two principal components. The PCA was conducted with the use of the average values presented in Table 1.
PCA analysis provided a) clustering of the samples into two distinct groups, b) first three PCA describe 80% of the chemical variability, and c) the major chemical and physical factors influencing water quality. Thus, PCA1 showed that ph and dissolved oxygen (DO) distinguish the karstic water from different origin water (karstic water has negative values for PCA1). PCA2 showed positive values for water bodies which have high values for nutrients like (N-NO 3-, PO 4 2-, TKN, N-NH 4),mainly katoxori spring in alluvial deposits, Keramianos ephemeral river in alluvial deposits and Anavreti karstic perched spring. The fact that Anavreti spring was among that group suggests a mixing of karstic water and alluvial water. Finally the PCA3 (not depicted in Figure 1) showed high correlation of Mn, Si, Mg, K, Ti related to Kampi well. Kampi well is situated in alluvial deposits originated from the weathered products of metamorphic schists. Zourbos karstic spring has been affected from sea water,depicted by high concentrations of Mg 2+, Na +,Cl - etc. 3.2 Temporal variability in karstic water chemistry The temporal variability of chemical species in the karstic system of Stylos spring showed rapid response to different water inflows from different areas of the karst. The White Mountains (Lefka Ori) karstic system comprises a fast and a slow responding reservoir as it has been already described through hydrological modeling (Moraetis et al., 2010, Kourgialas et al., 2010, Nikolaidis et al., 2013). The juxtaposition of the geological setting and the modeling results, reveal that the fast responding reservoir is the lower black dolomitic bed in the autochthonous Mani unit which has at least 300m thickness. The exploration by different speleological teams showed the development of deep karstic systems which originated mainly in areas with the lowermost black karstified dolomite (Figure 3). Figure 3. The lowermost black dolomitic bed of the autochthonous nappe in White Mountains (Lefka Ori) in Crete (Photo from the report of Catamaran Speleology group 2008). The temporal distribution of the chemical and physical parameters at Stylos spring is presented in Figure 4. There are two periods with clear changes in the sum of chemical species corresponding to February-March (black arrows), whereas after this period all chemical species showed a period of decreased concentration (bllue arrows). The first period coincided with the wet period which is distributed as rain in elevation lower than
1000-1200 m and as snow in higher elevation. The second period coincided with minimum rainfall ( and mainly of snowmelt. The third and fourth period represent the dry period (red arrows) and the first rainfalls (yellow arrows). The sum of the chemical species followed the conductivity variability. The chloride concentration showed characteristic peaks during the wet period, whereas for the other periods is relatively constant. The silicon and magnesium concentration showed peaks during wet period and reduced concentration during snowmelt. During dry period the magnesium and silicon concentration showed a recovery in the concentration values. The DO showed lower values during the wet period and higher during snowmelt and the dry period. Figure 5 depicts a synthetic diagram of the hydrological conditions during a period of a year according to the observations in water quality and the geological setting as it is presented in Figure 1. The lower most geological bed, the black dolomites, are the most karstified and exhibit a system of chaotic shafts and sinkholes which have direct response to the Stylos spring (Figure 3 and Figure 5). Thus, first rainfalls during September, October and November may have direct response in the Stylos spring when the rainfall is distributed both in low and high elevation (Figure 5a, Figure 4-yellow arrows). 14 Cl*10 mmole/l Sum mmole/l Ca*2 mmole/l DO mg/l Mg*10 mmole/l Cond/100 (μs/cm) Si*200 (mmole/l) 12 10 8 6 4 2 0 21-Feb-10 22-May-10 20-Aug-10 18-Nov-10 16-Feb-11 17-May-11 15-Aug-11 13-Nov-11 11-Feb-12 11-May-12 9-Aug-12 7-Nov-12 5-Feb-13 6-May-13 Figure 4. Sum of chemical species, electrical conductivity (Cond), dissolved oxygen (DO), chloride, calcium, magnesium and silicon concentration at Stylos spring. The same is observed in the slightly increase in chloride concentration during October 2012 which is related to intense rainfall in high elevation and wash out of chloride from the limited soil cover and rocks and slight decrease of electrical conductivity due to overall decrease of the other chemical species like Ca 2+, Si 4+, Mg 2+ as a consequence of fresh water intrusion from higher elevation with limited soil cover in the black dolomitic bed. During the wet period, the conditions are freezing in high elevation and snow cover spreads above 1500 m whereas below 1500 m there is intense rainfall which ranges from 500-1000 mm. Thus the pile of nappes in lower elevation like the autochthonous dolomites and marbles, the recrystallized limestones with silex and the allochthonous Trypali units, supply water in the karstic system (Figure 5b).The water quality corresponds mainly to soil infiltration. The soil development below 1000 m is orders of magnitude more abundant compared to higher elevation. The peaks in electrical conductivity and the chloride content are clearly identified during December, February
and March 2011 and 2012 and are related to soil infiltrates which gradually reach the deeper karstic layers. The dissolved oxygen has slightly lower values during the wet months and that is related to water move through less oxygenated environment like this in small conduits and limestone matrix. On the contrary, the more oxygenated water moves through the lower most dolomitic bed within large karstic pipe system (Figure 5c). The snow melt period is characterized by dilution effects during April, May and June with decrease in all parameters apart from dissolved oxygen (Figure 4-blue arrows). Finally the dry period showed that most chemical species recover in concentration with increase in silicon and magnesium whereas calcium is rather steady (Figure 4-red arrows). The dry period is characterized with the water move in the lower most nappe (black dolomitic bed) which has more time to react with the aquifer due to lower water pressure since most of the winter water has been drained and probably some residual flow from the pile of nappes from lower elevation which has also more time to react due to lower dip of the rock beds.. 4. Conclusions The spatial variability of Koiliaris CZO hydro-chemistry shows influence of the agriculture activities in shallow aquifers, whereas karstic water is mainly of lower chemical load. During the wet period there is an intrusion of soil infiltrates in the karstic system which increases the concentration of several cations and anions. The fast and slow response of the karstic system as presented by Moraetis et al. (2010), has been clearly defined from hydrochemical observations as follows: a) Fast response of the lowermost karstic system, the black dolomitic bed in snowmelt conditions (April-June) and rainfall in high elevation during first flush events (October-November) b) Medium to slow response of the pile of the limestone nappes in low elevation which is supplied mainly by soil percolation of precipitation. c) Slow response of the lowermost black dolomitic bed in low flow conditions due to low water supply, lower dip of the limestone beds and probably the residual water which runs through the pile of nappes from lower elevation compared to the outcrops in higher elevation of the black dolomites. REFERENCES 1. Aquilina L., Labouche B., Dörfliger N. (2006), Water storage and transfer in the epikarst of karstic systems during high flow periods, Journal of Hydrology 327, 472-485. 2. Bicalho C.C., Batiot-Guilhe C., Seidel J.L, Van Exter S., Jourde H., (2012) Geochemical evidence of water source characterization and hydrodynamic responses in a karst aquifer, Journal of Hydrology, 450 451 206 218. 3. Catamaran speleology group report (2008), L expédition spéléologique «LEVKA 2008» est organisée par le Groupe Spéléologique CATAMARAN de Montbéliard affilié à la Fédération Française de Spéléologie du 15/07 au 15/08 sur l île de Crête (In France). 4. Kourgialas, N.N., Karatzas, G.P., Nikolaidis, N.P., (2010), An integrated framework for the hydrologic simulation of a complex geomorphological river basin. Journal of Hydrology, 381, 308 321. 5. Kourgialas, N.N., Karatzas, G.P., Nikolaidis, N.P., (2010), An integrated framework for the hydrologic simulation of a complex geomorphological river basin. J. Hydrol. 381, 308 321. 6. Moraetis, D., Efstathiou, D., Stamati, F., Tzoraki, O., Nikolaidis, N.P., Schnoor, J.L., Vozinakis, K., (2010), High frequency monitoring for the identification of hydrological and biochemical processes in a Mediterranean river basin. Journal of Hydrology, 389, 127 136. 7. Nikolaidis N.P., Bouraoui B., Bidoglio G., (2013), Hydrological and geochemical modelling of a karstic Mediterranean watershed. Journal of Hydrology, 477, 129-138. 8. Papanikolaou D, Vassilakis E, (2010) Thrust faults and extensional detachment faults in Cretan tectono-stratigraphy: Implications for Middle Miocene extension. Tectonophysics, 488:233 247.
9. White, W.B., 1999. Conceptual models for karstic aquifers. In:Palmer, A.N., Palmer, M.V., Sasowsky, I.D. ŽEds.., Karst Modeling. Karst Waters Institute Special Publication, vol. 5 Acknowledgements The extensive sampling and analyses obtained in this study was financially supported from the European Commission FP 7 Collaborative Project Soil Transformations in European Catchments (SoilTrEC) (Grant Agreement no. 244118).
Table 1. Average values for the physical and chemical parameters in spring, river and well water at Koiliaris River CZO, (the first parenthesis indicates the number of samples and the second the sample code in Figure 1). Gournes (karstic (4) (G5) Stylos (karstic (24) (S1) Armeni (karstic (24) (S3) Zourbos (salted karstic (24) (S4) Gauging station (river) (24) (R1) Anavreti (karstic (3) (S5) Kampi well (alluvial) (24) (G3) Xaritakis (Neogene rocks) (24) (G2) Katoxori spring (alluvial) (24) (S2) Keramianos ephemeral river (7) (R2) Rain Temp. (Celcius) 13.3 12.4 13.1 12.9 14.1 12.5 16.9 18.4 17.5 15.0 * ph 8.4 8.1 8.1 7.9 8.1 8.2 7.6 7.3 7.0 8.1 5.9 DO (ppm) 9.8 11.3 10.3 10.6 10.1 11.3 1.4 2.3 7.7 11.4 * E.C. (μs/cm) 208.8 223 259 1104 239 265 1044 988 739 605 48.8 N-NO3 (ppm) 0.5 0.5 0.4 0.9 0.5 0.7 1.0 9.2 6.6 4.2 0.8 N-NH4 (ppm) 0.0 0.2 0.2 0.1 0.2 1.0 0.4 0.3 0.1 0.7 0.1 P-PO4 (ppm) 0.1 0.1 0.1 0.2 0.2 0.4 0.3 0.4 0.2 0.2 Cl - (ppm) 11.3 11.4 11 225.6 14.7 13.6 73 55.9 44 54.4 4.3 TKN (ppm) 1.7 1.1 1.2 1.4 1.1 0.4 9.7 9.0 7.4 4.4 - SO4 2- (ppm) 2.0 2.5 3.1 40.5 3.7 11.1 122 62.0 28.3 47.7 1.9 Na (ppm) 0.4 1.7 0.4 10.4 0.4 0.5 3.0 2.4 1.9 1.9 0.6 Mg (ppm) 1.0 5.2 5.7 21.6 5.5 3.7 45.0 10.0 5.3 5.3 0.9 Si (ppm) 0.7 1.0 1.1 1.2 0.97 1.4 5.6 5.1 3.8 3.8 11.6 K (ppm) 0.5 0.2 0.3 4.2 0.24 0.5 5.5 0.0 1.2 2.0 1 Ca (ppm) 41 37 35 49 38 36 115 141 124 86 4.1 Ti (ppb) 1.9 1.2 1.4 1.3 1.34 2.8 2.0 1.7 2.0 1.3 - Mn (ppb) 0.2 1.2 0.2 0.1 1.2 2.2 8.8 0.7 1.0 1.4 - Fe (ppb) 3.2 5.7 3.1 4.5 7.3 1.9 8.8 8.7 7.3 23.4 - Sr (ppb) 85 86 86 222 91 95 247 257 595 231 7.3 Ba (ppb) 48 29 38 57 74 60 37 225 58 59 14 B (ppb) 7 6 7 44 12 5 41 62 26 14 - -: Not detected, *: Not measured
Figure 5. Diagram of the White Mountains (Lefka Ori) cross section with the hydrological conditions during a) first rainfalls (September-November) b) wet period (December, January, February, mid March) c) Snowmelt period (April, May, June) d) summer dry period. Double blue arrows show fast response in the karstic spring