Geochemistry, sedimentology, contamination assessment and submerged aquatic vegetation (SAV) mapping using unmanned surface vehicle (USV) : the Gialova lagoon case study

Thumbnail Image



Παπακωνσταντίνου, Μαρία

Journal Title

Journal ISSN

Volume Title



Gialova lagoon, suited in SW Peloponnesus (Greece), constituted the subject of this study, where a multidisciplinary methodological approach was followed. The main objectives of this study were to identify the morphology and characteristics of the lagoon’s floor, to visualize the spatial distribution of sedimentological characteristics of the lagoon, to examine possible correlations amongst acoustic characteristics of the lagoon’s floor and granulometry, to calculate the elemental concentration of surface sediments and finally to examine potential heavy metal enrichment by validated techniques. The purposes of the study were fulfilled on the basis of two sampling campaigns carried out on August and on December 2020, during which geophysical and sedimentological data were collected, respectively. The reconnaissance survey revealed the geophysical and natural characteristics of the lagoon, using a high resolution side scan sonar (SSS) on board an Unmanned Surface Vehicle (USV). A total area of 0.85 km2 was insonified (almost the 36 % of the 2.37 km2 in total) and after the processing of the SSS data, the Gialova lagoons’ floor surface mosaic was revealed. The E-W and the N-S running ridges have also been recorded on the mosaic. A detailed bathymetry map was also acquired from the single beam echosounder collected data, with the additional information obtained by the two indices extracted values, E1 (roughness) and E2 (hardness), all of which showed that Gialova lagoon consists of an extremely shallow water body, with a minimum and maximum water depth of 0.40 m and 0.70 m, respectively. Three morphological parts were defined: (i) the western shallow basin with maximum water depth of 0.65 m, (ii) the eastern shallow basin with slightly higher depth (0.70 m) and (iii) the central and coastal part with water depth less than 0.55 m. It is found that, the spatial distribution of the backscatter coincides well with the three morphological parts and the subaqueous vegetation of the lagoon. The examination of the side scan sonar mosaic, through a manual classification process, showed that six (6) distinct acoustic types (AT) were identified in the lagoon floor. The manual acoustic classification of lagoons’ floor was based on the reflectivity (low or high) and on the alternation of the reflectivity (fragmented or continuous) within the respective AT. Along with the in situ ground truthing results, the submerged aquatic vegetation (SAV) is found to cover each of the six (6) AT. The ATs 1, 2 and 3 are representing areas covered by SAV in a percentage of <1 %, 5-12 % and (13 %)-25-50 %, respectively, each of which occupying the 12.55 %, 17.77 % and 19,99 % of the total surveyed area, correspondingly. Similarly, ATs 4 and 5, which are representing areas covered by SAV in a percentage of 50-75 % and >75%, respectively, hold the 38.94 % and the 9.13 % of the total surveyed area, correspondingly. AT 6 holds the 1.61 % of the total surveyed area and it represents the E-W ridge. With the use of SonarClass software and based on: (i) the extracted tonal grey level statistics (1st order), (ii) the textural statistical parameters (2nd order), extracted by the Grey Level Co-occurrence Matrices (GLCMs), and (iii) the ground-truthed training samples, the unsupervised classification results were revealed, regarding the overall SSS mosaic dataset. The unsupervised classification results are correlated very well with the AT spatial distribution, as defined by the manually classification procedure. Further, a cloud free imagery from the archive of MAXAR Technologies was selected for the mapping of the SAV in Gialova lagoon, with emphasis on the percentage cover. The selected image was an 8-band multispectral WorldView II, collected at 27/06/2020. The classification scheme was based on the analysis of the hydroacoustic data, mainly the SSS mosaic and the delineation of the percentage of cover of the SAV in each region. The overall accuracy of the resulted pixel based classification was 54 % and the F1 score was 49.2 %. The SAV coverage spatial distribution based on the satellite image processing, found to be very well correlated with the SAV spatial distribution, as it emerged from the manual and unsupervised classification procedures. Further, on the basis of the bathymetric map and the backscatter properties of the lagoon floor, sediment samples were collected from 28 pre-defined sampling stations, for the investigation of the spatial distribution of the lithological characteristics, the major/ trace elements and the organic carbon. Sediment samples were subjected to macroscopic description, granulometry and chemical analysis. Initially, all sediment samples were visually described in terms of the color scale of Munsell and by visual inspection of sediments textural features. Grain size analysis was performed using a Malvern Mastersizer 2000 particle analyzer, while prior to analysis, Hydrogen peroxide (H2O2) treatment was applied for the elimination of organic matter. Also, textural evaluation of the sediments (0.063 – 1 mm size range) was made with the use of Leica binocular microscope. Based on the grain size analysis, the macroscopic observations and the textural evaluation, the sediment samples of Gialova lagoon are classified in 4 sediment types. Type A is consisted of very dark grey (5Y 3/1) mud with rare – sparse presence of bioclasts (sand to gravel sized bivalve shells) and there is rare presence of plant residues. Type B is comprised of very dark grey (5Y 3/1) silt with increased clay and water content. In this type, there is rare to sparse presence of bioclasts and sparse to common presence of plant residues. Type C is consisted of very dark grey (5Y 3/1) sandy silt with common presence of plant residues and bioclasts (gastropods and bivalve shells fragments) whereas Type D contains very dark grey (5Y 3/1) silt with common to dominant presence of bioclasts and plant residues. It is also noteworthy that, a thin (mm scale) dark grayish brown (2.5Y 4/2) mud layer found to cover the Gialova lagoon bottom sediments. The definition of the main four grain size statistical parameters, mean (Mz), sorting (σi), skewness (Sk) and kurtosis (KG) were calculated with the GRADISTAT program, while sediment classification was based on Folk (1974) nomenclature. The investigation of the surface sediments granulometry showed that it coincides well with the morphology of the lagoons’ floor. More specifically, the western basin consists of very fine-grained sediments, the eastern basin is covered by slightly coarser sediments, whereas the central shallower part of the lagoon is covered by coarser sediments compared to the two basins. The central part displays the highest sand mean percentage, high silt and the lower clay percentage. The spatial distribution of clay proportion clearly reflects the three different morphological parts of the lagoon. The coarser sediments associated with the higher sand proportion were obtained along the southwestern coast of the lagoon, just north of the sandy lagoon barrier separating the lagoon from Navarino Bay and near the inlet. Total Organic Carbon (TOC) was measured using a LECO Carbon analyzer after HCL (25%) pre-treatment to remove carbonates. The spatial distribution of the TOC seems to be affected by the morphology of the lagoons’ floor. The higher TOC percentages were measured in the two shallow basins where the very fine sized sediments dominate, while low concentrations were obtained along the sand barrier at the southwestern part of the lagoon and at the central part. The concentrations of selected major (wt%) and trace elements (ppm) in the surface sediments of Gialova Lagoon were determined by four acid digestion followed by ICP-MS finish. The spatial distribution of sediment characteristics was visualized using the Surfer 9 and ArcGIS software. The whole geochemical dataset (Ag, Al, Ba, Be, Ca, Cd, Ce, Co, Cr, Cs, Cu, Fe, Ga, Ge, Hf, In, K, La, Li, Mg, Mn, Mo, Na, Nb, Ni, P, Pb, Rb, S, Sb, Sc, Se, Sn, Sr, Ta, Th, Ti, Tl, U, V, W, Y, Zn, Zr and Corg) was subjected to Multivariate statistical treatment. R-mode Factor Analysis (FA) was used for revealing the dominant geochemical processes in the lagoon and Q-mode FA, for detecting the interrelationships of the sediment samples, in terms of their geochemical composition. Five (5) different geochemical processes occurring in the lagoon bottom sediments were identified based on R-mode FA. There are processes defined by the presence of terrigenous aluminosilicates associated with Corg which present a clear antipathetic relation with autochthonous biogenic carbonates, processes driven by agricultural drainage, processes oriented by phosphates and selenium element and processes driven by the association of Y adsorbed and/ or co-precipitated with amorphous Mn oxyhydroxides. Through Q-mode Factor Analysis plot, three (3) surface sediment samples clusters were identified: the sediment samples included in cluster A represent the terrigenous aluminosilicates associated with Corg which is characterized by elevated concentrations of many metals, the sediment samples which are classified in cluster B represent the biogenic carbonates with the highest Calcium (Ca) and Strontium (Sr), whereas sediment samples of cluster C, located in the midpoint of factor plot, show intermediate element concentrations compared to the end-member samples (Cluster A and B). The Q factor plot can be considered that portrays the entire spectrum of contamination of the surface sediments of the lagoon from the most contaminated to less contaminated. For the assessment of the level of trace metals contamination and the possible anthropogenic impact on the sediments of Gialova lagoon, three (3) of the most used pollution indices were applied: Geo-accumulation Index (I-geo) (Muller, 1979), Enrichment Factor (EF) (Kemp et al., 1976) and Pollution Load Index (PLI) (Tomlinson et al., 1980). For the calculation of the indices, the global average shale data from Turekian and Wedepohl (1961) were used as background. I-geo values indicated that the Gialova lagoon is characterized from unpolluted to moderately polluted sediments for the elements Mo, Na, Ni, Pb and Se and from strongly polluted sediments in S element. EF values showed that elements with moderate enrichment are: Mn, Mg, Cr, Ni, Pb, Mo and Sr, whereas, elements with significant enrichment are: Na, Se and S. PLI exhibiting values lower than one (1), therefore the sediments are unpolluted to slightly polluted.



Lagoon, Side scan sonar, Unmanned surface vehicle, Sediments, Heavy metals, Pollution, Geoaccumulation index, Pollution load index, R-mode factor analysis, Q-mode factor analysis