Search

Coastal water quality management is a public health concern, as poor coastal water quality can harbor pathogens that are dangerous to human health. Tourism-oriented countries need to actively monitor the condition of coastal water at tourist popular sites during the summer season. In this study, routine monitoring data of $Escherichia\ Coli$ and enterococci across 15 public beaches in the city of Rijeka, Croatia, were used to build machine learning models for predicting their levels based on environmental parameters as well as to investigate their relationships with environmental stressors. Gradient Boosting (Catboost, Xgboost), Random Forests, Support Vector Regression and Artificial Neural Networks were trained with measurements from all sampling sites and used to predict $E.\ Coli$ and enterococci values based on environmental features. The evaluation of stability and generalizability with 10-fold cross validation analysis of the machine learning models, showed that the Catboost algorithm performed best with R$^2$ values of 0.71 and 0.68 for predicting $E.\ Coli$ and enterococci, respectively, compared to other evaluated ML algorithms including Xgboost, Random Forests, Support Vector Regression and Artificial Neural Networks. We also use the SHapley Additive exPlanations technique to identify and interpret which features have the most predictive power. The results show that site salinity measured is the most important feature for forecasting both $E.\ Coli$ and enterococci levels. Finally, the spatial and temporal accuracy of both ML models were examined at sites with the lowest coastal water quality. The spatial $E. Coli$ and enterococci models achieved strong R$^2$ values of 0.85 and 0.83, while the temporal models achieved R$^2$ values of 0.74 and 0.67. The temporal model also achieved moderate R$^2$ values of 0.44 and 0.46 at a site with high coastal water quality.

The history of metal pollution in the semi-enclosed and human-influenced marine system of the Boka Kotorska Bay (Southern Adriatic) was studied considering geological composition of the surrounding catchment, the sedimentation rate and the mineralogical and early diagenetic processes in the recent sediments. The determination of background concentrations of metals, undertaken for the first time in this environment, proved to be particularly important for Ni and Cr, which are naturally enriched in the sediments of the southern Adriatic. The results showed widespread moderate contamination with Pb and Sn since the 1970s, while the upper layers of sediments near shipyards, marinas and urban areas were more contaminated with Sn, Cu, Pb, Zn, Cd, As, Sb and Mo. The transport of material through the narrow straits separating different parts of Boka Kotorska Bay resulted in a different geochemical composition of the smaller bays and a limited distribution of contaminated sediments from local sources.

This paper is conceived as a contribution to understanding the tectonic setting in the Jabuka region in the Central Adriatic Sea, one of the seismically active regions within the Mediterranean. This seismicity is the result of the compressional tectonics within the Adria microplate, accompanied by halokinetic activity. The Jabuka Islet and the Jabuka Shoal appear to be Triassic igneous bodies extruded by diapir tectonics. Sedimentological, mineralogical and micropaleontological analyses of the surface sediments were done to provide insights into the depositional environment around these igneous outcrops. Coarse-grained gravelly sands surrounding the Jabuka Islet and the Jabuka Shoal are composed predominantly of biogenous detritus, containing small amounts of particles derived from igneous rocks. This highly carbonate sediment cover has been formed in the low-energy environment, below wave base and without terrigenous supply from remote areas. A peculiar finding of reworked Lower Miocene foraminiferal assemblage in the recent surface sediment accompanied with re-interpreted seismic profile was used to propose a conceptual geodynamic model for the Jabuka diapir protrusion. It suggested Miocene age for the earliest protrusion of the Jabuka diapir, followed by further Plio-Quaternary growth and shaping. Together with mostly Paleogene fossil calcareous nannoplankton, fossil foraminifera were presumed to have been weathered from the pierced and extruded overburden rocks and admixed to recent, slowly forming carbonate skeletal association, typical for cool-temperate depositional environments.

In order to contribute to the debate on the role of fluvial erosion in the shaping of karst, two nearby areas with different karstic landscapes were compared. Areas A and B are located relatively close to each other on the southern side of the Krk Island (Adriatic Sea, Croatia). Both areas are composed of similar limestone with a very high CaCO 3 content. Area A is a typical doline or polygonal type (“classical”) of karst with numerous dolines (up to 57/km 2 ) covered with terra rossa (red soil) and Mediterranean maquis shrubland. Dolines are located in zones which correspond to the strike of the main geological structures. Dry karstic valleys are visible only on gently inclined coastal slopes bordering the karstic plateau. In contrast, area B is typical of a bare karst landscape with a strong (palaeo)fluvial imprint. The dolines are absent, and the bedrock is only sporadically covered with terra rossa. Palaeogene marls have been observed in a few elongated depressions and in the coastal zone of area B. Along steep coastal slopes, valleys (up to 460 m/km 2 ) are cut into the carbonates. The traces of episodic surface flows are visible in some of these valleys, in contrast to the valleys in area A. Remnants of a disrupted ancient fluvial network are clearly visible on the elevated karstic plateau in area B. Differences in the recent morphology are attributed mainly to varying thicknesses of the Palaeogene impermeable marly cover, and the intensity of tectonics in the two areas.

The aim of this research was to explore the possibility of a successful and balanced integration of fish farming installations into an ecosystem dominated by Posidonia oceanica (L.) Delile species. We selected light, temperature, seabed topography, sediment characteristics, meadow density, bottom coverage, maximum leaf length and lower depth limit as principle components in assessing the influence of the fish farm. All P. oceanica descriptors showed significant correlation with light deprivation effect while sediment organic matter content revealed slightly higher values than normal, increasing with distance from the cages. The results point to a conclusion that in such lightly nutrient enriched ecosystems, the seagrass growth and distribution are principally controlled by the shadow that cages cast on the seabed below, and that when carefully planned, fish farms do not necessarily degrade the health status of the surrounding area, but in fact facilitate a transition into a secondary stable state.

The recent shape of the coast in the Kvarner area is primarily a consequence of the drowning of karst relief due to sea-level rise during the Holocene. Further sea-level rise and increased storminess in this coastal area is expected in the future. In order to assess coastal vulnerability, two stretches of the mainland coast and the coast of four main islands in Kvarner have been examined. Most of the coast in the Kvarner area has a low risk of wave-induced marine erosion and has low vulnerability to predicted sea-level changes due to its rocky origin. Carbonate rocky coasts make up most of the coasts, whereas coasts formed in siliciclastic (flysch) rocks are less common. The largest proportion of fine grained beaches (over 20%) is found on Rab Island, due to a higher presence of less resistant flysch rocks. In order to develop mitigation and adaptation plans for future coastal changes, an integrated coastal zone management is proposed.