Seasonal variability and Po River plume influence on biochemical properties along western Adriatic coast

Authors


Abstract

[1] The influence of the Po plume on the northern Adriatic Sea was observed during two seasons in 2003 under distinct physical forcing regimes. During the winter, the plume was cool, low in both salinity and chlorophyll, but with higher chlorophyll concentrations occurring along the plume boundary. The plume mixed deeply in the water column in response to the strong wind forcing. The northern Adriatic and the Po plume cooled significantly during the observational period, and therefore salinity alone was the best discriminator of water mass variability. In contrast to the strong forcing of the winter period, the late spring was characterized by weak wind forcing, and below-average Po River discharge (∼600 m3/s) which was about one third of the typical discharge for this period. As in winter, salinity was again the best discriminator of water mass variability. The Po plume advected southward along the Italian coast and in some locations portions of the coastal plume were transferred offshore in filament-like features. However, theone observed filament was quite low in chlorophyll and was quite thin vertically, extending downward less than 5 m from the surface. The spring observations provide a distinct contrast in the effects of the physical forcings of river flow and wind stress from two different seasons. The strong winter forcing resulted in deep mixing of the plume despite its low salinity and buoyancy, whereas the weak summer flow under weak winds resulted in a very shallow plume (<5 m) that was high in chlorophyll.

1. Introduction

[2] The Adriatic Sea is the most continental basin of the Mediterranean Sea. It lies between the Italian peninsula and the Balkans and is elongated longitudinally, with its major axis (about 800 km by 200 km) in NW-SE direction. The basin shows clear morphological differences along both the longitudinal and the transversal axes and has been divided into northern, middle and southern subbasins [Artegiani et al., 1997a]. The Adriatic Sea has complicated morphology and bathymetry. The western coast is low and generally sandy, while the eastern coast is rugged, with multiple islands and coves. The northern subbasin, extending from the northernmost coastline to the 100 m isobath, is extremely shallow (mean depth ∼30 m) with a very gradual topographic slope along its major axis. It is characterized by strong river runoff; indeed, the Po and the other northern Italian rivers are believed to contribute about 20% of the whole Mediterranean river runoff [Hopkins, 1992]. The middle Adriatic is a transition zone between northern and southern subbasins, with the three Jabuka depressions reaching 270 m depth. The southern subbasin is characterized by a wide depression about 1200 m in depth. Water exchange with the Mediterranean takes place through the Otranto Straits, which has an 800 m deep sill. The present study focuses on the northern and central continental Adriatic margin, where circulation is mainly controlled by wind stress and river discharge. Two currents dominate circulation in the Adriatic: the West Adriatic Current (WAC) flows toward southeast along the western (Italian) coast, and the East Adriatic Current (EAC) flows northwest along the eastern (Croatian) coast [Artegiani et al., 1997a, 1997b]. Being a continental basin, the Adriatic Sea circulation and water masses are strongly influenced by atmospheric forcing [Orlic et al., 1992, 1994]. The major winds blowing over the Adriatic Sea are Bora and Sirocco. Bora winds are generally from the northeast and are associated with a high-pressure system over central Europe [Dorman et al., 2007]. Bora is a cold and dry wind where air spills through gaps in the Dinaric situated along the Adriatic's eastern shore, resulting in intense wind jets at specific points along the Adriatic eastern coast due to catabatic effects [Poulain and Raicich, 2001]. The Bora wind system causes the free sea surface to rise near the coast and this intensifies a coastal current toward the south (WAC). Historical data and numerical simulations have demonstrated that Bora winds can cause the formation of a double gyre structure consisting in a larger cyclone in front of the Po River delta and a smaller anticyclone to the South [Poulain et al., 2001]. In winter, cold, dry Bora winds cause strong heat loss in the Northern Adriatic and formation of the Northern Adriatic Deep Water (NAdDW). Another factor influencing NAdDW formation is the water flux, mainly governed by the Po River runoff, that can lower the salinity, and hence the density, of the NAdDW. Vilibic [2003] demonstrated the relationships between NadDW formation, heat flux and autumn Po River runoff.

[3] The Sirocco wind is generally from the southeast, and is associated with a low-pressure system over the Tyrrhenian Sea. Sirocco is a warm and humid wind and often causes flooding events in the shallow lagoons along the Adriatic coast including Venice. Coming from the southeast over the sea, the Sirocco is less subject to local variations than the Bora, but it does show some geographical variations due to the coastal orography. It tends to be southerly in the Strait of Otranto and off the Istrian Peninsula (Pula), and more easterly at some places along the northern Adriatic near Ravenna and Pesaro [Poulain and Raicich, 2001].

[4] River runoff is particularly strong in the northern basin and affects the circulation through buoyancy input and the ecosystem by introducing large fluxes of nutrients [Zavatarelli et al., 1998]. Freshwater is discharged into the northern Adriatic from major rivers along the North and Northwest coasts. The Po River represents the major buoyancy input with an annual mean discharge rate of 1500∼1700 m3 s−1, accounting for about one third of the total riverine freshwater input in the Adriatic [Raicich, 1996]. Runoff is also responsible for making the Adriatic a dilution basin. The riverine water discharged into the northern Adriatic forms a buoyant coastal layer that flows southward along the Italian coast. Since the Po River is the main source of this water, the coastal layer is predominantly south of the Po River delta and is named Western Coastal Layer (WCL) [Poulain et al., 2001]. It is associated with a strong near-surface current, which flushes the nutrient-rich water out of the northern Adriatic along the Italian coast [Hopkins et al., 1999; Marini et al., 2002; Campanelli et al., 2004]. The principal compensating inflow occurs along the eastern boundary (Eastern Adriatic Current-EAC) where warm, high-salinity modified Levantine Intermediate Water (LIW) is advected northward. Kourafalou [1999] elucidated the role of the major Adriatic rivers in creating buoyancy-driven coastal currents that are essential in maintaining the cyclonic circulation.

[5] The intent of this study is to examine the Po River plume influence on nutrients and hydrological properties along the western Adriatic coast under varying conditions of river discharge and wind stress during winter and spring.

2. Material and Methods

[6] The data for this study were gathered in the central and northern Adriatic Sea during two oceanographic cruises aboard the R/V Knorr (Woods Hole Oceanographic Institution) during the periods from 31 January to 24 February 2003 (winter period) and from 26 May to 15 June 2003 (late spring) and during an overlapping cruise aboard the R/V G. Dallaporta (ISMAR-CNR) in the winter period from 12 February to 20 February 2003. Observations were gathered using three primary modes of sampling: underway mapping of near-surface seawater utilizing the ship's uncontaminated seawater distribution system, vertical profiling with a CTD/rosette system or a bio-optical profiler, and three-dimensional mapping using a towed undulating vehicle equipped with a CTD and bio-optical sensors. The results from the bio-optical profiler will be presented in a separate paper in preparation. Near-surface mapping utilized the R/V Knorr's uncontaminated seawater distribution system. The intake for the system was located approximately 5 m beneath the vessel's waterline. As part of the ship's Improved Meteorological System (IMET), temperature, conductivity and chlorophyll fluorescence were measured continuously in the water sampled by this system, along with ship's position, and meteorological variables. Temperature and conductivity were measured with a Seabird Seacat CTD sensor. Chlorophyll fluorescence was measured with a Wetlabs WetStar flow-through fluorometer. During underway operations, surface samples were obtained with a PVC sampling bucket and analyzed for nutrients and chlorophyll a.

[7] Three-dimensional mapping was performed with a modified SeaSoar vehicle equipped with a Seabird 9/11plus CTD, Wetlabs C-Star transmissometer (25 cm pathlength), Wetlabs Wetstar chlorophyll and CDOM fluorometers, and a Wetlabs AC9 attenuation/absorption meter. The vehicle was towed at a horizontal speed of 13∼15 km per hour while undulated between near the surface and about 5 meters above the bottom. Vertical ascent and descent rates for the tow vehicle were about 1 meter per second. Near-surface mapping was carried with the ship's underway system and bucket sampling during three-dimensional mapping efforts.

[8] Vertical profiles were obtained with a CTD/rosette profiler equipped with a Seabird 9/11plus CTD, Seabird SBE 43 oxygen sensor, Wetlabs C-Star transmissometer, and Wetlabs ECO-AFL fluorometer. The rosette sampler carried 24 10 L Niskin bottles. The 24 Hz CTD data were processed according to U.N. Educational, Scientific and Cultural Organization [1988] standards, obtaining pressure-averaged data at 0.5 db intervals. Beam attenuation coefficient, c, was calculated from the beam transmission measured with the transmissometer by the equation c = −ln (%transmission/100)/pathlength(m) and is correlated primarily with suspended particulate material in the water. Nutrient samples were filtered (GF/F Whatman®) and stored at −22°C in polyethylene vials, or analyzed on board immediately after collection (winter cruise). Nutrient concentrations (ammonium-NH4, nitrite-NO2, nitrate-NO3, orthophosphate-PO4 and orthosilicate-Si(OH)4) were measured using a Technicon TRAACS 800 autoanalyzer. The elaboration of the data analyses were carried out with an appropriate software (AACE®) supplied by Bran+Luebbe. Nutrient analyses utilized modifications of the procedures developed by Strickland and Parsons [1972]. Determination of ammonium utilizes the Berthelot reaction, in which a blue-green colored complex is formed which is measured at 660 nm. A complexing agent is used to virtually eliminate the precipitation of calcium and magnesium hydroxides. Sodium nitroprusside is used to enhance the sensitivity of this method.

[9] Nitrite is measured by reacting the sample under acidic conditions with sulfanililamide to form a diazo compound that then couples with N-(1-Naphthyl)-ethilenediamine to form a reddish purple azo dye that is measured at 550 nm. Nitrate is first reduced to nitrite at pH 8 in a copper-cadmium redactor then reacted in the same way as nitrite. Nitrate is corrected for nitrite contribution by correcting for the efficiency of nitrate reduction and subtracting the nitrite concentration measured in an unreduced portion of the sample.

[10] Orthophosphate determination is a colorimetric method in which a blue compound is formed by the reaction of phosphate, molybdate and antimony followed by reduction with ascorbic acid. The reduced blue phosphor-molybdenum complex is read at 880 nm.

[11] The procedure for the determination of soluble orthosilicates is based on the reduction of a silico molybdate in acid solution to molybdenum blue by ascorbic acid. Oxalic acid is introduced to the sample to minimize interferences from phosphate. The absorbance is measured at 660 nm.

[12] Dissolved inorganic nitrogen (DIN) was calculated as the sum of the NH4, NO2 and NO3 concentrations.

[13] Chlorophyll a concentrations were measured fluorometrically. 100 ml. samples were filtered through a 25 mm Whatman GF/F filter, extracted for 24 hours in 90% acetone at −4°C, then analyzed on a Turner Designs 10-005 fluorometer that was calibrated with Sigma™ pure chlorophyll. Chlorophyll concentration was calculated according to Holm-Hansen et al. [1965].

[14] Contoured sections were plotted using data that was gridded using kriging (software Surfer 8.0).

[15] During each cruise, the sampling strategy was intended to focus on responses to strong physical forcing of the basin. During winter, Bora winds and the Po River freshwater input were the two major sources of forcing. The spring cruise was planned for the period when climatologically a freshet of the Po River flow occurs (Figure 1). The ships tracks/mapping grids were based on the analyses of meteorological data and on the location of specific features including the Po River plume observed from AVHRR and SeaWiFS imagery (Figures 2, 3, 5, and 6). Figure 1 shows the daily average Po River flow for the year of 2003 and the 14-year means of the daily average flow from 1989–2002.

Figure 1.

Daily averages of the Po River flow for the period 1989–2002. The solid line is the annual average for the period 1989–2002 and the dashed line is the daily flow for 2003 (DV1, period of the wintertime cruise; DV2, period of the springtime cruise).

Figure 2.

Sea surface temperature map from the AVHRR-NOAA-12, NRL (19 February 2003).

Figure 3.

SeaWiFS image of chlorophyll concentration for 20 February 2003 (provided by NRL).

3. Results

3.1. Physical Distribution of the Water Masses

[16] During the winter cruise (31 January to 23 February 2003) two sequential Bora wind events occurred between 11 and 19 February resulting in distinct circulation patterns in the northern Adriatic [Lee et al., 2005; Dorman et al., 2007]. Remotely sensed surface temperature and ocean color showed a strong front in the northern part of the Adriatic extending from Ravenna on the Italian coast to the northwestern corner of the Istrian Peninsula and southward along the western (Italian) boundary of the Adriatic Sea (Figures 2 and 3). Intense wind stress associated with Bora jets from Trieste and Senj drew a cold, fresh plume of Po River water across to northern basin. The Bora winds caused the plume to expand northeastward toward the Istrian coast between a northern cyclonic gyre and an anticyclonic circulation to the south [Lee et al., 2005, Figure 1]. Another front that extended westward from the southern tip of Istria separated the smaller anticyclonic gyre from a larger cyclonic gyre to the south. In situ measurements indicated cooler, fresher water to the north and warmer, saltier water to the south of the front located southern tip of Istria (Figures 4a and 4b).

Figure 4.

Winter surface field of: (a) temperature (°C), contour interval is 0.5; (b) salinity, contour interval is 0.5; (c) DIN (μM), contour interval is 2; (d) orthosilicate (μM), contour interval is 1; and (e) orthophosphate (μM), contour interval is 0.05. (f) Rectangles indicate the Po (A) and Pesaro (B) transects. The dots represent the sampling points.

[17] During the late spring cruise (26 May to 15 June 2003), wind forcing was weak and volume flux from the Po River was about one third of its 14-year average discharge for this period (Figure 1). Despite the low discharge flux, the Po plume remained a significant feature in the northern and western Adriatic. Satellite images (Figures 5 and 6) showed a strong color front along the western boundary that separates the higher-chlorophyll coastal water from the more oligotrophic midbasin and eastern boundary Adriatic waters (Figure 6). Offshore from the mouth of the Po River, the surface layer was characterized by low salinity and high temperature (Figures 7a and 7b). The Po plume extended much more eastward in late spring than in winter because the vertical mixing is reduced [Poulain et al., 2001]. However, the plume extended southward along the Italian Coast.

Figure 5.

Sea surface temperature map from the NOAA-17, GOS-ISAC-CNR (4 June 2003).

Figure 6.

SeaWiFS image of chlorophyll concentration for 4 June 2003 (provided by GOS-ISAC-CNR).

Figure 7.

Late spring surface field of: (a) temperature (°C), contour interval is 0.5; (b) salinity, contour interval is 0.5; (c) DIN (μM), contour interval is 2; (d) orthosilicate (μM), contour interval is 1; and (e) orthophosphate (μM), contour interval is 0.05. (f) Rectangles indicate the Po (C) and Vasto (D) transects. The dots represent the sampling points.

3.2. Large-Scale Surface Distribution of Biochemical Properties

[18] During the winter observations, DIN concentrations were high along the western boundary and decreased rapidly toward the east forming a strong front along the western side of the Adriatic (Figure 4c). This frontal structure extends southward from the Po River discharge consistent with the pattern of the seasonal circulation. The thermohaline circulation [Poulain and Cushman-Roisin, 2001] advects the nitrate-rich river plume water southward, constrained to a relatively narrow band along the Italian coast. The orthosilicate distribution is similar to that of nitrate (Figure 4d). On the eastern side of the basin, south of the Istrian Peninsula a strong temperature front and silicate gradient are present, probably resulting from the eastward extension of the Po River plume in response to the Bora winds [Lee et al., 2005]. The phosphate distribution does not show such a clear gradient in the region south of the Istrian Peninsula (Figure 4e) because minimum concentrations are sometimes found in the coastal waters and increase toward offshore. Thus, the northern Adriatic basin tends to be phosphorus limited despite the large river input.

[19] During late spring the temperature and salinity patterns (Figure 7a, Figure 7b) show a broader warm, low-salinity river plume band extending southward along the Italian coast. The central northern basin was better sampled during the spring cruise than during the winter cruise. Relatively low nutrient concentrations were observed in the offshore and eastern boundary regions of the northern basin (Figures 7c and 7d).

[20] The overall pattern was similar between winter and late spring with the nutrient-rich surface waters along the western boundary advected southward. Some mixing of this coastal water with middle Adriatic surface waters was evident; in particular during the winter a nutrient grading decreased toward offshore while in late spring filaments were present.

3.3. Vertical Distributions

3.3.1. Winter

[21] In addition to the surface mapping and towed vehicle mapping carried out on this expedition several transects with CTD/rosette profiling were obtained. Two sections each from the winter cruise and from the spring cruise are shown in Figures 8 and 9(winter) and Figures 10 and 11 (spring). The locations for the winter transects (indicated in Figure 4f) were across Po River plume (Figure 8) and off Pesaro (Figure 9). The spring cruise transects (locations indicated in Figure 7f) were in the Po River plume (Figure 10) and in a filament extending eastward off Vasto (Figure 11). Nutrient concentrations and chlorophyll a concentrations are shown as color shading and salinity is overlaid as contour lines.

Figure 8.

Po transect during winter cruise (21 February 2003) of DIN, orthosilicate, and chlorophyll a concentration (colored shading). The black contours represent the salinity (contour interval 0.1), and the dots represent the sampling points. The position of the transect is plotted in Figure 4f.

Figure 9.

Pesaro transect during winter cruise (18 February 2003) of DIN, orthosilicate, and chlorophyll a concentration (colored shading). The black contours represent the salinity (contour interval 0.1), and the dots represent the sampling points. The position of the transect is plotted in Figure 4f.

Figure 10.

Po transect during late spring cruise (8 June 2003) of DIN, orthosilicate, and chlorophyll a concentration (colored shading). The black contours represent the salinity (contour interval 0.1), and the dots represent the sampling points. The position of the transect is plotted in Figure 7f.

Figure 11.

Vasto transect during late spring cruise (13 June 2003) of DIN, orthosilicate, and chlorophyll a concentration (colored shading). The black contours represent the salinity (contour interval 0.1), and the dots represent the sampling points. The position of the transect is plotted in Figure 7f.

[22] The core of the Po River plume during the winter is indicated by the near-surface salinity minimum midway in the transect (Figure 8). Salinity induced stratification is strong in the core of the plume and relatively weak at the either end of this transect. The nutrient contribution from the Po plume to the northern basin is evident in the high DIN (∼10 μM) and orthosilicate concentrations (∼5 μM) associated with the salinity minimum of the plume (Figure 8, top and middle). Except for a single sample near surface in the core of the plume, chlorophyll concentrations were lower within the plume and higher at the boundaries where salinity was greater than 38. The chlorophyll distribution was confirmed by the towed vehicle mapping that showed lower chlorophyll concentrations within the plume and higher chlorophyll values in the higher-salinity water masses to either side of the plume (not shown).

[23] The Pesaro transect was located about 100 km south of the Po River delta (Figure 4). During the winter cruise, the water column is well mixed vertically at each of the four stations in the transect (Figure 9). The salinity shows that the cross-shelf section was somewhat complex with lower salinities found both near shore and at the third station (∼18 km offshore), indicative of interleaving of water masses in the region. The lowest salinities (<38.1) and highest beam attenuation values (>5 m−1, graph not shown) indicative of suspended particulate matter occur nearest to the coast indicating the influence of river input. DIN concentrations are higher (DIN ∼1.5–1.8 μM) near the coast and lower (DIN ∼ 0.5–0.6 μM) offshore, consistent with the overall salinity gradient and the riverine source of the nutrients (Figure 9). Silicate concentrations were highest at the third station offshore (∼18 km in Figure 9) at middepth where salinities were less than 33.2. Nutrient concentrations are much lower than those observed nearer the mouth of the Po River, presumably due mixing and perhaps uptake by phytoplankton. The highest chlorophyll concentrations occur within the near shore, low-salinity water in this section. Chlorophyll a concentrations are vertically homogeneous in each of the stations and decrease progressively from 1.2 μg/L near shore and decrease to about 0.6 μg/L at the offshore station.

3.3.2. Spring

[24] During the spring cruise another transect was made through the Po plume, but this time perpendicular to the coast and along the axis of the plume. Wind forcing is much weaker during the spring cruise, and as a result stratification is much stronger. During the period of sampling the Po plume transect the median shipboard wind speed was about 2.4 m s−1. The distribution of nutrient concentrations (Figure 10) shows that, with strong water column stratification, DIN concentrations are highest (10–12 μM) near surface where the lowest salinities are observed. Moderate DIN concentrations are also observed near bottom at one station (Figure 10, ∼8 km).

[25] Orthosilicate distributions show the opposite pattern with concentrations generally less than 2 μM in the surface layer and high values in the lower half of the water column (Figure 10, middle). The highest values of >15 μM are observed where salinity is >38. The distribution of orthosilicate does not appear to be controlled by river inputs but by the active consumption by phytoplankton as reported by Cozzi et al. [2002].

[26] The highest chlorophyll a concentrations from bottle samples, nearly 15 μg L−1, occur near surface in the low-salinity water and decrease monotonically offshore (Figure 10, bottom). The near-surface region where chlorophyll is high corresponds with the region where orthosilicate concentrations are quite low. Moderate concentrations of 1–2 μg L−1 are observed near the bottom at the offshore end of the transect. It is not clear from this hydrographic section whether this near-bottom chlorophyll results from sinking from the surface, or perhaps it is the edge of a subsurface chlorophyll maximum that is typical of regions away from the direct influence of the Po plume.

[27] About 330 km southeast from the Po River delta in the southern part of the central basin, a small filament distinguishable in ocean color (Figure 6) was observed extending offshore toward the east from the coastal region. A hydrographic section was obtained through the feature to determine its characteristics (Figure 7f). In the section, the filament is evident as a near-surface salinity minimum where salinity is less than 38.5 (Figure 11). The thickness of this low-salinity filament is less than 15 meters. Despite the evidence from the ocean color image, near-surface chlorophyll within the filament is low, <0.5 μg/L and maximum chlorophyll of up to 0.9 μg/L is observed in the chlorophyll maximum at depths of 70–75 m (Figure 11). Higher near-surface concentrations of DIN and orthosilicate concentrations are associated with the low-salinity core of the filament. However, the highest concentrations of DIN and orthosilicate are observed below 100 m. There is some complexity to this distribution with higher DIN present near surface outside the filament in high-salinity surface at the station located at about 22 km, the southern end of the transect. Below 100 m the silicate and DIN distributions also differ. Highest orthosilicate concentrations occur at either end of the transect, whereas DIN concentration is lower at the southern end of the transect.

4. Discussion

[28] The general patterns that were observed in the winter and late spring of 2003 demonstrate the responses of the northern Adriatic to wind forcing and freshwater inputs along the western coast that have been described previously [e.g., Zore-Armanda, 1963; Artegiani et al., 1997a; Zavatarelli et al., 1998].

4.1 Winter Characterization of Northern and Western Adriatic Sea

[29] In winter, in the northern Adriatic Sea, Bora winds result in frequent, intense, narrow sea surface wind jets coming from the northeast through the mountain passages along the eastern side [e.g., Poulain et al., 2001]. During the winter cruise, two sequential Bora events occurred during 11–19 February [Lee et al., 2005; Dorman et al., 2007].

[30] In order to describe the simultaneous influence of Bora wind and Po plume on the distribution of the water masses in the northern and western part of the Adriatic Sea, three regions have been compared: Po plume area, Pula area and Pesaro area.

[31] To understand the relationship of biogeochemical variables with water mass variability, nutrient and chlorophyll concentrations are superimposed on temperature/salinity plots (Figure 12). Temperature/salinity data are included from the ship's underway uncontaminated water system (blue dots), and the tow vehicle's CTD (green dots and red dots) to provide a full perspective of the T/S variability within the northern Adriatic during the cruise period. The broad-scale survey (green dots) spanned the width of the basin and extended along basin from the northern tip of the Istrian Peninsula to the Jabuka depressions that extends approximately northeast from Pescara. This survey began late on 31 January and ended on 6 February 2003. A second tow vehicle transect on 23–24 February, referred to as the Along Basin transect in Figure 12, followed the Bora events between 11 and 19 February [Dorman et al., 2007] and extended down the axis of the northern basin over the same along-basin extent as the broad-scale survey. The negative heat fluxes associated with the Bora events [Dorman et al., 2007] are evidenced in the decrease in water temperature by approximately 2°C between these two surveys, and is most clearly seen where salinity is less than about 37.5.

Figure 12.

Temperature/salinity diagram from the winter cruise period of 31 January to 24 February 2003. Temperature and salinity data are shown in blue for the ship's underway thermosalinograph, as green dots for the tow vehicle CTD for 1–6 February, and as red dots for the tow vehicle CTD for 23–24 February. Si(OH)4(μM), DIN (μM), DIN/Si(OH)4 (μM), and chlorophyll (μg L−1) ranges are indicated by the black symbols. These concentrations are indicated for three comparative regions: near the Po delta (circles), south of Pula (diamonds), and off Pesaro (inverted triangles). Concentration ranges for each variable are indicated in the legend and are represented by the symbol size.

[32] In this period it was possible to describe some biochemical characteristics of the water masses. In the zones investigated, two water masses were distinguished. Salinity was used to discriminate water masses in the three regions investigated. Using this parameter two different water masses in each regions were found (Table 1 and Figure 12). Water unaffected by river runoff generally had salinities equal to or greater than 35.5, which characterizes North Adriatic Deep Water [e.g., Artegiani et al., 1997b] and water influenced by the river runoff. The water mass influenced by river runoff was fresher, colder, richer in orthosilicates, DIN, DIN/orthosilicates ratio and chlorophyll a, and confined to the northern Adriatic, between the Po Delta and the Istrian peninsula, and along the western Adriatic coast. The more saline water mass was warmer and less nutrient-rich water and was found in the more offshore regions. The two water masses, always present in the three surveys, were compared and the results were summarized in Table 1 and showed in Figure 12. The northern region where salinity was less than 37.5 was strongly impacted by Po River runoff, in particular high values of DIN, orthosilicates and DIN/orthosilicates ratio were found. A DIN/orthosilicate ratio close to 1 is good for diatom phytoplankton growth as observed by Redfield et al. [1963] and Brzezinski [1985]. The high surface values of nutrients concentration were evident in vertical section across the Po river plume (Figure 8) in correspondence with low salinity up to 20 m depth in the center of the plume. The maximum of chlorophyll a was at the edge of the plume owing to higher chlorophyll concentrations in waters outside the plume. The lower chlorophyll values within the plume may be due to cold temperature and low light concentrations within the plume, inhibiting growth of phytoplankton within the plume. South of the Po plume and east of Pesaro, DIN and orthosilicate concentrations decrease and temperature and salinity increase (Table 1). In Figure 9, DIN, orthosilicates and chlorophyll a concentrations decreased from the cost toward offshore while the salinity increased. The influence of Po River discharge is quite evident in the northern Adriatic area and along the western coast in front of Pesaro (Figures 2, 3, and 4). DIN concentrations decrease threefold from north to south and orthosilicates decrease approximately twofold. Both temperature and salinity increase eastward and southward from the Po delta, whereas orthosilicates and DIN concentrations decrease (Table 1). South of the Istrian Peninsula, a strong front separated colder, fresher, and more DIN-rich water on the north from warmer, saltier, less DIN-rich water south of the front (Table 1). This front marks the eastern extension of the Po plume driven by strong Bora winds on either side of the plume. Orthophosphate concentrations do not show much difference between the zones, indicating little or no contribution from the Po River runoff.

Table 1. Winter Cruise (31 January to 24 February 2003) Parameters for Distinguished Two Water Masses in Three Areasa
AreasTemperature, °CSalinitySi(OH)4, μMDIN, μMPO4, μMDIN/Si, μMChl a, μg L−1
  • a

    Number of data given in parentheses, ±standard deviation.

Po Survey (surface)
S < 37.57.06(12) ± 0.3437.02(12) ± 0.324.03(12) ± 1.126.51(12) ± 3.690.17(12) ± 0.121.51(12) ± 0.590.88(12) ± 0.20
S > 37.58.71(21) ± 0.5437.91(21) ± 0.112.20(21) ± 0.901.50(21) ± 0.830.11(21) ± 0.060.74(21) ± 0.420.79(20) ± 0.19
Pula Survey (Surface)
S < 38.1510.08(13) ± 0.4637.84(13) ± 0.143.77(13) ± 0.511.97(13) ± 1.070.18(13) ± 0.120.53(13) ± 0.310.63(13) ± 0.05
S > 38.1512.17(12) ± 0.2738.32(12) ± 0.031.95(13) ± 0.240.97(13) ± 0.290.15(13) ± 0.110.50(13) ± 0.150.45(13) ± 0.04
Pesaro Survey (Surface)
S < 38.29.62(15) ± 0.7637.94(15) ± 0.182.30(15) ± 0.942.17(15) ± 1.220.24(15) ± 0.071.07(15) ± 1.050.91(15) ± 0.32
S > 38.211.39(13) ± 0.5538.38(13) ± 0.071.60(13) ± 0.521.06(13) ± 0.600.22(13) ± 0.070.76(13) ± 0.610.63(13) ± 0.13

4.2. Late Spring Characterization of Western Adriatic Coast

[33] During the late spring cruise strong stratification characterized the water column of the northern Adriatic (Figure 10). A SeaWiFS image of chlorophyll from 4 June shows the southward extension of the Po plume along the western boundary within the flow of the WAC (Figure 6). Despite below average flow from the Po River, buoyancy driven flow was evident along the western boundary of the basin consistent with results from models and observations [e.g., Kourafalou, 1999; Poulain, 2001].

[34] In order to evaluate the influence of Po River runoff on biochemical properties of surface waters from three regions along the western Adriatic are compared: the Po plume area and the coastal regions off Pescara and Vasto (Figure 7). As in winter, salinity was used to discriminate between the different water masses. Similar to the winter case, temperature/salinity structure indicates that temperature is highly variable (Figure 13). The T/S plots in Figure 13 include a similar set of observations as in Figure 12 to provide an overall sense of the T/S variability in the system. The blue dots indicate temperature and salinity data from the ship's near-surface uncontaminated water system, the green dots show data from the towed vehicle's broad-scale survey (similar area to the winter broad-scale survey) on 26–29 May, and the red dots indicate profile data from the CTD/rosette profiler during the Po plume hydrographic transect on 8 June 2003. The effects of solar insolation on upper layer heating are apparent in the increase of temperature in the low-salinity plume water. For example, the temperature at salinity of 32 in the Po plume increased by ∼4°C from the period of the broad-scale survey in late May to time of the Po plume transect on 8 June (Figure 13). As in winter, two water masses are identified from the late spring cruise: fresher water where salinity is generally less than about 38.2 is indicative of the influence of river inputs along the western boundary, and more saline water, greater than 38.2, is present offshore (Figures 7 and 13 and Table 2). The area influenced most directly by the Po River discharge is characterized by salinities less than 37.5 (Figure 13). As in winter, salinity increased toward the east, but in contrast to the winter, strong vertical stratification constrained this mainly to the surface layer.

Figure 13.

Temperature/salinity diagram from the spring cruise period of 26 May to 15 June 2003. Temperature and salinity data are shown in blue for the ship's underway thermosalinograph, as green dots for the tow vehicle CTD for 26–29 May, and as red dots for the Po plume hydrographic section on 8 June. Si(OH)4 (μM), DIN (μM), DIN/Si(OH)4 (μM), and chlorophyll (μg L−1) ranges are indicated by the black symbols. These concentrations are indicated for three comparative regions: near the Po delta (circles), south of Pula (diamonds), and off Pesaro (inverted triangles). Concentration ranges for each variable are indicated in the legend and are represented by the symbol size.

Table 2. Late Spring Cruise (26 May to 15 June 2003) Parameters for Distinguished Two Water Masses in Three Areasa
AreasTemperature, °CSalinitySi(OH)4, μMDIN, μMPO4, μMDIN/Si, μMChl a, μg L−1
  • a

    Number of data given in parentheses, ±standard deviation.

Po Survey (surface)
S < 35.723.95(11) ± 1.9434.18(11) ± 1.971.37(11) ± 0.862.74(11) ± 2.890.19(11) ± 0.214.19(11) ± 6.796.01(9) ± 4.49
S > 35.723.00(18) ± 0.8136.58(18) ± 0.461.75(17) ± 1.531.93(17) ± 1.980.13(17) ± 0.0327.41(17) ± 71.602.97(2) ± 1.89
Pescara Survey (Surface)
S < 37.521.61(3) ± 1.1236.21(3) ± 0.832.50(3) ± 0.584.64(3) ± 1.080.20(3) ± 0.021.88(3) ± 0.390.81(3) ± 0.31
S > 37.520.29(6) ± 0.4838.74(6) ± 0.020.15(6) ± 0.182.95(5) ± 2.060.09(6) ± 0.04130.15(5) ± 121.690.15(6) ± 0.03
Vasto Survey (Surface)
S < 37.4224.40(7) ± 1.7036.54(7) ± 0.872.43(7) ± 1533.36(7) ± 1.860.14(7) ± 0.121.72(7) ± 1.180.50(3) ± 0.02
S > 37.4225.52(9) ± 0.3937.85(9) ± 0.431.86(7) ± 1.281.59(7) ± 1.580.12(7) ± 0.090.87(7) ± 0.470.32(4) ± 0.11

[35] In the fresher water of the Po plume (salinity <35.7, Table 2) nutrient concentrations were variable. The highest concentrations of nutrients were found at intermediate salinities where S > 35.7 (Figure 13). DIN/orthosilicates ratio were variable, higher in the more saline water than in the fresh water. However, DIN/orthosilicate ratios were much higher in summer than in winter period when DIN/Si ratios were nearly all less than 2.5. The much larger DIN/Si ratios in the spring are most likely due to greater consumption of orthosilicate by diatom phytoplankton groups in response to high light availability near the surface and stratification. Diatoms were a dominated the phytoplankton community of the Po plume (I. Cetinic, personal communication, 2005). In the bottom plot in Figure 10 the maximum surface chlorophyll a of about 15 μg L−1 corresponded to a minimum of orthosilicate concentration as observed by Socal et al. [2002].

[36] The Pescara area showed larger differences in the water properties between the surface coastal water and the surface offshore water (Table 2). Orthosilicate and DIN concentrations decreased from near the coast toward offshore and DIN/orthosilicates ratio increased. In contrast to winter when concentrations of nutrients decreased southward from the Po River plume, concentrations off Pescara were higher than concentrations in the Po plume area.

[37] In the most southern region off Vasto, a filament extending from the coast toward offshore was evident in the SeaWiFS image (Figure 6). In situ measurements across the filament (Figure 11) indicate that it contains less saline surface water than the water into which it advects (Table 2, average S = 36.54 ± 0.87). Although the filament is generally characterized by water less than saline than 38.55, the salinity of hydrographic samples from the filament were <37.42. Outside of the filament the where salinity was greater and temperatures warmer; the mean concentrations of orthosilicates and DIN decreased (Table 2).

[38] In all areas chlorophyll a concentration was higher in the lower salinity coastal water than in the more saline water offshore and decreased from the area of Po plume southward along the coast. A deep chlorophyll a maximum at about 70–75 m depth was present in the offshore higher-salinity water where deep penetration of the light field coincided with the top of the nutricline [Boldrin et al., 2002].

4.3. Characterization of the Seasonal Biochemical Variations

[39] In general the basin is characterized by decreasing nutrient gradient in the surface layer from the western boundary eastward. Nutrient levels in the northern Adriatic result from river input not only from the Po River, but from other smaller rivers along the Italian coast. Wintertime DIN and orthosilicates concentrations were on average twice as high as late spring concentrations in the same region (Tables 1 and 2). That is probably due also to both low Po River discharge and high phytoplankton uptake of nutrients during the spring, indicated by the relatively high chlorophyll concentrations within the Po plume.

[40] During the winter cruise the Po plume was more clearly defined extending northeastward toward the Istrian Peninsula in response to strong Bora winds from Trieste and Senj, and higher, but typical, river discharge rates. During the late spring cruise, although the Po plume spread more broadly, the area south of the Istrian Peninsula appeared less influenced by the Po plume (Figures 7c and 7d) perhaps because of below average river discharge rates and the absence of strong wind forcing. The middle Adriatic shows less influence from the Po River. During springtime local river inputs contribute to the nutrient concentrations along the coast [Marini et al., 2002; Campanelli et al., 2004]; in particular the Pescara and Vasto areas are characterized by nutrient concentrations similar to the Po plume area (Table 2). Nutrient and chlorophyll concentrations are highest in the western coastal areas of the Adriatic during both seasons.

5. Conclusions

[41] The biochemical characteristics of water masses in the northern Adriatic and the western boundary of the Adriatic have been presented. Because temperature is highly responsive to seasonal heat fluxes in the shallow northern region of the Adriatic, salinity is a better discriminator of water mass variability and mixing in this region.

[42] During winter the extent and shape of the Po plume appeared to respond to Bora winds, extending northeastward toward the Istrian Peninsula carrying highs concentrations of DIN and orthosilicate. In general, nutrient concentrations were negatively correlated with salinity, nutrients increasing with decreasing salinity. Little accumulation of phytoplankton biomass was observed within the Po plume, and nutrient concentrations tended to be transported offshore with the freshwater. The Western Coastal Layer, observed in the Pesaro transect, showed a nutrient concentrations decreasing and salinity increasing toward offshore. Coastal advection transported freshwater, nutrients, and suspended material southward, dominated by physical mixing, and with limited phytoplankton growth.

[43] In late spring, the coastal waters along the western boundary were characterized by lower salinity water near the coast and saltier water offshore, as in winter period. Unlike the winter period, the onshore-to-offshore gradient is confined vertically mainly to the surface layer because of strong stratification. During the spring cruise, low Po River discharge (∼625 m3/s) and weak wind forcing resulted in a broadly spreading, vertically stratified river plume, where high phytoplankton abundance contributed to rapid depletion of nutrients. In the central part of the Adriatic basin a filament extended offshore from the coast. In situ measurements showed that this filament was characterized by lower salinity and temperature, and higher concentrations of orthosilicates and DIN relative to the surrounding water. Alongshore advection of the plume does occur, but because of the weak mixing, relatively slow advection, nutrients are relatively low in the advected plume, phytoplankton biomass is elevated relative to the offshore water, but decreases rapidly alongshore and in offshore filaments because of the lack of nutrients to sustain phytoplankton growth.

[44] DIN/orthosilicate ratios were much typically less than 2–3 during the winter when phytoplankton growth was small. Low phytoplankton growth rates were probably the result of cold temperatures, low incident light, high attenuation of light in the plume, and deeper mixing than in spring. During the late spring the DIN/orthosilicate ratios were much higher. The very high ratios in spring reflected the phytoplankton uptake under conditions of high available light and stratification.

Acknowledgments

[45] We are grateful to the crews of R/V Knorr, R/V G. Dallaporta and Paola Fornasiero for their assistance with sample collection. The research was supported by the “Dynamics Of Localized Currents and Eddy Variability In The Adriatic” (DOLCEVITA) program funded by the U.S. Office of Naval Research (award N000140210854 to B. Jones), NATO, the Croatian Ministry of Science and Technology and the Italian Ministry of the Environment and Ministry of Universities and Research. Winter SST and SeaWiFS chlorophyll were provided by Robert Arnone (Naval Research Laboratory, Stennis Space Center).

Ancillary

Advertisement