Control mechanisms on the reactive silicate fluxes by using on board resuspension experiments in the eutrophicated coastal environment

The Gulf of Izmir is a large natural bay in the Mediterranean, located between latitudes 38°20’N and 38°40’N and between longitudes 26°30’E and 27°10’E. It has a semi-closed basin and extends along the west coast of Turkey into the Aegean Sea. The average depth of the Bay is 15 meters and its total area is 500 km². It has been contaminated with domestic and industrial waste for the last 30 years. The bay waters have therefore eutrophic features. The quality of the water and the sediment of the inner bay is seriously threatened by industrial waste and pollutants discharged into rivers. In terms of its hydrographic and topographic structure, the Gulf is divided into three parts: the inner, the middle and the outer bay (Güven & Akıncı 2008, Ozkan 2012). The area of our study is limited to the inner bay rather than the entire Gulf of Izmir (Ozkan et al. 2008). The city of Izmir is extending both into the inner and outer bays and is exposed to a wide variety of pollutants, such as domestic and industrial waste. The eastern and northern parts of the inner bay, which constitute the research area, form the river storage shores and the southern part the rift (Başoğlu 1975).

Sediment samples were taken from 16 stations seasonally 4 times between February 2015 and November 2015. Field studies in the winter and spring periods were carried out with the PiriReis RV; during the summer period with the Dokuz Eylül1 RV and the autumn period with the Dokuz Eylül3 RV.

There is a mixture of materials, both terrestrial from the north and south at the western border of the inner bay and from different sources in the inner harbor (Ozkan et al. 2012). Sampling stations are as shown in Fig. 1.

Figure 1

The CO₃^-2 values of the east coast (stations 1, 2, 4, 5, 9, 11) and the south coast (station 13) are more than two times the west coast (stations 14, 15, 16) in winter (Fig. 2). A maximum CO₃^-2 value of 10% was observed at sta. 4. In the western region CO₃^-2 value decreases to < 2.5%. It is observed that in the spring stations 3, 4, 9, 10, 11, 12, 13 and 16 have relatively high CO₃^-2 concentrations as far as the north and south coasts compared to the winter period (Fig. 2).

Figure 2

While there is less than 4.4% CO₃^-2 in stations 7, 11 and 15 in the interior of the study area during summer, there are higher CO₃^-2 values in other stations (coastal stations) surrounding these stations. In the stations on the north coast especially CO₃^-2 values have reached more than 6.6% (Fig. 2). The amount of CO₃^-2 increased at the stations on the northern coast in autumn compared to summer and CO₃^-2 values greater than 9% were observed. At stations 3, 6, 7, 12 and 15, the CO₃^-2 values were less than 5.5%.

As a result, while a max CO₃^-2 deposit is observed in the east in winter, this deposit expands towardsthe north and the south in spring. While CO₃^-2 deposits reaches max values on the northern shores in summer, a similar trend is observed in autumn (Fig. 2).

In winter at stations 9 and 10 in the eastern and southern parts of the study area high orgC values of > 5% (Fig. 3) were recorded. The high values of OrgC in spring were observed in the northern, southern and eastern parts of the study area. In summer in the eastern and southern regions high orgC values were unchanged, while in the northern and western regions, there was a reduction in orgC values. OrgC values of < 3.4% at stations 2 and 6 in the north of the study area are worthy of note. In autumn the orgC values in the northern reaches of the study area are < 3.2%. In the south orgC values were found at < 2.4% only in station 10. The highest orgC value was observed at station 4 to the east of the study area (5.2%) (Fig. 3). When we compare Figures 2 and 3 in stations where CO₃^-2 is increased in general, the orgC value is lower. OrgC takes control of the silicate flux in stations with the same CO₃^-2 values.

Figure 3

Mn concentrations in sediment are relatively high at stations 4, 8, 11, 12 and 13 during winter (Fig. 4). Mn values are lower in spring than in winter (except for station 4). The high Mn value of up to 2217 ppm only at station 4 is remarkable. Station 4 therefore seems to be the most important source of Mn for the sediments of the Gulf of İzmir Inner Bay. It is understood that mn spreads from this station into the environment. This is only valid for station 4 in spring. High Mn values at stations 6, 7, 12 parallel to the east coast in summer indicate the manganese that has been mobilized from inner part of the sediment into the water, possibly due to the reduction of sediment conditions (Burdige, 1993). Mn values are relatively low at the east and west stations in summer (Fig. 4). Mn distribution in autumn shows a similar situation to that of summer. It may be said that the majority of Mn accumulates in a line closer to the shore due to water movement.

Figure 4

The BSi distribution patterns are similar throughout the four seasons (Fig. 5). While BSi values were observed quantitatively at their highest values in winter, they reached the lowest values in summer. It may be said that high BSi values in winter are caused by benthic diatom blooms. Stations 13, 14, 15 and 16 in the west have the lowest BSi values during all seasons (Fig. 5). The sediment of station 9 has the highest BSi value in all seasons, so it is a source of RSi (reactive silicate) for the water column for diatom blooms.

Figure 5

The distribution patterns of chl-a values throughout the seasons are similar in sediment. It is also closely related to BSi (Fig. 6). This situation may be explained in two ways: first, Chl-a is likely to show the content of live benthic microalgae. The second possibility is that chl-a may have been preserved in the sediment by carbonate precipitation (Strickland & Parsons, 1972). As a matter of fact, in determining chl-a in sea water, an MgCO₃ solution is added in order to prevent chl-a from turning into feopigment and filter paper is provided. With the grazing of zooplankton, chl-a also turns into feopigment in the acidic conditions of the digestive system. Precipitation of calcium and magnesium carbonate in the basic environment may therefore stop a decrease in the pH. Chl-a values, which increase in autumn, reach their maximum in winter. Spring chl-a values decrease and reach the lowest value in summer (Fig. 6) and nitrogen determines the growth rate of phytoplankton in the inner bay (Sunlu et al. 2005) due to low productivity in summer, so the amount of organic matter that sinks to the sediments and the decomposition of organic matter is also less. This will limit the development of benthic microalgae.

Figure 6

The BSi and chl-a distributions in the sediment are similar. Controlled entry of dissolved silicate by CO₃^-2 in the study area ensures that the silicate remains longer in the region and supports the productivity of benthic diatom. Due to the primary productivity of benthic microalgae in the bottom water/sediment interface, CO₃^-2 precipitates onto the sediment surface. The CO₃^-2 that sinks crystallizes on silica skeletons, reducing the solubility of the silica. In summer especially, when river mouths are low, the sediment supports the productivity of diatom by providing dissolved silicate to the water column. In winter high primer productivity (PP) and high chl-a biomass in the study area may reflect relatively weak feeding activity with benthic suspension. Hily (1991) states, however, that the feeding activity of benthic grazers from Brest Bay might explain the paradox between high PP and low chl-a biomass in summer. It was also concluded that dissolved silica from diatom skeletons in these pellets was slowly diffused into the water column, leaving the fecal pellets of the gastropod Crepidula fornicata with benthic microalgae. It was therefore emphasized that the productivity of diatom was supported in summer.

When looking at the dark net Si flux, the direction of flux is from water to sediment in stations 5 and 6 in winter and in stations 6, 8, 11, 12, 13, 14, 15 and 16 in autumn (Fig. 7). In these stations the silicate flux calculated from the control samples containing only bottom water comes mainly from the dissolution of the diatom skeletons in the bottom water. The change of net silicate flux to negative values should therefore result from the uptake of silica by benthic microalgae. In spring and summer the net dark silicate flux direction is from sediment to water. In the control trials conducted in these seasons, negative silicate flux values were determined at certain stations. This may be explained by dark silicate uptake of the diatoms in the overlying water, because dissolved silicate is not scavenged (Koschinsky et al. 2001). The dark silica fluxes determined in the resuspension event were found approximately between 4 and 10 times higher than the values obtained by Büyükışık et al. (2015) without resuspension.

Figure 7

A negative linear relationship was observed between the silicate concentrations in pore water and reactive silica (RSi) fluxes from the sampling conducted in January 2008 in still weather conditions (Büyükışık et al. 2015). The main factor determining the silicate flux from the sediment was the permeability of the cement material (clay, fine silt) in the sediment (Fig. 8). Increased pore water silicate concentrations indicate reduced permeability.

Figure 8

Where the permeability is minimum the pore water silicate value is 300 μM (Fig. 8). This 300 μM value increases with temperature. It is impossible to differentiate other variables that may control the flux under stagnant weather conditions. The variation of the data obtained from silica fluxes from sediment in stagnant weather conditions is higher than the variation of silica flux data obtained from resuspension of the sediment made in this study (Tengberg et al. 2003).

In this study resuspension of the sediment in stormy weather conditions eliminates permeability, which provides control of the silicate flux. Each sediment particle is equilibrated separately with the water column. In order to simulate stormy conditions, the intensity of the air bubbling was increased and the sediment upper layer (0-5 cm) was allowed to mix into the water column in the core liner.

An increasingly positive linear relationship between BSi and CO₃^-2 was observed in the sediment in most stations during winter. As the abundance of benthic diatoms increases, consuming CO₂ for photosynthesis causes carbonate precipitation (Fig. 9). At the same time, the increase in the abundance of the benthic and planktonic diatom communities increases the inorganic carbonate deposits as it makes the marine environment more alkaline. It is therefore expected that a linear relationship between these two parameters in the winter season be observed, a process we may define as bio-mediated inorganic carbonate precipitation. Due to the positive relationship between these two parameters, it may be assumed that the source of calcium carbonate in the sediment is marine because the biogenic silica is of marine origin. In Fig. 9, while the carbonate values at stations 7, 8 and 9 are almost the same (5%, 5.33% and 4.7% respectively), BSi values vary significantly. Scrippsiella trochoidea, Emilliania huxleyii, Chrysochromulina in the inner bay of the Gulf of Izmir was reported by Koray, (2001). S. trochoidea accumulates CaCO₃ on the cyst, while E. Huxleyii and Chrysochromulina form CaCO₃ on the scales on the outer surface of their cells. These carbonate structures therefore constitute different marine carbonate sources.

Figure 9

Carbonate values independent of biogenic silica at stations 9, 13, 14, 15 and 16 during spring may indicate a different marine source. This was clarified by the relationship between Ca-Ti below (see Fig. 11). The negative relationship observed between the two parameters at other stations (1, 2, 3, 4, 5, 6, 7, 8, 10, 11 and 12) may indicate that the biogenic silica in the sediment dilutes the carbonate. S. trochoidea cysts also prevent CO₃^-2 accumulation due to the productivity of diatom, as it reduces CO₃^-2 content in sea water by precipitating it as CaCO₃ (Janofske 2000, Shin et al. 2013, Koray et al. 1992, Jantschke et al. 2019). These processes may be referred to as biogenic carbonate precipitation. Relationships between biogenic silica and carbonate are extremely weak in summer and autumn, seasons that may be considered to have different dominant species from winter and spring.

Strong linear regressions between BSi-chl-a are noted for all four seasons. BSi explains the changes in chl-a in between 98% and 99% per season (Fig. 10). Increased CaCO₃ precipitation with an increase in biogenic silica should occur with live diatom. Dead diatom cannot produce primary productivity. Since benthic diatoms make an important contribution to primary productivity, 99% of BSi should probably be living diatom. Increased pH may also prevent the decay of chl-a in the diatom following carbonate precipitation. The regressions are extremely close to each other for all four seasons. The intercepts explaining the chl-a of dinoflagellate are between 1.1 and 2.5 (Fig. 10). Dinoflagellat pigments add a bias to the diatom/chl-a relationships.

Figure 10

The fact that Ti-Ca relations are strong-negative in all four seasons explains that terrestrial Ti is diluted by the autigenic Ca (Fig. 11). The negative relationship observed in each graph shows that Ca and Ti sources are of different origins. The fact that Ca is of marine origin explains that CaCO₃ precipitates in the water column or overlying water.

Figure 11

In winter there is a strong correlation between CO₃^-2 and Mn except at station 1. Changes in carbonate may explain changes in Mn by 79.66% (Fig. 12, Table 3). Station 4 appears as the center of carbonate precipitation and also the source of Mn in winter and spring. The strong relationship observed may be explained by the participation of Mn in the CaCO₃ crystal lattice. The master variable is bio-mediated carbonate, since Mn levels are between 400 ppm and 700 ppm, while carbonate levels are between 2% and 11% (10000 ppm = 1%). Manganese is only an indicator of carbonate precipitation. In other words, it may be thought that CaCO₃ and MnCO₃ co-precipitate in sediment.

Figure 12

The highest value of manganese in spring (2217 ppm) was seen at station 4 (Fig. 12). At other stations the Mn value was around 400 ppm and was almost constant. This situation may be attributed to the reduced state of the sediment with increasing productivity in the water column and the mobility of manganese towards the sediment surface. The Mn/ CO₃^-2 relationship was not observed in this season. Negative relationships between Mn/CO₃^-2 in summer and autumn may be interpreted as Mn and CO₃^-2 and have different sources or are in separate solid phases (Mn: terrestrial/sediment origin, CO₃^-2: marine origin). In other words, CO₃^-2 dilutes Mn. CO₃^-2 may be said to be of biogenic origin.

The weak relationships observed between BSi and Ti during all four seasons may be the result of nutrients and other physicochemical parameters acting on diatom growth (Table 1). BSi is therefore controlled by factors such as nutrient levels, carbon dioxide and light intensity in each station. For this reason, Ti and BSi are independent of each other in all four seasons, especially in winter and autumn.

The relationship Ti/CO₃^-2 is negative in all seasons except spring. In these three seasons the source of CO₃^-2 should therefore be different from the source of Titanium (terrestrial source). We may therefore say that most CO₃^-2 is in the form of CaCO₃ and is of marine origin. As a matter of fact, Table 2 shows us that Ca⁺² is of marine origin. In spring the relationship Ti/CO₃^-2 is not observed. It is likely that CO₃^-2 is secreted from dinoflagellate cysts or by haptophyte members.

Relationships between dark silicate flux and BSi, CO₃^-2, OrgC, Mn were examined according to seasons. The reason silicate flux experiments are not emphasized in the light is the simplification of the processes (photodegradation, photosynthesis, etc.) affecting the sediment.

The correlation between the RSi flux and both BSi and CO₃^-2 in a dark environment is negative at all stations (Figs 13a and 13b) with the exception of station 6 (Fig. 13b) in winter. It may be either a feedback mechanism or an indirect mechanism that biogenic silica controlled silica flux. The decrease in flux with BSi, which increases as a result of the proliferation of larger benthic diatom species, may be due to the decrease in the surface areas of diatom species per cell volume. As a feedback mechanism, with increasing alkalinity in overlying water via primary productivity, carbonate precipitation increases except at stations 7, 8 and 9 (see Fig. 9, winter). Stations 7, 8, 9 have carbonate levels much lower than expected from the linear model, as a result of low levels of benthic productivity. The low carbonate levels in stations are provided by control of the organic coating as a result of increased exuding of orgC due to a reduction in other nutrients with the exception of silica. In Fig. 13b the difference in silica flux between the value modeled and the observed at station 6 may be explained by BSi levels. In Fig. 13b, three separate negatively sloped and almost parallel lines may therefore represent three distinct diatom size groups. In general, diatom size groups add bias to CO₃^-2/JSi regressions changing the intercept of the line. Nature reduces laterally dissolved silica export (loss of silica) by vertical sinking of diatom skeletons from the bay due to the sufficiently dissolved silica from point or nonpoint sources throughout the region’s rainy winter. As a matter of fact, it causes a slow dissolution of the silica skeletons from the fecal pellets left by the benthic gastropods, Crepidula fornicata, grazing on the benthic microalgae in Brest Bay. In the dry and silicate-poor summer the silicate in the water column therefore remains in the Brest Bay (Fouillaron et al. 2007). Similarly, the existing diatoms in the water column are transported to the sediment in the form of fecal pellets either by sinking or grazing in the inner bay. This mechanism in the water column also reduces the lateral export of the silicate.

Figure 13

The main mechanism that controls the RSi flux is CO₃^-2 precipitation. BSi controls the flux indirectly because it controls both diatom surface area (Fig. 13a) and carbonate (see Fig. 9) according to their biological activities.

Fig. 14a is similar to Fig. 13a. because BSi and OrgC represent biomass. Organic carbon controlling silica flux is similar to BSi and yields a negative linear relationship (Table 3). There are negative relationships between organic carbon and silica flux in winter. These negative relationships indicate that a significant part of the biogenic silica is living diatom (see Figs 10 and 14a). In Fig. 14a the organic matter may be thought to control the silica flux because the CO₃^-2 values are very close to each other at stations 7, 8, 9 and 10 (5%, 5.33%, 4.70% and 4.1% respectively). Organic carbon cannot clearly explain the silica flux at stations 1, 2, 3, 11, 13 and 14. It may be said that carbonate and BSi rather than organic carbon better explains the silica flux at these stations. The unit of Mn is ppm and the unit of CaCO₃ is %. There is homogeneous dispersion of MnCO₃ in the CaCO₃ crystal lattice in bio-mediated carbonate precipitation. For these reasons Mn is not a control mechanism of silica flux, but just a flux control indicator (Fig. 14b) in winter.

Figure 14

Fig. 15a shows three different situations between dark silica flux and BSi, two negative and one positive relationship (Table 3). A separate pattern with a positive linear relationship was distinguished with the pattern defined by two different equations with BSi and dark silica fluxes that are negatively related in spring (Fig. 15a). As may be understood from the two negative equations, as the BSi increases, so the dark silicate flux decreases as in winter. This decrease of the dark silica flux shows both the increase of the diatom size and the decrease of diatom surface area. There is a positive relationship between dark silicate flux and BSi at stations 2, 8, 9, 10, 13, 14 and 16 in the lower graph of Fig. 15a showing a decrease in diatom size with increase in BSi or biomass. Increase in total diatom surface area related to diatom size decrease means increasing silica flux. Another explanation of this positive correlation suggests the dilution of biogenic silica by biogenic carbonate of marine origin as a cause. Biogenic carbonate is unrelated to dark silica flux. Carbonate added to sediment reduces the biogenic silica content. In these stations it may also be that Scripsiella trchoidea cysts, which are a dinoflagellate, are dominant or the inclusion of scales formed by haptophyte members in the sediment may provide this dilution (Koray 2001). In this case the CaCO₃ formed by both cysts and haptophyte scales may block the carbonate precipitation of diatom productivity by lowering the concentration of dissolved carbonate in the bottom water to under saturation.

Figure 15

In this study, direct microscopic cell observations were not made. The groups that produced CaCO₃ are not only S. trochoidea but also E. huxleyii (Koray 2001), and some flagellates (ex. genus Chrysochromulina) have been reported to form CaCO₃ scales (Raymond 1980).

There are two different there are two differences between silicate flux and carbonate in the spring (upper graph of Fig. 15b). The first are low dark silica flux values, where the carbonate value is less than 7% and the high dark silica flux values, where the carbonate value is greater than 7%. The second is the control line showing that dark silica flux reduces carbonate values between 3% and 7% (lower graph of Fig. 15b). A single group of diatom sizes may therefore be mentioned in which silica flux is controlled by bio-mediated carbonate in spring (Fig. 15b).

It is worthy of note that three separate lines with a negative slope are statistically strong between OrgC and silicate flux (Fig. 15c). Organic carbon, BSi and Chl-a values are a measure of diatom biomass in surface sediment. OrgC also has a controlling mechanism on silica flux.

Only in bio-mediated carbonate precipitation, manganese precipitates together with CaCO₃ and joins the CaCO₃ lattice (Sturm et al. 1982). No relation was therefore observed between manganese and silicate flux, since dinoflagellate cysts or Haptophyt members secreting calcium carbonate include no manganese in the lattice (Fig. 15d).

During summer there is no correlation between BSi and silica flux at all stations and low silica fluxes were observed except at station 6. The highest silica flux was observed as 94353 μmol/m²day at station 6. Two separate negatively inclined linear lines were obtained between carbonate and silica fluxes at all stations except station 6 (Fig. 16a and Table 3). In the graph between silicate flux and CO₃^-2 the fact that there is almost no change in silicate flux versus a 1% change in carbonate values at stations 4 to 7 and 9 and 10 may indicate that part of CaCO₃ is of biogenic origin (Fig. 16a). The slopes of the two lines shown in Fig. 16a have -7006.8 and -7044.4 μmol/m² day CO₃^-2 % values and are very close to each other. This may indicate that each line describes the dominant diatom size groups. The statistically strong presence of three separate negative sloped lines between the OrgC and silicate flux shows that OrgC is one of the mechanisms that control the silicate flux (Fig. 16b). Both Org C values and silica flux values at stations 4 to 7 and 9 and 10 are very close to each other. The factor controlling the silicate flux here is organic carbon rather than carbonate (Fig. 16b). The introduction of biogenic carbonate into the environment in summer means that there was no relationship between manganese and silica flux due to.

Figure 16

There is a negative relationship between the silicate flux and BSi in autumn at stations 2, 3, 5, 6, 10, 12, 15 and 16 (Fig. 17a, Table 3). Diatom skeletons may be said to dominate in autumn as in other seasons. Statistical analysis does not apply for stations 1 to 4 and 9 to 16, due to lack of data. Two lines with negative slopes may be distinguished between the silicate flux and CO₃^-2 (Fig. 17b). Two dominant size groups are observed. Three separate negative lines were obtained between OrgC and silicate flux (Fig. 17c). On the other hand, when looking at the relationship between OrgC and BSi, a strong positive linear relationship was observed at stations 4, 5, 6 and 16 (R² = 0.963) and stations 1, 2, 3, 8, 9 and 10 (R² = 0.997). Similarly, strong positive linear relationships (R² = 0.947 for stations 4,5,6 and 16; R² = 0.992 for stations 1, 2, 3, 8, 9 and 10) were obtained between OrgC and Chla for the same station groups. OrgC is therefore understood to explain living diatom content such as protoplasm, chla and organic template on the silica skeleton. There are two strong negative linear relationships between Mn and Silicate flux (Fig. 17d). Carbonate (CaCO₃ and MnCO₃) may be said to precipitate on diatom valves together with Mn, as in winter. Here the major variable is CaCO₃.

Figure 17

When the relations between silicate flux and BSi are examined, statistically linear negative relationships are consequently observed in winter and autumn, while one positive and two negative strong linear relationships are observed in spring. There was no relationship between these two parameters in the summer season. The growth rate of small phytoplankton is known to be higher than that of large phytoplankton (Stenuite et al. 2007). The smaller size individuals may be said to be more abundant as biomass, with the reduction in diatom size group (which increases the surface area of diatom per unit volume) at stations 2, 8, 9, 10, 13, 14 and 16, where a positive linear relationship is observed in spring. At stations where each negative linear relationship is observed between BSi and silicate flux in spring, there is a diatom size spectrum.

There is at least one and at most three negative linear relationships between the silicate flux and CO₃^-2 in all seasons. The stations on each line that form negative relationships may therefore be said to contain diatoms belonging to a single size group. The CaCO₃ secreted from both Scripsiella trochoidea cysts and haptophytes scales clearly emphasizes that there is no relationship between the Mn and silicate flux at stations 1, 3, 5, 7, 9, 10, 11, 12, 13 and 14, where there is no relationship between silicate flux and carbonate, especially in spring. CaCO₃ forms a separate phase independent of diatom and does not control silicate flux. Scripsiella trochoidea may be said to reproduce vegetatively in the water column in other seasons (Ishikawa&Taniguchi 1996, Nuzzo&Montresor 1999).

While two negative linear relationships were observed between silicate flux and OrgC in winter, three negative linear relationships were found in other seasons. The strong relationship of orgC with both BSi and chla also explains that almost the majority of benthic diatoms are alive.

No relationship was observed between the Mn/ Silicate flux in spring and summer. In summer Mn is therefore of terrestrial origin due to its very strong positive relationships with Ti (R²=0.992; R²=0.883). In winter and autumn at least two, at most three statistically significant negative linear relationships between the Mn and Silicate flux were observed. Negative linear relationships in both seasons, while orgC forms a significant benthic microalgae protoplasm, mean it is possible to include Mn in the protoplasm with benthic microalgae uptake. Mn therefore seems to play the role of indicator in this control mechanism, while OrgC controls the silicate flux, especially at stations with very close carbonate values.