Estuarine, Coastal and Shelf Science 157 (2015) 42e50
Contents lists available at ScienceDirect
Estuarine, Coastal and Shelf Science
journal homepage: www.elsevier.com/locate/ecss
Variations in the release of silicate and orthophosphate along a salinity gradient: Do sediment composition and physical forcing have roles?
* Bhanu Paudel a, , Paul A. Montagna b, 1, Leslie Adams b, 2 a Academy of Natural Science of Drexel University, 1900 Benjamin Franklin Pkwy, Philadelphia, PA 19103, USA b Harte Research Institute for Gulf of Mexico Studies, Texas A&M University-Corpus Christi, 6300 Ocean Drive, Unit 5869, Corpus Christi, TX 78412, USA article info abstract
Article history: It was hypothesized that sediment composition, i.e. organic matters and minerals, and physical forcing Received 8 September 2014 can influence retention and release of silicate (SiO4) and orthophosphate (o-PO4) along salinity gradients. Accepted 18 February 2015 An experiment was performed to measure nutrient release by using treatments with and without Available online 26 February 2015 sediment organic matter from the Guadalupe and Nueces Estuaries at five different salinities. The sample mixtures were shaken at intervals over the course of 48 h to simulate wind and river forcing. The release Keywords: of silicate from sediments increased with time from 2 min to 48 h in all five salinities. The added silicate orthophosphate concentration was adsorbed in most of the sediment containing organic matter and orthophosphate calcium-rich shells from both estuaries. From the sediments without organic matter, the release of quartz m calcite orthophosphate was as high as 52 mol/L. The sediment minerals quartz and calcite were abundant in organic matter both estuaries. The average quartz to corundum peak intensities ratio were 14.04 and 13.36 and the average calcite to corundum peak intensities ratio were 3.06 and 1.32 in the Guadalupe and Nueces Estuaries respectively. The average organic matter in the Guadalupe and Nueces estuaries were 10.67% and 13.39% respectively. The retention and release of orthophosphate from the sediments may have been caused by the bonding with organic matter and calcite in the sediments. These findings indicate that sediment composition was a significant contributor in the low dissolved orthophosphate concentration in the estuaries. The release of silicate from the sediments containing quartz, and organic matter, when shaken, indicate that the combined forcing of river and wind may have been maintaining the estuaries silicate concentrations. © 2015 Elsevier Ltd. All rights reserved.
1. Introduction According to previous studies, transported dissolved silicate decreased in concentration as it moved from a river to an estuary The fluctuation of silicate and phosphate concentrations along salinity gradient (Morris et al., 1981; Conley and Malone, 1992; Eyre estuarine salinity gradients might be due to the interaction be- and Balls, 1999). Different explanations have been suggested for the tween sediment composition and overlying water (Morris et al., decrease in silicate along river-mouth to estuary-ocean salinity 1981; D'Elia et al., 1983; Anderson, 1986). The sediment composi- gradients. Those include coagulation of a colloidal form of silica tion includes sediment organic matter and minerals that interact in (Krauskopf, 1956), retention of silicate on clay minerals as it moves the presence of salt water electrolytes. This, in turn, may affect from freshwater to the salt water interface (Williams and Crerar, retention-release, precipitation-dissolution and flocculation- 1985), and silicate uptake by diatoms (Conley and Malone, 1992). deflocculation of dissolved silicate and phosphate. A sudden increase in silicate concentration in oligohaline (salinity of 0.5e5, Venice Classification system; Anonymous, 1958) regions from riverine freshwater concentrations may be a result of the
* Corresponding author. dissolution of biogenic silica present in river water (Anderson, E-mail addresses: [email protected] (B. Paudel), paul.montagna@ 1986). tamucc.edu (P.A. Montagna), [email protected] (L. Adams). Previous studies have revealed two types of phosphate behavior 1 Tel.: þ1 361 825 2040. in estuarine salinity gradients. The first one is the decrease in 2 Tel.: þ1 361 825 2033. http://dx.doi.org/10.1016/j.ecss.2015.02.011 0272-7714/© 2015 Elsevier Ltd. All rights reserved. B. Paudel et al. / Estuarine, Coastal and Shelf Science 157 (2015) 42e50 43 dissolved phosphate concentration from fresh to saline water gra- 2.2.1. Organic matter dients with most of the losses at the confluence of a river and an The organic matter content in the sediment sample was deter- estuary. There are several potential causes of this sudden decrease mined by H2O2 digestion (Schumacher, 2002). H2O2 is an oxidant; in concentration; a change in pH (Garcia-Luque al. 2006), precipi- its addition oxidizes organic matter and decreases the pH (up to tation of phosphate-colloid (Bale and Morris, 1981), and retention 3e4). The decrease in pH could dissolve calcite and calcium car- of phosphate onto suspended clay particles (Garcia-Luque al. 2006). bonate shells that can release calcite bonded phosphorus The second type of phosphate behavior was the increase in phos- (Staudinger et al., 1990). Care was taken to avoid excessive frothing phate concentration in the oligohaline regions because of the and dissolution of calcite mineral. Our main goal on using H2O2 was release of phosphorus from the iron and aluminum oxide organic to digest organic matter and observe nutrient release from complex as the river water mixes with the estuary water sediments. (Sundareshwor and Morris, 1999). Flocculation of organic matter A10e15 g sediment sample was weighed and dried overnight at inhibits release from organic complexes as sediment is transported 60 C in an aluminum pan. After drying, the samples were to mesohaline (salinity of 5e18, Venice Classification system; reweighed and each placed in a beaker containing 50 ml of 30% Anonymous, 1958) and polyhaline (salinity of 18e30, Venice Clas- H2O2. The sediments from both estuaries contained clam and oyster sification system; Anonymous, 1958) regions, and result in the shells, which might get corroded by H2O2 solution. Effort was made decrease in phosphate concentration in the oceanic end of an es- to preserve sediment characteristics by not keeping the sediment tuary (Howarth et al., 1995; Nielson et al., 2001). for too long in the solution. After the digestion of organic matter the Recent research has found a salinity of 35 can dissolve silica gel supernatant was decanted and an aliquot of deionized (DI) water and release silicate in the water (Tanaka and Takahashi, 2013). was added. The beaker was stirred and allowed to settle until the Additionally, Spagnoli and Bergamini (1997) identified the release supernatant was clear. This process was repeated two more times. of phosphate during resuspension. The interaction between sedi- The sample was then filtered by pouring the beaker contents ment composition and water in the estuaries could be important in through a 0.7 mm filter paper suspended in a glass funnel. The the retention and release of silicate and orthophosphate from the sediment collected on the filter paper was oven dried at 60 C sediment. A weekly nutrient study in Nueces and Corpus Christi Bay overnight. The samples were again weighed. Sediment organic identified very low orthophosphate (o-PO4) concentrations matter was calculated by the difference in weight before and after (mean ¼ 0.12 mmol/L), and high silicate (SiO4) concentration organic matter digestion. From here on samples digested with H2O2 (mean ¼ 127.5 mmol/L) in all samplings (Turner, 2014). Phosphate refers to without organic matter treatment. can absorb onto clay particles resulting in reduced orthophosphate concentration (Garcia-Luque al. 2006), the low orthophosphate 2.2.2. Mineral content concentration might be because of the absorption onto clay parti- Sediment mineral has a role in fluctuating inorganic nutrient cles. Furthermore, in the Guadalupe and Nueces estuaries salinity concentration by retention and release into the mineral surface. and resuspension induced by wind or river flow may be the major The sediment without organic matter was ground using a mortar factors that cause variation in sediment nutrient release. Therefore and pestle, and 2 g of sample was mixed with 0.5 g of corundum (a- it is important to identify the effects of estuarine sediment Al2O3) in 4:1 ratio (Naehr et al., 2000). A few drops of ethyl alcohol composition at various salinities and the retention and release of were also added to the sample-corundum mixture to help in the SiO4 and o-PO4 concentrations due to resuspension. Thus, the homogenization. The prepared sample was then placed in a sample present study is to characterize silicate and orthophosphate release holder of a Rigakut Ultima III X-ray diffractometer and sediment from sediment in a laboratory setting by varying sediment organic mineral content was identified. In the x-ray diffractometer, scans matter, salinity and agitation rate. Agitation was used to mimic were run from 2 to 80 2q with scanning speed of 0.01 2q/s. The river flow and wind forcing. crystal structure of each mineral is composed of sets of planes responsible for ‘d’ values (distance between array of crystal in 2. Materials and methods mineral), and which is responsible for the variations in the peak. The peak, obtained as intensity count per second of the mineral 2.1. Sediment sample collection sites and bottom water quality type was compared with the corundum peak, and the mineral content was identified based on the mineral to corundum ratio. The top 5 cm of sediment were collected using cores from sta- MDI Jade e 7 software (Bishop et al., 2011) was used to identify tion ‘A’ of Guadalupe and Nueces Estuaries during July 2012, mineral types and peak intensities. The software can automatically October 2012, and January 2013 quarterly samplings (Fig. 1). Data identify the mineral content in the sample based on the location of collected from station A of both estuaries was used because it was the peak. However, the user has to differentiate the mineral type closer to the river mouth, i.e. near to the head of the estuaries, and based on the ‘d’ value and one main peak of the mineral. was therefore assumed to have higher organic loads. In the field, sediment samples were kept on ice and stored frozen upon transfer 2.2.3. Preparation of nutrients spike solution to the lab in order to slow organic matter degradation. Organic Primary standards concentration of 10,000 mmol/L were pre- contents, mineral contents and nutrient release experiments were pared using 1.36 g of potassium phosphate monobasic (KH2PO4) performed by using these sediment samples. and undried 2.12 g of sodium metasilicate pentahydrate (Na2SiO3). Temperature, pH, salinity, dissolved oxygen (DO), were collected Each chemical was weighed out, placed in 1000 ml volumetric using YSI sonde in the bottom water (20 cm above the sediment flask, dissolved and brought up to 1000 ml with freshly drawn surface) of the two estuaries. Water samples were also collected for Milli-Q water. From these refrigerated primary standards, a 100 ml nutrient analysis. mixed stock standard was prepared daily containing 500 mmol/L, o- PO4 and 5000 mmol/L SiO4. That solution was then used to ‘spike’ 2.2. Sediment analysis and laboratory experiments the solute with additional nutrients in the following experiments.
The laboratory experiment was conducted at room temperature. 2.2.4. Experimental design for laboratory study Before the sediment was used for laboratory experiment, organic All the laboratory experiments to identify sediment nutrient matter, and mineral analyses it was carefully homogenized. release were performed within two weeks of sample collection. The 44 B. Paudel et al. / Estuarine, Coastal and Shelf Science 157 (2015) 42e50
Fig. 1. Map of the study area.
laboratory experiment was conducted using sediments with and five salinity solutions of 0 (DI water), 7.5, 15, 25, and 35 were pre- without organic matter treatments at five salinity solutions over a pared using low nutrient sea water (LNSW) and DI water. During period of 48 h. A total of 6 laboratory experiments (3 with each experiment the sediment to salinity solution ratio was 1:30 by organic þ 3 without organic) using the Nueces Estuary and 6 lab- mixing 4 g of sediment in 120 ml of solution in a 250 ml poly- oratory experiments (3 with organic þ 3 without organic) using the carbonate Erlenmeyer flask. The experiment was conducted with Guadalupe Estuary sediment samples were conducted (Fig. 2). The two replicates for each salinity solutions from both estuaries for a
Fig. 2. Experimental Design for the laboratory experiment, and only Guadalupe Estuary (GE) is showing. Design repeated for Nueces Estuary (NC). T1 ¼ 1 min, T2 ¼ 15 min, T3 ¼ 50 min, T4 ¼ 360 min, T5 ¼ 48 h, ¼ replicates. B. Paudel et al. / Estuarine, Coastal and Shelf Science 157 (2015) 42e50 45 total of 20 flasks. Each flask was supplemented with the spike concentrations, whereas salinity was significantly higher in the nutrient solution to provide an additional concentration 5 mmol/L Nueces Estuary compared to the Guadalupe Estuary. Silicate con- of o-PO4 and 50 mmol/L of SiO4. centration in both estuaries stations near (Stations A and B) to the The spiked salinity solutions were carefully mixed prior to river was greater than 100 mmol/L (Table 1). adding the weighed sediment to each flask. After sediment addi- e fl tion, 3 5 ml sample was collected from each ask, using a large 3.2. Silicate and orthophosphate concentrations in the laboratory bore plastic syringe, to be used as the baseline measurement (T1). experiment Additional water samples for nutrient analysis were then collected at 10 min (T2), 50 min (T3), 360 (T4) mins and 48 h (T5) after the In the lab, SiO4 concentration was the highest in the treatments addition of sediment to the flasks. The flasks were shaken vigor- without sediment organic matter. The difference in SiO4 concen- ously using an oscillating shaker between each sampling time. The tration between Time-T5 and Time-T1 was higher in the DI treat- samples were shaken for approximately 50 s at the following in- ments without sediment organic matter compared to salinity tervals: every 2 min between T1 to T2, every 5 min between T2 to T3, treatments without sediment organic matter (Fig. 3). The average every 15 min between T3 to T4 and every hour between T4 to T5. The highest SiO4 concentrations of 120 mmol/L from the Nueces Estuary samples were allowed to settle for 10 h overnight and the agitation and 140 mmol/L from the Guadalupe Estuary were observed in the protocol was resumed the following day. DI treatments without sediment organic matter at Time-T4. The A control was prepared using the nutrient spike solution at the concentration of SiO4 was significantly different with respect to same salinity and nutrient concentration as in the experimental time, salinity, and organic content as variables, whereas it was not fl asks but without the sediment addition. The control was designed significantly different between the two estuaries (Table 2). Addi- to identify and measure the degradation of nutrients in the solu- tionally the SiO concentrations were affected more with time and fl 4 tions over the time of experiment without sediment in uence. less with organic matter and salinity. The dissolved orthophosphate The depicted SiO4 and o-PO4 concentrations were the concen- (o-PO4) concentrations in all saline treatments with sediment trations after subtracting control's concentrations at that given organic matter were below detection levels regardless of time, time. Hence the negative and positive values indicate retained and whereas this was not the case with silicate. However, the o-PO4 in released concentration to and from the sediments. some DI treatments with sediment organic matter did show detectable levels of o-PO4. Four out of ten average o-PO4 concen- 2.2.5. Inorganic nutrients analyses trations in the DI treatments with sediment organic matter were fi Three to ve milliliters of nutrient sample was taken out from greater than 0 mmol/L, i.e. ranges from 0.6 to 2.7, of which only one fl each polycarbonate Erlenmeyer ask using a syringe and each was was from the Nueces Estuary treatment (Fig. 3). The average o-PO fi m fi 4 ltered through polycarbonate 0.45 m lter paper. Each sample concentration was the highest in the DI treatments without organic was kept frozen in a labeled vial until analysis. An O.I Segmented matter. The average concentration identified in the DI treatments Flow Autoanalyzer was used for the nutrient analysis. Manufacturer without sediment organic matter from the Guadalupe Estuary and recommended applicable range of method detection limit (MDL) the Nueces Estuary were 30.2 and 24.1 mmol/L respectively (Figs. 4 are 0.02e10 mmol/L and 0.35e35 mmol/L for orthophosphate (o- and 5). In the saline treatments without organic matter, the o-PO4 PO4) and silicate (SiO4). Higher concentrations were determined by m concentration was higher in the Guadalupe Estuary treatments sample dilution. Five and 0.10 mol/L were the lower detection compared to the Nueces Estuary treatments (Figs. 4 and 5). The o- limit of silicate and orthophosphate concentrations used for pre- PO4 concentrations in the treatments without organic matter in all paring standard curve respectively. salinities increased over time (Fig. 3). The o-PO4 concentrations were significantly different with respect to organic matter, time, 2.3. Statistical method salinity, and estuary (Table 2).
Analysis of variance (ANOVA) was performed to identify the effects of salinity (SAL), time, organic matter content (OM) and 3.3. Sediment mineral and organic composition estuaries (EST) on SiO4 and o-PO4 concentration. Before the anal- In both estuaries quartz was the dominant mineral, based on the ysis, SiO4 and o-PO4 data were log transformed. ANOVA was per- formed using PROC GLM procedure in SAS version 9.3. mineral to corundum ratio. The average quartz to corundum ratio ® All graphs were created using SAS statistical analysis. In SAS was 14.04 and 13.86 in the Guadalupe and Nueces Estuaries different procedures of statistical and graphical analysis are per- respectively (Fig. 6). Besides quartz (SiO2) and calcite (CaCO3), e e formed using PROC syntax command. The PROC SGSCATTER and feldspar (KAlSi3O8 Na, AlSi3O8 Ca, Al2Si3O8) was also found in the PROC COMPARE procedures were performed to compare different variables. The nutrients data were checked for normality using Table 1 HISTOGRAM statement with density type ¼ kernel. Then the data Average salinity, phosphate and silicate concentrations during three field samplings were log transformed in order to obtain normal distribution of in the Guadalupe (GE) and Nueces (NC) Estuaries. Abbreviations: DL ¼ detection ¼ ¼ residuals. PROC GCHART was performed to get bar diagrams and limit, o-PO4 orthophosphate, and SiO4 silicate.
PROC GREPLAY was used to obtain a panel of different bar diagrams. Estuary Station Latitude Longitude Salinity o-PO4 SiO4 (ppt) (umol/L) (umol/L) 3. Results mean mean mean GE A 28.36878 96.71878 11.06 2.75 183.05 3.1. Silicate and orthophosphate concentrations in the Nueces and GE B 28.37248 96.72080 15.80 1.73 160.14 Guadalupe estuaries GE C 28.31027 96.68002 22.33 1.04 117.02 GE D 28.31397 96.67832 19.75 1.22 119.25 NC A 28.39352 96.77240 34.24 2.16 149.46 In the Nueces Estuary, salinity was above 30 during all three NC B 28.34777 96.74573 35.61 1.35 113.66 samplings, whereas in the Guadalupe Estuary the lowest salinity of NC C 28.24618 96.76488 36.69 0.35 45.79 9 was identified in July 2012. The two estuaries did not vary NC D 28.30210 96.68435 36.75