Study Sites

The Maryland Coastal Bays, part of the mid-Atlantic coastal lagoon system in the United States, are characterized by shallow waters (<3 m) and relatively long water residence times (Wazniak, Wells, and Hall, 2004) due to limited exchange with the open Atlantic Ocean through small channels (Figure 1). Circulation within the bays is controlled by wind and tides. The river inputs are fairly low due to the area's flat landscape and sandy soils, and groundwater is a major pathway for the introduction of freshwater and nutrients to the bays (Wazniak, Wells, and Hall, 2004). Water and surface sediment samples were collected at 13 preselected sites with variable characteristics (Table 1). Site 8, which is near the Ocean City inlet, showed higher mixing of bay water and seawater than other sites, as indicated by high bottom stress and low clay/silt content in sediments (type I). Sites 4, 6, 10, 12, and 13 are located near stream/river mouths, where nutrient input is usually stronger, and sediments at these sites are muddier than at other sites (type III). Other sites (type II) are generally considered open bay sites with characteristics between type I and type III sites.

Figure 1. 

Map of sampling sites in the MCB. Sites 1–7 are located in the southern bays, while sites 8–13 are located in the northern bays. Mixing is greatest at sites 8 and 9, due to their proximity to the Ocean City inlet.

Table 1. 

Salinity, mud content, sediment texture, and nutrients of the sampling sites.

a 

Average of salinities measured from June 2011 to July 2013.

b 

Sediment size distribution of dried sediment was measured with a set of six sieves; the mud (silt + clay) content is the fraction which passes the 0.039 mm sieve.

c 

Nutrient condition refers to nutrient input from stream.

d 

Type 1 = Low clay/silt content and high bottom stress, Type 2 = Medium clay/silt content, Type 3 = High clay/silt content.

Sample Collection/Processing

Approximately 1 L of water was collected from 1.0 m below the surface at each site using a peristaltic pump and was stored on ice in a cooler. In the laboratory, the water samples were filtered through a Whatman GF/F 47 mm glass fiber membrane (GE Healthcare Bio-Sciences, Pittsburgh, Pennsylvania, U.S.A.) with a Fisherbrand vacuum microfiltration system (Fisher Scientific, Pittsburgh, Pennsylvania, U.S.A.). The filtrates were used for nutrient analysis. The filters were then placed in 5–10 mL of 100% acetone in a 15 mL polypropylene centrifuge tube, sonicated, and extracted in darkness at 4°C overnight. The pigment extracts were then centrifuged and the supernatants collected and dried by exposing to a nitrogen stream. To redissolve the residual, 200 μL methanol/water (9:1 v/v) was added and 20 μL solution was injected to the HPLC.

Surface sediment samples were collected with a bottom grab sampler, stored in glass bottles, and stored on ice in a box. In the laboratory, about 100 g sediment were freeze dried, and approximately 1 g freeze-dried sediment was weighed into a 10 mL test tube. Pigments were extracted with 5 mL acetone, mixed thoroughly with a vortex mixer, sonicated, and allowed to stand in the dark at 4°C overnight. Test tubes were placed in a centrifuge at 2500 rpm for 5 minutes, and 3 mL extract was filtered with a syringe filter (0.2 μm). The extract was then dried with a nitrogen stream. To the residual, 200 μL methanol/water (9:1 v/v) was added and 20 μL solution was injected into HPLC.

Nutrient Analysis

Total dissolved nitrogen (TDN) in the filtrate was analyzed by high-temperature combustion method using a TNM-1 total nitrogen analyzer module (Shimadzu Scientific Instruments, Columbia, Maryland, U.S.A.) coupled to a TOC-V_CSH total organic carbon analyzer (Schimadzu Scientific Instruments) (Álvarez-Salgado and Miller, 1998). Standard calibration for TDN was achieved using potassium nitrate. A DR4000 spectrophotometer (Hach Company, Loveland, Colorado, U.S.A.) was used to measure dissolved reactive phosphorus (DRP; Hach method 8048) and dissolved silica (DSi, Hach method 8185). TDN was chosen to represent the availability of nitrogen nutrients since concentration of nitrates is usually very low year-round and represents only a minor contribution to the total dissolved nitrogen pool (Glibert et al., 2014).

HPLC Analysis of Pigments

The separation of pigments was conducted on an 1100 Series model G1322A HPLC with a binary pump (Agilent Technologies, Santa Clara, California, U.S.A.) following the method described in Zapata, Rodríguez, and Garrido (2000). The two eluents used were methanol/acetonitrile/0.25 M aqueous pyridine (50:25:25 v/v/v) and acetonitrile/acetone (80:20 v/v). All organic solvents were HPLC grade and prepared as described in Zapata, Rodríguez, and Garrido (2000). Analytical separations were performed using a Symmetry C₈ column (150 × 4.6 mm, 3.5 μm particle size, 100 Å pore size; Waters Corporation, Milford, Massachusetts, U.S.A.). The gradient was a version modified from Zapata, Rodríguez, and Garrido (2000) that achieves partial separation between zeaxanthin (Zeax) and lutein (Lut).

Briefly, after injection of 20 μL of sample, a gradient program began with 100% A, which ramped to 40% B in 18 minutes, then ramped to 100% B in 4 minutes, remained isocratic for 13 minutes, and then returned to 100% A in 3 minutes. The flow rate remained constant at 1 mL/min. An Agilent model G1316A diode array detector (DAD, Agilent Technologies) and model G1321A fluorescence detector (FLD, Agilent Technologies) were used to detect pigments. Chromatograms at 432 nm and 666 nm were recorded for quantitative analysis of carotenoids (Lut, Zeax, etc.) and pheopigments (Phide-a, Phtin-a, and pPhtin-a), respectively. Chromatograms from FLD (excitation = 440 nm, emission = 660 nm) were used for quantification of Chl-a and chlorophyll b (Chl-b). ultraviolet/visible spectra from 350 to 700 nm were also recorded by DAD for qualitative analysis. Pigments were identified based on the retention time and ultraviolet/visible spectra characteristics (Table 2). Chl-a, Chl-b, and β-carotene (β-Car) standards were purchased from Sigma-Aldrich (St. Louis, Missouri, U.S.A.), while other carotenoids were purchased from DHI (Hørsholm, Denmark). The Phtin-a standard was prepared by acidifying the Chl-a standard. The responding factor of Phtin-a was used for quantification of other pheopigments (Phide-a and pPhtin-a).

Table 2. 

Spectral and chromatographic information of chlorophylls, accessory pigments, and Chl-a decomposition products.

Phytoplankton Community Characteristics in the MCB

In March 2013, fucoxanthin (Fuco) was the dominant accessory pigment at all sites except site 10, where peridinin (Peri, 3.90 nMol/L) was higher than Fuco (Figure 2). The highest concentration of Fuco (7.49 nMol/L) was detected at site 6, where relatively high concentrations of TDN, DRP, and DSi were observed. The unusual nutrient condition at site 10, with the highest TDN and DSi among all sites (Table 1), may explain the high abundance of dinoflagellates, indicated by high concentration of Peri at this nearshore site. Excluding sites 6 and 10, Chl-a had a moderate negative relationship with both TDN (R = −0.64, p < 0.05) and DRP (R = −0.73, p < 0.05) and a positive relationship with DSi (R = 0.50, p < 0.05), indicating that the primary production was likely limited by DSi throughout the bay except at sites where nutrients were supplied by streams.

Figure 2. 

Photosynthetic pigments in the water column in March 2013 showing spatial variations of Chl-a (sum of normal and epimer of Chl-a) and major accessary pigments.

The average concentration of Fuco (2.32 ±0.27 nMol/L) was higher than the other diagnostic pigments by approximately one order of magnitude. In addition, the average Fuco concentration in March 2013 was highest among all sampling periods since 2012, when we started monitoring the photosynthetic pigments in the MCB; this is consistent with a moderate spring diatom bloom. Another diagnostic carotenoid detected in water samples at all sites was alloxanthin (Allox), a biomarker for cryptophytes; at most sites, Allox was the second-highest accessory pigment (Figure 2). Peri, Zeax, Lut, and Chl-b were only detected in the water column at a few sites (6, 10, 12, 13) where higher nutrients from riverine inputs were expected (Glibert et al., 2014).

Spatial Variation of Chl-a and Accessory Pigments in Surface Sediments

Strong spatial variation in Chl-a and accessory pigments was observed in surface sediments (Figure 3). The average concentration of fresh Chl-a (0.065 ±0.043 nMol/g) in sediments is generally low throughout the coastal bays. The highest concentration of Chl-a was observed at site 10 (0.12 nMol/g) followed by Sites 2, 7, 8, and 9 (Figure 4). Fuco, Zeax, Lut, Allox, Peri, and Chl-b were detected in surface sediment at all sites, and prasinoxanthin (Pras) was detected at most sites. Neoxanthin, violaxanthin, 19′-butanoyloxy-fucoxanthin, and 19′-hexanoyloxy-fucoxanthin were only detected at a few sites, and their concentrations were also much lower than the other pigments. Average concentration of Zeax (0.13 ±0.15 nMol/g) was the highest among accessory pigments, followed by Fuco (0.080 ±0.062 nMol/g), Lut (0.045 ±0.048 nMol/g), Allox (0.044 ±0.054 nMol/g), Peri (0.027 ±0.018 nMol/g), β-Car (0.023 ±0.027 nMol/g), and Pras (0.017 ±0.022 nMol/g). This suggested that the main sources of phytoplankton-derived organic matter included cyanobacteria, diatoms, cryptophytes, dinoflagellates, and prasinophytes. However, caution must be used when interpreting photosynthetic pigments as biomarkers of sources of phytoplanktonic material, since stability varies substantially among pigments. For example, Fuco is much more labile than Zeax (Abele-Oeschger, 1991; Burford, Long, and Rothlisberg, 1994; Louda et al., 2004; Repeta and Gagosian, 1982). Therefore, the abundance of accessory pigment in sediments cannot be directly used to evaluate the relative magnitude of contribution of phytoplankton groups represented by these pigments in sediments.

Figure 3. 

Spatial distribution of major accessory pigments in surface sediments in the MCB. Pigment concentrations were higher in the northern bays than in the southern bays due to higher nutrient input from the surrounding creeks.

Figure 4. 

Spatial distribution of Chl-a and its decomposition products in surface sediments of the MCB, indicative of the source of organic matter either through senescence or zooplankton grazing. Chl-a, Chlide-a, Phide-,a and Phtin-a in this figure represent the sum of normal chloropigments and their epimers, respectively.

The spatial distributions were very similar for most accessory pigments, with the exception of Fuco and Peri. For other accessory pigments, relatively high concentrations were observed in the northern bays (sites 10, 12, and 13). The similar spatial distribution for most accessory pigments except Fuco and Peri was also evidenced by the strong linear relationships between Allox, Zeax, Lut, and β-Car (Table 3), indicating that cyanobacteria, cryptophytes, and some chlorophytes shared similar niches where nutrient inputs were usually high. Spatial variation of Peri was not as remarkable as the other accessory pigments, indicating that dinoflagellates were more evenly distributed throughout the bays.

Table 3. 

Correlations (R) between pigments (n = 13) from Pearson correlation analysis. Significant relationships (p < 0.05) are given in bold. Abbreviations of pigments are listed in Table 2.

Spatial variation of Fuco in sediments seems to follow an opposite trend when compared with other accessory pigments. The highest concentration of Fuco was observed at site 8 (0.25 nMol/g), where other accessory pigments were very low or not detected. Relatively high concentrations of Fuco were also detected at sites 7 and 9, the two sites closest to site 8. This distribution is different from the water column Fuco distribution, which actually peaked at site 6 and remained relatively high throughout the bays (Figure 3). We speculate that the high Fuco concentration in surface sediment at site 8 might be caused by sinking of diatoms at locations with relatively stronger mixing of bay water and seawater. In addition, the bottom stress at site 8 was also the highest, as indicated by very low clay/silt contents (Table 1). High bottom stress may also prevent deposition and burial of organic matter derived from other phytoplankton groups due to lack of silicate shells such as diatoms.

Spatial Variation of Chl-a Decomposition Products in Surface Sediments

The major Chl-a decomposition products in surface sediments are Phtin-a (mean 0.083 ±0.114 nMol/g) and Phide-a (mean 0.083 ±0.097 nMol/g) at all sites (Figure 4). There were strong spatial variations in Chl-a decomposition products. Phtin-a showed a very different pattern than Chlide-a, Phide-a, and pPhtin-a. The highest concentration of Phtin-a was observed at site 8 (0.44 nMol/g), where the highest concentration of Fuco was observed. The highest concentration of Phide-a (0.17 nMol/g) was observed at site 13. Chlide-a and pPhtin-a generally followed the pattern of Phide-a, and this is further supported by regression analysis (Table 3). In this table, the concentrations of Chlide-a, Phide-a, and Phtin-a include their respective epimers.

Statistical Analysis

Pearson correlation analysis was conducted for sedimentary photosynthetic pigments and Chl-a decomposition products with SPSS Statistics 20 (IBM Corporation, Armonk, New York, U.S.A.) to provide information on sources of phytoplankton-derived organic matter in benthic sediments and Chl-a decomposition pathways in the MCB (Table 3). We observed strong correlations between Fuco and Phtin-a (R = 0.81, p < 0.05), some pheopigments (Chlide-a, Phide-a and pPhtin-a), and also some carotenoids (Allox, Zeax, Lut, and β-Car).

Sources of Photosynthetic Pigments in Surface Sediments

Photosynthetic pigments in MCB waters have been monitored monthly since February 2012. No strong spring blooms have been observed. In both 2012 and 2013, average Chl-a concentrations increased moderately from February (<1 nMol/L) to March (1–2 nMol/L) and quickly decreased to very low levels (<0.2 nMol/L; O. Oseji, unpublished data) in April. The highest average Chl-a concentration, which may reach >5 nMol/L at certain sites in some years, usually occurs in June or July and is caused by a “brown tide” (Trice et al., 2004). Diatoms are dominant species year-round, and their dominance is more remarkable in winter and early spring. Chl-a concentration remains relatively high in summer and decreases gradually in fall and winter.

The main pathway for pelagic diatoms to reach surface sediments is through zooplankton grazing and direct deposition of dead/senescent diatom cells (Itoh et al., 2007; Roy, 1989). There is strong evidence that diatoms were deposited directly at site 8 near the Ocean City inlet, where stronger mixing of coastal bay water with seawater occurred. First, Phtin-a and Fuco dominated Chl-a decomposition products and accessory pigments at site 8, indicating that the most likely source of Phtin-a at this site was the deposition of dead/senescent diatoms. This trend is further evidenced by the strong linear relationship between Fuco and Phtin-a (r² = 0.81, p < 0.01; Table 3). The concentrations of other accessory pigments in surface sediment were very low at site 8, probably due to low grazing activity or high bottom stress, which prevents the deposition and burial of organic matter derived from phytoplankton. Furthermore, correlation analyses also suggested that the main source of Chlide-a, Phide-a, and pPhtin-a was similar—most likely grazing, as indicated by strong linear relationships between these biomarkers (Table 3). However, no accessory pigments showed any clear linear relationship with Chlide-a, Phide-a, and pPhtin-a, likely due to spatial variation in the composition of phytoplankton.

Another source of photosynthetic pigments and decomposition products is the phytobenthos. Wazniak (2004) measured Chl-a of viable algal cells in surface sediments and found that its concentration varied remarkably (from 8.5 to 259 mg/m²). Wazniak (2004) suggested that the phytobenthos may represent an important primary producer within the system. The pigments in our samples included both pelagic and benthic sources, and the benthic contribution to photosynthetic pigments may vary significantly. For example, at site 8, a remarkably high concentration of Fuco coincided with the highest concentration of Phtin-a among all sites. Chl-a was not higher at site 8 than the adjacent sites (7 and 9), indicating the pigments at this site were mainly derived from deposition of senescence/death of pelagic diatoms. In contrast, at the four sites located near the river mouths with high concentrations of nutrients and finer sediments (sites 6, 10, 12, and 13), Zeax concentrations were overwhelmingly higher than other carotenoids, while Zeax concentration in the water column was very low, suggesting that benthic cyanobacteria may contribute remarkably to the overall pigment budget (Chl-a and Zeax). However, the relatively low concentration of fresh Chl-a and high concentrations of Chl-a decomposition products at these sites indicated that viable phytobenthos was not likely the major source of photosynthetic pigments.

Selective Grazing on Phytoplankton in the Water Column

Selective grazing of zooplankton on phytoplankton has been observed in many previous studies (Burkill et al., 1987; Granéli and Turner, 2002; Huang et al., 2008; Liu et al., 2010; Stoecker, Guillard, and Kavee, 1981). A recent study on the grazing impact of microzooplankton on phytoplankton in Xiamen Bay, China, revealed that microzooplankton preferentially grazed on nano- and picophytoplankton and avoided harvesting large phytoplankton species such as diatoms in the cold season (Huang et al., 2008). The peak grazing pressures for dinoflagellates and chlorophytes occurred in early spring, and for cryptophytes and cyanobacteria in winter (December). In contrast, peak grazing pressures for diatoms occurred in summer, presumably due to the larger zooplankton species such as copepods. If the seasonal variations of zooplankton grazing in the MCB follow the same pattern, the relatively low standing biomass of nano- and picophytoplankton (prasinophyte, chlorophyte, cyanobacteria, and cryptophyte) in winter can be interpreted as preferential grazing by microzooplankton, and the remarkable dominance of standing biomass of diatoms in the cold season can be attributed to the lack of grazing pressure.

High concentrations of Zeax, Lut, and Allox coupled with high concentrations of Chlide-a, Phide-a, and pPhtin-a in sediments at sites 6, 10, 12, and 13 with high nutrients in the water column are consistent with selective grazing at these sites. The strong linear relationships among Zeax, Allox, and Lut, and the linear relationships among Chlide-a, Phide-a, and pPhtin-a further support this view (Table 3), indicating that the main source of phytoplankton-derived organic matter at these sites were from fecal materials from microzooplankton grazing. Phytobenthos may contribute to the total chlorophyll and carotenoids, and benthic grazers and suspension feeders may contribute to the production of Chl-a decomposition products. However, since the diets of benthic grazers and suspension feeders are not selective or are less selective compared with zooplankton/protozoa, the inference about selective grazing based on the pigment compositions in the water column and sediments holds true. In addition, the low temperature during the sampling period may subdue the feeding activities of benthic grazers and suspension feeders.

Chl-a Decomposition Pathway

The results from this study also shed light on the Chl-a decomposition pathway in the MCB. Two main pathways can be proposed based on the current results. The presence of maximum concentrations of both Fuco and Phtin-a at site 8 and the strong linear relationship between Fuco and Phtin-a support the production of Phtin-a through the senescence of diatoms at that site. The low concentrations of Chlide-a and pPhtin-a, and the lack of linear relationship between dephytylized derivatives (Chlide-a and Phide-a) and both Phtin-a and Fuco, indicate that the autolytic decomposition of Chl-a in the dead/senescent diatoms catalyzed by the enzyme chlorophyllase (Jeffrey, 1974; Jeffrey and Hallegraeff, 1987) was, at least, not significant at this site. However, this does not rule out the production of Phide-a in dead/senescent cells through enzymatic dephytylization at this site. As a matter of fact, a moderately high concentration of Phide-a (0.09 nMol/L), lower only than Phtin-a (0.44 nMol/L), was observed at site 8 (Figure 4), indicating that production of Phide-a through enzymatic autolytic decomposition of Phtin-a or demetalization of Chlide-a is also an important pathway in cell senescence.

Furthermore, strong evidences support another decomposition pathway of Chl-a leading to production of Chlide-a, Phide-a, and pPhtin-a through microphytoplankton grazing on nano- and picophytoplankton and feeding of benthic grazers and suspension feeders. (1) High concentrations of Chlide-a, Phide-a, and pPhtin-a occurred at sites with high abundance of nano- and picophytoplankton due to high nutrient concentrations. As mentioned earlier, high concentrations of Allox, Zeax, Lut, and β-Car in surface sediments at the same sites were most likely caused by selective grazing of nano- and picophytoplankton. (2) A strong linear relationship between Chlide-a and Phide-a and between Phide-a and pPhtin-a indicated that they were derived from the same processes, most likely microzooplankton grazing on nano- and picophytoplankton and benthic grazing and suspension filter feeding. (3) The lack of linear relationship between Phtin-a and Chlide-a/Phide-a/pPhtin-a indicated that Chlide-a/Phide-a /pPhtin-a were produced through a distinctly different pathway than Phtin-a. The presence of significant amounts of Phide-a is usually considered an indication of a high contribution of zooplankton fecal pellets in sinking particles (Lorenzen, 1967). The high abundance of Phide-a (comparable to Phtin-a) in this study is consistent with the selective grazing of nano- and picophytoplankton species mentioned earlier. Therefore, the dominant decomposition pathway of Chl-a at most sites (e.g., sites 6, 10, 12, and 13) was the formation of Chlide-a, Phide-a, and pPhtin-a through microzooplankton grazing on nano- and picophytoplankton species and feeding of benthic grazers and suspension feeders.