Microorganisms 2014, 2, 33-57; doi:10.3390/microorganisms2010033 OPEN ACCESS microorganisms ISSN 2076-2607 www.mdpi.com/journal/microorganisms Article Emergence of Algal Blooms: The Effects of Short-Term Variability in Water Quality on Phytoplankton Abundance, Diversity, and Community Composition in a Tidal Estuary Todd A. Egerton 1,*, Ryan E. Morse 2,†, Harold G. Marshall 1 and Margaret R. Mulholland 3,† 1 Department of Biological Sciences, Old Dominion University, Norfolk, VA 23529, USA; E-Mail: [email protected] 2 Graduate School of Oceanography, The University of Rhode Island, Narragansett, RI 02882, USA; E-Mail: [email protected] 3 Department of Ocean, Earth and Atmospheric Sciences, Old Dominion University, Norfolk, VA 23529, USA; E-Mail: [email protected] † These authors contributed equally to this work. * Author to whom correspondence should be addressed; E-Mail: [email protected]; Tel.: +1-757-683-3595; Fax: +1-757-683-5283. Received: 24 September 2013; in revised form: 28 October 2013 / Accepted: 28 November 2013 / Published: 8 January 2014 Abstract: Algal blooms are dynamic phenomena, often attributed to environmental parameters that vary on short timescales (e.g., hours to days). Phytoplankton monitoring programs are largely designed to examine long-term trends and interannual variability. In order to better understand and evaluate the relationships between water quality variables and the genesis of algal blooms, daily samples were collected over a 34 day period in the eutrophic Lafayette River, a tidal tributary within Chesapeake Bay’s estuarine complex, during spring 2006. During this period two distinct algal blooms occurred; the first was a cryptomonad bloom and this was followed by a bloom of the mixotrophic dinoflagellate, Gymnodinium instriatum. Chlorophyll a, nutrient concentrations, and physical and chemical parameters were measured daily along with phytoplankton abundance and community composition. While 65 phytoplankton species from eight major taxonomic groups were identified in samples and total micro- and nano-phytoplankton cell densities ranged from 5.8 × 106 to 7.8 × 107 cells L−1, during blooms, cryptomonads and G. instriatum were 91.6% and 99.0%, respectively, of the total phytoplankton biomass during blooms. The cryptomonad bloom developed following a period of rainfall and concomitant Microorganisms 2014, 2 34 increases in inorganic nitrogen concentrations. Nitrate, nitrite and ammonium concentrations 0 to 5 days prior were positively lag-correlated with cryptomonad abundance. In contrast, the G. insriatum bloom developed during periods of low dissolved nitrogen concentrations and their abundance was negatively correlated with inorganic nitrogen concentrations. Keywords: algal blooms; cryptomonads; dinoflagellates; Chesapeake Bay; Gymnodinim instriatum 1. Introduction In estuarine systems, phytoplankton communities are highly variable, affected by numerous environmental and ecological factors including water temperature, salinity, light intensity, nutrient availability, inter- and intra-specific competition among the algae, and predation [1–4]. Many of these environmental factors vary on short time scales in estuaries due to tidal and diel fluctuations in physical/chemical parameters and episodic nutrient inputs from precipitation events [5–8]. Because of their short generation times, phytoplankton populations can respond rapidly to environmental and ecological forcing [9–11] Consequently, changes in algal community composition and diversity can occur over relatively short time periods in response to environmental variability [12–14] and this can affect ecological function [15,16]. Algal blooms usually involve rapid changes in phytoplankton community composition, where in phytoplankton communities become dominated by a single (or a few) species over the course of days, resulting in nearly monospecific assemblages that can then persist for weeks to months [17,18]. Such monospecific algal blooms appear to be increasing in frequency and magnitude, and nutrient over-enrichment has been implicated as a causal factor [19,20]. However, linking blooms to a proximate trigger has proven difficult because blooms are generally sampled only after they become visible when cell densities are already high enough to discolor water. As a result, the environmental conditions during bloom initiation are usually unknown. In addition, bloom organisms can be transported from sites of initiation to the areas where algal biomass accumulates and blooms are observed [21]. Because environmental conditions and phytoplankton communities can change rapidly in estuaries due to physical and meteorological forcing, monthly monitoring is not sufficient to document bloom initiation in response to short-term environmental variability. Recent studies aimed at identifying causal factors promoting bloom formation sampled more frequently and demonstrated that rapid changes in algal biomass as blooms developed and then dissipated associated with short-term variability in water quality [7,8,22,23]. Continuous monitoring and high resolution mapping has further highlighted the rapid fluctuations and spatial variability in nutrient and dissolved oxygen concentrations, chlorophyll biomass, temperature and salinity over diurnal timescales [7,21,24]. Daily sampling studies have also been used to document the relationships between water quality parameters and algal community composition that occur over short periods of time [7,12,25]. These studies suggested that meteorological forcing was important in driving changes in algal community structure. Microorganisms 2014, 2 35 The objectives of this study were to identify short-term changes in phytoplankton species composition and diversity and relate these with water quality parameters and meteorological forcing, the development of mono-specific blooms, and algal species diversity in a tidal estuarine system during spring, when rainfall is usually frequent and the algal community diverse. 2. Experimental Section 2.1. Study Site The Lafayette River, located in Norfolk, VA, USA, is a tributary of the Elizabeth River that flows into the lower James River near its confluence with the Chesapeake Bay. It is a tidal river, approximately 8 km in length, with a mean depth of 1.3 m, and a maximum channel depth of 7.6 m [26]. The river is surrounded by residential and commercial development, within an urban watershed of 43.28 km2, and a shoreline that includes bulk headed regions, marinas, private docks and wetland marshes of Spartina alterniflora [26–29]. Freshwater input is delivered through precipitation and shoreline drainage that includes 13 storm sewers and overflow drains [27,30]. Seasonal dinoflagellate blooms common in this river include: Prorocentrum minimum (early spring) and Akashiwo sanguinea and Cochlodinium polykrikoides (summer and fall) [22,31–33]. The river has been identified as an initiation site for regional dinoflagellate blooms dominated by C. polykrikoides during summer and fall [20]. 2.2. Methods and Materials Surface water samples were collected once a day during the incoming tide, approximately 2 h after low tide, from a stationary floating dock on the Lafayette River between 20 April and 25 May 2006, as described by Morse et al. [7]. The mean water depth at the station was 0.9 m. Water temperature, salinity, pH, and dissolved oxygen were measured daily just before collecting water samples using a Hydrolab Data Sonde 4a water quality multiprobe (Hach Company, Loveland, CO, USA). Rainfall and air temperature were recorded at Norfolk International Airport, <10 km from the Lafayette River station. Samples (25–50 mL) were collected onto Whatman GF/F filters (pore size ~0.7 μm) and frozen for later analysis of chlorophyll a (Chla). Chla was measured fluorometrically after extraction in acetone within 2 weeks of sample collection [34]. Samples were filtered through 0.2 μm Supor filters and the filtrate frozen for later analysis of dissolved nutrient concentrations. Dissolved nitrate, nitrite, urea, phosphate, and silicate were measured colorimetrically using an Astoria Pacific nutrient autoanalyzer according to the manufacturer’s specifications. Ammonium was analyzed colorimetrically using the phenolhypochlorite method [35]. Nano- and microphytoplankton samples (500 mL) were collected from the surface (<1 m), preserved with Lugol’s solution (1% final concentration), and quantified using an inverted microscope (Nikon TS100) at 150–600× magnification following a modified Utermöhl settling and siphoning protocol [36]. Autotrophic picoplankton samples, collected at the same time and depth were preserved with gluteraldehyde (2%) and quantified using epifluorescence microscopy (Nikon E600) at 1000× magnification [37]. Phytoplankton cell volume was calculated based on observed cell dimensions and phytoplankton carbon (C) biomass calculated using established biovolume to biomass relationships [38]. Dinoflagellate species identities were positively confirmed using scanning electron Microorganisms 2014, 2 36 microscopy (SEM). Samples for SEM were fixed with gluteraldehyde and osmium tetroxide, dehydrated through an ethanol series, dried using a critical point drier, sputter-coated with gold-paladium, and analyzed using a LEO 435VP (LEO Electron Microscopy Ltd., Thornwood, NY, USA) [39]. Phytoplankton diversity was calculated daily using both species richness (number of species per sample) and the Shannon index (H′; Equation (1)), the latter incorporates the relative abundance of each species and therefore is commonly used as a measure of species evenness [40]. (1) ′ 𝐻 =−�(𝑝𝑖log𝑝𝑖) p is the proportion of the total algal biomass of species i. i The daily sampling regimen was designed to measure phytoplankton species abundance and nutrient concentrations prior to, during, and following algal blooms. Phytoplankton species abundance and diversity were compared with corresponding environmental data using Pearson correlation analysis. Because algal growth rates are on the order of days, we anticipated a lag response of phytoplankton abundance relative to nutrient concentrations and associated meteorological forcing [7]. The lag correlation analyses conducted here compared nutrient concentrations at one day intervals over an 11 day window encompassing the period before and after observed blooms of a cryptomonad and the dinoflagellate, Gymnodinium instriatum. Because biological interactions such as competition and predation are known to influence phytoplankton composition, we also used a lag correlation analysis of species richness, H′, and the abundance of other dominant phytoplankton groups relative to dinoflagellate and cryptomonad abundance. Regression analysis was used to examine the relationship between species diversity (both richness and H′) and total algal biomass. These results were compared to regression analyses of phytoplankton diversity and biomass data collected by the authors from nearby sites during the same time period as part of the Virginia Chesapeake Bay monitoring program (n = 26). Because previous studies identified both linear and non-linear (unimodal) relationships between the variables (e.g., [41]), analysis of variance was conducted to test for significant linear and quadratic relationships using regression models (SPSS 20; IBM). If both regression models were significant for a particular analysis, a partial F test was used to determine if the quadratic model significantly improved the explanation of the data relative to the linear model [42,43]. 3. Results and Discussion 3.1. Meteorological and Physical Parameters Over the 34-day sampling period, mean daily air temperatures ranged from 11.7 to 21.7 °C, and water temperatures ranged from 15.1 to 24.0 °C (Figure 1a). Average daily wind speeds were variable and ranged from 8 to 32 km h−1 with gusts exceeding 48 km h−1 (30 miles h−1) on 9 days; maximum wind gusts of 69 km h−1 were observed on May 1 (Figure 1b). During the sampling period there were 8 rain events recording 0.5 cm or more of precipitation (Figure 1c). Salinity at the sampling site decreased over the sampling period, with a maximum of 20.2 observed on 20 April and a minimum of 17.5 on 18 May. Salinity decreased following periods of rainfall (Figure 1d). The average pH at our study site was 8.31, but pH ranged from 7.98 to 8.79 (Figure 1e). Dissolved oxygen concentrations ranged from 5.0 to 7.8 mg L−1; this was 61.6% to 98.1% saturation (Figure 1f). Microorganisms 2014, 2 37 Figure 1. Daily measurements of physical and chemical parameters in the Lafayette River from 20 April to 25 May 2006. Water temperature (°C) was measured at the sampling site using the Hydrolab and mean daily air temperatures were measured at Norfolk International Airport (ORF) (a). Mean daily wind speed and maximum daily speed of wind gusts (miles h−1) were measured at Norfolk International Airport (ORF) (b). Daily cumulative precipitation (cm) was also measured at ORF (c). Salinity (d), pH (e), dissolved oxygen (mg L−1), and percent saturation (f) were measured using the Hydrolab and chlorophyll a measurements (µg L−1) were made daily on surface water samples (g). 25 a e mperaturoC)(1250 water temperature Te 10 air temperature d speed-1)m h467505 b amvagx. .w winindd s gpuesetd Win(K30 15 0 3 on c ecipitati(cm) 12 Pr 0 21 d y 20 alinit 19 S 18 17 9 e H 8.5 p 8 Dissolved Oxygen-1(mg l)7.03695 f DDOO m%g s/alturation 0120000Dissolved Oxygen(% saturation) 100 g ophyll a-1)g l 6800 hlor(µ 40 C 20 0 4/20 4/22 4/24 4/26 4/28 4/30 5/2 5/4 5/6 5/8 5/10 5/12 5/14 5/16 5/18 5/20 5/22 5/24 3.2. Phytoplankton Abundance, Composition and Diversity Chlorophyll a (Chla) concentrations ranged from 5.54 to 97.6 µg L−1 over the 34-day study period, but were less than 20 µg L−1 for all but 8 of the days (Figure 1g). Elevated Chla concentrations, 20.8–30.7 µg L−1 observed between 24 April and 1 May, were associated with a cryptomonad bloom , (Figure 1g). High Chla concentrations observed between 16 and 24 May (35.2–97.6 µg L−1) were associated with high abundances of the dinoflagellate Gymnodinium instriatum (Figure 1g). Nano- and microphytoplankton cell densities ranged from 5.8 × 106 to 7.8 × 107 cells L−1 throughout the study period (Figure 2a); picoplankton abundances ranged from 3.7 × 106 to 1.3 × 109 cells L−1 (data not shown). Microorganisms 2014, 2 38 Figure 2. Nano and micro phytoplankton abundance and biomass (a), and phytoplankton diversity (Shannon diversity index H′) and species richness (b) in daily samples collected from the Lafayette River between 20 April and 25 May 2006. 9.E+07 7.E+04 total abundance 8.E+07 total biomass 6.E+04 Cell abundance-1)(cells l234567......EEEEEE++++++000000777777 a 2345....EEEE++++00004444 -1Biomass (µgCl) 1.E+07 1.E+04 0.E+00 0.E+00 35 3 30 2.5 hness 25 2 x (H') Species ric 112050 b 11.5 versity inde Di 5 species richness 0.5 diversity index (H') 0 0 4/20 4/22 4/24 4/26 4/28 4/30 5/2 5/4 5/6 5/8 5/10 5/12 5/14 5/16 5/18 5/20 5/22 5/24 While dominated by a single species during blooms, the phytoplankton community consisted of 65 taxa from 8 major taxonomic groups, with 41 taxa present on 5 or more days (Table 1). There were 37 species of diatoms, 17 dinoflagellate species, 3 cyanobacteria, 2 silicoflagellates, 2 chlorophytes, and 1 each of cryptomonads, euglenophytes and prasinophytes. While diatoms were the most diverse group, consisting of mainly centric species (e.g., Skeletonema costatum and Chaetoceros spp.), they never represented more than 49% of the total phytoplankton present, and were generally much less abundant than the phytoflagellates. Table 1. Summary statistics of phytoplankton abundance data (cells L−1) of the 41 taxa observed in the Lafayette River samples at least 5 times during the study. The two bloom taxa are identified in bold. Abundance Phytoplankton Taxa Mean Value Minimum Value Maximum Value Diatoms unidentified Centrales 10–30 µm 2.0 × 106 1.0 × 103 5.3 × 106 unidentified Centrales 30–60 µm 1.0 × 104 2.6 × 102 1.1 × 105 Chaetoceros pendulus 6.7 × 102 2.6 × 102 1.0 × 103 Chaetoceros sp. 1.2 × 105 7.7 × 102 4.3 × 105 Cocconeis sp. 2.8 × 102 2.6 × 102 5.1 × 102 Coscinodiscus sp. 5.7 × 102 2.6 × 102 1.3 × 103 Cyclotella sp. 1.1 × 105 5.1 × 102 4.3 × 105 Cylindrotheca closterium 5.3 × 102 2.6 × 102 1.5 × 103 Dactyliosolen fragilissimus 2.9 × 104 5.1 × 102 3.2 × 105 Gyrosigma fasciola 3.1 × 102 2.6 × 102 5.1 × 102 Leptocylindrus minimus 7.7 × 104 7.7 × 102 5.4 × 105 Navicula sp. 7.3 × 102 2.6 × 102 3.8 × 103 Microorganisms 2014, 2 39 Table 1. Cont. Nitzchia sp. 2.6 × 102 2.6 × 102 2.6 × 102 unidentified Pennales 10–30 µm 2.6 × 105 2.6 × 102 8.7 × 105 unidentified Pennales30–60 µm 8.9 × 103 2.6 × 102 1.1 × 105 unidentified Pennales > 60 µm 9.0 × 102 2.6 × 102 2.8 × 103 Pleurosigma sp. 2.6 × 102 2.6 × 102 2.6 × 102 Rhizosolenia setigera 1.2 × 103 2.6 × 102 3.8 × 103 Skeletonema costatum 1.0 × 105 1.0 × 103 9.7 × 105 Thalassiosira sp. 6.7 × 102 2.6 × 102 1.3 × 103 Dinoflagellates Akashwio sanguinea 3.0 × 103 2.6 × 102 1.8 × 104 Cochlodinium polykrikoides 1.1 × 104 5.1 × 102 3.7 × 104 unidentified dinoflagellate 1.8 × 105 5.1 × 102 5.4 × 105 Dinophysis punctata 5.4 × 102 2.6 × 102 2.0 × 103 Diplopsalis lenticula 3.1 × 102 2.6 × 102 5.1 × 102 Gymnodinium sp. 8.9 × 104 2.6 × 102 8.7 × 105 Gymnodinium instriatum 3.7 × 106 2.6 × 102 3.1 × 107 Heterocapsaro tundata 6.6 × 105 1.1 × 105 3.9 × 106 Heterocapsa triquetra 1.0 × 104 2.6 × 102 1.1 × 105 Polykrikos kofoidii 5.9 × 103 1.0 × 103 4.5 × 104 Prorocentrum micans 9.3 × 102 2.6 × 102 5.9 × 103 Prorocentrum minimum 2.2 × 104 2.6 × 102 4.3 × 105 Protoperidinium sp. 7.5 × 102 2.6 × 102 1.8 × 103 Scrippsiella trochoidea 7.3 × 102 2.6 × 102 2.3 × 103 Cryptomonads Cryptomonas sp. 1.5 × 107 5.4 × 105 7.6 × 107 Cyanobacteria Lyngbya sp. 5.3 × 105 5.1 × 102 2.3 × 106 Chlorophytes Ankistrodesmus falcatus 8.7 × 103 2.6 × 102 1.1 × 105 Chlamydomonas sp. 3.6 × 105 1.1 × 105 8.7 × 105 Euglenoids Euglena sp. 7.6 × 104 2.6 × 102 4.3 × 105 Eutreptia lanowii 2.0 × 103 5.1 × 102 5.6 × 103 Prasinophytes Pyramimonas sp. 7.2 × 104 2.6 × 102 3.2 × 105 Cryptomonad taxonomic identification is notoriously problematic due to the cells’ sensitivity to chemical fixatives and the small number of morphological features that distinguish them from one another [44,45]. The morphology and size of the cryptomonads appeared consistent throughout the course of the study. The cells were comma-shaped, with a round anterior and a reflex curved pointed antapex with an average length of 18.3 μm and an average maximum width of 8.3 μm. While consistent morphological features were observed during the sampling period, the cryptomonad bloom was conservatively identified as Cryptomonas spp., indicating the possible presence of multiple species. Gymnodinium instriatum was recognized by its morphological features including the Microorganisms 2014, 2 40 displacement of the cingulum and the shape of the apical groove (Figure 3) [46] and identified using the most recent nomenclature [47]. Figure 3. Scanning electron micrograph of a Gymnodinium instriatum vegetative cell, collected at the study site on 18 May 2006, during a bloom of this dinoflagellate. Scale bar = 10 μm. Phytoflagellates, specifically cryptomonads and dinoflagellates, were the dominant algae throughout the study. At the beginning of the study period in April, the algal community was dominated by cryptophytes and diatoms but also contained substantial populations of dinoflagellates and other species. The most abundant taxon was Cryptomonas spp., which reached a maximum density of 7.7 × 107 cells L−1 on 27 April. At its peak, this group represented 96.1% of the total phytoplankton abundance and 91.6% of the phytoplankton biomass (Figure 4b). Cryptomonas spp. concentrations decreased to 4.0 × 106 cells L−1 on May 5 and then increased again having a second smaller peak in abundance of 2.6 × 107 cells L−1 on May 13. As the Cryptomonas spp. abundance declined, the densities of Gymnodinium instriatum rose dramatically beginning May 15. G. instriatum reached a maximum density of 3.0 × 107 cells L−1 on May 18 (Figure 4a). This represented 89.8% of the phytoplankton abundance and 99.0% of the total phytoplankton biomass (Figure 4b). G. instriatum Microorganisms 2014, 2 41 abundance and chlorophyll a concentrations decreased on May 19 following a rainfall event (Figure 1c) and then increased again to 1.9 × 107 cells L−1 on May 21. Figure 4. Biomass (µg C L−1) of the major taxonomic groups within the Lafayette River from 20 April to 25 May shown as absolute algal biomass for each major taxonomic group (a) and each taxonomic group as a percentage of the total algal biomass (b). 7.0E+04 Others 6.0E+04 Cryptophytes -1gCl)5.0E+04 Cyanobacteria µ omass (4.0E+04 Dinoflagellates bi n o kt Diatoms n pla3.0E+04 o micr no/2.0E+04 Na 1.0E+04 a 0.0E+00 100% 90% 80% 70% mass 60% o ve Bi 50% Relati 40% 30% 20% 10% b 0% 4/20 4/22 4/24 4/26 4/28 4/30 5/2 5/4 5/6 5/8 5/10 5/12 5/14 5/16 5/18 5/20 5/22 5/24 Estimates of phytoplankton biomass made using cell abundance and biovolume were highly correlated with Chla concentrations (r = 0.95, p = 0.000). Calculated nano-and microphytoplankton biomass ranged from 609 to 65,819 µg C L−1, with the highest biomass measured during the Gymnodinium bloom from May 16 to 24 (Figure 4a). Picoplankton always contributed less than 1% of total phytoplankton biomass throughout the study ranging from 0.5 to 181 μg C L−1 throughout the study. Species richness was low in the Lafayette River during this study, ranging from 16 to 32 with a mean of 21 taxa identified per sample compared to an average of 32 taxa identified in samples collected from the nearby Chesapeake Bay Monitoring Program station located in the Elizabeth River Microorganisms 2014, 2 42 (SBE5) during the same time period. The Shannon diversity index (H′), which includes a measure of species evenness, ranged between 0.03 and 2.57 (Figure 2b) and was lowest during the Cryptomonas spp. and G. instriatum blooms when these species dominated the phytoplankton populations. However, even when Cryptomonas spp. and G. instriatum were at their maximum abundance and represented 96.1% and 99.0% of the biomass, respectively, there were still about 20 other phytoplankton species present and species richness did not vary during bloom and non-bloom periods. Diversity rapidly increased again after blooms dissipated (Figure 2b) and there was a significant negative linear relationship between phytoplankton biomass and diversity (H′) over the 34 day study (R2 = 0.637, p < 0.0001) (Figure 5a). A negative relationship between phytoplankton biomass and species diversity was also observed during the same time period at other locations within the lower Chesapeake Bay. No significant relationship was observed between species richness and phytoplankton biomass (p > 0.05) (Figure 5b). Figure 5. Scatterplots of phytoplankton biomass and phytoplankton diversity expressed as species richness (a) and Shannon diversity index H′ (b). Black circles represent measurements of biomass and diversity recorded daily from samples collected in the Lafayette River from 20 April to 25 May 2006. White circles represent algal biomass and diversity measurements recorded at 14 Chesapeake Bay Program Monitoring stations in Virginia during April and May 2006. There was a significant negative linear relationship between species diversity and phytoplankton biomass for both datasets (p < 0.0001) (b). The solid line shows the relationship for the Lafayette River samples and the dashed line shows the relationship for the Chesapeake Bay Program data. 45 40 35 ss 30 e n ch 25 es ri 20 ci e p 15 s 10 a 5 A 0 4 1 10 100 1000 10000 100000 3.5 3 2.5 H' 2 1.5 1 b 0.5 B 0 1 10 100 1000 10000 100000 Phytoplankton biomass (mg C l-1)