A comparison of phytoplankton communities of the deep chlorophyll layers and epilimnia of the Laurentian Great Lakes

JGLR-01092; No. of pages: 10; 4C: 7, 8 Journal of Great Lakes Research xxx (2016) xxx–xxx

Contents lists available at ScienceDirect

Journal of Great Lakes Research journal homepage: www.elsevier.com/locate/jglr

A comparison of phytoplankton communities of the deep chlorophyll layers and epilimnia of the Laurentian Great Lakes Andrew J. Bramburger ⁎, Euan D. Reavie Natural Resource Research Institute, University of Minnesota Duluth, Duluth, MN 55812, USA

a r t i c l e

i n f o

Article history: Received 27 December 2015 Accepted 2 July 2016 Available online xxxx Communicated by Joseph Makarewicz Index words: Great Lakes Long-term monitoring Phytoplankton Algal ecology

a b s t r a c t Phytoplankton biomass and primary productivity within Great Lakes deep chlorophyll layers (DCL) remain largely uninvestigated. Consequently, the taxonomic makeup of DCL phytoplankton communities, as well as the mechanisms regulating their formation and maintenance, is poorly understood. We examined 6 years of phytoplankton compositional characteristics of Great Lakes summer DCL and epilimnetic communities as well as spring communities from isothermal water columns. DCLs were regularly observed during summer stratification in all lakes with the frequent exception of Lake Erie. Relative compositions of summer chlorophyte and cryptophyte assemblages were not different between the epilimnion and DCL, but DCL phytoplankton communities from other algal groups were distinct from their epilimnetic counterparts and comprised an integration of phytoplankton from the overlying epilimnetic assemblages and relict taxa characteristic of spring. Summer epilimnetic communities were characterized by higher abundances of cyanophytes, and centric diatom communities were dominated by Cyclotella sensu lato (i.e. species within Cyclotella and closely related genera). Cyclotella species exhibited distinct patterns of vertical distribution, with small-bodied taxa being partitioned heavily into the epilimnion, while larger-bodied forms tended to occupy the DCL. Vertical size partitioning was exemplified by larger mean individual cell sizes in epilimnetic siliceous algae (diatoms and chrysophytes) in the DCL compared to the epilimnion, while the opposite pattern was exhibited by cyanophytes. These findings demonstrate the importance of stratification intensity to vertical structuring of summer phytoplankton communities and imply that changing stratification regimes (such as that due to recent climate change) may exert profound effects on Great Lakes primary producer communities. © 2016 International Association for Great Lakes Research. Published by Elsevier B.V. All rights reserved.

Introduction In water bodies that exhibit seasonal or permanent stratification, deep chlorophyll layers (DCL) (Fahnenstiel et al., 1984; Moll et al., 1984) often occur in the water column below the thermocline (Brooks and Torke, 1977; Pilati and Wurtsbaugh, 2003). While the DCL may represent a large portion of the water column chlorophyll and can be responsible for much of the primary production of a lake, the mechanisms governing the general composition and dynamics of the DCL remain largely uninvestigated (Camacho, 2006; Moll and Stoermer, 1982; Pilati and Wurtsbaugh, 2003). There remain several potentially valid hypotheses regarding mechanisms that influence the formation and maintenance of the DCL. These include active processes such as in situ production in the metalimnion and hypolimnion (Cullen, 1982; Fasham et al., 1985; Venrick, 1982) and decreased gazing pressure below the thermocline (Fee, 1976), as well as active light and/or

⁎ Corresponding author. E-mail address: [email protected] (A.J. Bramburger).

predation avoidance by motile taxa (Campbell et al., 2009; Fiedler, 1982; Saros et al., 2005). Alternatively, DCL formation can be driven by passive mechanisms, including formation of a relict community following stratification and differential sinking of phytoplankton from the epilimnion (Kiefer and Kremer, 1981). When they occur, DCLs can vary considerably in their taxonomic composition and structure (Cullen, 1982; Cullen and Eppley, 1981), and this can confound indirect measures of DCL productivity. Chlorophyll a concentrations estimated by in situ fluorescence, not necessarily a reliable indicator of phytoplankton biomass (Falkowski and Kolber, 1995), may be affected by several factors and can exhibit substantial heterogeneity both through space and among taxa (Yilmaz et al., 1994). Cullen (1982) cautioned that chlorophyll a profiles provide limited information regarding mechanisms that regulate DCL formation and maintenance. Taxonomic investigations of phytoplankton communities in both the DCL and overlying waters are necessary in order to understand the role of the DCL in vertical community structure and function. Comprehensive taxonomic studies can provide insight into the importance of DCLs in contributing to overall water column productivity

http://dx.doi.org/10.1016/j.jglr.2016.07.004 0380-1330/© 2016 International Association for Great Lakes Research. Published by Elsevier B.V. All rights reserved.

Please cite this article as: Bramburger, A.J., Reavie, E.D., A comparison of phytoplankton communities of the deep chlorophyll layers and epilimnia of the Laurentian Great Lakes, J. Great Lakes Res. (2016), http://dx.doi.org/10.1016/j.jglr.2016.07.004

2

A.J. Bramburger, E.D. Reavie / Journal of Great Lakes Research xxx (2016) xxx–xxx

as well as food web function and potentially as an indicator of climate change effects on water column stratification. Although relationships between algal productivity, carbon uptake, and DCL algal communities have been investigated in marine systems (e.g., Jochem and Zeitzschel, 1993; Shulenberger and Reid, 1981; Veldhuis et al., 1997), relatively little work has been conducted on DCL productivity in lakes (Fee, 1976). Planas (1990, 1973) showed that metalimnetic carbon assimilation rates can often be higher than those observed in the epilimnion. Current understanding of what taxa are responsible for DCL productivity is minimal (Camacho, 2006). Several authors have described deep, maximal abundances of eukaryotic algal groups (e.g., Barbiero and Tuchman, 2004, 2001; Fahnenstiel et al., 1989; Pick et al., 1984; Wolin and Stoermer, 2005) within the DCL, while cyanophytes have been implicated as the primary component of DCL communities in other systems (Craig, 1987; Gervais et al., 2003; Kasprzak et al., 2000). Low grazing pressures in the metalimnion (Work and Havens, 2003) can provide refugia for palatable algal taxa capable of existing under low-light conditions (Gasol et al., 1992) and favor biomass accumulation of these forms in the DCL during stratification (Naselli-Flores and Barone, 2003). Whether these mechanisms exert sufficient influence to constrain the development of a DCL-specific algal community across multiple lakes is unknown. The existence of DCL-specific assemblages could provide a useful indicator of prolonged stratification periods that could be linked to data from paleolimnological reconstructions. For instance, increases in Cyclotella sensu lato (including taxa from the genus Cyclotella and closely related genera) taxa in the Great Lakes (Chraïbi et al., 2014) and in other northern lakes (e.g., Leavitt et al., 2009; Rühland et al., 2008) appear to be related to increasing atmospheric temperatures that are changing the physical characteristics of lake stratification. This group includes species from the genus Cyclotella and other closely related genera. Examples from the Laurentian Great Lakes include Cyclotella comensis Grunow, Discostella pseudostelligera (Hustedt) Houk and Klee, and Cyclotella cf. delicatula Reavie and Kireta. This paleolimnological shift may be related to changing assemblage characteristics of Great Lakes DCLs, but to date, no evaluation supports such a hypothesis. Deep chlorophyll layers have been reported from the Great Lakes (Putnam and Olson, 1966; Watson et al., 1975) and have been studied primarily within Lakes Superior (Barbiero and Tuchman, 2001; Putnam and Olson, 1966; Watib et al., 1975; White and Matsumoto, 2012), Michigan (Fahnenstiel and Scavia, 1987; Scavia and Fahnenstiel, 1987), and Huron (Barbiero and Tuchman, 2001; Fahnenstiel and Carrick, 1992). To date, investigations of DCLs within the Great Lakes have been limited to single lakes and short temporal durations. Fahnenstiel and Scavia (1987) provided a synopsis of the DCL community of Lake Michigan and its temporal trends from 1982 to 1984, while Twiss et al. (2012a) described growth and loss rates in phytoplankton communities in Lake Ontario. Barbiero and Tuchman (2001) broadly summarized physical, chemical, and biological properties of DCLs in the Great Lakes based on a single season dataset (1998). We compared and contrasted the composition and structure of phytoplankton communities from the spring isothermal water column and summer epilimnia and DCLs of the Great Lakes during the period spanning 2007–2012, and evaluated dissimilarities between epilimnetic and DCL phytoplankton assemblages at the basin scale in order to determine whether a characteristic DCL community exists within the Great Lakes. We also described general biovolume and abundance characteristics for epilimnetic and DCL assemblages in order to provide initial insight into the relative contributions of DCL assemblages to the overall Great Lakes phytoplankton community. We hypothesize that the phytoplankton assemblages of Great Lakes DCLs are compositionally distinct from corresponding epilimnetic assemblages. We further anticipate that the same suite of taxa contributes to this dissimilarity across lakes. Additionally, we hypothesize that size differences exist between conspecific occupants of the DCL and epilimnion.

Methods Sampling site locations and sample collection A total of 1034 phytoplankton samples were collected from 71 stations within the Great Lakes during a series of twice-annual cruises (April and August 2007–2012) by the R/V Lake Guardian as part of the USEPA-GLNPO Monitoring Program (Fig. 1). Water quality parameters (temperature, specific conductivity, pH, irradiance, dissolved oxygen, turbidity, chlorophyll a by fluorescence) were measured in situ using a SeaBird 911 CTD equipped with auxiliary sensors. Additional parameters (total phosphorus, nitrates + nitrites, silica) were measured according to methods described in detail by in the USEPA (2010) standard operating procedure. Phytoplankton samples were collected simultaneously with water quality measurements via Niskin bottle rosette. Integrated samples collected in spring (381) were produced by combining samples from discrete depths through the water column. Summer integrated epilimnetic samples (385) were produced by combining samples from discrete depths above the thermocline (surface, 5 m, 10 m, 20 m), while summer DCL samples (268) were taken from a single discrete depth associated with the fluorescence-inferred chlorophyll a maximum below the thermocline at each site (USEPA, 2010). When no DCL was detected at summer stations, only integrated epilimnetic samples were collected. Spring sampling cruises occurred annually in April, while summer cruises took place annually in August. This study is based on data from samples collected during the 2007– 2012 cruises. Sample preparation and algal enumeration Whole-water phytoplankton samples were preserved with Lugol's iodine solution and returned to the laboratory for taxonomic analysis. Subsamples for soft-bodied algal analysis were loaded into Utermöhl (1958) counting chambers for inverted light microscope (LM) analysis. Diatom samples were subjected to digestion with heated 30% H2O2. Cleaned diatom material was mounted on coverslips and counted under LM. Diatom and soft algae (all non-diatom and non-siliceous groups) samples were enumerated along transects until a total count of 250 entities for soft algae or 500 diatom valves was achieved. Both diatoms and soft algae were identified to the lowest taxonomic level possible. For diatoms, identification was to species or variety, while identification was to species, and occasionally genus for soft algae. Up to 10 individuals of each taxon were measured (length, width, depth, diameter as applicable) in order to determine taxon-specific individual biovolume (cell size) (Reavie et al., 2010). Count and measurement data were used to calculate cell density, species-relative abundance, biovolume, and individual cell biovolume. These counting methods follow the standard GLNPO phytoplankton enumeration techniques outlined by USEPA (2010). Additional details of sample processing are provided by Reavie et al. (2014a). Statistical analysis Paired-sample t-tests were used to examine differences in water quality parameters between summer epilimnetic and DCL samples. A series of one-way analyses of variance (ANOVA) were used to evaluate differences in mean phytoplankton taxonomic richness, density, and biovolume among spring integrated (SprINT), summer epilimnetic (SumEPI), and summer DCL (SumDCL) phytoplankton samples. We employed non-metric multidimensional scaling (NMDS), coupled with analysis of similarity (ANOSIM) in order to visualize and quantify dissimilarities among spring and summer epilimnetic and summer DCL phytoplankton assemblages within each lake. Similarity percentages (SIMPER; per Clarke, 1993) were used to evaluate species' contributions to assemblage dissimilarities. We used repeated-measures analysis of variance (rANOVA) and paired-sample t-tests in order to examine

Please cite this article as: Bramburger, A.J., Reavie, E.D., A comparison of phytoplankton communities of the deep chlorophyll layers and epilimnia of the Laurentian Great Lakes, J. Great Lakes Res. (2016), http://dx.doi.org/10.1016/j.jglr.2016.07.004

A.J. Bramburger, E.D. Reavie / Journal of Great Lakes Research xxx (2016) xxx–xxx

3

Fig. 1. Map of the Great Lakes showing EPA-GLNPO phytoplankton/water quality sampling stations. Each station is sampled twice annually (early spring and mid-summer) by the R/V Lake Guardian.

differences in species-specific individual biovolume (cell-size) among spring integrated and summer epilimnetic and DCL assemblages. Results General observations Deep chlorophyll layers were observed with varying frequency, and at different depths across the Great Lakes (Table 1). In general, physical and chemical water quality parameters had few significant differences between mean values for summer epilimnetic and DCL samples, with the exceptions of temperature and photosynthetic irradiance, which are both higher, unsurprisingly, in the epilimnion (Fig. 2A). Mean chlorophyll a is typically higher in the DCL than in the epilimnion but only significantly so in Lake Michigan and Lake Huron (Fig. 2B). Lower pH was typical in the DCL, significantly so in Lake Ontario and Lake Erie. Dissolved oxygen was higher in the DCL, with the exception of Lake Erie, which showed no difference. Turbidity was significantly higher in Lake Erie's epilimnion with no difference apparent in the other lakes. Higher nitrates + nitrites occurred in the DCLs for Lake Michigan and Lake Ontario. The upper Great Lakes (Superior, Michigan, Huron) display different patterns of phytoplankton species richness and density than the lower Lakes (Erie, Ontario) (Fig. 3). Mean taxonomic richness was lower in summer samples (SumEPI and SumDCL) than in spring samples (SprINT) for the upper lakes (Superior F = 22.04, p b 0.001, Michigan F = 7.40, p = 0.008, Huron F = 76.97, p b 0.001). No significant Table 1 Occurrence frequency and depth of DCLs sampled. Lake

DCL occurrences

Mean DCL depth (m)

Min DCL depth (m)

Max DCL depth (m)

Superior Michigan Huron Erie Ontario

183 116 124 45 86

30.40 31.40 36.55 17.91 19.55

11 12.7 9 11.1 8.3

52 46.7 49.9 29.1 39.8

differences in richness were detected between SumEPI and SumDCL samples in any of these lakes (p N 0.05). In the lower lakes, richness was higher in the summer than spring samples (Erie F = 5.29, p = 0.006, Ontario F = 31.44, p b 0.001). In Lake Erie, SumDCL samples exhibited mean richness values intermediate to spring and summer epilimnetic samples (Fig. 3). Mean phytoplankton densities were higher within summer samples (SumEPI and SumDCL) than in spring samples in all lakes (F N 17.76, p b 0.001, Fig. 3). Summer epilimnetic phytoplankton densities were typically higher than densities in the DCL. Lake Michigan, however, exhibited higher mean phytoplankton density in the DCL than in the summer epilimnion (Student's t, p = 0.002). Phytoplankton biovolume generally followed a trend similar to that of density (Fig. 3), with the exception of Lake Erie, which exhibited no differences in mean biovolume among spring (SprINT) and summer (SumEPI and SumDCL) samples (F = 2.81, p = 0.062). Increases in density and biovolume were accounted for primarily by increases in soft-bodied algal abundance during the summer sampling season. A repeated-measures ANOVA showed that significant differences existed in mean taxon-specific individual biovolume (cell size) among spring integrated samples, summer DCL, and summer epilimnetic samples for species that were observed in all three sample types (F = 8.68, p = 0.0002). Post-hoc Student's t-tests showed that taxon-specific cell sizes were larger in spring integrated samples than in either summer DCL (p = 0.023) or summer epilimnetic samples (p b 0.0001). Cell sizes in summer DCL samples were larger than conspecifics in summer epilimnetic samples but not significantly so (p = 0.061). When this analysis was performed for specific algal divisions, significant differences among spring, summer DCL, and summer epilimnetic samples were observed in the chlorophytes (F = 4.57, p = 0.013), chrysophytes (F = 8.74, p = 0.0003), cryptophytes (F = 101.04, p b 0.0001), and cyanophytes (F = 3.26, p = 0.048). Among the chlorophytes and cryptophytes, cell sizes were larger in spring samples than in either summer sample group. Among chrysophytes, summer DCL samples contained larger-celled individuals than either the spring integrated or summer epilimnetic sample. The opposite pattern was observed in the cyanophytes (Fig. 4).

Please cite this article as: Bramburger, A.J., Reavie, E.D., A comparison of phytoplankton communities of the deep chlorophyll layers and epilimnia of the Laurentian Great Lakes, J. Great Lakes Res. (2016), http://dx.doi.org/10.1016/j.jglr.2016.07.004

4

A.J. Bramburger, E.D. Reavie / Journal of Great Lakes Research xxx (2016) xxx–xxx

Of course, not all species occurred in all three sample sets. A pairedsample t-test examining differences in taxon-specific cell size in taxa that occurred in both summer DCL and epilimnetic samples (regardless of presence in spring samples) indicated no significant differences (t = − 0.629, p = 0.53). Rather, cells within a given species were the same biovolume in both the summer epilimnion and DCL. However, among the major algal divisions, cell size was larger in the DCL than in the epilimnion for siliceous algae groups (centric diatoms t = 12.10, p b 0.001; pennate diatoms, t = 57.28, p b 0.001; chrysophytes, t = 49.85, p b 0.001) (Fig. 4). In contrast, mean cell size was lower in DCL than in epilimnetic samples for cyanophytes (t = 48.22, p b 0.001), and no significant difference in cell size was observed between DCL and epilimnetic samples for chlorophyte (t = 0.75, p = 0.39) and cryptophyte (t = 1.12, p = 0.29) taxa.

Relative biovolume contributions of major algal groups differed between spring and summer and between epilimnetic and DCL samples during the summer for all lakes (Fig. 5). Notably, cryptophytes and pyrrophytes represented large portions of the biovolume of phytoplankton in the upper lakes (Superior, Michigan, Huron) during the unstratified spring season. Centric diatoms, especially Aulacoseira islandica represented the majority of Lake Erie's spring biovolume. In all lakes (except Huron), relative cyanophyte biovolume increased in the summer and was partitioned primarily into the epilimnion. Pennate diatoms also became more important during the stratified period in Lakes Michigan, Erie, and Ontario, and were partitioned primarily into the DCL (Fig. 5). NMDS, ANOSIM, and SIMPER techniques revealed assemblage dissimilarities among algal assemblages from spring and summer

Fig. 2. A. Summary of general water quality parameters for Great Lakes summer epilimnetic (INT) and DCL samples from 2007 to 2012. Columns represent mean values. Error bars represent the 95% confidence interval of the mean. A * denotes significant difference between INT and DCL samples (paired-sample t-test, p ≤ 0.05). B. Summary of nutrient and turbidity concentrations for Great Lakes summer epilimnetic and DCL samples from 2007 to 2012. Columns represent mean values. Error bars represent the 95% confidence interval of the mean. A * denotes significant difference between INT and DCL samples (paired-sample t-test, p ≤ 0.05).

Please cite this article as: Bramburger, A.J., Reavie, E.D., A comparison of phytoplankton communities of the deep chlorophyll layers and epilimnia of the Laurentian Great Lakes, J. Great Lakes Res. (2016), http://dx.doi.org/10.1016/j.jglr.2016.07.004

A.J. Bramburger, E.D. Reavie / Journal of Great Lakes Research xxx (2016) xxx–xxx

5

Fig. 2 (continued).

integrated epilimnetic samples, and summer DCL samples. Pronounced seasonal shifts in epilimnetic algal community composition were observed in all lakes, along with frequent differences in vertical distribution of taxa during the stratified season. During the spring, the epilimnion was typically dominated by centric diatoms across all lakes. Species within the genera Aulacoseira and Stephanodiscus were particularly abundant in these assemblages. Summer epilimnetic assemblages showed higher relative abundance of taxa within the genus Cyclotella sensu lato and pennate diatom taxa including Fragilaria crotonensis, Synedra filiformis, and Synedra radians. SumDCL assemblage samples tended to cluster between SprINT and SumEPI samples in NMDS ordinations and contained taxa characteristic of both epilimnetic assemblages. A more in-depth description of lake-specific patterns is given below. Lake Superior In Lake Superior, the most abundant taxa in spring integrated samples (by biovolume) were Gymnodinium helveticum, Cryptomonas reflexa, Cryptomonas erosa, Cyclotella comta, and Rhodamonas lens. During the summer, Cyclotella comta, Cyclotella cf. delicatula, Cryptomonas reflexa, Gymnodinium helveticum, and Tabellaria flocculosa were the most abundant taxa in the epilimnion, while Cyclotella comta, Gymnodinium helveticum, Cryptomonas reflexa, Oscillatoria limnetica, and Asterionella Formosa were the most abundant taxa in the DCL. Significant dissimilarities existed among SprINT, SumEPI, and SumDCL sample groups in Lake Superior (ANOSIM Global R = 0.447, p = 0.001). Pairwise comparisons (SprINT–SumEPI, SprINT–SumDCL, SumEPI–SumDCL) also exhibited significant dissimilarities (0.238 ≤ R ≤ 0.684, p = 0.001). NMDS ordination illustrated complete separation between spring and summer integrated epilimnetic samples, while the summer DCL samples exhibited some overlap with both epilimnetic sample groups (Fig. 6).

SIMPER analysis demonstrated that SumDCL samples were slightly more similar to SprINT (mean dissimilarity = 54.55) than to SumEPI (mean dissimilarity = 50.31) samples, while SumEPI samples were most dissimilar to SprINT samples (mean dissimilarity = 64.80). Dissimilarities among SumEPI and SprINT samples, were accounted for primarily by increased abundances of Aphanocapsa spp. (26.18%), Cyclotella comensis “rough center with process” (also known as Cyclotella cf. delicatula, Reavie and Kireta, 2015) (8.59%), Discostella pseudostelligera (6.19%), and Cyclotella ocellata (5.08%), and by lower abundances of Synedra filiformis var. exilis (4.12%) in the summer samples. Similarly, increased abundances of Aphanocapsa spp. (21.59%), C. ocellata (6.36%), C. cf. delicatula (3.17%), and C. comensis (4.75%) during the summer months accounted for a large proportion of the dissimilarity between SprINT and SumDCL samples. Dissimilarities between SumEPI and SumDCL samples were accounted for largely by higher abundances of Aphanocapsa spp. (27.58%), C. cf. delicatula (6.06%), and C. ocellata (5.79%), as well as by lower abundances of S. filiformis and S. filiformis var. exilis in the epilimnetic samples. Lake Michigan In Lake Michigan, the most abundant taxa in spring integrated samples were Stephanodiscus parvus, Gymnodinium helveticum, Stephanodiscus alpinus, Stephanodiscus hantzschii, and Cryptomonas reflexa. During the summer, Ceratium hirudinella, Fragilaria crotonensis, Anabaena flos-aquae, Cryptomonas reflexa, and Diatoma tenue var. elongatum were the most abundant taxa in the epilimnion, while Diatoma tenue var. elongatum, Fragilaria crotonensis, Cryptomonas reflexa, Oscillatoria limnetica, and Ceratium hirundinella were the most abundant taxa in the DCL. Significant dissimilarities existed among SprINT, SumEPI, and SumDCL assemblages (Global R = 0.557, p = 0.001), as well as in all pairwise comparisons (0.258 ≤ R ≤ 0.796, p = 0.001). Similar to Lake

Please cite this article as: Bramburger, A.J., Reavie, E.D., A comparison of phytoplankton communities of the deep chlorophyll layers and epilimnia of the Laurentian Great Lakes, J. Great Lakes Res. (2016), http://dx.doi.org/10.1016/j.jglr.2016.07.004

6

A.J. Bramburger, E.D. Reavie / Journal of Great Lakes Research xxx (2016) xxx–xxx

Fig. 4. z-normalized mean individual biovolume (cell size) differences among spring integrated (SprINT), summer DCL (SumDCL), and summer epilimnetic samples (SumEPI) for dominant algal groups in the Great Lakes. Error bars represent 95% confidence interval of the mean. Bars not connected by the same letter are significantly different (p b 0.05).

low abundances of S. filiformis (7.54%) in the epilimnetic samples, relative to SumDCL samples. Lake Huron

Fig. 3. Mean taxonomic richness, density, and biovolume differences among spring integrated (SprINT), summer DCL (SumDCL), and summer epilimnetic samples (SumEPI) representing phytoplankton communities from the Great Lakes, 2007–2012. Error bars represent the 95% confidence interval of the mean. Bars not connected by the same letter are significantly different (p b 0.05).

Superior patterns, NMDS ordination showed that SprINT samples were distinct from SumEPI samples. Here, however, SumDCL samples clustered more closely to SumEPI samples and were also largely separated from SprINT samples (Fig. 6), indicating a stronger seasonal separation in assemblages than that observed in Lake Superior. SIMPER results for Lake Michigan reflected the ordination patterns. SprINT samples were most dissimilar to SumEPI samples (mean dissimilarity = 81.26), followed by SumDCL samples (mean dissimilarity = 73.37). SumDCL samples were most similar to SumEPI samples (mean dissimilarity = 39.11). The dissimilarity between SprINT and SumEPI samples was due mainly to higher abundances of Aphanocapsa spp. (14.96%) and Fragilaria crotonensis (8.31%), and lower abundances of Stephanodiscus sp. #51 (also described as an unknown species of S. parvus by Reavie and Kireta (2015); 11.31%) and S. parvus (8.316%) in the summer samples. High abundances of S. sp. #51 (12.16%) and S. parvus (9.72%), coupled with relatively low abundances of S. filiformis (6.80%) and Aphanocapsa spp. (6.68%) in SprINT samples compared to SumDCL samples accounted for a large proportion of the dissimilarity between these sample groups. The dissimilarity between Lake Michigan SumEPI and SumDCL samples was driven primarily by high abundances of Aphanocapsa spp. (16.11%), F. crotonensis (6.84%), and Oscillatoria minima (5.56%), and

In Lake Huron, the most abundant taxa in spring integrated samples were Rhodamonas lens, Gymnodinium helveticum, Cryptomonas reflexa, Aulacoseira islandica, and Fragilaria crotonensis. During the summer, Fragilaria crotonensis, Cyclotella cf. delicatula, Chrysosphaerella longispina, Synedra filliformis, and Cyclotella comta were the most abundant taxa in the epilimnion, while Fragilaria crotonensis, Cyclotella comta, Asterionella formosa, Cyclotella cf. delicatula, and Cryptomonas reflexa were the most dominant taxa in the DCL. As with the other upper lakes, Lake Huron exhibited significant dissimilarities among SprINT, SumEPI, and SumDCL samples (Global R = 0.691, p = 0.001). Significant dissimilarities were also observed in all pairwise comparisons (0.480 ≤ R ≤ 0.874, p = 0.001) and NMDS ordination indicated little overlap among groups (Fig. 6). SIMPER analysis indicated the widest divergence between SprINT and SumEPI samples (mean dissimilarity = 74.84), which was driven largely by high abundances of Aphanocapsa spp. (23.02%) and C. comensis var. 1 (12.19%) in summer samples and S. filiformis (12.34%), A. formosa (8.61%), and N.acicularis (7.06%) in spring samples. The SumDCL sample group was roughly equally dissimilar to both the SprINT (mean dissimilarity = 64.09) and SumEPI (mean dissimilarity = 60.92) samples. High abundances of Aphanocapsa spp. and C. comensis in SumDCL samples and S. filiformis, A. formosa, and Nitzschia acicularis in SumEPI samples contributed strongly to the dissimilarity between these sample groups (13.66%, 10.27%, 10.17%, 5.77%, 6.06%, respectively). Among summer samples (SumEPI and SumDCL), strong partitioning of C. comensis var. 1 (12.44%) and S. filiformis (8.00%) to the epilimnetic samples, as well as high abundances of Aphanocapsa spp. (22.91%) and C. comensis (12.64%) in SumDCL samples, accounted for most of the overall dissimilarity. Lake Erie In Lake Erie, Aulacoseira islandica, Stephanodiscus alpinus, Surirella ovata, Stephanodiscus binderanus, and Stephanodiscus parvus were the most abundant taxa in spring samples. In the summer, Microcystis aeruginosa, Aulacoseira granulata, Fragilaria crotonensis, Ceratium hirudinella, and Aphanizomenon flos-aquae were the dominant taxa in the epilimnion, while Fragilaria crotonensis, Aphanizomenon flos-aquae,

Please cite this article as: Bramburger, A.J., Reavie, E.D., A comparison of phytoplankton communities of the deep chlorophyll layers and epilimnia of the Laurentian Great Lakes, J. Great Lakes Res. (2016), http://dx.doi.org/10.1016/j.jglr.2016.07.004

A.J. Bramburger, E.D. Reavie / Journal of Great Lakes Research xxx (2016) xxx–xxx

7

Fig. 5. Mean relative biomass contributions of various phytoplankton types to spring integrated epilimnetic (INT), summer epilimnetic, and summer DCL samples. BAC = centric diatoms, BAP = pennate diatoms, CHL = chlorophytes, CHR = chrysophytes, CRY = cryptophytes, CYA = cyanophytes, EUG = euglenoids, PYR = pyrrhophytes, UNI = unidentified algae.

Ceratium hirudinella, Cryptmonas reflexa, and Cryptomonas erosa were the most abundant taxa in the DCL. Due to its shallower depth and strong vertical mixing, Lake Erie does not always develop a DCL. While ANOSIM indicated significant dissimilarities in the overall Erie sample set (Global R = 0.852, p = 0.001), and SprINT algal assemblages were significantly dissimilar to those of SumEPI and SumDCL samples (R = 0.916, p = 0.001; R = 0.866, p = 0.001, respectively), SumDCL samples, when collected, were also significantly dissimilar to SumEPI samples (R = 0.271, p = 0.019). Ordination of these samples using NMDS clearly illustrated the distinction between spring and summer algal assemblages, and to a lesser extent, SumEPI and SumDCL assemblages within the lake (Fig. 6). Dissimilarity between spring integrated and summer epilimnetic assemblages in Lake Erie (SprINT and SumEPI) were strongly influenced by high abundances of A. islandica (18.78%) and S. parvus (8.98%) in spring samples and high abundances of Aphanocapsa spp. (24.79%), Aphanothece spp. (4.45%), F. crotonensis (4.37%), and several species of Cyclotella in summer samples (11.99%, collectively). The same suite of species contributed consistently to dissimilarities between SprINT and SumDCL samples. When a DCL did form, dissimilarities between epilimnetic and DCL samples were driven largely by high abundances of Aphanocapsa spp. (22.44%) and Microcystis aeruginosa (5.46%) in epilimnetic samples and F. crotonensis (14.84%), C. comensis var. 1 (6.68%), and D. pseudostelligera (6.58%) in the DCL samples. Diatom abundances were typically higher in DCL assemblages. Lake Ontario In Lake Ontario, Stephanodiscus parvus, Gymnodinium helvaticum, Stephanodiscus alpinus, Stephanodiscus hantzschii, and Cryptomonas reflexa were the most abundant taxa in spring assemblages. During the summer, Ceratium hirundinella, Fragilaria crotonensis, Anabaena flos-aquae, Cryptomonas reflexa, and Diatoma tenue var. elongatum were the most abundant taxa in the epilimnion, while Diatoma tenue var. elongatum, Fragilaria crotonensis, Cryptomonas reflexa, Oscillatoria limnetica, and Ceratium hirundinella were the most abundant taxa in the DCL. Algal assemblages in Lake Ontario samples were less dissimilar to one another than those in other lakes (ANOSIM Global R = 0.307, p = 0.001). Pairwise comparisons demonstrated significant dissimilarities among spring integrated, summer epilimnetic, and summer DCL

samples (0.047 ≤ R ≤ 0.459, 0.001 ≤ p ≤ 0.019). NMDS ordination showed that most SprINT samples were well separated from summer samples and that some close association existed between SumEPI and SumDCL sample sets (Fig. 6). SIMPER analysis indicated the greatest dissimilarity between SprINT and SumEPI samples (mean dissimilarity = 88.02), which was accounted for by high abundances of S. parvus (18.59%) and Cyclotella atomus “fine form” [also described as C. atomus var. 1 (Reavie and Kireta, 2015) 7.13%] in the SprINT samples and F. crotonensis (18.63%), C. comensis var. 1 (9.92%), and Synedra ostenfeldii (8.00%) in the SumEPI samples. Dissimilarities between SprINT and SumDCL samples were accounted for by elevated abundances of F. crotonensis and Diatoma tenue in SumDCL samples (19.64%, 10.13%, respectively), as well as by higher abundances of Stephanodiscus parvus and C. atomus var. 1 in SprINT samples (18.54%, 7.27%, respectively). Discussion Deep chlorophyll layers are known to form in all of the Great Lakes (Putnam and Olson, 1966; Watson et al., 1975). With the exception of Lake Erie, we observed the formation of DCLs regularly in all lakes during periods of summer stratification. Patterns of phytoplankton community structure were surprisingly consistent and revealed interesting vertical distribution characteristics, especially with respect to diatoms and cyanophytes. While spring integrated and summer epilimnetic assemblages were compositionally distinct, summer DCL assemblages were often more similar to spring assemblages than to summer epilimnetic samples. This finding suggests that DCL assemblages are composed, at least partially, of remnants of the algal community that was present during isothermal conditions in the spring. Spring phytoplankton communities in the Great Lakes were frequently dominated by diatom taxa, primarily from the genera Aulacoseira and Stephanodiscus. In general, cyanophyte taxa exhibited slightly higher abundances in summer epilimnetic samples, compared with spring epilimnetic samples, and summer DCL samples. The shift to dominance by cyanophytes during the stratified period is well understood in lentic systems (e.g., Sommer, 1985). High relative abundances of cyanophytes in the summer epilimnion, compared to the DCL, suggest that the inherent buoyancy of cyanophytes allows them to maintain their position in the euphotic zone better than other taxa (Reynolds et al., 1987; Walsby

Please cite this article as: Bramburger, A.J., Reavie, E.D., A comparison of phytoplankton communities of the deep chlorophyll layers and epilimnia of the Laurentian Great Lakes, J. Great Lakes Res. (2016), http://dx.doi.org/10.1016/j.jglr.2016.07.004

8

A.J. Bramburger, E.D. Reavie / Journal of Great Lakes Research xxx (2016) xxx–xxx

Fig. 6. NMDS ordinations of phytoplankton assemblages in the five Great Lakes. Symbols represent sample scores (SprINT = spring integrated epilimnetic; SumEPI = summer integrated epilimnetic; SumDCL = summer DCL). Vectors illustrate the relative direction and magnitude of species' contribution to dissimilarities among samples. Species shown were correlated to among-group differences with a Pearson correlation coefficient ≥ 0.5. Bray–Curtis dissimilarity was used as the distance metric, and distances were calculated based on species relative abundances.

et al., 1997). Advantages conferred by enhanced buoyancy bear important implications for the formation of cyanophyte blooms under the predictions of climate change models (Reynolds et al., 1987; Wagner and Adrian, 2009). While these colonial cyanobacteria and other autotrophic picoplankton represented only a small portion of the overall phytoplankton biovolume of the lakes, their importance to the ecological function of the photosynthetically active epilimnion is an avenue

worthy of investigation in light of ongoing oligtrophication and intensified stratification within several Laurentian Great Lakes basins. The effects of incipient climate change have also resulted in community reorganization at finer levels of taxonomic organization. Several authors have proposed that Cyclotella abundances in lakes have increased globally as an indirect result of elevated atmospheric temperatures (e.g. Chraïbi et al., 2014; Rühland et al., 2008), although the

Please cite this article as: Bramburger, A.J., Reavie, E.D., A comparison of phytoplankton communities of the deep chlorophyll layers and epilimnia of the Laurentian Great Lakes, J. Great Lakes Res. (2016), http://dx.doi.org/10.1016/j.jglr.2016.07.004

A.J. Bramburger, E.D. Reavie / Journal of Great Lakes Research xxx (2016) xxx–xxx

mechanism responsible for Cyclotella increases in warmer waters has not been determined. In this study, summer epilimnetic assemblages were characterized by higher abundances of taxa within Cyclotella sensu lato, particularly C. comensis and C. comensis var. 1, than their springtime and summer DCL counterparts. NMDS illustrated the taxonomic composition of the DCL assemblages as intermediate communities composed of elements of both the spring and summer epilimnetic communities (Fig. 5). It is apparent that settling from the epilimnion is an important source of propagules to the DCL. In certain lakes, there are variations in water quality (e.g., nutrients) that might drive differences in phytoplankton between epilimnia and DCLs, but the consistency across lakes in the relationship between spring and summer DCL assemblages indicates that the chemical variables we present were less important determinants of the unique DCL community. Further, spring epilimnetic assemblages are more similar to DCL assemblages than they are to summer epilimnetic assemblages, suggesting that passive settling affects taxa differently; rather, sinking throughout the late spring and summer as stratification sets up is at least partly responsible for the DCL phytoplankton assemblage. Fahnenstiel and Scavia (1987) noted that DCL phytoplankton assemblages in Lake Michigan were similar to spring assemblages but suggested that in situ production was also important to maintenance of these assemblages. We suggest that the DCL community is a relict integration of propagules from both spring and summer epilimnetic assemblages, while the summer epilimnetic community represents a distinct assemblage. Unfortunately, temporal sampling resolution is too low to address vertical community structuring in the spring water column. As such, we are currently unable to confirm whether components of the summer DCL are descendants of sunken propagules from spring (i.e., active DCL maintenance) or simply remnants of the spring phytoplankton community (i.e., passive DCL formation). Further, due to the constraints of the USEPA sampling regime, we are also unable to assess rates of photosynthesis and primary production within the DCL and epilimnion, and are unable to comment on the relative contributions of these strata to total water column production. The infrequency of summer DCL formation in Lake Erie despite strong thermal stratification is not surprising, considering the relative shallowness of the lake and the high epilimnetic turbidity and susceptibility to meteorological drivers (Lick et al., 1994). When DCLs are observed within Lake Erie, they occur predominantly within the central basin and occur at shallower mean and maximum depths than in the other lakes (although mean DCL depths are similar to Lake Ontario). Differences between spring and summer algal assemblages in Lake Erie were driven primarily by high spring abundances of Aulacoseira islandica, which is known to grow under the ice in Lake Erie (Saxton et al., 2012; Twiss et al., 2012b), and large blooms of this taxon occur regularly in the spring within Lake Erie (Reavie et al., 2014b). After the onset of stratification in the summer months, Lake Erie's epilimnetic phytoplankton community becomes increasingly dominated by cyanophytes, including the potentially harmful Microcystis. When a DCL does form in Lake Erie, the DCL phytoplankton community is composed primarily of pennate and centric diatoms, despite overlying epilimnetic assemblages being dominated by cyanophytes, suggesting that low diatom biovolume in the summer epilimnion in Lake Erie is due in part to losses via sinking. Higher cyanophyte abundances within the summer epilimnion are likely supported in part by the inherent buoyancy of many cyanophyte taxa (per Dokulil and Teubner, 2000; Wynne et al., 2013). In all of the lakes, summer epilimnetic and DCL phytoplankton communities were significantly dissimilar to one another, suggesting that the summer epilimnetic community is distinct. With the exception of Lake Erie, whose summer epilimnetic community is characterized by high abundances of cyanophytes, vertical dissimilarity is driven largely by differences in diatom assemblages between the epilimnion and the DCL. Large, Fragilaria-like pennate diatoms including Fragilaria crotonensis, Synedra radians, and Synedra ostenfeldii are typically more

9

abundant in the DCL than in the epilimnion. These findings are consistent with results from Lake Michigan during the early 1980s (Fahnenstiel and Scavia, 1987), where the upper portions of the DCL were also dominated by Fragilaria crotonensis and other similar pennate taxa. Across the entire basin, centric diatom communities shift from being dominated by Stephanodiscus species in the spring to Cyclotella sensu lato species, particularly C. comensis and its varieties, in the summer. Interestingly, Cyclotella sensu lato exhibits differential vertical distributions in the summer water column. Typically, taxa with smaller cell sizes including C. comensis var. #1 and C. pseudostelligera are more strongly represented in the epilimnion than in the DCL. Conversely, larger-celled Cyclotella taxa such as C. comensis, C. ocellata, and Cyclotella tripartita are more abundant in the DCL. These results are in contrast to data presented by Barbiero and Tuchman (2001), who suggested that C. comensis and C. comta were more abundant in the epilimnion than in the DCL in Lakes Superior and Huron during the 1998 stratified season. Cyclotella taxa that exhibit a broad range of individual biovolumes are distributed approximately evenly between the epilimnion and DCL, although larger individuals of these taxa tend to be favored within the DCL and vice versa, suggesting that inter-species variability in water column position is likely driven by differential sinking rates. The distribution of different-sized conspecific individuals across a depth gradient is not limited to Cyclotella. In general, siliceous algae (diatoms and chrysophytes) exhibited higher mean cell sizes in the DCL than in the epilimnion, suggesting that larger, heavier individuals were more prone to sinking out of the epilimnion. In contrast, cyanophytes were typically larger in the epilimnion than in the DCL, implying a buoyancy effect among larger individuals. The lack of significant vertical distribution patterns among other groups is likely due to the confounding influences of individual motility (cryptophytes and dinoflagellates) and variable colony shape and size (chlorophytes). The opposite patterns of vertical distribution displayed by heavy, siliceous taxa and largely unicellular, buoyant small taxa suggest that patterns of vertical community structure are regulated at least in part by differential sinking rates. This implies that summer stratification intensity imparts an important structuring influence on algal communities within the Great Lakes and promotes the existence of an epilimnion-specific phytoplankton assemblage. In summary, summer epilimnia and DCLs within the Laurentian Great Lakes support distinct phytoplankton communities. While summer epilimnetic assemblages are characterized by taxa typically found in warm, stratified waters, DCL assemblages contain a mixture of individuals representing both spring and summer epilimnetic components. In fact, DCL assemblages are often more similar to spring communities than to their summer epilimnetic counterparts, suggesting that passive sinking from the spring community plays an important role in DCL development. Further, within-species size differences between DCL and summer epilimnetic assemblages illustrate the importance of buoyancy and loss due to sinking in the development of vertically structured communities in stratified water columns. The development of a distinct epilimnetic and DCL algal communities during summer stratification in the Great Lakes provides modern context for paleolinmological indicators thought to represent periods of stratification consistent with climate-driven water column warming. Acknowledgments This project was financially supported through the US Environmental Protection Agency Great Lakes National Program Office (GLNPO) Surveillance and Monitoring program, under Cooperative Agreement GL-00E23101-2. This document has not been subjected to the EPA's required peer and policy review and therefore does not necessarily reflect the view of the agency, and no official endorsement should be inferred. Michael Agbeti supported algal assessments of the phytoplankton samples The authors would like to acknowledge field and laboratory personnel Kitty Kennedy and Lisa Estepp and the crew of the R/V Lake

Please cite this article as: Bramburger, A.J., Reavie, E.D., A comparison of phytoplankton communities of the deep chlorophyll layers and epilimnia of the Laurentian Great Lakes, J. Great Lakes Res. (2016), http://dx.doi.org/10.1016/j.jglr.2016.07.004

10

A.J. Bramburger, E.D. Reavie / Journal of Great Lakes Research xxx (2016) xxx–xxx

Guardian. We also thank MacKenzie Waller for GIS and mapping assistance. This is contribution number [issued on acceptance] of the Center for Water and the Environment, Natural Resources Research Institute, University of Minnesota Duluth. References Barbiero, R.P., Tuchman, M.L., 2001. Results from the US EPA's biological open water surveillance program of the Laurentian Great Lakes: II. Deep chlorophyll maxima. J. Great Lakes Res. 27, 155–166. Barbiero, R.P., Tuchman, M.L., 2004. The deep chlorophyll maximum in Lake Superior. J. Great Lakes Res. 30, 256–268. Brooks, A.S., Torke, B.G., 1977. Vertical and seasonal distribution of chlorophyll a in Lake Michigan. J. Fish. Res. Board Can. 34, 2280–2287. Camacho, A., 2006. On the occurrence and ecological features of deep chlorophyll maxima (DCM) in Spanish stratified lakes. Limnetica 25, 453–478. Campbell, R.G., Sherr, E.B., Ashjian, C.J., Plourde, S., Sherr, B.F., Hill, V., Stockwell, D.A., 2009. Mesozooplankton prey preference and grazing impact in the western Arctic Ocean. Deep-Sea Res. II Top. Stud. Oceanogr. 56, 1274–1289. Chraïbi, V.L.S., Kireta, A.R., Reavie, E.D., Cai, M., Brown, T.N., 2014. A paleolimnological assessment of human impacts on Lake Superior. J. Great Lakes Res. 40, 886–897. Clarke, K.R., 1993. Non-parametric multivariate analyses of changes in community structure. Aust. J. Ecol. 138, 143–157. Craig, S.R., 1987. The distribution and contribution of picoplankton to deep photosynthetic layers in some meromictic lakes. Acta Acad. Aboensis. 47, 55–81. Cullen, J.J., 1982. The deep chlorophyll maximum: comparing vertical profiles of chlorophyll a. Can. J. Fish. Aquat. Sci. 39, 791–803. Cullen, J.J., Eppley, R.W., 1981. Chlorophyll maximum layers of the Southern-California Bight and possible mechanisms of their formation and maintenance. Oceanol. Acta 4, 23–32. Dokulil, M.T., Teubner, K., 2000. Cyanobacterial dominance in lakes. Hydrobiologia 438, 1–12. Fahnenstiel, G.L., Carrick, H.J., 1992. Phototrophic picoplankton in Lakes Huron and Michigan: abundance, distribution, composition, and contribution to biomass and production. Can. J. Fish. Aquat. Sci. 49, 379–388. Fahnenstiel, G.L., Scavia, D., 1987. Dynamics of Lake Michigan phytoplankton: the deep chlorophyll layer. J. Great Lakes Res. 13, 285–295. Fahnenstiel, G.L., Scavia, D., Schelske, C.L., 1984. Nutrient-light interactions in the Lake Michigan subsurface chlorophyll layer. Verh. Internat. Verein. Limnol. 22, 499–508. Fahnenstiel, G.L., Chandler, J.F., Carrick, H.J., Scavia, D., 1989. Photosynthetic characteristics of phytoplankton communities in Lakes Huron and Michigan: PI parameters and end-products. J. Great Lakes Res. 15, 394–407. Falkowski, P.G., Kolber, Z., 1995. Variations in chlorophyll fluorescence yields in phytoplankton in the world oceans. Funct. Plant Biol. 22, 341–355. Fasham, M.J.R., Platt, T., Irwin, B., Jones, K., 1985. Factors affecting the spatial pattern of the deep chlorophyll maximum in the region of the Azores front. Prog. Oceanogr. 14, 129–165. Fee, E.J., 1976. The vertical and seasonal distribution of chlorophyll in lakes of the Experimental Lakes Area, northwestern Ontario: Implications for primary production estimates. Limnol. Oceanogr. 21, 767–783. Fiedler, P.C., 1982. Zooplankton avoidance and reduced grazing responses to Gymnodinium splendens (Dinophyceae) 1. Limnol. Oceanogr. 27, 961–965. Gasol, J.M., Guerrero, R., Pedrós-Alió, C., 1992. Spatial and temporal dynamics of a metalimnetic Cryptomonas peak. J. Plankton Res. 14, 1565–1579. Gervais, F., Siedel, U., Heilmann, B., Weithoff, G., Heisig-Gunkel, G., Nicklisch, A., 2003. Small-scale vertical distribution of phytoplankton, nutrients and sulphide below the oxycline of a mesotrophic lake. J. Plankton Res. 25, 273–278. Jochem, F.J., Zeitzschel, B., 1993. Productivity regime and phytoplankton size structure in the tropical and subtropical North Atlantic in spring 1989. Deep-Sea Res. II Top. Stud. Oceanogr. 40, 495–519. Kasprzak, P., Gervais, F., Adrian, R., Weiler, W., Radke, R., Jäger, I., Riest, S., Siedel, U., Schneider, B., Böhme, M., Eckmann, R., 2000. Trophic characterization, pelagic food web structure and comparison of two mesotrophic lakes in Brandenburg(Germany). Int. Rev. Hydrobiol. 85, 167–189. Kiefer, D.A., Kremer, J.N., 1981. Origins of vertical patterns of phytoplankton and nutrients in the temperate, open ocean: a stratigraphic hypothesis. Deep Sea Res. Part A 28, 1087–1105. Leavitt, P.R., Fritz, S.C., Anderson, N.J., Baker, P.A., Blenckner, T., Bunting, L., Catalan, J., Conley, D.J., Hobbs, W.O., Jeppesen, E., Korhola, A., 2009. Paleolimnological evidence of the effects on lakes of energy and mass transfer from climate and humans. Limnol. Oceanogr. 54, 2330–2348. Lick, W., Lick, J., Ziegler, C.K., 1994. The resuspension and transport of fine-grained sediments in Lake Erie. J. Great Lakes Res. 20, 599–612. Moll, R.A., Stoermer, E.F., 1982. Hypothesis relating trophic status and subsurface chlorophyll maxima of lakes. Arch. Hydrobiol. Suppl. 94, 426–440. Moll, R.A., Brache, M.Z., Peterson, T.P., 1984. Phytoplankton dynamics within the subsurface chlorophyll maximum of Lake Michigan. J. Plankton Res. 6, 751–766. Naselli-Flores, L., Barone, R., 2003. Steady-state assemblages in a Mediterranean hypertrophic reservoir. The role of Microcystis ecomorphological variability in maintaining an apparent equilibrium. Hydrobiologia 502, 133–143.

Pick, F.R., Nalewajko, C., Lean, D.R.S., 1984. The origin of a metalimnetic chrysophyte peak. Limnol. Oceanogr. 29, 125–134. Pilati, A., Wurtsbaugh, W.A., 2003. Importance of zooplankton for the persistence of a deep chlorophyll layer: a limnocorral experiment. Limnol. Oceanogr. 48, 249–260. Planas, M.D., 1973. Composición, ciclo y productividad del fitoplancton del Lago de Banyoles. Oecol. Aquat. 1, 3–106. Planas, M.D., 1990. Factores de control de la distribución espacial y temporal de la producción primaria del fitoplancton del Lago de Banyoles. Sci. Gerundensis 16, 193–204. Putnam, H.D., Olson, T.A., 1966. Primary productivity at a fixed station in western Lake Superior. In Proc. 9th Conf. Great Lakes Res, pp. 119–128. Reavie, E.D., Kireta, A.R., 2015. Centric, Araphid and Eunotioid Diatoms of the Coastal Laurentian Great Lakes. Gebr. Borntraeger Verlagsbuchandlung, Stuttgart. Reavie, E.D., Jicha, T.M., Angradi, T.R., Bolgrien, D.W., Hill, B.H., 2010. Algal assemblages for large river monitoring: comparison among biovolume, absolute and relative abundance metrics. Ecol. Indic. 10, 167–177. Reavie, E.D., Heathcote, A.J., Chraïbi, V.L.S., 2014a. Laurentian Great Lakes phytoplankton and their water quality characteristics, including a diatom-based model for paleoreconstruction of phosphorus. PLoS One 9, e104705. Reavie, E.D., Barbiero, R.P., Allinger, L.E., Warren, G.J., 2014b. Phytoplankton trends in the Great Lakes, 2001–2011. J. Great Lakes Res. 40, 618–639. Reynolds, C.S., Oliver, R.L., Walsby, A.E., 1987. Cyanobacterial dominance: the role of buoyancy regulation in dynamic lake environments. N. Z. J. Mar. Freshw. Res. 21, 379–390. Rühland, K.M., Paterson, A.M., Smol, J.P., 2008. Lake diatom responses to warming: reviewing the evidence. J. Paleolimnol. 54, 1–35. Saros, J.E., Interlandi, S.J., Doyle, S., Michel, T.J., Williamson, C.E., 2005. Are the deep chlorophyll maxima in alpine lakes primarily inducedby nutrient availability, not UV avoidance? Arct. Antarct. Alp. Res. 37, 557–563. Saxton, M.A., Arnold, R.J., Bourbonniere, R.A., McKay, R.M.L., Wilhelm, S.W., 2012. Plasticity of total and intracellular phosphorus quotas in Microcystis aeruginosa cultures and Lake Erie algal assemblages. Front. Microbiol. 3, 1–9. Scavia, D., Fahnenstiel, G.L., 1987. Dynamics of Lake Michigan phytoplankton: mechanisms controlling epilimnetic communities. J. Great Lakes Res. 13, 103–120. Shulenberger, E., Reid, J.L., 1981. The Pacific shallow oxygen maximum, deep chlorophyll maximum, and primary productivity, reconsidered. Deep Sea Res. Part A 28, 901–919. Sommer, U., 1985. Seasonal succession of phytoplankton in Lake Constance. Bioscience 35 (6), 351–357. Twiss, M.R., Ulrich, C., Zastepa, A., Pick, F.R., 2012a. On phytoplankton growth and loss rates in the epilimnion and metalimnion of Lake Ontario in mid-summer. J. Great Lakes Res. 38, 146–153. Twiss, M.R., McKay, R.M.L., Bourbonniere, R.A., Bullerjahn, G.S., Carrick, H.J., Smith, R.E.H., Wilhelm, S.W., 2012b. Diatoms abound in ice-covered Lake Erie: an investigation of offshore winter limnology in Lake Erie over the period 2007 to 2010. J. Great Lakes Res. 38, 18–30. USEPA, 2010. Sampling and analytical procedures for GLNPO's open lake water quality survey of the Great Lakes. United States Environmental Protection Agency, Great Lakes National Program Office. Chicago, Illinois. EPA 905-R-05-001. http://www. epa.gov/glnpo/monitoring/sop/ (Accessed 15 February 2016). Utermöhl, H., 1958. Zur vervollkommnung der quantitativen phytoplankton-methodik. Mitt. Int. Ver. Theor. Angew. Limnol. 9, 1–38. Veldhuis, M.J., Kraay, G.W., Van Bleijswijk, J.D., Baars, M.A., 1997. Seasonal and spatial variability in phytoplankton biomass, productivity and growth in the northwestern Indian Ocean: the southwest and northeast monsoon, 1992–1993. Deep-Sea Res. I 44, 425–449. Venrick, E.L., 1982. Phytoplankton in an oligotrophic ocean: observations and questions. Ecol. Monogr. 52, 129–154. Wagner, C., Adrian, R., 2009. Cyanobacteria dominance: quantifying the effects of climate change. Limnol. Oceanogr. 54, 2460–2468. Walsby, A.E., Hayes, P.K., Boje, R., Stal, L.J., 1997. The selective advantage of buoyancy provided by gas vesicles for planktonic cyanobacteria in the Baltic Sea. New Phytol. 136, 407–417. Watib, N.H.F., Thomson, K.P.B., Elder, F.C., 1975. Sub thermocline biomass concentration detected by transmissometer in Lake Superior. Verh. Internat. Verein. Limnol. 19, 682–688. Watson, N.H.F., Nicholson, H.F., Culp, L.R., 1975. Chlorophyll a and primary production in Lake Superior, May to November 1973. Fish. and Marine Service, Techn. Rep. 525. Great Lakes Biolimnology Laboratory, Canada Center for Inland Waters. White, B., Matsumoto, K., 2012. Causal mechanisms of the deep chlorophyll maximum in Lake Superior: a numerical modeling investigation. J. Great Lakes Res. 38, 504–513. Wolin, J.A., Stoermer, E.F., 2005. Response of a Lake Michigan coastal lake to anthropogenic catchment disturbance. J. Paleolimnol. 33, 73–94. Work, K.A., Havens, K.E., 2003. Zooplankton grazing on bacteria and cyanobacteria in a eutrophic lake. J. Plankton Res. 25, 1301–1306. Wynne, T.T., Stumpf, R.P., Tomlinson, M.C., Fahnenstiel, G.L., Dyble, J., Schwab, D.J., Joshi, S.J., 2013. Evolution of a cyanobacterial bloom forecast system in western Lake Erie: development and initial evaluation. J. Great Lakes Res. 39, 90–99. Yilmaz, A., Ediger, D., Basturk, O., Tugrul, S., 1994. Phytoplankton fluorescence and deep chlorophyll maxima in the Northeastern Mediterranean. Oceanol. Acta 17, 69–77.

Please cite this article as: Bramburger, A.J., Reavie, E.D., A comparison of phytoplankton communities of the deep chlorophyll layers and epilimnia of the Laurentian Great Lakes, J. Great Lakes Res. (2016), http://dx.doi.org/10.1016/j.jglr.2016.07.004