LIGHT INTENSITY INFLUENCES on ALGAL PIGMENTS, PROTEINS and CARBOHYDRATES: IMPLICATIONS for PIGMENT-BASED CHEMOTAXONOMY by Cidya Grant

LIGHT INTENSITY INFLUENCES ON ALGAL PIGMENTS, PROTEINS AND CARBOHYDRATES: IMPLICATIONS FOR PIGMENT-BASED CHEMOTAXONOMY by Cidya Grant

A Dissertation Submitted to the Faculty of

The Charles E. Schmidt College of Science

in Partial Fulfillment of the Requirements for the Degree of

Doctor of Philosophy

Florida Atlantic University

Boca Raton, FL

December 2011

ACKNOWLEDGEMENTS

Special thanks to my research advisor Dr. J. W. Louda, for his guidance and support during this dissertation research. To the members of my dissertation committee:

Drs. J. E. Haky, C. Parkanyi and S. Hagerthey, for answering pertinent questions and steering me on the right path to fulfilling the objectives and goals of this research.

To the FAU-Harbor Branch Oceanographic Institute for NMR sample analyses: special thanks to Dr. Amy Wright for granting permission for instrument use and to her post- doctoral associate Dr. P. Winder for her assistance with experiment set-up.

To the West natural products research group at FAU, particularly Dr. L. West, his post-doctoral associate Dr. P. Gupta and graduate student T. Vansach: thank you for the technical assistance with LC-MS analyses and NMR interpretation.

To my teaching supervisors and mentors at FAU: Drs. D. Chamely-Wiik and E.

Rezler, thank you for always challenging me to reach the highest academic standards, in research and teaching. The encouragement and assistance were all greatly appreciated.

Funding for this material is based in part upon work supported by the National

Science Foundation under Grant no. DGE: 0638662. Any opinions, findings and conclusions or recommendations expressed in this material are those of the author and do not reflect the views of the National Science Foundation.

iii

ABSTRACT

Author: Cidya Grant

Title: Light Intensity Influences on Algal Pigments, Proteins and Carbohydrates: Implications for Pigment-Based Chemotaxonomy

Institution: Florida Atlantic University

Dissertation Advisor: Dr. J. W. Louda

Degree: Doctor of Philosophy

Year: 2011

Phytoplankton Chlorophyll a (CHLa), total protein, colloidal carbohydrates, storage carbohydrates and taxonomic pigment relationships were studied in two cyanophytes (Microcystis aeruginosa and Synnechococcus elongatus), two chlorophytes

(Dunaliella tertiolecta and Scenedesmus quadricauda), one cryptophyte (Rhodomonas salina), two diatoms (Cyclotella meneghiniana and Thalassiosira weissflogii) and one dinophyte (Amphidinium carterae) to assess if algal biomass could be expressed in other indices than just chlorophyll a alone. Protein and carbohydrates are more useful currencies for expressing algal biomass, with respect to energy flow amongst trophic levels. These phytoplankton were grown at low light (LL = 37 µmol photons m-2 s-1),

medium light (ML = 70-75 µmol photons m-2 s-1), and high light (HL= 200 µmol photons m-2 s-1). Even though pigment per cell increased with increasing light intensity,

statistically light had very little effect on the CHL a: taxonomic marker pigment ratios, as

they covaried in the same way. Protein, colloidal carbohydrates and storage

carbohydrates per cell all increased with increasing light intensity, but they did not co-

vary with CHLa. Statistical data showed that light intensity had a more noticeable effect

on protein: CHL a, colloidal carbohydrate: CHLa, storage CHO: CHLa, therefore a

general mathematical expression for these relationships cannot be generated. This study

showed that light intensity does have an influence on these biomass indices, therefore,

seasonal and latitudinal formulas may be required for meaningful algal biomass

estimation. However, more studies are needed if that goal is to be realized.

While studying the effects of light intensity on algal pigment content and

concentration, a new pigment was isolated from a cyanophyte (Scytonema hofmanii) growing between 300-1800 µmol photons·m-2·s-1 and from samples collected in areas of

the Florida Everglades. This pigment was characterized and structurally determined to

possess indolic and phenolic subunits that are characteristic of scytonemin and its

derivatives. In addition, the pigment has a ketamine functionality which gives it its unique polarity and spectral properties. Based on the ultra violet/visible absorbance data,

this pigment was postulated to be protecting the chlorophyll a and cytochrome Soret

bands as well as α and β bands of the cytochromes (e.g. cyt-c562) in the photosynthetic unit.

LIGHT INTENSITY INFLUENCES ON ALGAL PIGMENTS, PROTEINS AND CARBOHYDRATES: IMPLICATIONS ON PIGMENT-BASED CHEMOTAXONOMY

LIST OF TABLES ...... ix

LIST OF FIGURES ...... X I. INTRODUCTION ...... 1 The working hypothesis ...... 4 BACKGROUND ...... 4 Methods for estimating algal biomass ...... 4 Converting CHLa to biomass...... 10 Select algal metabolites which may serve as biomass indices ...... 17 Photosynthesis overview ...... 23 Novel sunscreen pigment ...... 30 Overall goals of this study ...... 33 II. MATERIALS AND METHODS ...... 34 Experimental organisms...... 34 Algal culturing ...... 36 Culture conditions ...... 37 Cell counting...... 38 Chemical Analyses...... 39 Algal protein extraction ...... 39 Algal protein measurement ...... 39 Algal colloidal and storage carbohydrate extraction ...... 40 Algal colloidal and storage carbohydrate measurement ...... 40 Algal total organic carbon (TOC) extraction ...... 41 Colorimetric determination of extracted TOC samples ...... 42

Nutrient analyses ...... 42 Pigment Analyses...... 43 Ultra Violet - Visible (UV/Vis) Analyses of Extracts ...... 45 High Performance Liquid Chromatography (HPLC) ...... 46 HPLC Data Calculations ...... 47 Statistical analyses ...... 49 Isolation and characterization of a new pigment...... 50 IR analysis ...... 52 Mass Spectrometry ...... 52 NMR analyses ...... 54 Acetylation reactions ...... 54 Deuterium exchange reactions ...... 55 III. RESULTS - STATISTICAL ANALYSES ...... 56 Significance of the algal species used in this study ...... 56 Analyses overview ...... 59 Synechococcus elongatus ...... 60 Microcystis aeruginosa ...... 70 Dunaliella tertiolecta ...... 78 Scenedesmus quadricauda ...... 87 Rhodomonas salina ...... 95 Cyclotella meneghiniana ...... 103 Thalassiosira Weissflogii ...... 111 Amphidinium carterae ...... 119 IV. DISCUSSION ...... 127 Growth patterns ...... 127 Phytoplankton protein as a biomass indicator ...... 128 Phytoplankton colloidal carbohydrate (CHO) as a biomass indicator ...... 137 Phytoplankton storage carbohydrate (CHO) as a biomass indicator ...... 139 Marker pigments as indicators of algal biomass ...... 140 Phytoplankton chlorophyll a, protein and carbohydrate relationships to biovolume . 145 vii

V. CONCLUSION: IMPLICATIONS FOR CHEMOTAXONOMY ...... 147 VI. CHARACTERIZATION OF NOVEL PIGMENT ...... 149 The ‘scytoneman’ skeleton ...... 149 New pigment – putative structure elucidation ...... 155 Mass interpretation ...... 159 IR analysis ...... 166 Ecological significance of the new pigment ...... 167 VII. APPENDICES ...... 170 I- Pigment calculation and data handling ...... 171 II. Select photoprotectorant and accessory pigments ...... 179 III- Spectroradiometric output ...... 181 IV- Calibration curves and equations ...... 185 V- Retention times and UV-Vis maximas ...... 187 VI-ANOVA tables ...... 190 VII- Cellular concentration of CHLa and photosynthates ...... 264 VIII- Typical chromatograms of species studied ...... 267 IX- Specific growth rate (µ) curves ...... 271 X – NMR SPECTRA ...... 275 XI – Mass Spectra ...... 284 VIII. REFERENCES...... 295

viii

LIST OF TABLES

Table 1: Methods for estimating algal biomass ...... 5 Table 2: Marker pigments having stoichiometric relationships with CHLa in biomass estimations ...... 12 Table 3: Colloidal and storage carbohydrate composition of the taxonomic groups studied...... 20 Table 4: Gradient program used in FAU OGG laboratory ...... 47 Table 5: Gradient program used for LC-MS runs ...... 53 Table 6: Cellular concentration of chlorophyll a and products of photosynthesis ...... 130

Table 7: Protein:CHLa (log10) ratios of the species as influenced by irradiance ...... 134

Table 8: Colloidal CHO/CHLa (log 10) ratios as a function of irradiance ...... 138

Table 9: Storage CHO/CHLa (log 10) ratios as a function of irradiance ...... 140 Table 10:Three new pigments isolated form Scytonema sp...... 152 Table 11: Scytonemin- a comparison of literature and observed values ...... 153 Table 12: 1H and 13C NMR data for putative structure of pigment ...... 157

LIST OF FIGURES

Figure 1: Structure of Chlorophyll a ...... 2 Figure 2:. Xanthophyll cycling in (a) Chrysophytes and (b) Chlorophytes...... 26 Figure 3: Structure and UV/Vis spectra of Scytonemin ...... 31 Figure 4: New pigments isolated form Scytonema sp...... 32 Figure 5: Flow Chart of Analytical Scheme...... 35 Figure 6: Schematic of inoculation procedure...... 36 Figure 7: Synechococcus elongatus Marker pigment/CHLa ...... 62 Figure 8: Synechococcus elongatus Protein/CHLa relationships ...... 64 Figure 9: Synechococcus elongatus Colloidal CHO/CHLa ...... 66 Figure 10: Synechococcus elongatus Storage CHO/CHLa ...... 69 Figure 11: Microcystis aeruginosa Markerpigment/CHLa ...... 71 Figure 12: Microcystis aeruginosa Protein/CHLa relationships ...... 73 Figure 13: Microcystis aeruginosa Colloidal CHO/CHLa ...... 75 Figure 14: Microcystis aeruginosa Storage CHO/CHLa...... 77 Figure 15: Dunaliella tertiolecta Marker pigment/CHLa ...... 79 Figure 16: Dunaliella tertiolecta Protein/CHLa relationships ...... 81 Figure 17: Dunaliella tertiolecta Colloial CHO/CHLa ...... 84 Figure 18: Dunaliella tertiolecta Storage CHO/CHLa ...... 86 Figure 19: Scenedesmus quadricauda Marker pigment/CHLa ...... 88 Figure 20: Scenedesmus quadricauda Protein/CHLa relationships ...... 89 Figure 21: Scenedesmus quadricauda Colloidal CHO/CHLa ...... 92 Figure 22: Scenedesmus quadricauda Storage CHO/CHLa ...... 94 Figure 23: Rhodomonas salina Marker pigment/CHLa ...... 96 Figure 24: Rhodomonas salina Protein/CHLa ...... 98 Figure 25: Rhodomonas salina Colloidal CHO/CHLa ...... 100 x

Figure 26: Rhodomonas salina Storage CHO/CHLa ...... 102 Figure 27: Cyclotella meneghiniana Marker pigment/CHLa ...... 104 Figure 28: Cyclotella meneghiniana Protein/CHLa relationships ...... 105 Figure 29: Cyclotella meneghiniana Colloidal CHO/CHLa ...... 108 Figure 30: Cyclotella meneghiniana StorageCHO/CHLa ...... 110 Figure 31: Thalassiosira weissflogii Marker pigment/CHLa ...... 112 Figure 32: Thalassiorira weissflogii : Protein /CHLa relationships ...... 113 Figure 33: Thalassiosira weissflogii Colloidal CHO/CHLa ...... 116 Figure 34: Thalassiosira weissflogii Storage CHO/CHLa ...... 118 Figure 35: Amphidinium carterae Marker pigment/CHLa ...... 120 Figure 36: Amphidinium carterae Protein/CHLa relationships ...... 121 Figure 37: Amphidinium carterae Colloidal CHO/CHLa ...... 123 Figure 38: Amphidinium carterae Storage CHO/CHLa ...... 125 Figure 39: The scytoneman skeleton ...... 150 Figure 40: Red Rock aerial - areas where samples, scraped off rocks, contain the visible light sunscreen pigment...... 151 Figure 41 HPLC of observed scytonemin and new pigment ...... 154 Figure 42: UV/VIS absorption spectra of the new pigment ...... 154 Figure 43: Scytonemin oxidized and new pigment – overlay ...... 155 Figure 44: Molecular structure of new pigment from 1H, HSQC, HMBC ...... 156 Figure 45: HR ESI-TOF MS of new pigment ...... 160 Figure 46: Initial dissociation of m/z 602 [M+H] + ion ...... 161 Figure 47: (+) ESI- MS/MS dissociation...... 162 Figure 48: Fragmentation patterns of new pigment ...... 163 Figure 49: HPLC/UV...... 164 Figure 50: Mass analysis of pigment after acetylation ...... 165 Figure 51: IR spectra of new pigment ...... 167 Figure 52: New pigment and chlorophyll a – overlay spectra ...... 168 Figure 53: Excitation, emission spectral overlay of new pigment ...... 168

I. INTRODUCTION

Phytoplankton constitute approximately 40-60 % of the primary production of the world’s aquatic environments (Antoine et al., 1996; Falkowski, 1994). During the last few decades, phytoplankton have been monitored in an increasing number of marine environments, where they have been and are used as indicators of environmental (Pybus,

1996; Edward’s et al., 2002 Boyce et al., 2010) and climatic (Moline and Prezelin 1996;

Hallegraeff, 2010; Marinov et al., 2010) changes. These changes include early warning signals of potentially harmful species becoming dominant in a population as well as the onset of algal blooms. Assessing and monitoring the biomass of different algal communities is therefore necessary as rapid indicators of ‘true’ biomass are currently not available. Thus, facile methods are needed for estimating phytoplankton biomass (algal biological material per unit area and/or volume).

The direct measurement of phytoplankton organic matter is not normally possible with rapid sampling to data turnaround times and biomass is therefore estimated using alternate methods. Cell counting, assessment of cellular biovolume by microscopy and determination of chlorophyll a (CHLa) concentrations are among the most commonly used methods (Sournia, 1978). The conversion of phytoplankton CHLa to cell number and/or biovolume has been reasonably done by linear regression (Gieskes et al., 1998;

Schulter and Havskum, 1997), simultaneous equations (Andersen et al., 1996; Tester et al., 1995; Letelier et al., 1993), advanced algorithms (Mackey et al., 1996; Mackey et al.,

1998; Higgins and Mackey, 2000) and Bayesian/MCMC estimation (Van den Meersche et al., 2008). Thus, routine ‘biomass’ assessment of phytoplankton communities are made using CHLa (structure shown in Figure 1) as biomass proxy. However, what does that equal in metabolizable biomass?

I II N N

Mg H N IV N III

V O H H O

COOCH3 O

H C 3 H H C 3 H

Figure 1: Structure of Chlorophyll a

The ecological significance of phytoplankton (algae) lies in the fact that they trap and convert to organic matter almost all of the energy used in the pelagic ecosystem. This study examined how changes in light intensity affected the relationships between CHLa to the following: taxonomically significant pigments, protein, two functional groups of carbohydrate (colloidal and storage) and organic carbon. These were measured in relation to cell numbers and biovolume as a way to estimate ‘true’ biomass in the context of the metabolites involved in food chains and energy flow in aquatic ecosystems. The resultant

data and conclusions should have implications and applications in algal biomass

estimations from CHLa, be they by spectrophotometry (Johnsen and Sakshaug, 1993),

fluorescence (Wilhelm et al., 1991), or HPLC (Millie et al., 1993; Wright et al., 1996)

analyses.

Additionally , besides total algal biomass from CHLa estimations, pigment-based chemotaxonomy using extracted pigments (see chapters in Jeffrey et al., 1997) and even by advanced spectral algorithms with satellite and airborne telemetry (O’Reilly et al.,

1998; 2001) should then yield not only cell number estimations from taxon-specific

CHLa estimations but biomass estimations including protein, carbohydrate and total organic carbon values. These results will have application in limnology, oceanography as well as pure algal cultures. The latter is important when dealing with bulk culture aimed at biomass conversion as fuel feedstock (Gouveia and Oliveira, 2009). The question here then becomes, is there a way to relate CHLa to the ‘food’ available in a particular aquatic ecosystem? Then will an ecological modeler be able to predict if a certain plankton group could support/provide food for the organisms in the trophic level?

Light has never been adequately factored into the CHLa based phytoplankton biomass estimations. That is, with different seasons there are different intensities of light seasonally, latitudinally and other parameters such that overall light conditions are not the same. In order to get a real understanding for this type of modeling, one will need large

numbers of species, enough light studies, enough temperature studies et cetera, such that so potential summer formulas, a winter onset formula, an early spring formula, et cetera can be generated.

The working hypothesis for this study would therefore be as follows:

Chlorophyll a – per se is not the ultimate descriptor of phytoplankton biomass. That is, large variations in ‘true’ biomass, defined here as metabolizable organic matter (proteins, carbohydrates, lipids) exist between phytoplankton groups (taxa) and within each taxon

by variations in light and/or nutrient availability. The null hypothesis would be:

Chlorophyll a alone perfectly describes phytoplankton biomass.

In the present study, correlations between CHLa and biomass parameters (protein,

carbohydrates, cell number, biovolume) under the influence of light intensity were

investigated in order to ascertain if that they can be used for ‘true’ biomass estimation.

This chapter will introduce methods used to determine algal biomass, methods for

measuring CHLa, mathematical methods of converting CHLa to biomass, marker

pigments used for estimating CHLa content of different taxonomic groups, significance

of other biomass parameters, a brief look at photosynthesis and roles of pigments in

photosynthesis. Additionally the author introduces a second project which involves the structure elucidation of a potential visible light sunscreen pigment isolated from a cyanobacteria (Scytonema hoffmanii), grown at high light conditions in the laboratory and

from samples collected in the Florida Everglades.

BACKGROUND Methods for estimating algal biomass:

Since phytoplankton carbon content in natural environments often has interference from detritus and other organic materials, alternate methods for biomass

‘estimation’ have been developed, as given in Table 1.

Microscopy: Microscopic assessment may be used to determine algal cell number and biovolume (Stevenson et al., 1985). Cells are mounted on microscope slides, which may be prepared in counting chambers, inverted microscopes or in different types of

media on regular microscopes (Palmer, 1962).

Table 1: Methods for estimating algal biomass – Adapted from Stevenson et al., 1985 Measurement Detail: advantages, disadvantages

Cell density Microscope generally required; good indicators of algal species composition, biovolume and biomass if size and mass of all cells assumed to be the same; variation in cell size lead to biomass error. Flow cytometry may also be used for cell density determinations. Biovolume Microscope used to accurately assess algal biomass; most time consuming; error due to cell vacuoles have to be accounted for. C,N,P Several analytical methods exist; may be used for assessing nutrient status of cells; for field samples, biomass of living and non-living matter included. Dry Mass Involves inexpensive gravimetry; biases arise when inorganic matter and non-algal organic matter are present. Ash free dry Simple laboratory heating and gravimetric procedures; field samples mass (AFDM) include living and detrital matter. Chlorophyll a Ubiquitous to all photosynthetic algae; several analytical methods exits; light and nutrient adaptations may bias biomass estimates.

Microscopy gives detailed information on the composition and diversity of microalgal

assemblages, to the species level, but due to the high level of expertise required, it can be

tedious and costly (Millie et al., 1993).

Algal cell volume measurement via microscopy is a relatively good indicator of

algal biomass if most of the algae of each species are of similar size and the mass of all cells is assumed to be the same (Smayda, 1978). The method involves measuring cell dimensions using an ocular micrometer and geometric formulae for calculating cell volume. Algal biovolume measurements will give very good estimates of algal biomass if

the assumption is made that the mass of algal cytoplasm is the same among each taxon.

Biovolume can correct cell density estimates of biomass by accounting for size

differences among species. When the vacuole size is estimated and subtracted from cell

volume to estimate cytoplasm volume, then algal biovolume becomes a relatively

accurate measure of algal biomass, especially when large variations in cell size exists in a

community.

However, species volume measurement is laborious and highly dependent on the

skills of the researchers. Samples usually have to be fixed in a solution or preserved until

microscopic analysis. The nature and concentration of this fixative has been shown to

alter cellular volume (Montagnes et al., 1994) and there is an increased level of uncertainty due to the small size of the organisms being analyzed. The development and use of such techniques as epifluorescence microscopy (Daley and Hobbie, 1975), electron

microscopy (Johnson and Sieburth, 1982), flow cytometry (Olson et al., 1985) and

immunofluorescence (Shapiro et al., 1989) have vastly improved the study of

phytoplankton. However, these methods are still very time consuming and do not allow

for the rapid spatial and temporal monitoring of phytoplankton.

Carbon, Nitrogen and Phosphorus: Particulate organic matter (POM) is an

important component in an ecosystem. POM includes particulate organic carbon (POC),

particulate organic nitrogen (PON), and particulate organic phosphorus (POP), among

other detritus. POM provides a primary food source for aquatic food webs. Dissolved

organic matter (DOM) on the other hand, consists of organic matter which is not in cells

per se and passes through a 0.45 μm filter. However a cut off of 0.22 μm is becoming

standard as well. DOM can contribute to the acidity of a water body and can increase 6

light attenuation, thus detrimentally affecting phototrophic organisms in an aquatic

environment (Eby, 2004; Hansell and Carlson, 2001).

POC and PON are generally determined by high temperature dry combustion using a carbon, hydrogen, nitrogen (CHN) analyzer. POP is often analyzed by wet

chemical oxidation using potassium peroxydopersulfate (Menzel and Corwin, 1967).

With advances in instrumentation, POP, PON and POC determination can now be done

simultaneously from the same filter (Raimbault, 1999). The procedure generally involves

collection of particulate matter sample by filtration; placement of the filters in digestion

flasks; elimination of inorganic carbon by acidification and bubbling; digestion in an

autoclave; automated analysis of C, N, and P species. The resulting data will typically be

biased by other living (e.g. bacteria and zooplankton) or non-living matter.

Dry mass and Ash-free dry mass: The dry mass (DM) or ash free dry mass

(AFDM) can then be determined with the use of glass fiber filters (GF/F), an analytical

balance, drying oven and/or muffle furnace (APHA, 1995). For measuring DM, the

sample is filtered on to a pre-weighed filter then reweighed to get the difference in wet

weight. The sample and filter are then dried overnight at low temperature (60ºC) then reweighed to get the dry weight by difference (U.S. EPA, 1995a). Samples for AFDM are filtered and frozen, dried (100 ºC) for twenty four hours, weighed (DM), ashed (450

ºC) for four hours (to remove all organic carbon), rehydrated with water, dried for twenty four hours and re-weighed (U.S. EPA, 1995a). These methods are relatively inexpensive, but DM and AFDM estimates of algal biomass may both be biased by inorganic and non- algal organic matter (detritus, bacteria, fungi, etc.) present in the sample (Steinman and

Lamberti, 1996). Therefore, these methods are poor indicators of algal biomass, 7

especially when inorganic deposition and other organisms constitute a significant portion of the sampling community. In the case of samples with a significant (5%) proportion of

the DM as carbonate, a decalcification step needs to be included. That is, treatment of the

sample with hydrochloric acid (HCl), done best in the gas phase, washing out salts, re-

drying to constant decalcified DM and then proceeding to AFDW (APHA, 1995).

Chlorophyll a (CHLa) measurement: Chlorophylls are the molecules in

photosynthetic bacteria and plants that capture light energy for carbon fixation and the

splitting of H2A (H2O or H2S). Chlorophyll a is the most widely used estimator of algal

biomass because it is relatively unaffected by non-algal substances. It is assumed to be

and accepted as a fairly accurate measure of algal ‘biomass’ (weight and volume) and can

serve to indicate interactions between nutrient concentration and a number of biological

phenomena in lakes and rivers (Berkman and Canova, 2007). Several methods exist for

measuring chlorophylls, as detailed below.

Spectrophotometry: Development of spectrophotometric analyses of chlorophyll

pigments began in the 1930’s – 1940’s (Weber et al., 1986). A trichromatic technique

was later introduced (Richards and Thompson 1952) for measuring chlorophylls a,b,c

and attempted to remove overlapping absorbance by the other chlorophylls at the

absorption maximum for each chlorophyll. Several modifications have been made to

these equations over the past decades and each claim to produce better estimates of the

chlorophylls (Jeffrey and Humphrey, 1975; UNESCO, 1966). However, when these

equations are compared with the concentration of the ‘alternate’ chlorophylls (-b, -c)

obtained via physical separation techniques (e.g. chromatography), the degree of

correspondence is low (Louda and Mongkhonsri, 2004). The trichromatic chlorophyll a, 8

for instance, is chlorophyll a, minus the interference wavelengths from the other

chlorophylls, but includes all of the degradation products of chlorophyll a, which share the primary absorbance maxima.

Fluorometry: Chlorophyll molecules fluoresce in the red region of the electromagnetic spectrum when exposed to blue light. Fluorometry is a highly sensitive method which has its own multi-chromatic fluorescence equations (Loftus and Carpenter,

1971). Even though fluorometry is more sensitive than spectrophotometry, it is not

typically recommended for routine work (Aminot, 2000). There are no independent

fluorometric chlorophyll attenuation coefficients, and each individual fluorometer must

be calibrated daily against spectrophotometric standards (Standard Methods, APHA,

1991).

High Performance Liquid Chromatography: This is an analytical method that

makes it possible to gain information about the community composition of phytoplankton

(Ansotegui et al., 2001; Gieskes and Kraay, 1998b), as well as correct chlorophyll a data.

The analysis of algal pigments using high performance liquid chromatography (HPLC)

allows the separation, identification and quantitation of taxon- specific, diagnostic

marker- pigments, in addition to the chlorophylls and their breakdown products (Millie et

al., 1993). In contrast to microscopic enumerations, analysis by HPLC is reproducible

and the method allows for rapid examination of phytoplankton composition. When an

autosampler is connected to the HPLC system, more than 40 samples can be analysed per

day. HPLC can therefore provide faster examination of the spatio-temporal dynamics of

phytoplankton populations than has been possible using enumeration of phytoplankton

under the microscope (Schulter et al., 2000). 9

Converting CHLa to biomass:

Pigment- based chemotaxonomy is a viable method for studying the assemblage of phytoplankton communities (Louda, 2008). The chemotaxonomic analysis of phytoplankton communities using marker pigments allows the calculation of the relative abundance of distinct algal taxa or groups (Millie et al., 1993; Jeffrey et al., 1997). A marker pigment is one that is found only in certain groups (taxa) of algae and has a distinctive relationship to that group. The major algal groups include: chlorophytes, prochlorophytes, cyanophytes, cryptophytes, dinophytes and diatoms. There is some need for caution with pigment-based chemotaxonomy, as some marker pigments are shared between several algal groups (Rowan, 1989), so the conversion to biomass estimates is not always straightforward.

Pigment per cell may change for a number of reasons including light intensity

(Grant and Louda, 2010), growth rate and nutritional state (Llewellyn and Gibb 2000).

However, it has been shown that the concentrations of chlorophylls and specific carotenoids in certain but not all taxa vary in a similar way (Goericke and Montoya,

1998), so ratios between them do not change very drastically. Therefore to estimate the contribution of each algal group to the total population, most methods use ratios of CHLa to a marker pigment, or the inverse, for that group (Gieskes et al., 1988, Mackey et al.,

1996, Wright et al., 1996). Examples of these marker pigments include: chlorophyll b for chlorophytes, echinenone for filamentous cyanophytes, zeaxanthin for coccoidal cyanophytes, alloxanthin for cryptophytes, peridinin for peridinin containing dinophytes and fucoxanthin for diatoms and other chrysophytes. Table 2 contains the structures of these pigments. The implementation of reversed phase high performance liquid 10

chromatography (HPLC: Mantoura and Llewelyn, 1983), combined with real time spectrophotometric detection (photodiode array – PDA, a.k.a. diode array detector or

DAD), along with improvements in pigment extraction procedures (Hagerthey et al.,

2006) has become central to the development of pigment-based chemotaxonomy. The basic method uses preliminary estimates for pigment ratios and then refines these values iteratively using measured chlorophyll a as a criterion, followed by calculation of the contributions of different groups of microalgae to total chlorophyll a from the optimized pigment ratios.

Experiments with cultures remain central to the understanding of microalgal responses to environmental variability and, as such, studies have been carried out with lab grown cultures to determine more reliable pigment ratios for use in the mathematical applications for determining algal biomass (Gieskes et al., 1998; Grant and Louda 2010; references in Jeffrey et al, 1997). The only disadvantage with the mathematical methods, and currently no solution exists, is that it has to be assumed that the ratios in the calculations reflect the same physiological state of the population being studied (Mackey et al., 1996). Thus, in addition to taxonomic structure, the physiological state of the community also needs to be addressed.

For biomass estimations utilizing ratios of CHLa to a marker pigment, these ratios are used in linear regression equations (Gieskes et al., 1998), simultaneous equations

(Tester at al., 1995), factor analysis and iterative methods (Mackey at al., 1996). Early linear regression equations (Gieskes et al., 1988) did not distribute CHLa evenly among algal groups and did not distinguish those with shared marker pigments, so other methods

Table 2: Marker pigments having stoichiometric relationships with CHLa in biomass estimations Taxa Marker pigment Chlorophytes

Chlorophyll b Cryptophytes

Alloxanthin Cyanophytes

Zeaxanthin – coccoidal cyanophytes

Echinenone – filamentous cyanophytes

Dinophytes

Peridinin – peridinin containing dinoflagellates Diatoms

Fucoxanthin

have been applied to determine the abundance of individual phytoplankton groups from the concentration of marker pigments (Mackey et al., 1996; Van den Meersche et al.,

2008).

All of these methods rely on some knowledge of ratios of CHLa to marker

pigment. However, when these ratios are applied to the mathematical models, the light intensity, depth of sampling, irradiance at depth, growth condition and nutritional state are rarely considered. Extensive knowledge of the influence of light intensity, light quality, and nutritional state on the CHLa to pigment ratios of different phytoplankton species is therefore needed and certain inroads have been made (Grant and Louda, 2010 and references therein).

In the present study, algal cultures grown in nutrient replete media will be used to better determine the influence of varying light intensities on the ratios of CHLa to pigment with the aim of producing more robust and reliable ratios for use in the mathematical applications and models for estimating algal biomass. Algal species were

selected based on their relevance with ecology and biogeochemical contexts. Working

with lab grown cultures is critical for understanding how microalgae respond to different

environmental variables (MacIntyre and Cullen, 2005).

Three mathematical methods currently being used in pigment-based chemotaxonomy

For the assessment of algal class abundance and community structure, some type

of mathematical relationship is pre-determined, refined and then used to describe real

systems. Here, three mathematical approaches: simultaneous linear equations (SLE),

CHEMTAX and Bayesian Compositional Estimator (BCE) will be briefly reviewed. In a separate study done at FAU in the Environmental Geochemistry Laboratory (see J. L.

Brown thesis Florida Atlantic University, 2010), these three methods were compared to see which, if any, could accurately enumerate the periphyton (phototrophic group of algae and cyanobacteria living attached to aquatic vegetation and sediments, often forming microbial mats) community composition. The methods were applied to artificial data sets, mixed lab cultures of known composition and Florida Everglades periphyton samples. All three methods gave somewhat accurate sample compositions for artificial and mixed lab cultures. SLE and CHEMTAX performed better than BCE.

Simultaneous Linear Equation (SLE): These models apply a series of straightforward equations to the problem: using one ratio of one biomarker to CHLa for each algal class in the sample. This method only has one possible answer. If the ratios and the algal classes used are accurate and complete for the sample, the answer will be correct. SLE is the model currently in use at the FAU Environmental Biogeochemistry

Laboratory and shows promising results when applied to marine phytoplankton in Florida

Bay (Louda 2008). Equation 1 shows the SLE used in our laboratory and evolved using data collected from cultures grown in our laboratory as well as samples collected from

Lake Okeechobee and Florida Bay. The refined ratios in the SLE given here reflect the influence that irradiance has on pigment concentration per cell (Grant and Louda 2010).

∑ CHLa = ((1.1 x ZEA) + (11 x ECHIN)) + (2.5 x CHLb) + (1.2 x FUCO) + (1.5 x PER)

Equation 1: Current SLE, used in our laboratory 14

SLE is sometimes called the fixed coefficient method since, unlike numerical

methods (such as CHEMTAX and BCE) which change coefficients within certain

parameters, the SLE coefficients cannot be altered during the calculations. Each coefficient is the estimated ratio of a biomarker pigment to CHLa, which is considered typical in a given class. SLE has no provision for shared pigments, so only unique or nearly unique biomarkers can be used. The amount of each biomarker pigment in the sample is multiplied by its respective coefficient to determine the estimated amount of

CHLa contributed by each algal class (taxon specific CHLa). Each of the estimated class contributions may then be divided by the sum of all estimated contributions to arrive at an algal class composition in percent form. An example of how algal species pigment calculation is done in our lab is shown in Appendix I.

CHEMTAX: This is a factor analysis program developed in 1996 (Mackey et al.,

1996) and licensed through CSIRO Marine Laboratories. It is written to run inside

MATLAB (The MathWorks, Inc. 2008). CHEMTAX works by evaluating groups of samples, with pigment data arranged in matrix form. Biomarker ratios to CHLa are also arranged in a matrix. Using (unknown) algal class composition of the samples as a third matrix, this forms a linear inverse problem which is solved by matrix factorization, using a straightforward algorithm to provide the least-squares solution (Mackey et al., 1996). In contrast to SLE, which takes only sample data and biomarker ratios as input, CHEMTAX allows input as to how much and which ratios are allowed to vary and how the data are weighted. Results from CHEMTAX include algal class composition of each sample,

revised ratio matrix, residuals (from the least-squares calculations), and breakdowns of pigments assigned to each algal class within each sample, as well as information regarding the iterative calculation process.

The Bayesian Compositional Estimator (BCE): This is a chemotaxonomic program, developed in 2007 by researchers at the Netherlands Institute of Ecology (Van den Meersche et al., 2008). BCE is implemented as a package (Van den Meersche and

Soetaert 2009) in the open source software R (R Development Core Team 2009). The

BCE program was designed in part to specifically address certain shortcomings in

CHEMTAX (Van den Meersche et al., 2008). Like CHEMTAX, BCE uses a ratio matrix and an unknown sample composition matrix to compose a linear inverse problem. BCE, however, uses Bayesian methods to fit a probability distribution to the data and find a maximum likelihood solution for the problem. BCE first finds a least-squares solution and uses it as a starting point for a Markov Chain Monte Carlo (MCMC) simulation. The program provides a number of diagnostic outputs in order to check the performance of the simulation, and this output must be inspected prior to acceptance of any results. This output includes the number of runs as well as plots which indicate the extent and randomness of the sampling of the solution space. Final results of the program include the algal class composition of each sample, a revised ratio matrix, standard deviations and covariance matrices for the ratio and class compositions (see Brown, 2010 for further information).

Select algal metabolites which may serve as biomass indices

Chlorophyll a is used as an index of biomass because it is unique to oxygenic photosynthetic organisms. However, changes in algal physiology are not confined to

CHLa: pigment ratios, but is also reflected in other indices of biomass such as proteins, carbohydrates and organic carbon. The determination of the protein and carbohydrate content of microalgae may provide important information for phytoplankton biomass assessment, which can in turn be used to investigate protein and carbohydrate dependent physiological processes in cells as well as with studies of nutritional value of phytoplankton (Clayton et al., 1988).

Algal carbohydrates: Carbohydrates are the major products of photosynthesis and are represented by polysaccharides and storage structure compounds - including cellulose, hemicellulose and pectin found in plants (Aspinall, 1983), as well as laminaran

and starch found in some algae and cyanobacteria (Stewart, 1974).

Carbohydrates play important roles in biogeochemical cycles in the water column

and water sediment interface (Hedges et al., 1994), in cellular metabolism and structure

(Granum and Myklestad, 2001; Handa, 1969), and are major storage compounds in

autotrophic organisms. Carbohydrates, particularly polysaccharides, contribute significantly to the organic matter of diatoms, green algae and cyanobacteria (Fernandez et al., 1992).

The two major groups of carbohydrates in microalgae are extracellular, loosely bound colloidal carbohydrates and intracellular storage polysaccharides (glucans and

starch). Several groups of microalgae have been shown to secrete copious amounts of

carbohydrates (Geesey, 1982). These secretions, though composed of some small sugar

moieties, are largely polysaccharide in composition and are thought to be involved in the

transfer of nutrients in lower food webs (Decho, 1990). These products of photosynthesis

can be excreted within a few hours of formation and are thought to be light dependent

(Underwood et al., 2004). Studies have shown that the concentration of secreted, loosely

bound carbohydrates in sediments is closely related to the biomass of diatoms

(Underwood et al., 1995). Colloidal carbohydrate fractions have been shown to contain mucopolysaccharides, extracellular polymeric substances (EPS), transparent exopolymers (TEP), and others, each with its own function (Thornton, 2002). However, these secretions have largely been ignored in studies regarding microalgal production and trophic energy transfer.

In the water column, EPS is now being observed within phytoplankton bloom sedimentation and other types of marine snow (Riemann, 1989; Alldredge et al., 1986).

Epipelic diatoms secrete mucopolysaccharides to facilitate movement. These secretions then represent sources of food for bacteria and invertebrates (Decho 1990, Goto et al.

2001). Mucopolysaccharides have been shown to be the main component of the mucous

matrix of algal colonies (Aldercamp et al., 2006). In addition to function in the mucous

matrix of diatoms, mucopolysaccharides have also been reported to serve as storage

polysaccharides (Lancelot & Mathot 1985). Few studies have been done to investigate

the influence of light on mucopolysaccharide production in phytoplankton. Studies

carried out on Cyanospira capsulate and Synechococcus strains grown under various

light/dark cycles showed that both produced smaller amounts of mucopolysaccharides in 18

comparison to control cultures grown under continuous light (Philips et al., 1989, De

Philippis et al., 1995). Decreased production of mucopolysaccharides equate to shorter light periods. Therefore it could be concluded that the synthesis and release of these polysaccharides is light dependent (Philips et al., 1989).

The storage polysaccharide in Phaeophyceae (macroalgae) is a β-1, 3-glucan and

has been classified as laminaran (Meeuse 1962). Laminarins are primarily composed of

D-glucose residues (Peat et al., 1958). Chrysolaminarin is also a β-(1, 3) glucan and is

also a major storage product of Chrysophyceae and diatoms, it is produced in the light

and consumed in the dark (Janse et al., 1996b). The quantity of storage glucans in algal

cells and thus their contribution to algal biomass largely depend on nutrient status, light

intensity and growth condition of the cells. For example, it has been observed in the

diatom Chaeotoceros that cellular glucan content accumulated markedly under nutrient

deficiency (Myklestad 1974). For Phaeocystis, an increase in the ratio of total

carbohydrate to total carbon was observed in nutrient limited batch cultures and at the

end of a spring bloom (Fernandez et al. 1992, Van Rijssel et al., 2000). Generally, algal

carbohydrate content changes in response to light variations over a diel cycle, and when

nutrients and irradiance are sufficient to sustain high photosynthetic rates (higher than

metabolic demands). Thus, the excess photosynthates are stored when photosynthetic

production (P) exceeds respiration (R) utilization (P>R). During this time, glucan

accumulates during the day and in the night it can be respired as an energy supply to

maintain cell metabolism and provide carbon and energy for protein synthesis (Lancelot

& Mathot 1985, Granum et al. 2002). Since environmental factors influence the

accumulation or degradation of storage polysaccharides (Heldt, et al, 1977), knowing 19

how storage carbohydrate levels are controlled will have potential impacts on the

estimation of the world’s primary productivity. Table 3 is a simple summary of the types

of carbohydrates typical of the two functional groups that are being investigated in this

study.

Table 3: Colloidal and storage carbohydrate composition of the taxonomic groups studied. Adapted from Stewart, (1974) Taxonomic Group Colloidal Storage carbohydrate carbohydrate Chlorophytes Polysaccharides; simple Starch (Meeuse, 1962) sugars (Hough et al., 1952; Fogg, 1952) Cyanophytes Polysaccharides; simple Glucans (Richardson et al., sugars (Lewin, 1956; 1968) Moore and Tisher, 1964) Diatoms Polysaccharides Chrysolaminarin (Meeuse, (Lewin, 1955) 1962) Dinophytes Polysaccharides Starch (Bursa, 1968) (McLaughlin et al,1960) Rhodophytes Polysaccharides Starch (Archibald et al., 1960) (Sieburth, 1969)

The phenol-sulfuric assay (Dubois et al., 1956) is a commonly used method for assessing algal carbohydrates (intra-cellular or secreted). The calorimetric method is

sensitive to a wide range of carbohydrates, including sugars, methylated sugars and both neutral and acidic polysaccharides. This study will only take into consideration the colloidal fraction in a broad sense along with storage carbohydrate fractions, as too many extraction techniques for the colloidal fraction components are currently being used to make reasonable comparisons.

Algal proteins: Proteins are essential, biomolecular components of cells and have the following roles: regulating metabolic activities, providing structural support and also 20

are the pre-cursors as well as end-products of macromolecular synthesis and catabolism

(Clayton et al., 1988). Therefore, knowledge of the quantity of total protein present is important for understanding a broad range of biological processes in phytoplankton cells.

Although bulk proteins of algae are not expected to differ much in their overall proportions of amino acids (Stewart, 1967), some algal cell walls contain appreciable

proportions of proteins, which may prove to be of taxonomic value (Thompson and

Preston, 1967). Several studies have been carried out to investigate different aspects of

protein metabolism in phytoplankton (Dortch et al., 1982; 1984). With respect to

biomass, the proteins in phytoplankton cells are also important to the secondary

consumers that feed on them and the benthic consumers that receive particulate organic

matter derived from phytoplankton residues in sediments. The quantitative information on the protein content in phytoplankton cells, as well as their relationships to chlorophyll a, will be important to a variety of studies that are directly and indirectly related to various aspects of cellular nitrogen metabolism as well as predictors of phytoplankton

dynamics and physiological state. To date, very few studies have reported generalized

relationships between algal protein and chlorophyll a. For example, a weight- to- weight ratio for protein/CHLa = 8.57:1 has been reported and is often used in the literature

(Meyers and Kratz, 1955). However, that work only focused on one species of blue-green alga, Anacystis nidulans.

Protein synthesis in algae is believed to be mainly a component of algal night (i.e. dark) metabolism (Morris et al,. 1974). These workers showed that ratios of labeled pools

of carbohydrate carbon: protein carbon changed during day/night experiments. They

postulated that these changes were due to the flow of carbon from storage polymers 21

through metabolites into protein. Many zooplankton migrate vertically to feed in surface

waters at night, therefore night time protein synthesis in algae will have implications on

zooplankton nutrition (Scott 1980). Environmental factors such as prior light intensity

history, nutrient status and the species composition of the population are determinants of

the algal dark (night) metabolism and growth (Cuhel et al., 1984).

Interestingly, in another study done on Dunaliella tertiolecta, CHLa accumulation in the photosynthetic apparatus was linked to the synthesis of apoproteins of pigment- protein complexes and a high ratio of protein to pigment in light harvesting and other complexes (Mortain-Bertrand et al,. 1990). It is therefore only assumed that in light saturating and nutrient limiting conditions, when CHLa concentrations decrease, the concentrations of the associated proteins will also decrease.

A number of methods exist for the extraction and determination of phytoplankton protein (Bradford, 1976; Lowry et al., 1951). Consideration of the various analytical methodologies used by different authors makes inter-comparison of micro-algal protein contents difficult (Berges et al., 1993; Clayton et al., 1988; Hach et al., 1987; Rausch,

1980). For the present study, we have chosen to adapt the warm sodium hydroxide extraction and micro biuret assay as developed, evaluated and standardized by Rausch and co-workers (Rausch, 1981).

Algal total organic carbon (TOC): Even though biomass is expressed in terms of CHLa, organic carbon concentration is normally what is desired (Cullen 1982).

However, organic carbon cannot be measured directly because of interference from zooplankton and non-living organic matter. Estimations have to be made based on some multiplying parameter or ratio of Carbon: CHLa (Banse, K., 1977). The carbon: CHLa 22

ratio (θ) is a poorly studied factor in phytoplankton growth and ecology (Geider, 1987).

The ratio has been assumed to be constant in ecological studies, for example,

recommendations of θ = 30 g C·g-1 CHLa for nutrient rich waters and (θ) = 60 g C·g-1

CHLa for nutrient poor waters have been made reported by Strickland (1960). However, due to phenotypic variation in chemical composition and rates of physiological processes,

a universal ratio cannot be utilized (Geider, 1987). In diatoms, the ratio can vary from 10-

200 g C·g-1 CHLa depending on light level, temperature or nutrient availability (Geider,

1984, Osborne and Geider, 1986). These variations are indicative of the physiological

plasticity of microalgae and the need for additional study.

We therefore intend to use pigment-based chemotaxonomy to gain a better

understanding of the relationships of chlorophyll a with algal functional carbohydrates and proteins as well as organic carbon under the influence of light and nutrient conditions. By investigating if correlations exist between CHLa and these components, then a novel approach may be able to be introduced where it may be possible to describe biomass in terms of colloidal and storage carbohydrates, proteins and organic carbon based on CHLa concentration. This will also represent a large step in the direction of the possible use of chemotaxonomy to relate to algal organic carbon, protein and carbohydrates in energy flow/food web models.

Photosynthesis overview

Photosynthesis is driven by visible light (400-700 nm), termed photosynthetically

active radiation (PAR). Photosynthesis in eukaryotic algae takes place on inner foldings

of chloroplasts called the thylakoid membranes as well as in the stroma, the cytoplasm of

the chloroplasts. The thylakoid membrane is folded upon itself, forming many discs called grana. The reactions of photosynthesis can be broken in a series of light-dependent and light independent reactions. The light dependent reactions occur on the thylakoid membranes, while the light independent reactions occur in the stroma. In prokaryotic organisms such as cyanobacteria, photosynthesis occurs within the cell membrane (Zak et al., 2001).

Light harvesting pigment-protein complexes form clusters called antennas on the thylakoid membranes. Algae contain various pigments, which can be classified into two major groups: The photosynthetic accessory pigments (PAP) and the photoprotectorant pigments (PPP). The most abundant chlorophyll is chlorophyll a (CHLa). CHLa molecules can act as light absorbers in the light harvesting or antenna complexes and as electron donors and receivers in the reaction centers of the two photosystems where photosynthesis occurs. Photosynthetic accessory pigments (PAP) absorb energy that

CHLa does not absorb and pass this energy to the other antenna pigments and finally to the reaction centers for photosynthesis. Accessory pigments include: CHLb ,

Chlorophylls c1/c2/c3 (CHLsc1/c2/c3), Fucoxanthin (FUCO), and Peridinin (PER), among others (Appendix II). Photoprotectorant pigments mainly protect the plant, cyanobacteria or algae from photo-oxidative damage. Photoprotectorant pigments are often taxon specific and include: Myxoxanthophyll (MYXO), Scytonemin (SCYTO), Echinenone

(ECH), Canthaxanthin (CANTHA), Lutein (LUT) and others (Appendix II). Carotenoids are a highly colored (red, orange and yellow) group of fat-soluble isoprenoid pigments.

Carotenoids comprise some of the photosynthetic accessory pigments and many are photoprotectorant pigments (PPPs). 24

The xanthophylls: These are a diverse group of oxygenated carotenoids with

various structures and multiple functions (Britton, 1995). The interconversion of violaxanthin (diepoxide), antheraxanthin (monoepoxide), and zeaxanthin (epoxide free) is termed the xanthophyll cycle and is found in both higher plants and green algae

(chlorophytes). This cycle is shown in Figure 2. Sapozhnikov et al. (1957) were the first to describe the xanthophyll cycle. The scheme of reactions that takes place in the light involves two de-epoxidation steps through which violaxanthin, via the intermediate antheraxanthin, becomes zeaxanthin, (Hager and Stransky, 1970). The role of the xanthophyll cycle in plants and algae was first thought to be linked to oxygen evolution

(Sapozhnikov et al., 1957). Hager (1980) later proposed a role in the electron transfer activity in the photosystems during photosynthesis. Krinsky (1971) was first to suggest a

role as a photodamage protection mechanism. That is, if more energy is harvested by

chlorophyll than what can be used in photosynthesis, then that excess energy has the

potential to cause damage to the intracellular components of photosynthetic organism.

The phenomenon that causes a reduction in photosynthetic efficiency due to exposure of

the photosynthetic apparatus to excess photons is termed photoinhibition (Powles, 1984).

Overall, the xanthophylls can function as accessory light-harvesting (antenna) pigments,

as structural entities within the antenna complex and as molecules required for the

protection of photosynthetic organisms from the potentially damaging effects of light

(Niyogi et al., 1997).

Carotenoid protection against photodamage is of paramount importance.

Demmig- Adams (1990) introduced one mechanism by which carotenoids function to

protect against photodamage. Here, specific xanthophylls are involved in the de-

excitation of singlet chlorophyll that accumulates in the light-harvesting (antenna)

complex. This accumulation occurs under conditions of excessive light. The de-excitation

is measured as nonphotochemical quenching of chlorophyll fluorescence, and is

dependent on a large trans-thylakoid proton gradient that becomes established in

excessive light. This nonphotochemical quenching was determined to correlate with the synthesis of zeaxanthin and antheraxanthin from violaxanthin via the xanthophyll cycle.

HO +O -O (DD)

OH a)

(DIATO)

(VIOLA) +O -O

(ANTH)

+O -O

b) (ZEA) Figure 2:. Xanthophyll cycling in (a) Chrysophytes and (b) Chlorophytes. 26

At low light intensities, or if all of the energy harvested is utilized for

photosynthesis, no zeaxanthin is formed. However, as light intensity is increased, and the

amount of light energy harvested starts to exceed that needed for photosynthesis, more

zeaxanthin is formed from violaxanthin. It is believed that zeaxanthin formation in

chlorophytes and higher plants aids the thermal dissipation of the excess energy within

the light harvesting system.

In algal divisions such as Chrysophyta (diatoms and golden-brown algae) and

Dinophyta (dinoflagellates), the xanthophyll cycle described above, is paralleled by a

xanthophyll cycle that alternates diadinoxanthin with diatoxanthin. Diadinoxanthin is

converted to diatoxanthin via a single epoxidation step (Figure 2). The formation of

diatoxanthin, like zeaxanthin, correlates with the nonphotochemical quenching of singlet

chlorophyll described earlier.

Photosynthesis: This very complex process is initiated when an antenna molecule

absorbs a photon (Govindjee and Braun, 1974). Absorption takes place in about a

femtosecond (Kok and Businger, 1956) and causes a transition from an electronic ground

state to an excited state. In a few seconds this excited state would decay by vibrational relaxation to the first excited singlet state. However, because of proximity to other antenna molecules, the excited state energy has a high probability of being transferred by resonance energy to a close neighbor (van Grondelle and Amesz, 1986). Photosynthetic antenna systems thus act as funnels, which efficiently transfer excitons to the reaction centers. Photosystem I (PS I) and photosystem II (PS II) comprise the two reaction centers of photosynthesis.

PS II is a dimeric chlorophyll-protein complex that absorbs maximally at 680 nm and is the site of the light dependent reactions of photosynthesis. When energy in the

form a photon arrives at the PSII reaction center, an electron (e-) in chlorophyll a

becomes excited. The electron can then travel through a series of redox reactions, by

electron carriers, such as pheophytin, cytochromes and plastoquinone. Water is oxidized

and plastoquinone is reduced in PSII. Water oxidation requires two molecules of water

and involves four turnovers of the reaction center (Kok et al., 1970). Each photochemical

reaction creates an oxidant that removes one electron and the net reaction results in the

release of one oxygen molecule. Four protons are deposited into the stroma and four

electrons are transferred to the plastoquinone pool, where they reduce two plastoquinone

molecules (Klein et al., 1993). Reduced plastoquinone debinds itself from the reaction

center and diffuses into the hydrophobic core of the membrane and travels to PSI to start

electron transport there. Tyrosine pulls one of the electrons produced from the oxidation

of water and uses it to replace the one that was lost from the reaction center. An oxidized

plastoquinone finds its way back to the quinine pool and the process is repeated.

Photosystem I is a chlorophyll-protein complex that absorbs maximally at 700

nm. The reactions of PSI can take place with or without light. At PS I, an excited electron

travels through a series of electron transfer components, such as special proteins

(A˚ and A1), three ferrodoxin proteins and then on to ferrodoxin. Ferrodoxin reductase

then facilitates the production of NADPH (nicotinamide adenine dinucleotide phosphate)

from NADP+. Along the electron transport pathway from water to NADP, a fraction of light energy is used to synthesize ATP (adenosine tri- phosphate) from ADP (adenosine

di-phosphate) and inorganic phosphate. With sufficient NADPH and ATP available, enzymatic reduction of CO2 to the carbohydrate level (Calvin-Benson cycle) becomes

possible. The equations (Figure 5) below represent the general reactions that take place

at the two reaction centers. As shown, the reactions are not stoichiometrically balanced.

+ H2O + NADP + ADP +Pi PS-II O2 + NADPH + ATP

+ CO2 + NADPH + H + ATP PS-I Glucose + NADP +ADP + Pi

Equation 2: Summary of two major reactions in the photosynthesis process.

Briefly, in the Calvin Benson cycle, CO2 combines with a ribulose 1, 5-

bisphosphate (RuBP) molecule to yield two molecules of a three carbon compound called

3-phosphoglycerate (PGA). In the presence of ATP and NADPH, PGA is reduced to 3-

phosphoglyceraldehyde (PGAL). PGAL is a 3-carbon sugar, and more than half of these

molecules are used to regenerate RuBP so the process can continue. The rest of the

PGAL molecules that are not recycled condense to form hexose phosphates. Hexose

phosphates yield sucrose, starch and cellulose. The sugars produced during this carbon metabolism go on to produce carbon skeletons that are used for other metabolic reactions, such as the production of amino acids and lipids. The carbohydrates, proteins and lipids produced from photosynthesis in algal cells serve as an energy source for the consumers in the next trophic level. Since light is a major factor driving photosynthesis, we believe that a link can be established between chlorophyll a, a major participant in photosynthesis

and some of the products of photosynthesis. 29

Novel sunscreen pigment isolated from Scytonema hofmanii grown at high light and from samples collected in the Florida Everglades.

Scytonemin is a known ultraviolet radiation screening pigment, produced in

cyanobacterial sheaths (Garcia-Pichel et al., 1992; Dillon and Castenholz, 1999).

Cyanobacteria can produce a variety of secondary metabolites, including notorious

toxins, some of which potential therapeutic agents. They are able to inhabit and thrive in

a variety of hostile environments, from intense radiation to intense dessicating conditions

(Flemming and Castenholz, 2007; Butel-Ponce et al., 2004). Scytonemin is a photostable dimeric pigment with indolic and phenolic subunits, as characterized by Proteau and

coworkers (1993). This structure is unique in nature and has been termed the

‘scytoneman’ skeleton. Two forms of this pigment exist: a yellow-brown oxidized form

and a red-brown reduced form. Reduced scytonemin is not considered to be biologically

active, but has been suggested to be a transformation product formed in reducing

environments (Garcia-Pichel and Castenholz, 1991). Both structures along with their

UV/Vis spectra are shown in figure 3.

Scytonemin reduced form

Scytonemin

Figure 3: Structure and UV/Vis spectra of Scytonemin (reduced form) and Scytonemin

Three new pigments (Figure 4), related to the scytonemin skeleton, were isolated and structurally identified in a study aimed at investigating plant succession in the

Mitaraka inselberg in French Guyana (Butel-Ponce et al., 2004). These molecules are derived from condensation of tryptophanyl- and tyrosyl-derived subunits with a linkage

Tetramethoxyscytonemin: Purple amorphous solid; UV: 212nm, 562 nm; + m/z [M+H] 671

Dimethoxyscytonemin: dark red amorphous solid; UV: 215 nm, 316 + nm, 422; m/z [M+H] 609

Scytonine: brown amorphous solid; UV: 207 nm, 225 nm, 270 nm; m/z + [M+H] 519

Figure 4: New pigments isolated form Scytonema sp. collected on Mitaraka Inselberg, French Guyana (Butel-Ponce et al., 2004).

between the units unique among natural. The compounds have been termed

tetramethoxyscytonemin, dimethoxyscytonemin and scytonine.

A new pigment, believed to be related in structure to the scytonemin was isolated from lab grown cultures of Scytonemin hoffmanii and from samples collected from areas of the Florida Everglades. Partial characterization of this pigment was done as a second

project in this study. The structural properties of this pigment, as well as a putative

structure characterization are presented and discussed in Chapter VI.

Overall goals of this study

The preceding pages were written to familiarize the reader with pigment-based

chemotaxonomy, the relevance of algal biomass parameters in pelagic communities and

the need for better assessment of these parameters. The study was therefore conducted to

investigate the possibility of partitioning algal protein and two functional classes of

carbohydrates that are being contributed from the various taxonomic groups in a population. This partitioning is suggested to be based on their relationship with chlorophyll a. In the same vein that taxonomic marker pigments are used to partition chlorophyll a among the different taxonomic groups and then applied to mathematical formulae for estimating algal class abundance, we want to determine if the same concept can be extended to the taxonomic estimation of algal proteins and carbohydrates.

The second project represents a fortuitous finding, as a novel sunscreen pigment was isolated and characterized. Postulations are forwarded regarding the physiological and ecological significances of this new pigment based on its UV/Vis absorbance spectral data.

II. MATERIALS AND METHODS

The overall experimental design for algal growth, cell counting, harvesting, pigment analyses, carbohydrate functional group analyses, and protein analyses is shown in Figure 5 and the batch culture preparation is shown in Figure 6.

Experimental organisms: The following fresh water and marine microalgal species were purchased from the Carolina Biological Supply Company (Burlington,

N.C.): Cyanobacteria; Synechococcus elongatus (marine), Microcystis aeruginosa

(fresh), Chrysophyta; Thalassiosira weissflogii (marine), Cyclotella meneghiniana

(marine), Chlorophyta; Scenedesmus sp. (fresh), Pyrrophyta, Dinophyceae; Amphidinium carteri (marine). Additionally, the following species were purchased from the University of Texas (UTEX) algal culture collection (Austin, TX): Rhodophyta; Rhodomonas salina

(marine), Chlorophyta; Dunaliella tertiolecta (marine).

The sample vials containing each unicellular culture were gently vortexed to achieve homogeneous distribution of the cells as soon as they arrived at the Florida Atlantic

University Organic Geochemistry laboratory. Approximately 5 mL quantities were taken from each vial for filtering and initial HPLC analysis, described below. This initial analysis served to verify that the samples had not reached the senescent/death stage of growth or were not overtly contaminated with another taxon.

Grow at 3 Inoculate, Monitor 8 ecologically irradiance test nutrients growth / significant levels algal species cell counts

Harvest: cell counting; collect samples on separate GF/F for pigment, protein, total organic carbon (TOC) analyses;colloidal and storage carbohydrate (CHO) analyses; nutrient tests of media filtrate.

Pigment extract: Protein extract: Carbohydrate TOC extract: 90% MADW & 0.5N NaOH, per extract: culture K2Cr2O7 & Rausch, 1980 volume centrifuged conc. H2SO4, per Walkley- Black, 1934 Phenol/H SO assay Extractant + Microbiuret 2 4 Ion pairing on supernatant- assay: extractant colloidal CHO +CuSO4/NaOH fraction Measure abs @ 610 nm; compare to RP HPLC Abs@310 nm, Pellet re- standard curves compare to standard suspended in curve warm water for storage CHO

PDA Filter, lyophilize, detector Phenol/H2SO4 assay; compare to standard curves

Concentrations used to determine relationships between CHLa/pigment, protein/CHLa, colloidal CHO/CHLa, : homoscedasticity and ANOVA tests performed Figure 5:. Flow Chart of Analytical Scheme.

unicellular algal stock culture

Batch 1

Batch 2 Batch 3 Batch 4 Batch 5 ……etc.

Figure 6: Schematic of inoculation procedure. Each species grown 5-7 times (inoculation, lag growth, exponential growth phase). Each batch is inoculated from stock culture, to prevent pseudo replication.

Algal culturing: All species were grown in 2 L batches in 4L cylindrical polycarbonate containers (CAMBRO. Huntington Beach, CA). Autoclaved (122 º C and 36

2 atm.) Zephyrhills ® Natural Spring Water was used for the freshwater cultures and the

media prepared according to Guillard’s (Guillard, 1975) f/2 medium.

The seawater for the marine cultures was collected from coastal water (FAU Gumbo

Limbo Environmental Complex and Nature Center, Boca Raton, Florida) and autoclaved

(122 º C and 2 atm.) after filtering. The addition of nutrients, including vitamins and trace

metals were also based on Guillard’s (Guillard, 1975) f/2 medium. Erdschreiber’s

(Schreiber, 1927) medium was used to prepare Dunaliella tertiolecta and Rhodomonas

salina.

Culture conditions: Light levels are given here as; high (180-200 µmol

photons·m-2·s-1), moderate (70-75 μmol photons·m-2·s-1), low (35-37 μmol photons·m-2·s-

1), and dim (10 μmol photons·m-2·s-1). Light conditions were achieved within three

temperature controlled (25oC) growth chambers: a Revco-Harris growth chamber was

used for the high light experiments, while two Precision low temperature Illuminator 818

growth chambers were used for the remaining light levels. All growth was at 25oC with a

12 Light: 12 Dark diurnal cycle. Temperature control was observed to be + 1.5oC. The

samples in the two Precision growth chambers were illuminated from the front only

(fluorescent tubes vertically attached to the inside door) with two 34W Econo (Philips)

120 cm long fluorescent tubes, covered by a diffuser screen for the medium light

experiments and without a diffuser screen for the low light experiments. Samples for the

high light experiments (Revco-Harris growth chamber) were illuminated from the top and

both sides with sunlight quality (Verilux Instant Sun™), full Spectrum™ (ValuTek) and

“aquarium” quality (Sylvania Gro-Lux™) fluorescent tubes. Three 8W (Westwek 20121)

cool white fluorescent tubes were attached horizontally on the inside door of the

Precision growth chamber, and used for illumination in the dim light experiments. Only

one of the species in the study: Synechococcus elongatus, grew in dim light level, and the

dim light experiments are eventually discontinued for the remaining species. Light

intensity (PAR radiation: 400 – 700nm) was measured with a 4π spherical radiometer and

Li-Cor LI-250 Light Meter. Spectroradiometric data for the fluorescent light sources used

in the three main light levels in this study are shown in Appendix III. Transmission of light through the culture flasks is also given in Appendix III.

Cell counting: Coulter Counter model ZM electronic cell counter was used for rapid cell counting. The method of counting and sizing used by the Coulter is based on the detection and measurement of changes in electrical resistance produced by a particle, suspended in a conductive liquid, traversing a small aperture.

Cell counting was carried out on the same day that the algal samples were to be harvested and every two to three days during growth to follow and plot logarithmic growth plots.

ISOTON® II diluent (electrolyte solution) was pipetted (20 mL) into the counting vial

(Fisher ‘Accuvette’) and 100μL of suspended algal cells was added. Each vial was placed in the counter for electronic counting. The three most consistent counts out of six were averaged and used as the corrected count. The dilution factor was determined, and the number of cells per milliliter was calculated by multiplying the corrected count by the dilution factor, as follows: Dilution Factor (DF) = (mL sample + mL electrolyte)/

(manometer setting x mL sample). DF x corrected counts = cells mL-1.

The number of cells per milliliter was determined from cell counting and the

concentration of each analyte (pigments, proteins, carbohydrates, organic carbon) could then be calculated.

Chemical Analyses: All analyses described below were performed on replicate aliquots collected at the same time.

Algal protein extraction: The procedure was adapted from Rausch (Rausch,

1981) with some modifications, and is as follows: 100 mL aliquots of algal culture were filtered on to pre-combusted glass fiber filters. The filters were then folded in halves, then quarters and refrigerated at -80 ˚C until analysis. Analyses were usually done one week after filtering. For extraction, samples were extracted in 0.5M sodium hydroxide (NaOH), by grinding the filters in 12 mL test tubes using tissue grinders (glass mortar with Teflon

® pestle e.g. Kontes Dwall). The tubes were next heated at 80 ˚C for 10 minutes to further extract the proteins. After this step, the tubes were quickly cooled to room temperature, and then centrifuged (Fisher Scientific, Centrific Model 228) for 5 minutes at approximately 2800 rpm. The supernatant was then transferred to 10 mL graduated tubes for subsequent protein analysis. A second extraction was then carried out on the remaining filter debris (extraction in 0.5 M NaOH at 80 ˚C for 10 minutes, followed by cooling and centrifugation), and the supernatants were combined in the 10 mL graduated tubes. A third extraction was carried out (0.5M NaOH at 100 ˚C for 10 minutes) for green algae and cyanobacteria, as prescribed by Rausch (1981). The combined supernatants were then made up to a definite volume (6-10 mL) with 0.5M NaOH and used for protein measurement.

Algal protein measurement: The micro-biuret method for estimating proteins as adapted from Itzhaki and Gill (1964) and was slightly modified. The procedure used is as follows: 2 mL of algal protein extract was assayed with 1 mL of 0.21% CuSO4.5H2O in

30% NaOH at 310 nm in a 1 cm quartz cuvette and another 2 mL of algal protein extract

was assayed with 1 mL of 30% NaOH at 310 nm in a 1 cm quartz cuvette. The

absorbance of the protein was obtained from the difference between the absorbance of the

sample in 30% NaOH and that following reaction in 0.21% CuSO4.5H2O in 30% NaOH.

All samples were measured against distilled water. Bovine serum albumin was used for calibration. See Appendix IV for the BSA calibration curve and equation.

Algal colloidal and storage carbohydrate extraction: The method was adapted

from Chiovitti et al. (2004) and developed with some modifications. The adapted

method is as follows: aliquots of approximately 50 mL algal cultures in the logarithmic stage of growth were collected in 50 mL centrifuge tubes and centrifuged (Dynac

Centrifuge, Becton Dickinson and Co, Parsippany N.J.) at 3300 rpm for 30 minutes. The supernatant was decanted to leave ~ 0.5-1 mL of wet cells, and 2 mL of the supernatant

was used for colloidal carbohydrate analysis. The remaining wet cells were re-suspended

in 30 mL ultra-pure water (Milli-Q® Ultra-pure water systems, Millipore Corporation)

and heated in a water bath for an hour, stirring every 10 minutes. The solutions were then

sonicated for 5 minutes (Burdloff et al., 2001), followed by pelleting the cells via

centrifugation for 30 minutes. The resulting supernatant containing mostly water soluble

(storage carbohydrates) was then filtered through 0.22 µm membrane filters (Fisher

Scientific). The filtrates were then lyophilized and the dried material used for

carbohydrate analysis.

Algal colloidal and storage carbohydrate measurement: The two extracted

carbohydrate fractions were analyzed using the phenol- sulfuric acid assay (Dubois et al,

1956). The lyophilized samples were dissolved in exactly 2 mL of ultra pure water and

pipetted into 10 mL disposable test tubes. For the colloidal fractions, exactly 2 mL of the

initial supernatant was pipetted into 10 mL disposable test tubes. Next, 0.05 mL of 80%

phenol was added to each tube followed by the rapid addition of 5 mL concentrated

H2S04. The tubes were allowed to stand for 10 minutes, after which they were placed for

approximately 20 minutes in a water bath at 25-30˚C with occasional shaking. The

resulting champagne - dark orange solutions were then measured at 485 nm against

distilled water in a Perkin Elmer UV/Vis Lambda 2 spectrometer. Alpha-D (+)-Glucose

was used for preparing calibration curves (see Appendix IV for calibration plots and equations).

Algal total organic carbon (TOC) extraction: The method used herein was adapted from Walkley and Black (Walkley, A and Black, I.A., 1934) as updated by Chan and coworkers (Chan et al,. 1995) and involves the rapid dichromate oxidation of organic matter according to the following equation:

2- 0 + 3+ 2Cr2O7 + 3C + 16H = 4Cr + 3CO2 + 8H2O

Equation 3: Dichromate oxidation of organic matter

This method is typically used for analyzing soils and sediments. Therefore, modifications

were carried out to allow for the application of the method to algal organic carbon.

Although the method is given here, it should be noted that it had to be abandoned as the

TOC was grossly overestimated. An external source was contacted for automated sample analyses, but at the time of this writing, no validation has been obtained as whether our samples could be accurately analyzed using the available analytical protocol for that instrument. The modified Walkley-Black method is as follows: known volumes (100-

200 mL) of algal cultures were collected on pre-combusted glass fiber filters (Whatman

GF/F, 0.7 micron pore size borosilicate glass fiber), folded in half, then quarters, wrapped

in aluminum foil and stored at -80ºC until extraction. For extraction, the glass fiber

filters are retrieved, and opened to reveal the collected algal cells. The cells are pretreated

with a 1 mL solution of 2N H2SO4/5%FeSO4 to remove any inorganic carbon present.

This solution is added in increments until any effervescence present stops. The GF/F

filters are cut into 1/8 pieces and combined in 50 mL Erlenmeyer flasks. Approximately

1-1.5 mL of 1/6 M K2Cr2O7 is added, followed by 5 mL concentrated H2SO4. To

overcome any possibility of incomplete digestion of organic matter, the sample and

extraction solutions are gently heated at 145ºC for 30 minutes (Mebius, 1960). The

temperature has to be strictly controlled as the acid-dichromate solution decomposes at

temperatures above 150 ºC (Charles and Simmons, 1986). Following heating, the flasks

are cooled to room temperature and 5 mL water is added to quench the reaction.

Colorimetric determination of extracted TOC samples: Quantitation of total

organic carbon is performed through the measurement of the color change that results

from the presence of Cr 3+ in solution. The digestate was poured into 12 mL disposable

centrifuge tubes and centrifuged, then filtered through a 0.45 μm pore filter attached to a

3mL syringe. The filtrate is placed in a 1 mL cuvette and the absorbance measured at 610 nm against distilled water in a Perkin Elmer UV/Vis Lambda 2 spectrometer.

Quantitation is performed by determining the concentration from a standard curve.

Potassium Hydrogen Phthalate (KHP) and sucrose are used for validation of the method

and for preparing standard curves. See Appendix IV for plots and equations.

Nutrient analyses: The nitrate and phosphate content of the enriched algal

growth media were determined before inoculation and at harvest to follow and ensure the

cultures were grown in nutrient replete conditions (this data will not be shown, as it was

only used to verify nutrient conditions of the batch cultures). The Hach DR 5000 spectrometer (Hach Company World Headquarters, P.O. Box 389. Loveland, CO) was

used for all analyses. All procedures used were per the Hach DR 5000 procedures

manual. The Chromotropic acid method (10020) was used for measuring high range (0.2-

30.0 mgL-1) nitrate; the Molydbovanadate method (8114) was used for high range (HR,

-1 3- 1.0 to 100.0 mgL ) phosphorus, namely reactive orthophosphate (SRP PO4 ); the

-1 3- PhosVer 3 (ascorbic acid) lower range method (8048) for 0.2 to 2.5 mgL PO4 was

3- used several times at culture harvest when the HR PO4 tests were not sensitive enough

to detect the small amounts of orthophosphate remaining in the growth media.

Pigment Analyses: All pigment analyses were carried out under dim yellow light conditions to prevent photo-oxidative alterations, such as pigment isomerization. For harvesting and for pigment monitoring during growth, culture volumes (25-100 mL), volume dependent on stage of growth/cell density) were filtered onto glass microfiber filters (Whatman GF/F, 0.7 micron pore size borosilicate glass fiber). The filters were removed from the filter funnel, folded in half and blotted between paper towels. The filters are then folded into quarters, re-blotted and wrapped in aluminum foil and then immersed in liquid nitrogen for quick freezing. The individual samples were removed from the liquid nitrogen and stored in a refrigerator at -80ºC until extraction.

For extraction, the filters were unwrapped from the aluminum foil, placed in pre- chilled glass tissue grinders (Kontes “Duall” 15 mL) and extracted by grinding (Barnant variable speed mixer Series 20, ~ 350 rpm) with 3 mL of an extraction solvent containing

a procedural internal standard. The extraction solvent used was a mixture of 43

acetone/methanol/dimethylformamide/water, 30:30:30:10. (90% MADW). This mixture

has been shown to give better extraction efficiency and peak resolution than those

previously used (Hagerthey et al., 2006). The internal standard (IS) used was Copper

Mesoporphyrin- IX Dimethyl Ester (CuMESO IX DME). This was dissolved in 90%

MADW and the absorbance readings at 394nm and 715nm baseline were taken. Using

Beer Lambert’s law (A= εcl), and εmM =305 (Fuhrhop and Smith, 1975) at 394 nm, the

concentration of the internal standard added was determined. The extraction solvent was

made up such that the IS absorbance at 394nm would not exceed 1.0 AU. Having the

internal standard in the extracting solvent allowed the monitoring of its recovery through

the extraction as well as the chromatography process. A system response factor was

applied to all the pigments based on the ratio IS added/IS detected, giving correction factors ranging from 1.2-1.5. During the HPLC analysis, the internal standard eluted as a sharp peak in the 394 nm integration but did not show in the 440 nm or slightly in the 410 nm integration where the pigments were detected and quantified. This allowed the internal standard peak and peak area to be readily identified and quantified without interference from other pigments. However, CuMESO IX DME did partially co-elute with canthaxanthin. Canthaxanthin (β, β-carotene- 4,4’- dione) could still be adequately quantified since only absorption of the internal standard is absent at 440 nm.

The extraction slurry was next sonicated in ice water for 30 seconds (tissue grinders in bath style sonicator) to further disrupt the cells in the samples. Sporadic sonication was shown to give good extraction (Hagerthey et al., 2006; Louda and Mongkhonsri 2004).

The extracts were then steeped for 1-2 hours at 4-6 ºC in a refrigerator. Following this, the extracts in the tissue grinders were then centrifuged (Centrific Cenrifuge Model 228,

Fisher Scientific) for 2-3 minutes and the supernatant was decanted into a vial. This was

labeled ‘Raw’ extract. The remaining slurry was then placed in a 2 mL centrifuge filter

(Ultra free-CL PVDM, 0.45 µm diameter) and re-centrifuged to recover remaining extract. The pooled ‘Raw’ extract was collected in a 3mL syringe and passed through a

0.45 μm pore diameter Cameo Syringe Filter into a second vial. This vial was labeled

‘Filtered’ extract. This procedure typically gave an overall total recovery of 90% (~

2.7/3.0 mL).

Ultra Violet - Visible (UV/Vis) Analyses of Extracts: The UV/Vis absorption

spectra (350-800 nm) of 1.0 mL aliquots of the filtered extracts were recorded on a

Perkin Elmer Lambda - 2 UV/Vis Spectrophotometer. This spectrophotometer was

calibrated for wavelength vs. holium oxide and absorbance vs. potassium chromate in

aqueous potassium hydroxide (Rao, 1967). The visible absorption spectrum of the extracts gives a rough spectrophotometric estimate of total chlorophyll and carotenoids

using the Beer- Lambert relationship and was used to determine if dilution of the extract

was required (e.g. A430 > ~ 1.2) prior to injection into the HPLC system.

1.0 mL of the filtered extract was then added to a pre-chilled vial containing 0.125 mL of

ion pairing solution (Mantoura and Llewellyn, 1983), giving a total of 1.125 mL. This

vial was labeled ‘Mix’. Next, 100 μL of this mixture (extract = 88.89 μL of the 100 μL)

was injected into the HPLC system. The ion pairing (IP) or ion suppression solution used

in the injectate solution consisted of 15.0 g tetrabutyl ammonium acetate, 77.0 g

ammonium acetate and nano-pure water to equal a final volume of 1L. Incorporation of

ion pairing agent allows the separation of highly polar substances on reversed phase

HPLC columns by masking such highly polar groups such as carboxylic acids (Poole and

Poole. 1991).

High Performance Liquid Chromatography (HPLC): The High Performance

Liquid Chromatography system consisted of a Consta Metric 4100 series Quaternary

Solvent Delivery Systems pump (Thermo Separation Products, Riviera Beach, Fl.), a

Rheodyne model 7125 syringe loading sample injector fitted with a 100 μL injection loop, a 250 mm long Waters Symmetry ® C18 column with an internal diameter of 4.6 mm (4µm spherical particle size, 100 ºA pore size, 335 m2 g-1 surface area) and a Waters

996 Photodiode Array Detector (PDA: 190-800 nm). This system was coupled to a Dell

PC using Millennium 32 software.

Gradient elution was carried out using a mixture of three solvents with the following solvent ratios (see Table 4): a combination of 60% solvent A (0.5M ammonium acetate in methanol: water, 85/15) and 40% solvent B (90/10 acetonitrile: water) for the first five

minutes. This combination provides a good ratio of polar and lipophilic solvents. Eluted

during this time are the solvent front of the injectate, the highly polar peridinin

derivatives P-468 and P-457, chlorophyllide-a, chlorophylls-c1/c2. The gradient is then

changed to 100% solvent C (100% ethyl acetate) from 5-10 minutes. The peaks eluted

here were the scytonemins, fucoxanthinol, pyrochlorophyllide- a, and peridinin. At 10

minutes, the gradient changes to 100% solvent B. This gradient changes gradually up to

35 minutes, where a ratio of 35% solvent B and 65% solvent C is reached. The

carotenoids neoxanthin, fucoxanthin, cis-fucoxanthin, violaxanthin, dinoxanthin,

antheraxanthin, astaxanthin, diadinoxanthin, myxoxanthophyll, diatoxanthin, lutein,

canthaxanthin, zeaxanthin, followed by chlorophylls-b, chlorophylls-a, and echinenone

eluted during this time. The gradient then changes to 30% solvent B and 70% solvent C

between 35 and 45 minutes. The pheophytins (-b/-a) plus beta-carotene are highly non-

polar and are eluted between 36 and 40 minutes. The gradient then changes to 100% solvent C for 2 minutes in order to flush any highly lipophilic compounds that may be left. The gradient then returns to the original 60% solvent A: 40% solvent B at 48

minutes, thus restoring the column to the required conditions for next use. A 100%

solvent D (85/15 methanol: water) is ran through the column for 15 minutes prior to

storage of the column in that solvent.

Solvents:

A = 0.5M Ammonium acetate in MeOH/water, 85:15 B = Acetonitrile/water, 90:10 C = Ethyl acetate, 100% D = Methanol/water, 85:15 (storage solvent)

Table 4: Gradient program used in FAU OGG laboratory

Time (min) Solvents A/B/C 0 60/40/0 5 60/40/0 10 0/100/0 40 0/30/70 45 0/30/70 46 0/0/100 47 0/100/0 48 60/40/0

HPLC Data Calculations: Pigment analysis was achieved 2 dimensionally based

on retention time and spectral absorbance from the photodiode array (PDA) detector. As

the sample extract partitioned between the stationary and mobile phase of the column, the pigments were separated based on their solubility in the changing solvent gradients and their affinity for the stationary phase. The integrated areas of the peaks from the 47

chromatogram were used in Beer-Lambert calculations to obtain the molar quantities and

weight of each compound.

The UV/Vis spectrum of each separated pigment was recorded with the PDA

detector from 300 – 800 nm. The conjugated C=C structure of the pigments is referred to

as the chromophore or the color bearing group of the compound. It is well known that the

number of conjugated double bonds (N), the conjugated end groups and the solvent

influence the absorption spectra of carotenoids. The position of the λmax of the absorption spectra is unique for individual carotenoids, with λmax being mainly influenced by the N value in carotenoids, increasing as the N value increases (Takaichi, 2000). Therefore, the

identity of carotenoids as well as chlorophylls can be determined by a combination of the

HPLC retention times and the absorption spectra, the so-called ‘2-D advantage’.

The retention times and spectral absorbance of the pigments were also compared

to those from the FAU Organic Geochemistry Group’s (FAU-OGG) library of standards.

Known standards are always required for any chromatographic system. Standards are

obtained either as pure compounds or as part of well-known accepted mixtures in

unicellular bacterial or algal cultures (Jeffrey et al., 1997). The members of our lab group

have obtained a large number (>80) of chlorophylls, carotenoids and their derivatives

through partial synthesis or derivatization and several have been purchased from VKI

(Denmark). The performance of these standard compounds (retention time and spectral

absorbance) was used to verify the identity of the algal pigments in this research. See

Appendix V for a detailed table of retention times and UV/Vis PDA spectral data for the

pigments typically encountered in our research group and in this study.

The HPLC software automatically integrated peaks on the chromatogram and

reported the areas under the peaks as time - based microvolt second (µV*s) units. This

gave a relation to the quantity of each compound. The µV*s data was previously

standardized versus AU· min-1 data from Waters 990 PDA and versus known

concentrations of pigments. The flow rate used was 1.00 mL/min, giving the integrated

peak area in microvolt second units·mL (µV*s ·mL). Manual integration was used to

separate pigment peaks that overlapped. Very small peaks were also integrated manually.

The µV*s data was next entered in an in-house (Florida Atlantic Organic Geochemistry

Group) generated spreadsheet called “PIGCALC” (pigment calculation). This spreadsheet contains standardized equations and specific absorption coefficients and is

used to calculate the quantity of each pigment, sums and ratios of pigments and then

converts that information into taxonomic divisions of algae, (see Appendix I for a

simplified pigment calculation example). The ratios of interest from each spreadsheet

were extracted and placed in another spreadsheet where they were pooled according to

the species grown at different light intensities over a particular growth period. The ratios

of interest at the different light levels were extracted and plots were made of

CHLa/biomarker as well as biomass parameter/CHLa and detailed analyses made of the

generated ratios and plots.

Statistical analyses: All statistical analyses were carried out using PASW

statistics software (SPSS Inc.). Data was tested for homoscedasticity (F-test) and

heteroscedastic data were log transformed (Miller and Miller, 2005) before analysis. The

means of the CHLa/marker pigment, protein/CHLa, colloidal CHO/CHLa and storage

CHO/CHLa ratios over the high, medium and low light levels were made using one-way

analysis of variance (ANOVA), followed by post hoc analysis (Tukey and Games-

Howell tests). See appendix VI for detailed tables of the outputs of these statistical tests.

Isolation and characterization of a new pigment:

Samples (periphyton) were obtained from those collected as part of the Florida

Everglades Comprehensive Everglades Restoration Plan (CERP). This same pigment had been seen before in Scytonema hoffmanii cultures grown in the laboratory at 300 -1800

μmol photons·m-2·s-1. Samples were frozen and lyophilized prior to extraction. The freeze dried samples were ground with a mortar, and then steeped in acetone for two to three days to allow for complete extraction of all pigments. The extract was collected in a

50 mL syringe and passed through a 0.45 μm pore diameter Cameo Syringe Filter. The extracts were pooled and concentrated by evaporation. The dried pigment film was then re-dissolved in 90% acetone (1-2 mL) and reversed phase low pressure HPLC (LP-

HPLC) was used for bulk pigment separation.

The LP-HPLC system consisted of an Autochrom Products Model 500 ternary gradient HPLC pump with a Model 2360 gradient programmer, a 85 mm long Michel-

Miller (ACE Glass, Vincland, N. J.) column with an internal diameter of 8mm, a 300 mm long Michel-Miller Chromatographic injection column with an internal diameter of 22 mm, and a micro flow cell in aSpectronic-20 UV/Vis spectrophotometer, set to 430 nm.

This system was coupled to a DELL PC using Peak Simple® Chromatography data software. Both columns were packed with C18 Silica Premium Rf, end-capped, with a pore size of 70 Å and particle sizes between 20-45 µm. Gradient elution was initially carried out using the following three solvent systems: 40% solvent A (90%

Acetonitrile/water) and 60% solvent B (5M Ammonium Acetate in Methanol: water

80/20) from 0-5 minutes, 50% solvent B and 50% solvent C (100% Ethyl Acetate) from

5-25 minutes, 30% solvent A and 70 % solvent C from 25-65 minutes, and 100% solvent

C from 65-90 minutes. The column returned to 100% solvent A for the last 5 minutes and for storage.

The extraction method above allowed for the identification of chlorophylls and carotenoids from the Scytonema sp. cultures, but failed to isolate the unknown and

scytonemin pigments in good purity from the field samples. The isolation plan was to

collect the unknown pigment in higher (~ mg) amounts. With this in mind, the initial

extraction method was changed to extraction with 100% Methanol (x3) to remove most

of the chlorophylls and some of the carotenoids, and then 100% Acetone (x3) to obtain

most of the scytonemin pigments. The gradient program was changed to an isocratic

elution using 100% solvent A (90% Acetonitrile/water) for 0-90 minutes, followed by

100% solvent C (100% Ethyl Acetate) from 90-120 minutes. The new pigment was

collected at approximately 60 minutes and scytonemin was collected at approximately 75

minutes. All other procedures remained the same.

The filtrate was next evaporated and the dark solid film re-dissolved in 100%

acetone (new pigment and scytonemin fractions) and UV/Vis absorbance reading (190-

800 nm) taken of a 1 mL aliquot of each to determine if the correct fraction had been

collected. For further verification, the 1 mL aliquot of the acetone/pigment solution was

evaporated, and the solid was dissolved in 90% MADW and IS. Ion pairing was added

and 100 µL of this was injected on to the main (Waters 996) HPLC - PDA system used

for separating the pigments of all the other algal species (described previously). The rest

of the fractions collected were evaporated to dryness, and purged under Argon gas before

storage at -80 ºC.

IR analysis: The pure, evaporated unknown pigment was dissolved in methylene chloride; a thin film of this solution was placed on a Sodium Chloride (NaCl) window to dry. Analysis was carried out on a Thermo Scientific iS5 Fourier Transform IR

Spectrometer, coupled to a PC using OMNIC ® software.

Mass Spectrometry: Four mass spectrometry methods were used to assist in

elucidating the structure of the new pigment. The first and second methods were conducted at Florida Atlantic University, Boca Raton and the other two were done at the

University of Florida, Gainesville.

Matrix-Assisted Laser Desorption Ionization – Time of Flight Mass

(MALDI-TOF) Spectrometry: The puified new pigment and purified scytonemin pigment were added to an a-cyano - 4- hydroxycinnamic acid (CHCA) matrix and mass analysis was carried out on a MALDI-TOF mass spectrometer (Applied Biosystems

Model), coupled to a PC using Data Explorer Software®. The analyte and matrix were layered on a stage and the stage was bombarded with a laser beam (matrix-assisted laser desorption), which ionized the analyte, spalled the ions off the stage and into the electrostatic lenses. The electrostatic lenses guided the ions into the tube of the time of flight mass analyzer. The ions are initially filtered to have a specific kinetic energy. The velocities of the ions in the tube then vary inversely with their masses. The lighter ions move faster and arrive at the detector first.

Liquid Chromatography –Mass Spectroscopy (LC-MS): analyses was carried out on an Agilent Technologies 1200 series liquid chromatographic system, coupled to a

single quadrupole 6120 LC/MS. Mass analysis was done in both positive and negative

mode, with a 100 -1000 mass range. Electrospray ionization was the ionization method used. Column separation was achieved on a Phenomenex Luna® C18 (2) column, which

was 150 mm long with an internal diameter of 4.60 mm (5µm spherical particle size, 100

Å pore size, 400 m2 g-1 surface area). Gradient elution, with a flow rate of 0.8 mL/min

was achieved with the following two solvent systems, according to table 4 below:

Solvent A = Water: formic acid (1000:1) Solvent B = Acetonitrile: formic acid (1000:1)

Table 5: Gradient program used for LC-MS runs Time (mins) Solvent % A:B 0.00 80/20 2.50 80/20 15.00 10/90 20.00 10/90 20.50 80/20 24.00 80/20 24.01 80/20

Mass analysis at the University of Florida’s mass spectrometry facility: High

resolution mass spectrometry (HR MS) was obtained using an Agilent 6210 time of flight

(TOF) mass spectrometer, using an electrospray ionization (ESI) source. Sodium ions were incorporated into the ionization source. Thus, the major ion and that of the sodium adduct was obtained in the spectra. The sample was injected directly into the ionization source.

HPLC - MSn: (further fragmentation of parent ion) This was conducted using an

Agilent (Palo Alto, CA) 1100 series HPLC system coupled to a Thermo Finnigan (San

Jose, CA) mass spectrometer, with an ESI source. The HPLC system consisted of a

G1322A degasser, G1312A binary pump, a Thermo scientific Hypurity C8 column (5µm;

2.1 x 100mm + guard column), an Agilent 1100 G1314A UV/Vis detector (set at 254 nm)

and a Rheodyne 7125 manual injector with a 25µL injection loop. A 25µL Hamilton

1702 gastight syringe was used for sample injection. Gradient elution was carried out

using 2mM ammonium acetate in water (solvent A) and HPLC grade methanol (solvent

B) at 0.25 mL/min according to the following gradient program: A:B (min) = 95:5 (0)

through to 5:95 (45-60).

The ESI- MSn collision - induced dissociation (CID) product spectra were

obtained with 5 u isolation of the precursor ion, using 42.5 percent CID energy (qCID 0.3

and 30 ms). Nitrogen was used for the sheath gas (N2 = 65) and auxiliary gas (N2 = 5).

The heated capillary temperature was set at 250 º C, the spray voltage at 3.3kV and heated capillary voltage set at +12.5V (positive mode) and -10V (negative mode).

NMR analyses: 1D and 2D NMR spectra were obtained on a JEOL nuclear magnetic resonance spectrometer with standard pulse sequences operating at 600 MHz.

For the analyses: approximately 3 mg of purified pigment was dissolved in deuterated dimethyl sulfoxide (DMSO- d6). Delta software was used for analyzing the spectra. The

NMR instrument was located at FAU-Harbor Branch Oceanographic Institute in Fort

Pierce, Florida. Permission for instrument use was obtained from Dr. Amy Wright. Dr.

Wright’s post- doctoral associate, Dr. Priscilla Winder, assisted with these analyses.

Acetylation reactions: The purified pigment sample (~0.5 mg) was first lyophilized. Acetic anhydride and pyridine were next added and the acetylation reaction was allowed to progress overnight. Following the reaction, the solvent was evaporated and the sample was again lyophilized. LC-MS was then used to assess the increased

sample mass. These acetylation reactions were repeated several times (using different

pigment sample each time).

Deuterium exchange reactions: Deuterated methanol (~600 µL) was added to

the pigment sample and 1D and 2D NMR analyses was done to investigate if previously

seen OH and possibly NH proton signals had been exchanged with the heavier deuterated proton.

III. RESULTS - STATISTICAL ANALYSES

Eight unicellular algal species were grown in nutrient rich batch cultures under

three light conditions: low light (LL = 37 µmol photons·m-2·s-1), medium light (ML = 70-

75 µmol photons·m-2·s-1), and high light (HL= 200 µmol photons·m-2·s-1). Between 4 and

6 culture batches were grown for each species at each light level.

Significance of the algal species used in this study:

Microcystis is a common unicellular colonial cyanobacteria found in freshwater environments. The species belong to the phylum Cyanobacteria, order Chroococcales and family Microcystaceae. The existence of intracellular structures such as gas vesicles provides cells with buoyancy. Microcystis aeruginosa, which was used in this study,

occurs in large amounts on the surface waters of lakes and reservoirs in spring and

summer months. It is one of the most damaging species, due to its toxicity to aquatic and

terrestrial organisms and is known to occur in many Florida Lakes (Bigham et al., 2009;

Phlips et al., 2002; Ross et al., 2006).

Synechococcus spp. are oxygenic phototrophs that can photolyze either H2O or

H2S. Synechococcus is the main source of primary production in oligotrophic, pelagic marine, open, warm waters. They have been known to cause harmful but not directly toxic algal blooms in Florida Bay (Phlips et al., 1999). Harmful here, is under the definition of Paerl (1997), which includes blooms leading to anoxia, disruption of socio-

economic function, environmental change, and the like. The dominance of this species in the center of the bay may be attributable to it physiochemical characteristics: small size, buoyancy and tolerance to high light intensity (Phlips et al., 1999).

Dunaliella is a unicellular, ovoid, biflagellate, naked green alga. Twenty eight species of Dunaliella are presently recognized (Jayappriyan et al., 2010). The cells are motile and have two equal, long smooth whiplash flagella which belong to the order

Volvacales, family Polyblepharidaceae and the class of Chlorophyceae. It was first identified by a French scientist Michael Felix Dunal in 1838 and later it was re- discovered by Teodoresco in 1905. The unique morphological feature of Dunaliella is that it lacks a cell wall. The cell is enclosed by a thin plasma membrane or periplast, which permits rapid changes in cell shape and volume in response to osmotic changes. To survive, these organisms have high concentrations of β-carotene to protect against the intense light and high concentrations of glycerol to provide protection against osmotic pressure.

Scenedesmus is a non-motile alga consisting of 2, 4, and 8 elongated cells, often with long spines on the terminal cell (Smith 1916), belonging to the order

Chlorococcales, family Scenedesmaceae and the class Chlorophyceae. This genus is very common in eutrophic freshwater ponds and as planktonic forms in rivers and lakes; it is reported worldwide in all climates and is rarely found in brackish water (Wehr and

Sheath, 2003). Some species are produced in mass culture and used as food because of their protein and mineral content, or used for other purposes in biochemical industry

(Krauss and Thomas, 1954).

Rhodomonas of the phylum Cryptophyta, class Cryptophyceae and order

Pyrenomonadales, are a small group of marine flagellates which contain chloroplasts.

The cells are ovoid and flattened in shape with an anterior groove; two slightly unequal

flagella are present for locomotion. They occur in marine and brackish water (Jeffrey and

Vesk, 1990). This species contain fragile cell membranes and have the name hidden-plant

(crypto-phyte). Cryptophytes can be detected in oceanic populations by the presence of the marker pigment, alloxanthin (Gieskes and Kraay, 1983). These species are very fragile and are often lost in fixed samples, thus, specific pigment markers (viz. alloxanthin) are essential for their identification.

Cyclotella is a small, centric diatom with cells only 3-5 µm in diameter. This alga belongs to the phylum Bacillariophyta and family Stephanodiscaeae. The valves are short and drum shaped; the cells have long chitinouos bristles that help decrease settling.

Cyclotella meneghiniana, used in this study, is perhaps the best known species and is widely used in growth experiments (Mitrovic et al, 2010; Finlay et al., 2002; Tedrow et al., 2002; Bourne et al., 1992; inter alia). Species belonging to this genus, particularly

Cyclotella choctawatcheeana and Cyclotella striata have been found in Choctawatchee

Bay, Florida (Prasad et al., 1990). Members of the genus are also reported to form a dominant part of the planktonic assemblage in Florida Bay USA (Prasad and Nienow,

2006).

Thalassiosira species grow primarily in marine waters and belong to the family

Thalassiosiraceae and order Biddulphiales. The fultoportulae, or strutted processes, secrete β-chitin, which is considered to offer resistance to settling (Johansen and Theriot,

1987). Some species within the genus are found in estuaries, high conductance waters 58

and rivers, polluted ponds, and other aquatic systems that have been impacted by human activities (Spaulding and Edlund, 2009). Species belonging to this genus are widely used in the shrimp and shellfish larviculture industry and are considered by several hatcheries to be the single best algae for larval shrimp (Jensen et al., 2006).

Amphidinium spp are brown tide organisms with species that forms harmful algal blooms – toxins, physical irritants and noxious events. Most species produce toxins that affect humans and also fish (ichthyotoxic). Amphidinium belongs to the class

Dinophyceae and family Gymnodiniaceae. Amphidinium carterae, used in this study, has a dorso-ventrally compressed body with a very small epitheca (Hulburt, 1957). This species, as well as others in the genus are CFP (ciguatera fish poisoning) producers

(Hallegraeff et al., 1993; Anderson et al., 1987; Yasumoto et al., 1987; inter alia).

Analyses overview

Protein, colloidal carbohydrate (CHO), and storage CHO analyses were performed in triplicate. Pigments were analyzed during growth and at also harvest.

Relationships between Chlorophyll a (CHLa)-to- taxon-specific marker pigment; protein- to- CHLa; colloidal CHO-to- CHLa; storage CHO-to- CHLa in relation to light treatments were analyzed for each species using one-way analysis of variance (ANOVA).

The ratios (relationships) were used as the dependent variable and the light treatments were used as the independent variable. All statistical analyses were carried out at the 0.05 alpha level, using PASW® statistics version 18 software. The ANOVA F-test is very robust to mild departures from homogeneous variances (Lentner 1993). However, all ratios (except CHLa: marker pigment) were log transformed in an attempt to limit

departures from homogeneity. Levene’s test for homogeneity of variance was conducted in combination with the one-way ANOVAs. In addition, post hoc follow-up tests were also conducted, to assess difference in treatment (group) means if the one-way ANOVAs were significant. Tukey’s honestly significant difference (HSD) post hoc follow-up test was used if the homogeneity assumption was not violated and Games- Howell post hoc follow-up test was used for samples with non-homogeneous sample variances. For all the species in this study, one-way ANOVA tested the null hypothesis that the group means are not significantly different, that is, the three light treatments have the same effect on the ratios being investigated. Pertinent results for each species are presented below. The software output from the one-way ANOVA analyses and other statistical tests are tabulated in appendix VI. Cellular concentrations of chlorophyll a, protein, and the two functional classes of carbohydrates, as well as their relationship to biovolume for each species, at each light level are tabulated in Appendix VII. Typical chromatograms of all the species in this study are shown in Appendix VIII. Throughout this section, in text and in figures, the acronyms LL, ML and HL will be used for low, medium and high light levels respectively.

Synechococcus elongatus (Cyanophyta; cyanobacteria): The pigments identified for this species were: polar myxoxanthophyll (MYXOL), myxoxanthophyll

(MYXO), zeaxanthin (ZEA), canthxanthin (CANTH), chlorophyll a allomer (CHLa allo), chlorophyll a (CHLa), chlorophyll a epimer (CHLa’), echinenone (ECHIN), beta carotene (BETA). The taxonomically significant pigment identified for coccoidal and filamentous cyanobacteria are ZEA and ECHIN respectively (Nichols, 1973). These pigments are typically photoprotectorant pigments which change in relation to light 60

intensity, as a result different ratios were seen at each light treatment. This species was the only one that grew successfully at the dim (DL, 10 µmol photons·m-2·s-1) light treatments. The DL treatments had CHLa/ZEA and CHLa/ECHIN ratios from 6.07 to

6.76 and 50.09 to 68.03 respectively. The LL treatments had CHLa/ZEA and

CHLa/ECHIN ratios ranging from 3.85-5.52 and 57.13 to 68.89 respectively. The ML experiments gave CHLa/ZEA and CHLa/ECHIN ratios between 2.86 and 3.4 and from

28.57 to 37.31 respectively. The HL experiments had CHLa/ZEA and CHLa/ECHIN ratios ranging from 0.64 to 1.01 and 10.2 to15.61. See plots in Figure 7 (a- b) of these relationships. These pigment ratios were not log transformed prior to statistical analyses.

Levene’s statistical test for homogeneity of variance for the CHLa/ZEA ratios for the four light groups was not violated, as F (3, 18) = 2.194 and p = 0.124. The overall

ANOVA F was significant, with F (3, 18) = 237.968 and p < 0.001, indicating that at least one of the means was significantly different from the others. Tukey’s (HSD) post hoc follow up tests showed that all of the group means were significantly different from each other at the 0.05 level. The Levene’s statistical test was significant for the

CHLa/ECHIN ratios for the four light groups, with F (3, 18) = 3.858 and p = 0.027. The

Brown- Forsythe robust test of equality of means and the overall ANOVA F were also significant. The Games-Howell post hoc follow tests showed that all the means except for the DL and LL were significantly different from each other.

(a)

8 6.49 6 4.73 /ZEA

a 4 3.06

CHL 2 0.87 0 0 50 100 150 200 250 Light intensity (μmol photons·m-2·s -1)

(b)

60 58.06 63.12

/ECHIN 40 a 34.18 20

CHL 13.62 0 0 50 100 150 200 250 Light Intens ity (μmol photons·m-1·s -1)

Figure 7: Synechococcus elongatus (a): CHLa/ZEA v Light Intensity; (b): CHLa/ECHIN v Light Intensity

Synechococcus elongatus – protein/CHLa relationships: Protein concentrations were assessed for the four light levels that batch cultures of this species were grown under, revealing the following: Concentrations of protein in the dim light (DL, 10 µmol photonsss·m-2·s-1) experiments ranged from 12.9 to 59.46 pg cell-1; The LL experiments showed concentration ranges from 25.5 pg cell -1 to 675 pg cell-1, while the ML and HL experiments had protein concentrations of from 31.5 pg cell -1 to 430 pg cell-1 and 600 pg cell-1 to 1986 pg cell -1 respectively. The protein/CHLa ratios for this experimental group

are as follows: DL experiments resulted in ratios between 3.07 and 43.96, the LL experiment ratios were from 110.43 to 136.28; ML ratios were between 161.04 and

195.69 and the HL ratios ranged from 371 to 596.10. All ratios were log transformed prior to statistical analyses to avoid possible violation of homogeneity of variances assumptions. Regression and dot plots are shown in Figure 8 (a - d).

Levene’s statistic gave an F value that was not significant, with F (3, 16) = 2.544 and p = 0.093. This indicated that the variances in the means were not significant.

However, the overall F ANOVA was significant, with F (3, 16) = 278.265 and p < 0.001, indicating that at least one of the groups of means is different from the others. Tukey’s

(HSD) post hoc follow-up tests showed the following: The DL group (M= 1.78) was significantly different from the LL group (M = 2.09), with a mean difference of -0.31 and a p value < 0.001. The DL group (M = 1.78) was significantly different from the ML group (M = 2.24), with a mean difference of -0.46 and a p value < 0.001. The DL group

(M = 1.78) was also significantly different from the HL group (M = 2.28), with a mean difference of -0.50 and a p value < 0.001. The LL group (M = 2.09) was also significantly different from the ML group (M = 2.24), with a mean difference of -0.16, with a p value

< 0.001. The LL group (M = 2.09) was also significantly different from the HL group (M

= 2.28), with a mean difference of -0.193 and a p value < 0.001. The ML (M = 2.24) was however, not significantly different from the HL group (M = 2.28), with a mean difference of -0.03 and a p value of 0.228.

(a) 2500 ) -1 2000 DL 1500 LL 1000 ML 500 HL cell (pg Protein 0 ‐50‐500 51015 CHLa (pg cell-1)

(b)

2000 y = 7.1073x ‐ 105.54 R² = 0.9282 )

-1 1500 1000

500

Protein (pg cell (pg Protein 0 0 50 100 150 200 250 ‐500 Light Intensity (µmol photons·m-2·s -1)

-1 8 6 4 2 CHLa (pg cell (pg CHLa 0 0 50 100 150 200 250 Light Intensity (µmol photons·m-2·s -1)

Figure 8: Synechococcus elongatus (a): Protein v CHLa; (b): Protein v Light Intensity; (c): CHLa v Light Intensity

(d)

2.4 DL (log) LL a 2 ML HL 1.6 protein/CHL 0 50 100 150 200 250

Light intensity (µmol photons·m-2·s -1)

Figure 8 contd.: Synechococcus elongatus; (d): Light treatment effect on protein/CHLa ratios (dot plot)

Synechococcus elongatus – colloidal carbohydrates/CHLa relationships: The

concentration of colloidal carbohydrates for the DL experiment group ranged from 2.22

to 7.48 pg cell-1 and colloidal CHO/CHLa ratios ranged from 4.43 to 5.98. The LL

experiments had colloidal carbohydrate concentrations between 1.82 and 60.3 pg cell-1,

with ratios from 9.54 to 10.87. The ML experiments gave colloidal carbohydrate

concentrations ranging from 1.97 to 24.1 pg cell-1, with colloidal CHO/ CHLa ratios from

10.01 to 11.25. The HL experiments had colloidal carbohydrate concentrations between

43.8 pg cell-1 and 114 pg cell-1 and colloidal CHO/CHLa ratios from 11.90 to 13.79. All

ratios were log transformed prior to statistical analyses. See regression and dot plots in

Figure 9 (a - c).

(a)

) 150 -1 DL 100 LL

ML 50 HL 0 ‐50 51015 ‐50 Colloidalcell (pg CHO CHLa (pg cell-1)

(b)

) 150

-1 y = 0.4469x ‐ 3.7532 100 R² = 0.8897

0 0 50 100 150 200 250 Light Intensity (µmol photons·m-2·s -1) Colloidal CHO (pg cell (pg CHO Colloidal

(c)

1.2 (log) DL

a 1.1 LL 1 ML 0.9 HL 0.8 0 50 100 150 200 250 Colloidal CHO/CHL Light intensity (µmol photons·m-2·s -1)

Figure 9: Synechococcus elongatus (a): Colloidal CHO v CHLa; (b): Colloidal CHO v Light Intensity; (c): Light treatment effect on colloidal CHO/CHLa ratios (dot plot)

Levene’s test for homogeneity of variance in the group means was not significant, with F (3, 16) = 1.805 and p = 0.187. This proved that the assumption of homogeneity of

variance had not been violated. However, the overall ANOVA F was significant, with F

(3, 16) = 12.487 and p < 0.001, which means that at least one of the experimental group

means is significantly different from the others.

Tukey’s (HSD) post hoc follow-up tests was then done to assess which of the

group means was significant. The following results were obtained: the DL group (M =

0.96) was not significantly different from the LL (M = 0.99), with a mean difference of -

0.04 and a p value of 0.503. The DL group (M = 0.96) was also not significantly different

from the ML group (M = 1.02), with a mean difference of -0.06 and a p value of 0.096.

The DL group (M = 0.96) was significantly different from the HL group (M = 1.10), with

a mean difference of -0.14 and a p value < 0.001. The LL group (M = 0.99) was not

significantly different from the ML group (M = 1.02), with a mean difference of -0.02

and a p value of 0.696. The LL group (M = 0.99) was also significantly different from the

HL group (M = 1.10), with a mean difference of -0.11 and a p value < 0.001.

Additionally, the ML group (M = 1.02) was significantly different from the HL group (M

= 1.10) with a mean difference of -0.08 and a p value of 0.008.

Synechococcus elongatus – Storage carbohydrate (CHO)/CHLa relationships: The

storage CHO concentrations and subsequent storage CHO/CHLa ratios were determined

for the four light levels used in the study for this species and are as follows: the DL

experiment group had storage carbohydrate concentrations between 23.6 and 93.5 pg cell

-1 and storage CHO/CHLa ratios from 39.51 to 45.17. The LL experiments showed

storage CHO concentrations from 10.6 to 337 pg cell-1 and storage CHO/CHLa ratios 67

between 56.42 and 62.17. The ML experiments had storage CHO concentrations between

10.4 pg cell-1 and 142 pg cell-1, and storage CHO/CHLa ratios from 76.38 – 106.26.

The HL experiment group had storage CHO concentrations from 161 to 590 pg cell-1 and

storage CHO/CHLa ratios between 105.22 and 178.98. The storage CHO/CHLa ratios

were log transformed prior to statistical analyses in an attempt to prevent violation of the

assumption of homogeneity of variance in the group means. See plots of these

relationships in Figure 10 (a - c).

Levene’s test for homogeneity of variance was not significant, with F (3, 16) =

0.566 and p = 0.646, indicating that the assumption had not been violated. The overall

ANOVA F was significant, with F (3, 16) = 32.465 and p < 0.001, indicating that at least one of the group means was significantly different from the others. Tukey’s (HSD) post hoc follow-up test was then carried out to identify which of the group means was significant. The following was determined: The DL experiment group (M = 1.62) was significantly different from the LL group (M = 1.77), with a mean difference of – 0.15 and p value < 0.001. The DL group (M = 1.62) was significantly different from the ML group (M = 1.74), with a mean difference of -0.11 and a p value < 0.001. The DL group

(M = 1.62) was also significantly different from the HL group (M = 1.74), with a mean difference of -0.12 and p value < 0.001. The LL group (M = 1.77) was not significantly different from the ML group (M = 1.74), with a mean difference of 0.03 and a p value of

0.136. The LL group (M = 1.77) was not significant from the HL group (M = 1.74), with a mean difference of 0.03 and a p value of 0.305. The ML group (M = 1.74)

(a) ) -1 800 DL 600 LL

400 ML 200 HL

0 Storage CHO cell (pg CHO Storage ‐50 51015 ‐200 CHLa (pg cell-1)

(b)

) 500 y = 1.87x + 9.9563 -1 R² = 0.852 400 300 200 100 0 Storage CHO (pg cell 0 50 100 150 200 250 Light Intensity (µmol photons·m-2·s -1)

(c)

1.9 DL

LL 1.7 ML HL

1.5

0 50 100 150 200 250 storage CHO/CHLa (log)

Light Intensity (µmol photons·m-2·s -1)

Figure 10: Synechococcus elongatus (a): Storage CHO v CHLa; (b): Storage CHO v Light Intensity; (c): Light treatment effect on storage CHO/CHLa ratios (dot plot)

was also not significantly different from the HL group (M = 1.74), with a mean

difference of -0.006 and a p value of 0.973.

Microcystis aeruginosa (Cyanophyta; blue-green algae): the pigments identified for this species are: polar myxoxanthophyll (MYXOL), myxoxanthophyll (MYXO), zeaxanthin (ZEA), canthxanthin (CANTH), chlorophyll a allomer (CHLa allo), chlorophyll a (CHLa), chlorophyll a epimer (CHLa’), echinenone (ECHIN), beta

carotene (BETA). The taxonomically significant pigment identified for coccoidal and

filamentous cyanobacteria are ZEA and ECHIN respectively (Nichols, 1973). These

pigments are typically photoprotectorant pigments which change in relation to light

intensity, as a result different ratios were seen at each light treatment (Grant and Louda

2010). The LL treatments had CHLa/ZEA and CHLa/ECHIN ratios ranging from 25.14

-29.12 and 15.32 – 22.92 respectively. The ML experiments gave CHLa/ZEA and

CHLa/ECHIN ratios between 16.88 – 21.12 and 17.78 – 27.31 respectively. The HL

experiments had CHLa/ZEA and CHLa/ECHIN ratios ranging from 9.21 – 12.34 and

13.26 – 16.23 respectively. These pigment ratios were not log transformed. See plots of

these relationships in Figure 11 (a - b).

The Levene’s statistical test for homogeneity of variance for the means of the

CHLa/ZEA ratios for the three light groups was not violated, as F (3, 13) = 0.477and p =

0.631. The overall ANOVA was significant, with F (3, 13) = 171.783 and p < 0.001,

indicating that at least one of the means was significantly different from the others.

Tukey’s post hoc follow up tests showed that all of the group means were significantly

different from each other at the 0.05 level. 70

(a)

25 26.67 20 19.35 /ZEA a 15 10 10.67 CHL 5 0 0 50 100 150 200 250 Light intensity (μmol photons.m-2.s -1)

(b) 25 22.01 20 18.68 15 14.6 10

CHLa/ECHIN 5 0 0 50 100 150 200 250 Light intensity (μmol photons .m-2.s-1)

Figure 11: Microcystis aeruginosa (a) CHLa/ZEA v Light Intensity; (b): CHLa/ECHIN v Light Intensity

Levene’s statistical test for homogeneity of variances for the CHLa/ECHIN ratios

over the three light levels was not significant since F (2, 13) = 2.672 and p = 0.107,

indicating that the homogeneity of variances assumption had not been violated. The

overall ANOVA F was significant, with F (2, 13) = 10.152 and p = 0.002. Tukey’s (HSD) post hoc follow up test showed that the LL (M = 18.68) group was not significantly different from the ML (M = 22.01) group, with a mean difference of -3.33 and a p value

of 0.170. The LL group was also not significantly different from the HL (M = 14.60)

group, with a mean difference of 4.03 and a p value of 0.068. The ML (M = 22.01) group

was however, significantly different from the HL (M = 14.60) group, with a mean

difference of 7.41 and a p value of 0.002.

Microcystis aeruginosa – protein/CHLa relationships: The LL experiments gave protein

concentrations between 3.36 pg cell-1 and 7.06 pg cell-1 and protein/CHLa ratios from

49.28 to 60.61. The ML experiments resulted in protein concentrations from 3.65 pg cell-

1 and 16.00 pg cell-1; with protein/CHLa ratios between 58.94 and 71.60. The HL

experiments gave protein concentration for this species ranging from 16.80 pg cell-1 to

85.90 pg cell-1 and protein/CHLa ratios between 72.11 and 95.89. All ratios were log transformed prior to statistical analyses. Regression and dot plots are shown in Figure 12

(a - d).

Levene’s statistic to test for homogeneity of variances in the group means for the

different light treatments was not significant as F (2, 14) = 0.005 and p = 0.995,

indicating that the assumption had not been violated. The overall ANOVA F was

significant, with F (2, 14) = 29.249 and p < 0.001, indicating that at least one of the group

means was significantly different from the others. Tukey’s (HSD) post hoc follow- up

tests showed that all the means were different from each other according to the following

(a)

100 ) -1 80 60 LL 40 ML 20

Protein (pg cell (pg Protein 0 HL 0 0.2 0.4 0.6 0.8 1 CHLa (pg cell-1)

(b)

60 y = 0.2788x ‐ 7.5557 )

-1 50 R² = 0.9626 40 30 20 10 Protein (pg cell (pg Protein 0 0 50 100 150 200 250 Light Intensity (µmol photons·m-2·s -1)

) y = 0.0031x ‐ 0.0367 -1 0.6 R² = 0.9512 0.4

0.2 CHLa (pg cell (pg CHLa 0 0 50 100 150 200 250 Light Intensity (µmol photons·m-2·s -1)

Figure 12: Microcystis aeruginosa (a): Protein v CHLa; (b): Protein v Light Intensity; (c): CHLa v Light Intensity

(d)

2.0

(log) LL

a ML 1.8 HL

Protein/CHL 1.6 0 50 100 150 200 250

Light Intensity (µmol photons·m-2·s -1)

Figure 12 contd.: Microcystis aeruginosa (d): Light treatment effect on protein/CHLa ratios (dot plot)

results: The LL (M = 1.73) group was significantly different from the ML (M = 1.80)

group, with a mean difference of -0.72 and a p value of 0.023. The LL (M = 1.73) group.

was also significantly different from the HL (M = 1.91) group, with a mean difference of

-0.18 and a p value < 0.001. The ML (M = 1.80) group was also significantly different

from the HL (M= 1.91) group, with a mean difference of -0.11 and a p value of 0.001

Microcystis aeruginosa– colloidal CHO/CHLa relationships: Colloidal carbohydrate

concentration in the LL experiments ranged from 0.132 pg cell-1 to 0.862 pg cell-1, the

ML had concentrations between 0.122 pg cell-1 and 0.849 pg cell-1, while those in the HL

experiments were between 1.19 pg cell-1 and 6.49 pg cell-1. The colloidal CHO/CHLa

ratios were between 5.09 and 6.56 for the LL experiments, between 5.79 and 8.34 for the

ML experiments and between 5.75 and 7.45 for the HL experiments respectively. The

colloidal CHO/CHLa ratios were log transformed before statistical analyses, in an

attempt to meet the homogeneity of variances assumption. Regression and dot plots are

shown in Figure 13 (a - c).

Levene’s test for homogeneity of variances, was not violated as F (2, 14) = 0.596 and p = 0.564. The overall ANOVA was significant, with F (2, 14) = 4.750 and p =

0.027, indicating that at least one of the group means was significantly different from the

others. Tukey’s (HSD) post hoc follow-test gave the following results:

(a) 8

) -1 6 LL 4 ML 2 HL 0

Colloidalcell (pg CHO 0 0.2 0.4 0.6 0.8 1 CHLa (pg cell-1)

(b) 5 y = 0.0231x ‐ 0.4487 )

-1 R² = 0.9654 4 3

2 1

Colloidal CHO (pg cell (pg CHO Colloidal 0 50 100 150 200 250 Light Intensity (µmol photons·m-2·s -1)

Figure 13: Microcystis aeruginosa (a): Colloidal CHO v CHLa; (b): Colloidal CHO v Light Intensity

0.9 LL

(log) 0.85

a ML 0.8 0.75 HL 0.7

0.65 0.6 Colloidal/CHL 0 50 100 150 200 250 Light Intensity (µmol photons·m-2·s -1)

Figure 13 contd.: Microcystis aeruginosa (c): Light treatment effect on colloidal CHO/CHLa ratios (dot plot)

The LL (M = 0.76) group was significantly different from the ML (M = 0.837) group,

with a mean difference of -0.078 and a p value of 0.048. The LL (M = 0.76) was also

significantly different from the HL (M = 0.841) group, with a mean difference of -0.814

and a p value of 0.037. However, the ML (M = 0.837) was not significantly different

from the HL (M = 0.841) group, with a mean difference of -0.0038 and a p value of

0.990.

Microcystis aeruginosa – storage CHO/CHLa biomass relationships: The LL

experiments gave storage carbohydrate concentrations between 2.19 and 10.85 pg cell-1;

The ML experiments gave concentrations between 4.50 pg cell-1 and 15.54 pg cell-1; The

HL experiments gave concentrations between 23.68 pg cell-1 and 95.62 pg cell-1. The

storage CHO/CHLa ratios were from 75.72 to 89.14 for the LL experiments, from 62.56

and 100.13 for the ML experiments and from 96.29 to 115.74 for the HL experiments

(a)

150 ) -1

100 LL ML 50 HL

0 0 0.2 0.4 0.6 0.8 1 Storage CHO (pg cell (pg CHO Storage CHLa (pg cell-1)

(b) y = 0.345x ‐ 8.0582 80 )

-1 R² = 0.9473 60 40 20 0 0 50 100 150 200 250 Storage CHO cell (pg CHO Storage Light Intensity (µmol photons·m-2·s -1)

1.5 050100150200250 -2 -1 Storage CHO/CHLa (log) Light Intensity (µmol photons·m ·s )

Figure 14 : Microcystis aeruginosa (a) Storage CHO v CHLa; (b): Storage CHO v Light Intensity; (c): Light treatment effect on storage CHO/CHLa ratios (dot plot)

respectively. All ratios were log transformed prior to statistical analyses. Plots of some of

these relationships are shown in Figure 14 (a - c) above.

The Levene statistic was not significant as F (2, 14) = 1.999 and p =0.172,

indicating that that the variances in the group means was equal. The overall ANOVA F

was significant, with F (2, 14) = 8.459 and p = 0.004. Tukey’s (HSD) post hoc follow-up

test gave the following results: the LL (M = 1.919) group was not significantly different

from the ML (M= 1.921) group, with a mean difference of -0.0021 and a p value of

0.997. The LL was significantly different from the HL (M = 2.03) group, with a mean

difference of -0.1078 and a p value of 0.010. The ML (1.921) group was also

significantly different from the HL (M = 2.03) group, with a mean difference of -0.1057

and a p value of 0.008.

Dunaliella tertiolecta (Chlorophyta; green algae): The pigment composition for this

species is as follows: chlorophyllide b (CHLideb), chlorophyllide a (CHLide a),

pyrochlorophyllide a (pCHLidea), neoxanthin (NEO), violaxanthin (VIOLA), antheraxanthin (ANTH), lutein (LUT), chlorophyll b (CHLb), chlorophyll a allomer

(CHLa allo), chlorophyll a (CHLa), chlorophyll a epimer (CHLa’), pheophytin (PHtin a), alpha carotene (ALPH), beta carotene (BETA). Typical chromatograms of the eight species are presented in appendix VIII. The most stable class/pigment group specific marker was identified to be chlorophyll b (Jeffrey and Vesk 1997). The molar ratios of chlorophyll a/ chlorophyll b (CHLa/CHLb) in the LL experiments ranged from 2.26 to

2.58, the ML ratios ranged from 2.24 to 2.92 and those for the HL experiments were

between 1.77 and 2.41. Figure 15 (a - b) shows the plots of these ratios per culture batch

over the light levels studied.

Levene’s test was done to check the assumption that the variances of the three

light levels are equal. The Levene’s test is not significant, F (2, 12) = 0.622, p = 0.553 and the homogeneity of variance assumption is not violated. However, the F ratio

(ANOVA) is significant at the 0.05 level: F (2, 12) = 4.542, p = 0.034.

(a)

3 2.54 2.5 2.41 2 2.14 1.5 1 CHLa/CHLb 0.5 0 0 50 100 150 200 250 Light Intensity (µmol photons·m-2·s -1)

(b) 4 LL b 3 ML

/CHL 2 a HL

CHL 1

0 02468 culture batches

Figure 15: Dunaliella tertiolecta (a): CHLa/CHLb v Light Intensity; (b): CHLa/CHLb per batch

As the means are significantly different, but homogeneity has not been violated, the

Tukey HSD (honestly significant difference) follow-up tests were done to determine which means differ from the other. The Tukey HSD test revealed that LL group (M =

2.41) is not significantly different from the ML group (M = 2.54), with a mean difference of -0.13 and a p value of 0.68. Also the LL group (M = 2.41) is not significantly different from the HL group (M = 2.14), with a mean difference of 0.27 and a p value of 0.19.

However, the ML group (M = 2.41) is significantly different from the HL (M = 2.14), with a mean difference of 0.39 and a p value of 0.03.

Dunaliella tertiolecta - protein/CHLa relationships: The LL experiments resulted in

protein content between 48.19 pg cell-1 and 62.32 pg cell-1 and protein/CHLa ratios from

69.91 to 73.02. The protein content of the ML experiments was between 33.07 pg cell-1

and 114.78 pg cell-1, with protein/CHLa ratios from 73.67 to 132.62. Additionally the HL

experiments showed protein content between 109.84 pg cell-1and 163.04 pg cell-1 and protein/CHLa ratios with a minimum of 376.24 and maximum of 482.36. The protein/CHLa ratios for the light levels were log transformed in order to meet the homoscedasticity assumption of the data. Regression and dot plots of some pertinent relationships are shown in Figure 16 (a - d).

Levene’s test was done prior to one-way ANOVA analysis to check the assumption that the variances of the three light level experiments are equal. The Levene’s test is significant: F (2, 12) = 14.618, p = 0.001 at the 0.05 alpha level. Thus, the assumption of homogeneity is not met. A more robust test for equality of means was carried out and it was also significant.

(a)

300 ) LL -1 200 ML HL 100

cell (pg Protein 0 00.511.5

-1 CHLa (pg cell )

(b)

200

) y = 0.6727x + 24.307 -1 150 R² = 0.985

100

50 Protein (pgcell Protein 0 0 50 100 150 200 250 -2 -1 Light Intensity (µmol photons·m ·s )

1.2 R² = 0.4358 )

-1 1 0.8 0.6 0.4 0.2 CHLa (pg cell (pg CHLa 0 0 50 100 150 200 250

Light Intensity (µmol photons·m-2·s -1)

Figure 16: Dunaliella tertiolecta (a): Protein v CHLa; (b): Protein v Light Intensity; (c): CHLa v Light Intensity

(d) 2.6

(log) a LL 2.1 ML HL

protein/CHL1.6 0 50 100 150 200 250 Light Intensity (µmol photons·m-2·s -1)

Figure 16 contd.: Dunaliella tertiolecta (d): Light treatment effect on protein/CHLa ratios (dot plot)

That is: Brown-Forsythe F (2, 9.557) = 10.158, p = 0.004.The one-way ANOVA, F ratio

is also significant: F (2, 12) = 34.08, p < 0.001. Since the assumption of homogeneity was

not met, even after log transformation, the Games-Howell post-hoc follow-up test was

done to assess which of the means from the three groups differed from each other. The

results are as follows: the LL group (M = 1.85) is significantly different from the ML

group (M = 2.04), with a mean difference of -0.19 and a p value of 0.006. The LL (M =

1.85) group is also significantly different from the HL group (M = 2.24), with a mean

difference of -0.39 and a p value < 0.001. Additionally the ML group (M = 2.04) is significantly different from the HL group (M = 2.24), with a mean difference of - 0.20

and a p value < 0.05.

Dunaliella tertiolecta - colloidal CHO/CHLa relationships: colloidal carbohydrate

content in the LL experiments ranged between 6.11 pg cell-1and 7.38 pg cell-1, while the

ML experiments ranged between 1.31 pg cell-1 and 5.09 pg cell-1, and the HL ranged

between 2.32 pg cell-1 and 7.78 pg cell-1. Additionally, the colloidal CHO/CHLa ratios

for the LL experiments were between 7.50 and 9.28, the ML experiments between 3.23 and 5.88, while the HL group had ratios between 3.99 and 7.03. The colloidal

CHO/CHLa ratios for the light levels were log transformed in order to meet the homogeneity assumption for the data. Regression and dot plots are shown in Figure 17 (a

- c).

Results from the Levene’s test show that it is not significant, with F (2, 12) =

3.748 and p = 0.054. Therefore, the assumption that the variances of the three light levels

are equal can be retained. However, one-way ANOVA gives an F ratio that is significant,

as F (2, 12) = 10.158 and p = 0.004. This resulted in a rejection of the null hypothesis that

the sample means from the light levels are equal. Tukey’s (HSD) post hoc follow-up test

showed the following results: The LL group (M = 0.93) is significantly different from the

ML group (M = 0.65), with a mean difference of 0.28 and a p value of 0.003. Also, the

LL group (M = 0.93) is significantly different from the HL group (M = 0.74), with a

mean difference of 0.19 and a p value of 0.033. However, the HL (M = 0.74) group is not

significantly different from the ML (M = 0.65), with a mean difference of 0.09 and a p value of 0.309.

(a)

) 10

-1 LL 8 ML 6 HL 4 2 0 00.511.5

cell (pg CHO Colloidal CHLa (pg cell-1)

(b) y = ‐0.0013x + 5.0693 )

-1 8 R² = 0.0035

4 2 0 0 50 100 150 200 250 Colloidal CHO (pg cell (pg CHO Colloidal -2 -1 Light Intensity (µmol photons·m ·s )

(log) 1 LL a ML 0.8 HL 0.6

0.4 Colloidal/CHL 0 50 100 150 200 250

Light Intensity (µmol photons·m-2·s -1)

Figure 17: Dunaliella tertiolecta (a): Colloidal CHO v CHLa; (b): Colloidal CHO v Light Intensity; (c): Light treatment effect on Colloidal CHO/CHLa ratios (dot plot)

Dunaliella tertiolecta - storage CHO/CHLa relationships: The LL experiments had storage carbohydrate content between 44.68 and 50.45 pg cell-1, while the storage

CHO/CHLa ratios were between 57.23 and 64.82. The ML experiments had storage carbohydrate content from 32.10 pg cell-1 to 84.28 pg cell-1 and the storage CHO/CHLa ratios were between 71.50 and 97.33. The HL experiments had storage carbohydrate content from a minimum of 79.01 pg cell-1 to a maximum of 116.56 pg cell-1, while the storage CHO/CHLa ratios ranged from 93.50 to 115.57. Plots of these relationships are shown in Figure 18 (a - c). The storage CHO/CHLa ratios were log transformed in order to meet the homogeneity assumption for the data.

Levene’s test indicated that assumption of homogeneity of variance had not been violated, that is: F (2, 12) = 1.080, p = 0.370. One - way ANOVA results showed that at least one of the group means was significantly different from the others, indicating rejection of the null hypothesis: F (2, 12) = 50.279, p < 0.001. The Tukey (HSD) post hoc analysis was done, and showed that all three group means were significantly different: the

LL group (M = 1.78) is significantly different from the ML group (M = 1.93), with a mean difference of - 0.16and a p value < 0.001. Also, the LL group (M = 1.78) is significantly different from the HL group (M = 2.04), with a mean difference of -0.26 and a p value < 0.001. The ML group (M = 1.93) is significant from the HL group (M =

2.04), with a mean difference of- 0.10 and a p value of 0.003.

(a)

150 ) LL -1 100 ML HL 50

0 00.511.5

cell (pg CHO Storage CHLa (pg cell-1)

(b)

y = 0.2981x + 37.284

) 120

l-1 100 R² = 0.9029 80 60 40 20 0

cel (pg CHO Storage 0 50 100 150 200 250 Light Intensity (µmol photons·m-2·s -1)

Storage CHO/CHL Light Intensity (µmol photons·m-2·s-1)

Figure 18 : Dunaliella tertiolecta (a): Storage CHO v CHLa; (b): Storage CHO v Light Intensity; (c): Light treatment effect on storage CHO/CHLa ratios (dot plot)

Scenedesmus quadricauda (Chlorophyta: green algae): The pigment composition for this species is as follows: chlorophyllide b (CHLideb), chlorophyllide a (CHLide a), pyrochlorophyllide a (pCHLidea), neoxanthin (NEO), violaxanthin (VIOLA), antheraxanthin (ANTH), lutein (LUT), chlorophyll b (CHLb), chlorophyll a allomer

(CHLa allo), chlorophyll a (CHLa), chlorophyll a epimer (CHLa’), pheophytin (PHtin a), BETA carotene (BETA). Typical chromatograms of the eight species are presented in appendix VIII. The most stable class/pigment group specific marker was identified to be chlorophyll b (Jeffrey and Vesk 1997). The molar ratios of chlorophyll a/ chlorophyll b

(CHLa/CHLb) in the LL experiments ranged from 2.52 to 2.88, ML experiment ratios ranged from 2.07 to 2.92 and those for the HL experiments were between 1.98 and 3.18.

Plots of these ratios per culture batch, over the light levels studied are shown in Figure 19

(a - b).

The Levene’s test was done prior to one-way ANOVA to test the homogeneity assumption. This test showed that the homogeneity assumption had not been violated: F

(2, 14) = 3.647, p = 0.053. Going further, the one-way ANOVA results showed that null hypothesis can be retained, as F (2, 14) = 0.355 and p = 0.708. That is, the CHLa: marker pigment ratios were not significantly different over the three light treatments.

(a)

3.5 3

b 2.74 2.5 2.69 2.57 2 /CHL

a 1.5 1 CHL 0.5 0 0 50 100 150 200 250 Light Intensity (µmol photons·m-2·s -1)

(b)

4 LL 3 b ML 2

/CHL HL a 1 CHL 0 02468 culture batches

Figure 19: Scenedesmus quadricauda (a): CHLa/CHLb v Light Intensity; (b): CHLa/CHLb ratios per batch

Scenedesmus quadricauda - protein/CHLa relationships: The LL experiments resulted in

protein concentrations between 13.48 pg cell -1 and 93.34 pg cell -1 and protein/CHLa

ratios between 488.57 and 854.70. The ML experiments gave protein concentrations between 50.50 pg cell -1 and 189 pg cell-1, with protein/CHLa ratios from 164.78 to

236.13. The HL experiments had protein concentrations between 489 pg cell -1 and 1239

(a) 1500

)

-1 LL 1000 ML HL 500

(pg Protein cell 0 0 5 10 15

CHLa (pg cell-1)

(b)

800 y = 4.2189x ‐ 139.71 ) -1 R² = 0.9854 600

400

200

Proteincell (pg 0 0 50 100 150 200 250 Light Intensity (µmol photons·m-2·s-1)

(c)

8 y = 0.0272x + 0.765 )

-1 R² = 0.5739 6

cell (pg CHLa 2 0 0 50 100 150 200 250

Light Intensity (µmol photons·m-2·s-1)

Figure 20: Scenedesmus quadricauda (a): Protein v CHLa; (b): Protein v Light Intensity (c): CHLa v Light Intensity

(d) 3.2

a LL 2.7 ML

2.2 HL

Protein/CHL 1.7 0 50 100 150 200 250 Light Intensity (µmol photons·m-2·s -1)

Figure 20 contd.: Scenedesmus quadricauda (d): Light treatment effect on protein/CHLa ratios (dot plot)

pg cell -1 and protein/CHLa ratios from 118.12 and 137.19. Plots of some of these relationships are shown in Figure 20 (a - d). Protein/CHLa ratios were log transformed prior to statistical analysis to test the homogeneity of variances assumption.

Levene’s test was significant, with F (2, 15) = 15.945 and p < 0.001, indicating that the assumption was violated. A robust test of equality of means was done using the

Brown-Forsythe statistic, but the null hypothesis that the variances are equal was still violated. This indicated that at least one of the group means was significantly different from the others. That is, the adjusted F (2, 5.449) equaled 81.411 and p < 0.001. One- way ANOVA also resulted in a rejection of the null hypothesis, as F (2, 15) = 99.450 and p < 0.001. Since the homogeneity of variances assumption was not met, the Games-

Howell post hoc follow up test was done to assess which of the group means was significant from the others. The test showed that the LL (M= 2.77) group was significantly different from the ML (M= 2.26) group with a mean difference of 0.52 and a

p value of 0.001. The LL (M=2.27) was also significantly different from the HL

(M=2.10) group with a mean difference of 0.67 and a p value of 0.001. The ML

(M=2.26) group was also significantly different from the HL (M=2.10) group, with a

mean difference of 0.15 and a p value of 0.001.

Scenedesmus quadricauda – colloidal CHO/CHLa relationships: The LL experiments

had colloidal carbohydrate concentration ranging from 2.85 pg cell-1 to 19.65 pg cell-1 and colloidal CHO/CHLa ratios from 51.31 and 179.91. The ML colloidal concentration range was between 1.88 pg cell-1 and 6.94 pg cell-1 and colloidal CHO/CHLa ratios from

6.12 to 10.06. The HL experiments gave colloidal CHO concentration between 30.06 pg

cell-1 and 74.55 pg cell-1 and colloidal CHO/CHLa ratios from 6.75 to 7.87. Plots of

pertinent relationships are given in Figure 21(a - c). The ratios were log transformed prior

to statistical analyses, to assess the assumption that variances in the group means (three light levels) were homogeneous.

The Levene’s test was significant, with F (2, 15) = 4.723 and p = 0.026, indicating

rejection of the null hypothesis. A more robust test for equality of means was done and

this was also significant. That is, the Browne- Forsythe analysis yielded F (2, 5.939) =

128.928 and p < 0.001. The overall one-way ANOVA F was also significant, with F (2,

15) = 152.592 and p < 0.001. This indicated that the null hypothesis that the three group

means are equal would be rejected. Since the homogeneity of variances had been violated, the Games- Howell post hoc follow-up test was done to assess which of the group means was significantly different from the others. The test showed that the LL (M

(a) 100 )

-1 80 LL 60 ML 40 HL 20 0 ‐20 024681012 Colloidal CHO (pg cell (pg CHO Colloidal CHLa (pg cell-1)

(b) y = 0.2215x ‐ 3.9731

) 50 R² = 0.9106 -1 40 30 20 10 0 Colloidal CHO (pg cell (pg CHO Colloidal 0 50 100 150 200 250 Light Intensity (µmol photons·m-2·s -1)

LL (log) 2 a ML 1.5 HL 1

0.5 050100150200250 -2 -1 Colloidal CHO/CHL Light Intensity (µmol photons·m ·s )

Figure 21: Scenedesmus quadricauda (a): Colloidal CHO v CHLa; (b): Colloidal CHO v Light Intensity; (c): Light treatment effect on colloidal CHO/CHLa ratios (dot plot)

= 1.996) group was significantly different from the ML (M= 0.863), with a mean

difference of 1.13 and a p value < 0.001. The LL (M =1.996) group was significantly

different from the HL (M = 0.868) group, with a mean difference of 1.13 and a p value <

0.001.The ML (M = 0.863) was however not significantly different from the HL (M =

0.868) group, with a mean difference of -0.005 and a p value of 0.992.

Scenedesmus quadricauda – storage CHO/CHLa relationships: The storage carbohydrate

concentrations in LL experiments ranged from 13.62 pg cell-1 to 71.51 pg cell-1 and had

storage CHO/CHLa ratios between 493.89 and 642.97. The ML experiments had storage

carbohydrate concentrations from 11.79 pg cell-1 to 35.40 pg cell-1 and storage

CHO/CHLa ratios ranging from 26.98 to 47.67. The HL experiments resulted in storage

carbohydrate concentration between 69.46 pg cell-1 and 1925.25 pg cell-1 and storage

CHO/CHLa ratios from 16.58 to 20.09. Ratios were log transformed before statistical analysis was carried out in order to meet the homogeneity of variances assumption. Plots of some of these relationships are shown in Figure 22 (a – c).

Levene’s test of homogeneity of variances was significant, with F (2, 15) = 5.098 and p = 0.020. This indicated a violation of the homogeneity of means assumption. A more robust test for equality of means was also significant. That is, the Brown-Forsythe model gave F (2, 10.107) = 93.945 and p < 0.001. The overall F ANOVA was also significant, as F (2, 15) = 86.701 and p < 0.001. This indicated that at least one of the group means was significantly different from the others. Games-Howell post hoc follow- up test indicated that all of the group means from the three light treatments were

(a) ) -1 250 200 LL ML 150 HL 100 50

Storage CHO (pg cell 0 ‐50 51015 CHLa (pg cell-1)

(b) 120 )

-1 y = 0.4183x + 17.118 100 R² = 0.8061 80 60 40 20

Storage CHO (pg cell 0 0 50 100 150 200 250 Light Intensity (µmol photons·m-2·s-1)

(c)

3.5

a 3 2.5 LL 2 ML 1.5 HL CHO/CHL Storage 1 01234 Light Intensity (µmol photons·m-2·s-1)

Figure 22: Scenedesmus quadricauda (a): Storage CHO v CHLa; (b): Storage CHO v Light Intensity; (c): Light treatment effect on storage CHO/CHLa ratios (dot plot)

different. That is, the LL(M = 3.03) group was significantly different from the ML (M =

1.74) group, with a mean difference of 1.29 and a p value < 0.001. The LL (M = 3.03) was also significantly different from the HL (M = 1.27) group with a mean difference of

1.76 and a p value < 0.001. The ML (M = 1.74) group was significantly different from the HL (M = 1.27) group with a mean difference of 0.47 and a p value of 0.015.

Rhodomonas salina (Cryptophyta, cryptomonad): The pigments identified from spectroscopic and chromatographic analyses for this species are: chlorophyllide a

(CHLidea), Chlorophylls c1/c2 (CHLs c1/c2) pyrochlorophyllide a (pCHLidea), alloxanthin (ALLO), monadoxanthin (MONADO) chlorophyll a allomer (CHLa allo), chlorophyll a (CHLa), chlorophyll a epimer (CHLa’), and alpha carotene (ALPH). The taxonomically significant pigment identified for cryptomonads is ALLO (Gieskes and

Kraay, 1983). The molar ratios of chlorophyll a / alloxanthin (CHLa/ALLO) in the LL experiments ranged from 2.37 to 3.03, ML experiment ratios ranged from 2.26 to 3.03 and those for the HL experiments were between 2.42 and 2.67. Plots of these ratios per batch culture and light level are shown in Figure 23 (a - b). Typical chromatograms of the species investigated in this study are shown in appendix VIII. These pigment ratios were not log transformed.

Levene’s statistical test for homogeneity of variance for the means of the

CHLa/ALLO ratios for the three light groups was not violated, as F (3, 14) = 1.455 and p

= 0.267.

(a)

3 2.59 2.61 2.54 2 /ALLO a 1

CHL 0 0 50 100 150 200 250 Light Intensity (µmol photons·m-2·s -1)

(b) LL 4 ML 3 HL 2 /ALLO a 1

CHL 0 02468

batch

Figure 23: Rhodomonas salina (a): CHLa/ALLO v Light Intensity; (b): CHLa/ALLO per batch

The overall ANOVA was not significant, with F (2, 14) = 0.163 and p = 0.851, indicating

that the group means from the three light treatments were the same. Tukey’s post hoc

follow up tests further showed that none of the group means were significantly different

from each other at the 0.05 level.

Rhodomonas salina – protein/CHLa biomass relationships: The LL experiments gave

protein concentrations between 7.05 pg cell-1 and 20.55 pg cell-1; The ML experiments gave protein concentrations from 24.76 pg cell-1 to 46.62 pg cell-1; The HL experiments

showed this species having protein concentrations between 122.87 pg cell-1 and 195.66 96

pg cell-1. The protein/CHLa ratios ranged from 265.96 to 299.52 for the LL experiments;

227.50 to 307.50 for the ML experiments and 307.31 to 414.01 for the HL experiments.

The ratios were log transformed prior to statistical analyses. Regression and dot plots of some of these biomass relationships are shown in Figure 24 (a - d).

Levene’s statistic for homogeneity of variances was not significant, with F (2, 13)

= 1.955 and p = 0.181, suggesting that the null hypothesis that the group mean variances are equal can be retained. The overall ANOVA F was significant, as F (2, 13) = 17.464 and p = 0.001, indicating that at least one of the group means was significantly different from the others. Tukey’s (HSD) post hoc follow-up test showed that the LL (M = 2.45) group was not significantly different from the ML (M = 2.44) group, with a mean difference of 0.010 and a p value of 0.913. The LL (M= 2.45) group was significantly different from the HL (M = 2.55) group, with a mean difference of -0.108 and a p value of 0.001. The ML (M = 2.44) group was also significant from the HL (M = 2.55) group, with a mean difference of -0.11 and a p value < 0.001.

(a) LL 250 ML

) -1 200 HL 150

100 50 (pg cell Protein 0 0 0.2 0.4 0.6 0.8 CHLa (pg cell-1)

(b) 200

) y = 0.9887x ‐ 31.175 -1 150 R² = 0.9907

100

Protein (pg (pg cell Protein 0 0 50 100 150 200 250

Light Intensity (µmol photons·m-2·s -1)

(c)

0.6

) y = 0.0027x ‐ 0.0631

-1 0.5 R² = 0.9984 0.4 0.3 (pg cell(pg

a 0.2

CHL 0.1 0 050100150200250 Light Intensity (µmol photons·m-2·s -1)

Figure 24: Rhodomonas salina (a) Protein v CHLa; (b): Protein v Light Intensity; (c): CHLa v Light Intensity

(d) 2.8 LL (log)

a 2.6 ML

HL 2.4

Protein/CHL 2.2 0 50 100 150 200 250 Light Intensity (µmol photons·m-2·s -1)

Figure 24 contd.: Rhodomonas salina (d): Light treatment effect on protein/CHLa ratios (dot plot)

Rhodomonas salina – colloidal CHO/CHLa relationships: The LL experiments had colloidal carbohydrate concentrations between 0.78 pg cell-1 and 2.50 pg cell-1, with colloidal CHO/CHLa ratios from 23.26 to 36.47. The ML experiments had colloidal carbohydrate concentrations ranging from 2.23 pg cell -1 to 3.97 pg cell-1 and colloidal

CHO/CHLa ratios from 23.51 to 26.59. The HL experiments showed concentration between 5.71 pg cell-1 and 12.38 pg cell-1, with colloidal CHO/CHLa ratios from 17.52 to

23.53. All ratios were log transformed prior to statistical tests. Regression and dot plots of some of these relationships are shown in Figure 25 (a - c).

Levene’s test for homogeneity of variances was not significant, with F (2, 13) =

1.815 and p = 0.202, indicating that the null hypothesis that the variances of the means from the three light experiments are equal can be retained. The overall ANOVA F was significant, as F (2, 13) = 10.929 and p = 0.002, indicating that at least one of the group means was different from the others. Tukey’s (HSD) post hoc follow-up tests showed that

(a)

) 15 -1 10 LL ML 5 HL

0 0 0.10.20.30.40.50.60.7 Colloidal CHO cell (pg CHLa (pg cell-1)

(b)

) 12 -1 y = 0.0494x ‐ 0.3851 10 R² = 0.9982 8 6 4 2 0 Colloidal CHO (pg cell 0 50 100 150 200 250 Light Intensity (µmol photons·m-2·s-1)

1.5 LL

1.4 ML HL 1.3

1.2

(log) CHO/CHLa Colloidal 0 50 100 150 200 250 Light Intensity (µmol photons·m-2·s -1)

Figure 25: Rhodomonas salina (a): Colloidal CHO v CHLa; (b): Colloidal CHO v Light Intensity; (c): Light treatment effect on colloidal CHO/CHLa ratios (dot plot)

100

the LL (M = 1.46) group was not significantly different from the ML (M = 1.38) group,

with a mean difference of 0.08 and a p value of 0.105. The LL (M = 1.46) was

significantly different from the HL (M = 1.29) group, with a mean difference of 0.16 and

a p value of 0.001. The ML (M = 1.38) group was not significantly different from the HL

(M = 1.29) group with a mean difference of 0.082 and a p value of 0.083.

Rhodomonas salina – storage CHO/CHLa relationships: The LL experiments had storage

carbohydrate concentrations between 4.14 pg cell-1 and 14.40 pg cell-1, with storage

CHO/CHLa ratios from 111.28 to 209.90. The ML experiments resulted in storage carbohydrate concentrations between 9.61 pg cell-1 and 21.27 pg cell-1 and storage

CHO/CHLa ratios from 82.96 – 140.15. The HL experiments had storage carbohydrate

concentrations between 47.57 pg cell-1 and 89.54 pg cell-1 and storage CHO/CHLa ratios

from 138.40 to 160.28. All ratios were log transformed before statistical analyses. Plots

illustrating some of these relationships are shown in Figure 26 (a - c).

Levene’s statistic to test the homogeneity of variances assumption was

significant, giving F (2, 13) = 4.292 and p = 0.037. This indicated that the null hypothesis

could be retained. The overall ANOVA F was significant, with F (2, 13) = 16.436 and p <

0.001, suggesting that at least one of the group means was significantly different from the others. Tukey’s (HSD) post hoc follow-up test showed that the LL (M= 2.004) was not significantly different from the ML (M= 1.956) group, with a mean difference of 0.048 and a p value of 0.663. The LL (M= 2.004) was significantly different from the HL (M =

2.233) group with a mean difference of - 0.229 and a p value of 0.002. The ML (1.956)

101

(a) LL 100 ML ) -1 80 HL 60 40

20 0 0 0.2 0.4 0.6 0.8 -1 Storage CHO (pg cell CHLa (pg cell )

(b) y = 0.4073x ‐ 12.027 80 ) R² = 0.9811 -1 60

Storage CHO (pg cell 0 50 100 150 200 250 Light Intensity (µmol photons·m-2·s -1)

c) LL 2.4 ML

(log) 2.2 HL a 2 1.8

1.6 0 50 100 150 200 250

Storage CHO/CHL Light Intensity (µmol photons·m-2·s -1)

Figure 26: Rhodomonas salina (a): Storage CHO v CHLa; (b): Storage CHO v Light Intensity; (c): Light treatment effect on storage CHO/CHLa ratios (dot plot)

102

was also significantly different from the HL (M= 2.233) group, with a mean difference of

- 0.277 and a p value < 0.001.

Cyclotella meneghiniana (Bacillarophyta; diatom): The pigment composition for this

species is as follows: chlorophyllide a (CHLidea), Chlorophylls c1/c2 (CHLs c1/c2) pyrochlorophyllide a (pCHLidea), fucoxanthinol (FUCOL), fucoxanthin (FUCO), cis- fucoxanthin (cis-FUCO), diadinoxanthin (Diad), diatoxanthin (Diato), phytylated-type chlorophyll c, (Phyt chlc), chlorophyll a allomer (CHLa allo), chlorophyll a (CHLa), chlorophyll a epimer (CHLa’), pheophytin a (pHtin a), beta carotene (BETA). Typical chromatograms of the eight species are presented in appendix IV. The most stable class/pigment group specific marker was identified to be FUCO (Stauber and Jeffrey,

1988). The molar ratios of CHLa/FUCO in the LL experiments ranged from a minimum of 1.02 to a maximum of 1.56, ML experiment ratios ranged from 1.07-1.20 and those for the HL experiments were between 0.99 and 1.28. Plots of these ratios per culture batch, over the light levels studied, are illustrated in Figure 27 (a - b).

Levene’s test for equality of variances was not significant, with F (2, 14) = 2.640 and p = 0.0.106. Thus the assumption that the variances in the means are homogeneous was not violated. The one-way ANOVA overall F ratio was not significant, as F (2, 14)

= 0.469 and p = 0.635. Thus, the null hypothesis that the means are equal was accepted.

103

(a) 2

1.5 1.18 1.12 1.11 1 0.5 CHLa/FUCO

0 0 50 100 150 200 250 Light Intensity (µmol photons·m-2·s -1)

(b) LL 2 ML 1.5 HL

/FUCO 1 a

CHL 0.5

0 02468 culture batches

Figure 27: Cyclotella meneghiniana (a): CHLa/FUCO v Light Intensity; (b): CHLa/FUCO per batch

Cyclotella meneghiniana – protein/CHLa relationships: the protein content at the LL

experiments for this species ranged from 7.23 pg cell-1 to 91.54 pg cell-1; the ML

experiments had protein content between 9.23 pg cell-1 and 155.76 pg cell-1 and the HL

experiments had a minimum of 14.76 pg cell-1 to a maximum of 91.54 pg cell-1 protein.

The protein/CHLa ratios for the LL experiments were from 90.20 – 105.06, those in the

ML experiments were from 43.85-106.08 and those in the HL experiments ranged from

122.57 -172.74. The protein/CHLa ratios for the three light levels were log transformed 104

in an attempt to prevent violation of the homogeneity assumption. See regression and dot plots in Figure 28 (a - d).

(a) 500

) -1 400 300 LL 200 ML 100

cell (pg Protein HL 0 0123 CHLa (pg cell -1)

(b) 250 y = 1.1459x ‐ 22.133

) R² = 0.9791 -1 200

150 100 50 (pgcell Protein 0 0 50 100 150 200 250 Light Intensity (µmol photons·m-2·s-1)

Figure 28: Cyclotella meneghiniana (a): Protein v CHLa; (b): Protein v Light Intensity

105

) R² = 0.9293 -1 1.5

1 0.5

cell (pg CHLa 0 0 50 100 150 200 250 Light Intensity (µmol photons·m-2·s -1)

Figure 28 contd.: Cyclotella meneghiniana (c): CHLa v Light Intensity; (d): Light treatment effect on protein/CHLa ratios (dot plot)

Levene’s test for homogeneity was not significant, with F (2, 15) = 0.049 and p =

0.952. Thus, the homogeneity of variance assumption was not violated. The overall one-

way ANOVA gave an F (2, 15) = 24.596 and p < 0.001 that is significant, indicating that the null hypothesis was to be rejected. Tukey’s (HSD) post hoc follow test showed that all the means were significantly different at the 0.05 level. That is, the LL group (M =

1.996) was significantly different form the ML group (M = 2.07), with a mean difference of - 0.08 and a p value of 0.009. The LL group (M = 1.996) was also significantly

106

different from the HL group (M = 2.17), with a mean difference of -0.17 and a p value <

0.001. Additionally, the HL group (M = 2.17) was significantly different from the ML

group (M = 2.07), with a mean difference of 0.09 and a p value of 0.003.

Cyclotella meneghiniana – colloidal CHO/CHLa relationships: The colloidal

carbohydrate content at for the LL experiments were from 0.63 pg cell-1 to 2.60 pg cell-1,

while the colloidal CHO /CHLa ratios ranged from 7.09 to 9.50. The ML experiments

had colloidal carbohydrate concentrations from 0.16 pg cell-1 to 16.23 pg cell-1, with colloidal CHO/CHLa ratios from 10.36 to 16.21. The HL experiments had concentrations from 11.50 pg cell-1 to 48.86 pg cell-1 and ratios ranging from a minimum of 19.89 to

28.33. Prior to statistical analyses, the colloidal CHO/CHLa ratios for the three light levels were log transformed in an attempt to prevent violation of the homogeneity assumption. Regression and dot plots are shown in Figure 29 (a - c).

Levene’s test to check the assumption that the variances from three light levels are equal was not significant, giving F (2, 15) = 0.520, p = 0.605. This indicated that the homogeneity assumption was not violated. However, the overall F ANOVA was significant, as F (2, 15) = 78.199 and p < 0.001. This indicated a rejection of the null hypothesis, since the means were different. Tukey’s (HSD) post hoc follow up tests showed that the LL group (M = 0.92) was significantly different from the ML group (M =

1.08), with a mean difference of -0.16 and a p value of 0.002. The LL group (M = 0.92) was significantly different from the HL (M = 1.38), with a mean difference of -0.47 and p

107

(a) ) -1 60 50 40 LL 30 ML 20 10 HL

0 ‐0.5 cell (pg CHO Colloidal 0 0.5 1 1.5 2 2.5 3 ‐10 CHLa (pg cell-1)

(b)

) 40

-1 y = 0.2005x ‐ 6.8628 30 R² = 0.9952 20

10 0 0 50 100 150 200 250 -2 -1 Colloidal CHO (pg cell (pg CHO Colloidal Light Intensity (µmol photons·m ·s )

Figure 29: Cyclotella meneghiniana (a): Colloidal CHO v CHLa; (b): Colloidal CHO v Light Intensity; (c): Light treatment effect on colloidal CHO/CHLa ratios (dot plot)

108

< 0.001 The ML group (M = 1.08) was also significantly different from the HL group (M

= 1.38), with a mean difference of -0.30 and a p value < 0.001.

Cyclotella meneghiniana – storage CHO/CHLa relationships: The storage carbohydrate

concentration for the LL experiments ranged from 0.11 pg cell-1 to 6.44 pg cell-1, with storage CHO/CHLa ratios from 37.99 to 80.32. The ML experiments showed storage carbohydrate concentrations ranging from 0.18 pg cell-1 to 204.33 pg cell-1 and storage

CHO/CHLa ratios from 52.78 to 109.62. The HL experiments had concentrations from

118.87 pg cell-1 to 575.01 pg cell-1 and ratios from 200.21 to 277.44. The storage

CHO/CHLa ratios were log transformed in an attempt to maintain the homogeneity of

variance assumption, prior to conducting statistical analyses. Regression and dot plots are

shown in Figure 30 (a - c).

Levene’s test was not significant, with F (2, 15) = 3.009 and p = 0.080, indicating

that the assumption of homogeneity of variance was not violated. The overall ANOVA F

was significant, with F (2, 15) = 65.101 and p < 0.001. This indicated that at least one of

the group means was significantly different from the others. Tukey’s (HSD) post hoc

follow up test was done to identify how different the group means were from each other.

The test showed that the LL group (M = 1.75) was significantly different from the ML

(M = 2.14), with a mean difference of - 0.39 and a p value < 0.001. The LL group (M =

1.75) was significantly different from the HL group (M = 2.38), with a mean difference

of -0.63 and a p value < 0.001. The ML group (M = 2.14) was also significantly different from the HL group (M = 2.38), with a mean difference of - 0.24and a p value of 0.001.

109

(a)

800 ) -1 600 LL 400 ML 200 HL 0 00.511.522.53 Storage (pgCHO cell CHLa (pg cell-1)

(b) 400 y = 2.1031x ‐ 81.291 ) -1 R² = 0.9912 300

200

100

Storage CHO cell (pg CHO Storage 0 50 100 150 200 250 ‐100 Light Intensity (µmol photons·m-2·s -1)

CHO/CHL storage 0 50 100 150 200 250 Light Intensity (µmol photons·m-2·s-1)

Figure 30: Cyclotella meneghiniana (a): Storage CHO v CHLa; (b): Storage CHO v Light Intensity; (c): Light treatment effect on storage CHO/CHLa ratios (dot plot)

110

Thalassiosira Weissflogii ( Bacillariophyceae; diatom): The pigment composition for this species was as follows: chlorophyllide a (CHLidea), Chlorophylls c1/c2 (CHLs c1/c2) pyrochlorophyllide a (pCHLidea), fucoxanthinol (FUCOL), fucoxanthin (FUCO), cis fucoxanthin (cis-FUCO), diadinoxanthin (Diad), diatoxanthin (Diato), phytolated-type chlorophyll c, (Phyt chlc), chlorophyll a allomer (CHLa allo), chlorophyll a (CHLa), chlorophyll a epimer (CHLa’), pheophytin a (PHtin a), beta carotene (BETA). Typical chromatograms of the eight species are presented in VIII. The most stable class/pigment group specific marker was identified to be FUCO (Stauber and Jeffrey 1988). The molar ratios of CHLa/FUCO in the LL experiments ranged from 1.11 to of 1.19, ML experiment ratios ranged from 1.14 to 1.20 and those for the HL experiments were between 1.12 and 1.19. See Figure 31(a - b) for plots of these ratios per culture batch over the light levels studied.

Statistical analyses was carried out determine if the variances in the ratios could be partitioned within the variances of the samples and between the different sample groups. That is, test the null hypothesis that the group means from all three light treatments are equal. All tests were carried out at the 0.05 level. Levene’s test is not significant, as F (2, 14) = 0.396 and p = 0.680. Thus the assumption of homogeneity of the group means was not violated. The one-way ANOVA overall F ratio was not significant, as F (2, 14) = 1.787and p = 0.219. Therefore, the null hypothesis that the means were equal was accepted.

111

(a) 2 1.5 1.14 1.17 1.15 1

0.5 CHLa/FUCO

0 0 50 100 150 200 250 Light Intensity (µmol photons·m-2·s-1)

(b) 1.4 LL ML /FUCO 1.2 a HL CHL 1 02468 Culture batches

Figure 31: Thalassiosira weissflogii (a): CHLa/FUCO v Light Intensity; (b): CHLa/FUCO per batch

Thalassiosira weissflogii – protein/CHLa relationships: The LL experiments gave protein concentration in this species between 25.89 pg cell-1 and 115.96 pg cell -1, while the protein/CHLa ratios ranged from 137.37 to 186.62. The ML experiments had protein concentration from 18.09 pg cell-1 to 124.46 pg cell -1, with protein/CHLa ratios between

166.40 and 195.76. The HL experiments had protein concentration ranging from 60.50 pg cell-1 to 489 pg cell-1 and protein/CHLa ratios from 192.03 to 238.06. All ratios were log transformed prior to statistical analysis. Plots are shown in Figure 32 (a - d). 112

(a) 600

500 ) -1 400 300 LL 200 ML 100 HL Protein (pg cell (pg Protein 0 0123 ‐100 CHLa (pg cell-1)

(b) 250

) y = 0.9343x + 18.087 -1 200 R² = 0.9052 150 100

50 Protein (pg cell (pg Protein 0 0 50 100 150 200 250 Light Intensity (µmol photons·m-2·s -1)

(c)

12 y = 0.0651x ‐ 3.0734 10 R² = 0.9441 )

-1 8 6 4 2 cell (pg CHLa 0 ‐2 0 50 100 150 200 250 Light Intensity (µmol photons·m-2·s -1)

Figure 32: Thalassiorira weissflogii (a): Protein v CHLa; (b): Protein v Light Intensity; (c):CHLa v Light Intensity;

113

Figure 32 contd.: Thalassiorira weissflogii (d): Light treatment effect on protein/CHLa ratios (dot plot)

Levene’s test for homogeneity of variance was not significant, as F (2, 13) =

1.231 and p = 0.324. This indicated that the null hypothesis that variance in the group

means over the three light treatment levels were the same. However, the overall ANOVA

F was significant, with F (2, 13) = 13.124 and p = 0.001. Tukey’s (HSD) post hoc follow-

up test was next done to assess which of the group means was significantly different from

the others and gave the following results: The LL (M= 2.21) group was not significantly

different from the ML (M= 2.25) group with a mean difference of -0.046 and a p value of

0.185. The LL (M= 2.21) was significantly different from the HL (M = 2.33) group, with

a mean difference of -0.118 and a p value of 0.001. The ML (2.25) group was also

significantly different from the HL (M = 2.33) with a mean difference of -0.072 and p

value of 0.022.

Thalassiosira weissflogii – colloidal CHO/CHLa relationships: The LL experiments

resulted in colloidal carbohydrate (CHO) concentrations of 1.15 pg cell-1 to 6.22 pg cell-1 and colloidal CHO/CHLa ratios between 6.31 and 10.56. The ML experiments had

114

colloidal CHO concentrations from 1.23 pg cell-1 to 7.53 pg cell-1 and colloidal

CHO/CHLa ratios between 11.33 and 12.29. The HL experiments gave colloidal CHO concentrations between 5.37pg cell-1 and 41.77 pg cell-1 and colloidal CHO/CHLa ratios from 14.26 to 17.68. All ratios were log transformed prior to statistical analyses. Plots of the ratio relationships are shown in Figure 33 (a - c).

(a) 50 ) -1 40 LL 30 ML 20 HL 10 0 0123 -1

Colloidal CHO (pg cell (pg CHO Colloidal CHLa (pg cell )

(b)

) 20 y = 0.0834x ‐ 1.0306 -1 R² = 0.9288 15 10 5

0 0 50 100 150 200 250 Colloidal CHO (pg cell Light Intensity (µmol photons·m-2·s -1)

Figure 33 : Thalassiosira weissflogii (a): Colloidal CHO v CHLa; (b): Colloidal CHO v Light Intensity

115

Figure 34 contd.: Thalassiosira weissflogii (c): Light treatment effect on colloidal CHO/CHLa ratios (dot plot)

Levene’s test was significant, as F (2, 13) = 5.685 and p = 0.017, indicating that the homogeneity of variance assumption had been violated. A more robust test of equality of means was done and it was also significant. That is, Brown-Forsythe gave F (2, 5.732)

= 41.465 and p < 0.001. The overall ANOVA F was also significant, with F (2, 13) =

43.920 and p <0.001, indicating that at least one of the group means was significantly different from the others. The Games-Howell post hoc follow-up test showed that all the group means were different, according to the following results: The LL (M= 0.892) group was significantly different from the ML (M=1.07) group, with a mean difference of -

0.1802 and a p value of 0.020. The LL (M= 0.892) group was also significantly different from the HL (M =1.202) group, with a mean difference of -0.310 and a p value of 0.001.

The ML (M = 1.07) group was significantly different from the HL (M = 1.202) group, with a mean difference of -0.1301 and a p value < 0.001.

Thalassiosira weissflogii – storage carbohydrate (CHO)/CHLa relationships: The storage carbohydrate concentrations for the LL experiments were between 12.24 and 90.84 pg

116

cell-1, and the storage CHO/CHLa ratios ranged from 88.19 to 121.55. The ML

experiments had storage CHO concentrations from 17.94 to 107.78 pg cell-1 and storage

CHO/CHLa ratios between 125.93 and 169.52. The HL light experiments had storage

CHO concentration from 73.40 to 238 pg cell-1 and storage CHO/CHLa ratios from

156.04 to 236.27. All ratios relationships were log transformed before statistical analysis

in an attempt to meet the homogeneity of variances assumption. Regression and dot plots

are shown in Figure 34 (a - c).

Levene’s test for homogeneity of variances was not significant, with F (2, 13) = 0.358 and p = 0.706, indicating that the assumption had not been violated. However, the overall

ANOVA F was significant, with F (2, 13) = 25.271 and p < 0.001. Tukey’s (HSD) post

hoc follow-up test was then done to identify which of the group means was significantly

different from the others. The results showed that the LL (M = 2.03) group was significantly different from the ML (M = 2.16) group, with a mean difference of -0.12

and a p value of 0.014. The LL (M= 2.03) group is also significantly different from the

HL (M = 2.29) group, with a mean difference of- 0.26 and a p value < 0.001. The ML (M

= 2.16) is significantly different from the HL (M = 2.29) group, with a mean difference of

-0.131 and a p value of 0.008.

117

(a) 800 ) -1 600 LL ML 400 HL 200 0

‐200 0123

Storage CHO cell (pg CHO Storage CHLa (pg cell-1)

(b) 250 ) y = 1.0524x ‐ 4.4881 -1 200 R² = 0.9382 150 100 50 0

0 50 100 150 200 250 Storage CHO cell (pg CHO Storage Light Intensity (µmol photons·m-2·s -1)

a ML 2.4 HL 2.2

Storage CHO/CHL 1.8 0 50 100 150 200 250

Light Intensity (µmol photons·m-2·s -1)

Figure 35: Thalassiosira weissflogii (a): Storage CHO v CHLa; (b): Storage CHO v Light Intensity; (c): Light treatment effect on storage CHO/CHLa ratios (dot plot)

118

Amphidinium carterae (Dinophyta; dinoglagellate): The main pigments identified from

the chromatographic plots and spectral analyses for this species were: two glycosylated carotenoids (P-457and P-468), Chlorophylls c1/c2 (CHLs c1/c2), Peridinin (PER),

Dinoxanthin (DINO), Diadinoxanthin (DD), Chlorophyll a allomer (CHLa allo),

Chlorophyll a (CHLa), Chlorophyll a epimer (CHLa’), and Beta Carotene (BETA).

Appendix IV shows typical chromatograms of the eight species investigated in this study.

The marker pigment for this species was determined to be PER (Jeffery et al., 1997). The

molar ratios of chlorophyll a/ peridinin (CHLa/PER) in the LL experiments ranged from

0.92 to 1.15, ML ratios ranged from 1.12 to 1.49 and those for the HL experiments were

between 0.80 and 1.48. Plots of these ratios per culture batch, over the light levels

studied, are shown in Figure 35 (a - b).

Levene’s test is not significant, with F (2, 14) = 2.982 and p = 0.083. Thus the

assumption that the variances in the means were homogeneous was not violated. The one-

way ANOVA gave an overall F ratio that was not significant, with F (2, 14) = 3.159 and

p = 0.074. Therefore, the null hypothesis that the means are equal was accepted.

Amphidinium carterae – protein/CHLa relationships: Protein concentrations in the LL

experiments ranged from 45.98 – 91.85 pg cell-1, ML concentrations were between 95.07

and 243.10 pg cell -1, while the HL experiments had had protein concentration ranges

from 124.32 – 216.28 pg cell-1. Protein/CHLa ratios for these light treatments were

between 176.46 and 268.84 for LL, 287.63 and 638.01for ML and 403.18 and 551.27 for

HL. The ratios were log transformed prior to statistical analysis to prevent

119

(a) 2 1.5 1.3 1 1 1.1 /PERI ratios /PERI a 0.5

CHL 0 0 50 100 150 200 250 Light Intensity (µmol photons·m-2·s -1)

(b)

2 LL 1.5

/PERI ML a 1

CHL 0.5 HL

0 02468 batch

Figure 36: Amphidinium carterae (a): CHLa/PERI v Light Intensity; (b): CHLa/PERI per batch

violation of the homogeneity of variances assumption. Pertinent regression and dot plots are shown in Figure 36 (a - d).

120

(a) 300 )

-1 250 200 150 LL 100 ML 50

Protein (pg cell (pg Protein HL 0 0 0.2 0.4 0.6

CHLa (pg cell-1)

(b)

200 ) -1 150

100 y = 0.5105x + 75.74 R² = 0.7285 50 Protein (pg cell (pg Protein 0 0 50 100 150 200 250 -2 -1 Light Intensity (µmol photons·m ·s )

-1 0.3 y = 6E‐05x + 0.3196 0.2 R² = 0.106

0.1 CHLa (pg cell (pg CHLa

0 0 50 100 150 200 250 Light Intensity (µmol photons·m-2·s -1)

Figure 37: Amphidinium carterae (a): Protein v CHLa; (b): Protein v Light Intensity; (c): CHLa v Light Intensity

121

(d)

3 2.8 (log)

a 2.6

2.4 LL 2.2 ML protein/ chl 2 HL 0 50 100 150 200 Light Intensity (µmol photons·m-2·s -1)

Figure 36 contd.: Amphidinium carterae (d): Light treatment effect on protein/CHLa ratios (dot plot)

Levene’s statistic was not significant, with F (2, 14) = 0.315 and a p value of

0.735. The overall F ANOVA was significant, with F (2, 14) = 26.215 and p ≤ 0.001,

indicating that at least one of the means was significantly different from the others. Since

the homogeneity of means assumption was not violated, Tukey’s (HSD) post hoc follow-

up test was done to assess the differences in the means. The LL group (M = 2.33) was

significantly different from the ML group (M = 2.68), with a mean difference of -0.34

and a p value < 0.001. The LL group (M = 2.33) was significantly different from the HL

group (M = 2.69), with a mean difference of -0.36 and a p value < 0.001. However, the

ML group (M = 2.68) was not significantly different from the HL group (M = 2.69), with

a mean difference of -0.02 and a p value of 0.938.

Amphidinium carterae – colloidal carbohydrates/CHLa relationships: The concentration

of the colloidal carbohydrate pool in the LL experiments ranged from 12.72 to 93.87 pg

cell -1, the ML had concentration ranges between 11.10 and 31.87 pg cell -1, and the HL

122

experiment showed concentrations between 30.67 and 38.26 pg cell -1. The colloidal

CHO/CHLa ratios were between 29.40 and 60.35 for the LL experiments, the ML ratios

were between 50.99 and 82.49, and the HL treatments showed ratios ranging from 92.35 and 113.92. These ratios were log transformed prior to statistical analysis to meet the

assumption of homogeneity of variances. Regression and dot plots are shown in Figure

37 (a – c).

(a) 50 ) -1 40 LL 30 ML 20 HL 10

0 0 0.2 0.4 0.6 0.8 Colloidal CHO (pg cell (pg CHO Colloidal CHLa (pg cell-1)

(b) 40 ) y = 0.1239x + 9.6736 -1 R² = 0.9999 30

cell (pg CHO Colloidal 0 50 100 150 200 250 Light Intensity (µmol photons·m-2·s -1)

Figure 38: Amphidinium carterae (a): Colloidal CHO v CHLa; (b): Colloidal CHO v Light Intensity

123

(c)

2.2 (log) 2 a 1.8 1.6 LL 1.4 ML 1.2 HL colloidal CHO/CHL colloidal 0 50 100 150 200 Light Level (µmol photons·m-2·s -1)

Figure 37 contd.: Amphidinium carterae (c): Light treatment effect on protein/CHLa ratios (dot plot)

Levene’s test for homogeneity of variance was not significant, with F (2, 14) =

3.20 and p = 0.072.Thus, the homogeneity of variance assumption had not been violated.

However, the overall ANOVA was significant, with F (2, 14) = 33.295 and a p value <

0.001. Tukey’s (HSD) follow up tests were done to determine which of the means were

significantly different from the others. The following results were obtained: The LL

group (M = 1.62) was significantly different from the ML group (M = 1.79), with a mean

difference of - 0.17 and a p value of 0.010. The LL group (M = 1.62) was significantly

different the HL group (M = 2.02), with a mean difference of - 0.39 and a p value <

0.001. Lastly, the ML (M = 1.79) group was also significantly different from the HL

group (M = 2.02, with a mean difference of - 0.22 and a p value of 0.001.

Amphidinium carterae – Storage carbohydrates/CHLa relationships: The storage

carbohydrate concentrations in the light treatments are as follows: LL concentrations

124

ranged between 49.09 and 76.42 pg cell-1, the ML experiments showed concentrations between 5.25 and 108.62 pg cell -1, and the HL experiments had concentrations between

93.49 and 195.07 pg cell -1. The storage carbohydrate/CHLa had ratios were between

157.07 and 203.09 for the LL experiments, 196.77 to 256.83 for the ML experiments and

317.89 to 492.64 for the HL experiments. All ratios were log transformed prior to statistical tests. Relevant plots are given in Figure 38 (a - c).

(a)

250 ) -1 200 LL 150 ML 100 HL 50 0 0 0.2 0.4 0.6 0.8 Storage CHO (pg cell CHLa (pg cell-1)

(b)

200 y = 0.4993x + 37.105 R² = 0.9758 150 100

Storage CHO (pg cell-1) 0 50 100 150 200 250 Light Intensity (µmol photons·m-2·s -1)

Figure 39: Amphidinium carterae (a): Storage CHO v CHLa; (b): Storage CHO v Light Intensity

125

(log) LL a 2.5 ML HL

2 0 50 100 150 200 250 storage CHO/CHL Light Intensity (µmol photons·m-2·s -1)

Figure 38 contd.: Amphidinium carterae (c): Light treatment effect on storage CHO/CHLa ratios (dot plot)

Levene’s test for homogeneity of variance was not significant, with F (2, 14) =

2.815 and p = 0.094, indicating that the variances in the group means were equal. The overall ANOVA was significant, with F (2, 14) = 41.663 and a p value < 0.001. This indicated that at least one of the means was significantly different from the others.

Since the homogeneity of variances assumption had not been violated, Tukey’s (HSD)

post hoc follow-up test was done to determine which of the means was significant.

These tests showed that the LL group (M = 2.26) was not significantly different from the

ML group (M = 2.35), with a mean difference of -0.09 and a p value of 0.073. The LL

group (M = 2.26) was however significantly different the HL group (M = 2.59), with a

mean difference of -0.33 and a p value < 0.000. The ML group (M = 2.35) was also

significantly different from the HL group (M = 3.59), with a mean difference of - 0.24

and a p value < 0.001.

126

IV. DISCUSSION

Growth patterns The specific growth rate constants (µ) were calculated for the cultures studied

from the slope of the linear portion of the semilog plot of growth versus time (Appendix

IX). Growth rate is a function of photon flux density (PFD) in nutrient sufficient cultures

at constant temperature (Geider, 1987). All species exhibited increasing specific growth rate constants with increasing light intensity. This indicated that the light intensities used did not limit or inhibit the growth of the algal cells. Of the two bacillariophytes investigated, Cyclotella meneghiniana had a higher µ at LL than Thalassiosira

weissflogii, but at ML and HL the growth rate constants for the former species were

lower. Conversely, Dunaliella tertiolecta grew at a faster rate than Scenedesmus

quadricauda at all three light levels.

Falkowski (1980) observed that CHLa cellular concentration varies as a linear function of irradiance in nutrient replete batch cultures of microalgae. It is expected that the concentration of CHLa and other light harvesting pigments will be higher at lower irradiance levels for capturing the limited supply of available photons. The opposite is expected for higher irradiance levels. That is, the concentrations of CHLa and the accessory pigments will be lower, while the photoprotecting pigments will be higher. The photoprotecting pigment concentration increases in order to prevent photoinhibition and

127

photodynamic action. In another paper, Falkowski (1981) also noted that increased CHLa

content appears to be a ubiquitous response by algae to decreased levels of incident light,

and cited previous works (Falkowski et al., 1980; Prezlin et al., 1978; Seneger et al.,

1978). Falkowski (1981) only investigated two chlorophyte species at two irradiance levels (30 and 600 μmol photon·m-2·s-1), with observations supporting the hypothesis. In

the study conducted by Grant and Louda (2010), chlorophytes, cyanophytes,

bacillariophytes, dinophytes and a chrysophyte were grown at 44.5 μmol photon·m-2·s-1,

108 μmol photon·m-2·s-1, 100-120 μmol photon·m-2·s-1, 300 μmol photon·m-2·s-1, 1600

μmol photon·m-2·s-1and 1800 μmol photon·m-2·s-1 irradiance levels. While the observations of CHLa concentration in that study largely followed the expected hypothesis, the concentration of CHLa did not start to decrease until the 300 μmol

photon·m-2·s-1 experiments. This leads us to think that photoinhibition and thus an

increase in the need for photoprotecting pigments by the algal species, occurred at 300

μmol photon·m-2·s-1 and higher light intensities. In this present study, the highest

irradiance level is 200 μmol photon·m-2·s-1, thus, observing increasing CHLa

concentration with irradiance, up to this light level is acceptable.

Phytoplankton protein as a biomass indicator

Protein is typically the major biological component of algae (Brown and Jeffrey

1992). Rapidly growing cells are characterized by high protein and low carbohydrate

content. When cells have reached the stationary phase, more carbon is incorporated into

carbohydrate and/ or lipids (Piorreck and Pohl, 1984). In this study, similar trends were

128

observed, as algal cultures were harvested during logarithmic growth phase or at very early transition to the stationary phase. Our data for the eight species investigated, showed that protein concentration per cell increased from the LL to the HL experiments while also being higher than colloidal carbohydrate concentration per cell for each particular species. Apart from a few anomalies, the protein concentration per cell was also higher than the storage carbohydrate concentration per cell at high light for the eight species investigated. Where storage carbohydrate concentration per cell was shown to be greater than the protein concentrations per cell, such as with Cyclotella meneghiniana at

HL, it is likely that cells had reached a stationary phase of growth, or a point where there was not sufficient nutrients to further facilitate production of proteins.

Concentration of proteins, chlorophyll a, colloidal carbohydrates and storage carbohydrates per cell and biovolume are tabulated in Table 6 below, and also in

Appendix VII. Regression plots of protein versus chlorophyll a (CHLa) in Chapter III gave positive correlation between CHLa and protein for each of the species studied.

Ratios of protein: CHLa were log 10 transformed in an attempt to satisfy such problems as skewed data, outliers, unequal variation, as well as for general easier statistical handling.

One-way analysis of variance (ANOVA) showed that light intensity does have an effect on the protein: CHLa relationships for all of the species investigated, though more for some than for others. Dunaliella tertiolecta, Cyclotella meneghiniana, Thalassiosira weissflogii, Microcystis aeruginose all exhibited significantly different effects to each light level, with lowest log10 ratios in the LL experiments and highest log 10 ratios in the

HL experiments, as shown in the dot plots in the results section and in Table 7.

129

Table 6: Cellular concentration of chlorophyll a and products of photosynthesis, (blank cells = run not performed; VL = 10 µmol photons·m-2·s-1; L = 37 µmol photons·m-2·s-1; ML = 70-75 µmol photons·m-2·s-1; HL = 200 µmol photons·m-2·s-1; 1µm3 = 1x10-9 µL) Genus Lt R 1 R 2 R 3 R 4 R 5 R 6 Biovol* Item/ biovol pg cell- pg cell- pg cell- pg cell- pg cell- pg cell -¹ fg (item) ¹ ¹ ¹ ¹ ¹ cell-¹ (µm³) µm³·cell A. carterae 432 CHLa L 0.44 0.26 0.31 0.37 0.29 0.1901 M 0.33 0.54 0.24 0.17 0.35 0.21 0.2333 H 0.33 0.27 0.33 0.4 0.35 0.1728 protein L 91.85 45.98 84.02 65.75 73.11 39.6792 M 95.07 243.1 130.4 111.2 169.9 110.6 73.4227 H 179.2 124.3 196.4 216.28 139.59 93.4330 colloidal L 13.04 93.87 12.72 22.25 14.48 40.5518 CHO M 16.85 31.87 14.08 11.1 21.37 17.64 13.7678 H 36.2 31.28 30.67 38.26 35.87 15.6384 storage CHO L 76.42 49.9 49.09 74.87 55.98 33.0134 M 65.03 108.62 5.25 43.54 78.04 54.91 46.9238 H 108.25 93.49 158.64 195.07 137.59 84.2702 C. 2720 meneghiniana CHLa L 0.17 0.99 0.14 0.07 0.37 2.6928 M 0.62 0.2 1.09 0.15 0.09 0.19 2.9648 H 1.17 0.44 1.32 2.15 2.45 1.16 6.6640 protein L 15.28 91.54 14.76 72.31 32.86 248.9888 M 85.21 24.59 155.76 17.55 9.23 23.15 423.6672 H 155.12 73.34 197.46 290.73 380.01 167.61 1033.6272 colloidal L 1.61 1.77 1.18 0.63 2.6 4.8144 CHO M 7.34 3.27 0.16 1.63 0.91 1.97 19.9648 H 31.8 11.5 32.95 43.3 48.86 32.82 132.8992 storage CHO L 6.43 0.16 0.11 4.11 0.16 17.4896 M 0.89 0.36 204.34 0.18 0.12 0.21 555.8048 H 235.17 118.87 368.16 496.26 575.01 268.58 1564.0272 T.weissflogii 2813 CHLa L 0.182 0.139 0.727 0.62 0.747 2.1013 M 0.164 0.173 0.636 0.109 0.205 1.7891 H 2.55 30.4 88.2 1.36 58.9 47.1 248.1066 protein L 31.42 25.89 99.8 99.25 115.96 326.1955 M 29.39 31.49 124.46 18.09 35.85 350.1060 H 489.62 60.5 210.01 272.04 133.66 103.31 1377.3011 colloidal L 1.15 1.47 6.22 4.4 5.34 17.4969 CHO M 1.86 2.13 7.53 1.23 2.5 21.1819 H 41.77 5.37 12.58 19.81 9.2 8.15 117.4990 (Table 6 continued)

130

storage CHO L 21.19 12.24 79.6 65.3 90.84 255.5329 M 20.62 22.84 107.78 17.94 26.79 303.1851 H 602 61.1 176 238 121 73.4 1693.4260 D. tertiolecta 43.42 CHLa L 0.69 0.86 0.78 0.79 0.0373 M 0.52 0.87 0.64 0.45 0.41 n/a 0.0378 H 0.93 1.25 0.7 0.85 0.91 0.87 0.0543 protein L 48.19 62.32 56.94 53.26 2.7059 M 47.64 114.78 84.91 33.07 54.15 n/a 3.6868 H 141.73 226.76 109.84 151.46 170.79 163.04 7.4157 colloidal L 6.11 6.42 6.72 7.38 0.3204 CHO M 2.39 5.09 3.72 1.69 1.32 n/a 0.2210 H 3.82 7.78 2.81 5.62 5.62 6.13 0.3378 storage CHO L 44.68 50.45 45.69 45.46 2.1905 M 43.5 84.28 53.94 32.1 38.38 3.6594 H 96.83 116.56 79.01 96.69 105.34 99.8 5.0610 S. 45 quadricauda CHLa L 0.028 0.055 0.096 0.109 0.126 n/a 0.0057 M 0.401 0.3 0.689 0.774 1.12 0.875 0.0394 H 10.5 4.19 4.45 6.79 4.41 3.83 0.3056 protein L 13.48 28.68 78.3 93.34 53.33 n/a 0.3321 M 94.7 50.5 114 135 189 144 8.5050 H 1239 575 567 871 551 489 55.7550 colloidal L 3.37 2.85 9.31 19.65 11.02 n/a 0.8843 CHO M 3.61 1.88 6.94 4.75 6.86 4.98 0.3123 H 74.55 30.77 30.06 52.96 32.63 30.12 3.3548 storage CHO L 13.62 29.85 51.83 70.22 71.51 n/a 3.2180 M 19.12 11.79 31.85 22.79 30.27 23.84 0.0057 H 1929.25 69.46 81.91 1250.72 88.54 73.74 56.2824 S.elongatus 4.2 CHLa DL 0.55 0.94 0.23 0.7 0.0039 L 0.19 2.64 0.76 1.9 5.73 0.0241 M 0.36 0.79 1.77 0.2 0.4 2.39 0.0100 H 9.25 7.88 5.8 3.18 0.1 0.0389 protein DL 35.2 59.46 12.99 41.72 0.1752 L 0.26 291.82 0.96 229.7 674.68 2.8337 M 0.58 137.02 346.61 0.31 0.75 430.16 1.8067 H 1706.69 1539.28 1095.9 599.76 1986.11 8.3417 colloidal DL 4.79 4.26 0.22 7.48 0.0314 CHO L 1.82 0.23 8.31 0.19 0.6 0.0349 M 3.59 8.61 0.18 1.97 1.54 0.24 0.0362 H 114.26 0.94 0.77 0.44 125.51 0.5271 storage CHO DL 0.24 0.37 0.93 0.32 0.0039 L 0.11 149.67 0.48 112.74 336.79 1.4145 M 0.2 0.4 0.93 0.1 0.22 142.21 0.5973 (Table 6 continued)

131

H 519.69 462.17 313.28 161.2 590.26 2.4791 M. 65 aeruginosa CHLa L 0.147 0.086 0.1 0.059 0.246 0.0160 M 0.0509 0.089 0.107 0.102 0.248 0.177 0.0161 H 0.826 0.896 0.207 0.216 0.533 0.899 0.0584 protein L 7.06 5.22 5.89 3.36 12.1 0.7865 M 3.65 5.25 7.33 5.79 16 11.1 0.7215 H 70.41 85.9 16.9 16.8 41.7 5.5835 colloidal L 0.862 0.438 0.577 0.39 0.132 0.0560 CHO M 0.389 0.589 0.62 0.849 0.155 0.122 0.0552 H 6.15 6.34 1.19 1.52 3.86 6.49 0.4219 storage CHO L 10.85 6.51 8.6 5.3 2.17 0.7053 M 4.5 8.92 9.16 9.47 15.54 13.4 0.1411 H 95.62 91.06 23.86 23.68 51.28 90.78 6.2153 R. salina 141 CHLa L 0.024 0.044 0.047 0.035 0.069 0.0097 M 0.095 0.097 0.152 0.14 0.151 0.0213 H 0.297 0.503 0.449 0.602 0.493 0.509 0.0849 protein L 7.05 11.61 12.88 9.41 20.55 2.8976 M 24.76 25.87 46.62 32.66 39.93 6.5734 H 122.87 195.66 168.54 185.01 158.27 180.8 27.5881 colloidal L 0.777 1.015 1.191 0.984 2.503 0.3529 CHO M 2.23 2.57 3.97 3.82 3.67 0.5598 H 5.71 9.14 8.9 12.38 11.6 8.92 1.7456 storage CHO L 4.14 6.08 5.27 5.73 14.4 2.0304 M 9.61 9.43 21.27 13.19 12.5 1.8598 H 47.57 69.62 68.42 89.54 70.79 78.21 12.6251 * Olenina et al., 2006

The positions on the dot plot for Scenedesmus quadricauda are reversed, with the LL having the highest log10 ratios and the HL having the lowest, possibly due to the fact this species grew very slowly at LL and the CHLa concentrations were very low. In

Amphidinium carterae, the ML and HL treatments had similar effects on the protein:

CHLa relationships when log10 transformed, as shown in the dot plot in Figure 36d.

This may be due to the fact that CHLa and protein concentrations only varied slightly in both experiments and therefore the ratios were very similar. In Rhodomonas salina, the

LL and ML experiments had similar CHLa and protein concentrations, as illustrated in 132

the regression plot (Figure 23a), therefore the dot plot showed a similar effect of light on these variables for these experiments (Figure 23d). In Synechococcus elongatus, the DL and LL experiments showed protein concentration increasing as CHLa increased.

However, the ML and HL experiments showed relatively steady concentrations of CHLa and protein as shown in the regression and dot plots in Chapter III (Figure 8a - 8d).

Interspecies variation was observed for the two chlorophytes investigated in this study. As shown in Table 7, Scenedesmus quadricauda has a higher concentration of

CHLa and protein per cell in the HL experiments when compared to Dunaliella tertiolecta at the same light intensity. The CHLa concentration per cell in Scenedesmus quadricauda was more than four times that of Dunaliella tertiolecta and the protein concentrations in the two species at HL showed approximately a two-fold difference.

Variation in the CHLa and protein concentrations was also observed for the two cyanophytes used in the study. While both species showed trends of increasing CHLa and protein with increasing irradiance, the CHLa concentration per cell at HL was significantly higher in Synechococcus elongatus than in Microcystis aeruginosa - a more than eight-fold difference in some instances. Three to four-fold differences in the protein concentrations per cell were also observed between the two species at the HL experiments. In the two bacillariophytes (diatoms) studied, similar trends were seen with

CHLa and protein per cell increasing with increasing irradiance levels.

133

Table 7: Protein: CHLa (log10) ratios of the species as influenced by irradiance Species Low Light Medium Light High Light Scenedesmus 2.774 ± 0.1381 2.258 ± 0.0592 2.104 ± 0.0209 quadricauda Dunaliella tertiolecta 1.849 ± 0.0175 2.039 ± 0.1185 2.238 ± 0.0401

Thalassiosira weissflogii 2.208 ± 0.0500 2.254 ± 0.0257 2.327 ± 0.0370

Cyclotella meneghiniana 1.996 ± 0.0368 2.078 ± 0.0461 2.167 ± 0.0358

Synechococcus elongatus 2.086 ± 0.0340 2.244 ± 0.0342 2.279 ± 0.0067

Microcystis aeruginose 1.732 ± 0.0374 1.804 ± 0.0380 1.911 ± 0.0421

Amphidinium carterae 2.330 ± 0.0852 2.676 ± 0.1176 2.694 ± 0.0648

Rhodomonas 2.446 ± 0.0241 2.436 ± 0.0288 2.554 ± 0.0450 salina Ratios represent means of replicate cultures.

The CHLa concentration per cell was significantly higher in the Thalassiosira weissflogii batches grown at HL than those of the corresponding Cyclotella meneghiniana batches.

However, apart from what may be an anomaly in one of the batches, the protein concentration in both species is arguably not significantly different. This similarity may be due to Cyclotella meneghiniana reaching the stationary phase of growth much earlier than

Thalassiosira weissflogii at the HL experiments; hence protein synthesis in the cells may have started to slow down instead of turn over. The interspecies variation reported here is evidence that a universal protein: CHLa ratio cannot be developed for a taxonomic group, as species in the same taxon respond differently to the same environmental condition.

However, for ecological modeling, if a community structure is known at the species level, then protein may be able to be adequately estimated from

pigment-based chemotaxonomic CHLa values.

134

Very few studies have attempted to determine protein: CHLa relationships. Myers

and Kratz (1955) studied the cyanobacterium Anacystis nidulans (a.k.a. Synechococcus),

and obtained estimates of cells with highest pigment content to contain 2.8% CHL a and

24% phycocyanin, thus deriving a phycocyanin: CHLa ratio of 8.57:1 (Myers and Kratz,

1955). They conducted their investigations of these relationships at 25 and 39 ºC and at

320 and 960 Watts PAR respectively. They reported that even though there was a three to four-fold variation in either component, there was only small variation in the phycocyanin: CHLa ratio. Phycocyanin is a light harvesting pigment, unique to cyanobacteria, from the phycobiliprotein family. Since this pigment is unique to cyanobacteria, it can be useful as a potential biomarker. It has been shown that CHL a does covary in the same way as specific biomarker pigments (Goericke and Montoya,

1998), so the report that the ratios only varied slightly over the light intensities studied is not surprising. Additionally, the reported phycocyanin: CHLa ratio would have only allowed the phycobiliprotein component of the cells to be estimated, and would therefore not allow for any estimation of total protein from CHLa. Thus, since the major objective

of our current investigation is to ultimately estimate algal cellular protein from CHLa

data, our protein: CHLa results cannot be reasonably compared to the results of the work

done by Myers and Kratz.

Muscatine and Marian (1982) reported a protein: CHLa ratio of 28:1 for algae

isolated from the tissues of Mastigias jellyfish that consumed phytoplankton. That study investigated the dissolved nitrogen flux in symbiotic and non-symbiotic medusa – they reported that the Mastigias had selective symbiotic relationships with nitrogen

135

assimilating dinoflagellates and would move up and down the water column to satisfy the

nutrient needs of the algal symbionts. The species of the dinoflagellates in the marine

lakes were not mentioned in the study, and apart from stating that the medusa had a high

aversion to light intensity and preferred more shaded habitats, irradiance levels were not

reported. If the ratio reported by Muscatine and Marian (1982) were to be log10

transformed, it would still be lower than what we obtained for the lowest irradiance

treatments for Amphidinium carterae, the only dinoflagellate investigated in our study

(Table 6). We know that protein and CHLa concentrations will vary among species of the same taxonomic group and since we have no knowledge of the dinoflagellate symbionts in the study, a true comparison cannot be made between the protein: CHLa ratios reported by Muscatine and Marian (1982) and ours. However, their results represents tentative confirmation that algal protein can be predicted from CHLa data, once prior information is obtained on the species involved and the conditions of the environment being investigated.

Although the relationship is not perfect, it is clear from our results that positive correlations exist between algal protein and CHLa as a function of light intensity. As this is one of the aims of this study, it is now apparent that algal protein content can be at least estimated from CHLa concentration, once prior studies are done on the environmental

conditions for that population under study. As discussed below, the physiological state of

the algae can also be predicted from the protein and carbohydrate concentrations. Our

work represents preliminary findings if pigment-based chemotaxonomy is to be taken to

this level.

136

Phytoplankton colloidal carbohydrate (CHO) as a biomass indicator

Studies have shown that positive correlations exist between colloidal

carbohydrate concentrations and chlorophyll a biomass of benthic microalgae and

cyanobacteria (Underwood et al, 1995; Fabiano & Danovaro, 1994). The extracellular

polymeric substances (EPS) are a major component of the colloidal carbohydrate fraction

of benthic diatoms (Hoagland et al, 1993) and are considered to be analogous to those

produced by planktonic diatoms (Welker et al., 2002). Studies of cultures under

controlled conditions have shown that factors such as nutrient availability and irradiance

affect the release of these colloidal fractions (Myklestad et al., 1989). Smith and

Underwood (1997) developed a model [(log (conc. Coll carbo. +1) = 1.40+1.02 (log (chla

conc. +1))] for predicting colloidal carbohydrate concentration from chlorophyll a data

based on the assumption that if close positive correlation exists between epipelic diatom

biomass and colloidal carbohydrate concentration, then it should be possible to predict

certain components of the colloidal fraction in diatoms in specific environments from

chlorophyll a concentration. Using the assumptions made about periphyton by Smith and

Underwood (1997) and, based on the results obtained from our culture studies, we feel

that once the environmental conditions are known, then it may also be possible to also

predict planktonic colloidal carbohydrate concentrations as well from chlorophyll a data.

In our study, Thalassiosira weissflogii, Cyclotella meneghiniana and

Amphidinium carterae all exhibited significantly different log10 ratios under the three

light treatments, indicating rejection of that the null hypothesis that irradiance had no

effect on the colloidal CHO: CHLa ratios. The regression plots in Chapter III for these

137

species also indicate close positive correlations between chlorophyll a concentration and

colloidal carbohydrate concentration. Light treatment did have effect on the remaining

species in the study, though not as significant as the diatoms and dinoflagellate studied.

Dunaliella tertiolecta, Scenedesmus quadricauda, Microcystis aeruginose, and

Rhodomonas salina all showed LL experiment log10 ratios significantly different from the

ML and HL experiments, though the latter two gave similar ratios, as shown in Table 8

below.

Chlorophytes are not known to secrete huge amounts of colloidal carbohydrates in

response to environmental changes, as also shown from the regression plots for the two

species investigated herein (Chapter III). Though chlorophyll a concentrations were

changing as irradiance increased, the colloidal carbohydrate concentration changed almost linearly and CHO/CHLa decreased with increased light.

Table 8: Colloidal CHO/CHLa (log 10) ratios as a function of irradiance Species Low Light Medium Light High Light Scenedesmus 1.996 ± 0.0200 0.863 ± 0.1031 0.868 ± 0.0249 quadricauda Dunaliella 0.932 ± 0.0398 0.651 ± 0.1183 0.744 ± 0.1081 tertiolecta Thalassiosira 0.892 ± 0.0872 1.072 ±0.0169 1.202 ± 0.0385 weissflogii Cyclotella 0.917 ± 0.0475 1.0773 ± 0.0701 1.3819 ± 0.0659 meneghiniana Synechococcus 0.993 ± 0.0387 1.017 ± 0.0215 1.102 ± 0.0288 elongatus Microcystis 0.760 ± 0.0438 0.837 ± 0.0581 0.841 ± 0.0409 aeruginose Amphidinium 1.624 ± 0.1224 1.793 ± 0.0690 2.016 ± 0.0354 carterae Rhodomonas 1.457 ± 0.0793 1.377 ± 0.0423 1.295 ± 0.0452 salina Ratios represent means of replicate cultures.

138

Phytoplankton storage carbohydrate (CHO) as a biomass indicator

As previously mentioned, light availability has been recognized as one of the

main factors regulating phytoplankton community growth (Welker et al., 2002).

Additionally, many phytoplankton species respond to nutrient limitation by producing

energy storage materials, and as a result, storage carbohydrates may accumulate inside

the cells (Myklestad, 1988/1989). In our study, the effect of irradiance on algal biomass

relationships was investigated using nutrient replete culture batches. Statistical tests

showed that irradiance did have an effect on storage carbohydrate to chlorophyll a log10 relationships for all the species investigated, though some more significantly than for others. Although the cultures were inoculated into nutrient replete media, the media was not replenished with nutrients, and nutrient test results at inoculation and at harvest showed that nutrients decreased but was not completely depleted during growth (data not shown). This decrease in nutrients could have facilitated the accumulation of cellular storage materials, as is a typical occurrence in many phytoplankton species (Borsheim et al., 2005; Granum et al., 2002; Myklestad 1988/1989). Amphidinium carterae,

Thalassiosira weissflogii, Cyclotella meneghiniana, Dunaliella tertiolecta all showed significantly different log10 ratios for the three light treatments, with the LL experiments

having the lowest ratios and the HL having the highest ratios, as shown in Table 9.

Irradiance had no effect on log10 ratios of storage CHO: CHLa for Rhodomonas salina

and Microcystis aeruginosa in the LL and ML experiments, but showed effect when these

ratios were compared to those in the HL experiments, even though there were changes in

the CHO and CHLa concentrations with increasing irradiance, as shown in the regression

plots (Figures26a - 26d and 14a - 14d respectively). For Synechococcus elongatus, even

139

though chlorophyll a concentration increased with irradiance (Figure 8c), storage carbohydrate increased from the DL to the ML experiments, but remained relatively stable from the LL to the ML experiments. Therefore, statistically, irradiance had no effect when the log10 ratios of the LL and ML experiments were compared (Figure 10c).

Table 9: Storage CHO/CHLa (log 10) ratios as a function of irradiance Species Low Light Medium Light High Light Scenedesmus 3.029 ± 0.2273 1.737 ± 0.3062 1.267 ± 0.0278 quadricauda Dunaliella 1.777 ± 0.0238 1.932 ± 0.0526 2.036 ± 0.1134 tertiolecta Thalassiosira 2.032 ± 0.0536 2.157 ± 0.0614 2.288 ± 0.0621 weissflogii Cyclotella 1.746 ± 0.1358 2.140 ± 0.0815 2.378 ± 0.0506 meneghiniana Synechococcus 1.768 ± 0.0172 1.735 ± 0.0251 1.741 ± 0.0236 elongatus Microcystis 1.919 ± 0.0337 1.921 ± 0.0729 2.026 ± 0.0330 aeruginose Amphidinium 2.263 ± 0.0457 2.353 ± 0.0468 2.588 ± 0.0826 carterae Rhodomonas 2.004 ± 0.0861 1.956 ±0.1272 2.233 ± 0.0244 salina Ratios represent means of replicate cultures.

Marker pigments as indicators of algal biomass

Numerous studies have reported the use of the relationship between algal taxonomic marker pigments and CHLa (Grant and Louda, 2010; Louda, 2008; Eker-

Develi et al., 2008; Hagerthey et al., 2006; Millie et al., 1992, among others) and their implications on algal biomass estimation. These relationships are typically incorporated in the computational methods (see Chapter I) for determining composition and abundance of phytoplankton populations (Everitt et al., 1990; Gieskes and Kraay, 1983a, 1986b;

140

Gieskes et al., 1988; Goericke and Montoya, 1988; Letelier et al., 1993; Mackey et al.,

1996; Wright et al., 1996; Pinckney et al., 1998; Van den Meersche et al., 2008; Van den

Meersche and Soetaert 2009).

These relationships are typically refined and re-investigated in an attempt to

address the limitations of the mathematical methods, the inability to address variations of

pigment content within various taxa, even at the species level. Thus, ratios and especially their variability have to be clearly defined in ways that will allow reliable/verifiable estimates of CHLa contributions for each taxon (Jeffrey et al., 1999; Mackey et al., 1996;

Peeken 1997). There is still a need for data describing the pigment ratios of major species

over a wide range of light and nutrient regimes (Jeffrey et al., 1999). This is especially

critical since past studies (Carreto et al., 2008) have shown a logarithmic decrease in pigments (e.g. CHL a per cell) and decreasing CHL a: marker pigment ratios with

increasing irradiance.

In the present study (cf. Grant and Louda 2010), we evaluated irradiance, an

easily measured laboratory and field parameter, as a driver for changes in CHLa: marker pigment ratios. The statistical tests (Chapter III and Appendix VI) showed that the irradiance levels investigated had little effect on the CHLa: marker pigment ratios for all the algal species, except the two cyanophytes investigated. This means that, although pigment (CHLa) per cell increased with irradiance, as shown in Table 6 and in Grant and

Louda (2010), CHLa and certain accessory biomarker pigments co-vary (cf. Goericke and Montoya 1988). That is, while CHLa concentration per cell is decreasing with increasing photon flux (300 μmol photon·m-2·s-1 and higher), the corresponding marker

pigment is decreasing. Thus, the ratios only varied slightly over the irradiance levels.

141

The CHLa: CHLb ratio for Dunaliella tertiolecta was approximately 2.4:1 and

that of Scenedesmus tertiolecta was approximately 2.7:1. The CHL a:CHL b ratio in

CHL b-containing organisms ranges from 2:1 to 3:1 (Halldal 1970, Strain et al., 1971,

Meeks 1974), the data presented here and that in Grant and Louda (2010) both confirm

this.

The two diatoms used in this study: Cyclotella meneghiniana and Thalassiosira

weissflogii, exhibited the characteristic pigments of diatoms with FUCO as the marker

pigment for CHLa divisional estimation. Although CHLc is a known accessory pigment in diatoms (Stauber and Jeffrey, 1988), it was not considered here as a diatom marker pigment, as it is also present in dinoflagellates, prymnesiophytes, cryptophytes, and others (Jeffrey & Vesk 1997, Jeffrey & Wright 2006), and would therefore give errors in taxonomic estimation if a mixed algal sample were being analyzed (see Grant and Louda

2010 for further reasoning). The molar ratios found during this study were about 1.1:1 for

Cyclotella meneghiniana and 1.2:1 for Thalassiosira weissflogii. Chl a: FUCO (molar converted) ratios of 2.34:1 (Gieskes et al., 1988), 1.21:1 (Wilhelm et al., 1991), and 1.8 to 2.6:1 (Garibotti et al., 2003) have been reported from studies on North Sea, Pacific and

Antarctic waters respectively.

The quantitative marker pigment for ‘peridinin-containing’ algae belonging to the division Dinophyceae is peridinin (PERI), (Jeffrey et al., 1975; Johansen et al., 1974).

The CHLa: PERI ratio was approximately 1.13:1 for Amphidinium carterae investigated at the three irradiance levels in this study. This ratio compares very well with that reported by Grant and Louda (2010) for the same species. They reported ratios for

Amphidinium carterae between 0.8:1 and 1.0:1, at irradiance levels of 30-45 through

142

1800 µmol photons·m-2·s-1. Although the ratios presented here are in good agreement with that of Grant and Louda (2010), they are both lower than the 1.5:1 conversion factor used by Louda (2008) to estimate the CHL a contributed from dinoflagellates in studies of Florida Bay, an intense light (~ 750 µmol photons·m-2·s-1) environment. CHLa: PERI ratios as high as 2.35:1 (Everitt et al., 1990; Ondrusek et al., 1991) and 3.96:1 (Barlow et al., 1995) have been reported.

Marker pigments for cyanobacteria are typically ZEA and/or ECHIN. Ratios of

CHLa/ZEA = 1.1:1 and CHLa/ECHIN = 11.0:1 have been used for estimating coccoidal or filamentous cyanobacteria, respectively, in Florida Bay (Louda 2008; Louda et al.,

2000), and Everglades (Hagerthey et al., 2006) studies. Similar ratios have been previously reported for samples grown without light or nutrient limitations (Barlow et al.,

1995; Wilhelm et al., 1991). Although the CHLa/ZEA ratios for the two cyanophytes investigated: Microcystis aeruginosa (LL: 26.67 to HL: 10.67) and Synechococcus elongatus (LL: 6.49 to HL: 0.87), were significantly different (Figures 11a and 7a:

Chapter III), both species showed a marked decrease in the ratios from the low light to the high light experiment levels. This decrease suggests that ZEA functions to protect the species from photodamage (cf. Paerl et al., 1983; Bidigare et al., 1989). Ratios based on

CHLa to photoprotective pigments generally decrease with irradiance (Ruivo et al.,

2011). It was previously found that Synechococcus sp. (elongatus?) had CHLa: ZEA ratios of 2.5:1 or 1.0:1 in the dark brown humic waters of Whitewater Bay or the clearer waters of Florida Bay proper, respectively (Louda 2008: both greater than 600 µmol photons ·m-2·s-1). In this study, with lower light levels, Synechococcus elongatus had

CHLa: ZEA ratios between 6.49:1 and 0.87: 1. Even though ZEA acts as a

143

photoprotective pigment (PPP) in these two cyanophytes, it is still required as a marker

pigment for coccoidal cyanobacteria lacking other carotenoids. ZEA does occur in the

chlorophytes as well but in very much reduced concentrations. The decrease in the ratio

CHLa: ZEA is explained wholly or partly by decreases in cellular CHLa contents as

previously reported for another cyanophytes: Anacystis nidulans (Allen 1968) and with

the largest decreases occurring above 300 µmol photons·m-2·s-1 (Utkilen et al., 1983).

Alloxanthin (ALLO) is the recognized marker pigment for phytoplankton

belonging to the cryptophyte algal division (Chapman, 1966). It is the opinion of J. W.

Louda (pers comm. 2011) that it may be difficult to correlate the microscopic and pigment-based (chemo) taxonomy of cryptophytes with natural communities. He suspects that the phycoerythrin-containing chloroplasts of ruptured cryptophytes cells may be mistaken for coccoidal cyanobacteria during microscopic exams, especially with phycobilin-based epifluorescence methods. Thus, ruptured fragile cryptophytes cells could decrease the cryptophytes count and, at the same time, increase the coccoidal cyanobacterial count. However, the cryptophytes estimate based on alloxanthin as the marker would remain.

The CHLa: ALLO ratios for Rhodomonas salina, the only cryptophyte in this

study, showed relative stability at about 2.6:1. This ratio compares very well with the

1.8:1 – 2.9:1 reported by Hendriksen (Hendriksen et al., 2002) for their study on the

effects of nutrient and light regimes on marine phytoplankton pigments, isolated from

Northern European waters. Their reported ratios are in agreement with ours and are from

a variety of cryptophytes, including Rhodomonas salina, harvested during the exponential

growth phase.

144

Phytoplankton chlorophyll a, protein and carbohydrate relationships to biovolume

Phytoplankton are at the base of the pelagic ecosystem, where they foster the flow of energy through the trophic levels. The understanding and modeling of these ecosystems are not possible without knowledge of species composition and biomass.

According to Paasche (1960), cell concentration data is inadequate for estimating a mixed

phytoplankton community, and for observations on a community containing a wide range

of size classes, biovolume will give a more accurate picture. Cell volumes can be

calculated from size and shapes using the appropriate geometric formulas. The use of a standardized species list with fixed size classes and biovolumes is a decisive method for

improving the quality of phytoplankton counting methods (Olenena et al., 2006).

In this study, the biovolume of the species investigated were obtained from a

standardized phytoplankton species biovolume list (Olenena et al., 2006) and used to

make relationships with the obtained chlorophyll a, protein and carbohydrate

measurements. Chlorophyll is believed to have an allometric relationship with

phytoplankton biovolume and has also been observed to be dependent on light intensity,

temperature and phytoplankton composition (Felip and Catalan, 2000). As shown in

Table 5, certain allometric trends were observed in our study, particularly with Cyclotella

meneghiniana and Thalassiosira weissflogii. These two diatoms have a reported high unit

volume per cell and also gave higher unit CHLa per cell volume than the other species.

Where allometric trends were not readily observed, unit CHLa per cell volume was noted

to generally increase with light intensity for the irradiance levels investigated.

A general trend of increasing unit protein per cell volume with increasing light

intensity was observed for all the species investigated in this study. Although allometric

145

relationships were observed between protein and cell volume, the species in this study with the lowest cell biovolume, Synechococcus elongatus, had a higher unit protein per cell volume when compared to Microcystis aeruginosa (the other cyanophyte investigated). The difference in cellular volume between the two species was more than

15 fold. Protein biovolume are also reported in Table 6.

As shown in Table 6, the species in this study that had large reported cell volumes

(the two diatoms) also showed colloidal carbohydrate having an allometric relationship with biovolume when compared with the other species used. The species in this study belonging to taxonomic groups known for secreting large amounts of colloidal carbohydrates in response to environmental variation, particularly light intensity, also gave higher unit colloidal carbohydrates per cell volume.

Storage carbohydrate per biovolume was noted to show a general increase as light intensity increased, though the increase was more significant for some species than others. This was particularly true for Amphidinium carterae, Cyclotella meneghiniana and Thalassiosira weissflogii. Additionally, these species were reported to have larger cell volumes (Olenena et al, 2006), thus confirming that an allometric relationship does exist between storage carbohydrates and biovolume.

146

V. CONCLUSION: IMPLICATIONS FOR CHEMOTAXONOMY

This study showed that light intensity has significant effects on protein/CHLa,

colloidal CHO/CHLa and storage CHO/CHLa ratio relationships, but a less significant

effect on the known and established marker pigment/CHLa ratios for the species

investigated. Interspecies variation was also observed for protein/CHLa, colloidal

CHO/CHLa and storage CHO/CHLa, and leads to the general conclusion that extensive

knowledge of the influence of light intensity on these parameters are needed before they can be applied to the methods used for chemotaxonomic assessments. Further, the biomass parameter/CHLa ratios to be used in the mathematical applications for estimating algal biomass should come from the more abundant phytoplankton species that are native to the communities studied. Universal ratios of biomass parameter/CHLa

cannot be determined. Seasonal ratios would need to be determined if these parameters

are to find use in the mathematical applications. That is, a set of ratios for the high light intensity summer months and another set of ratios for the low light, late autumn and winter months.

Algal cells are physiologically plastic. That is, cellular components, particularly pigments, proteins, colloidal and storage carbohydrates are altered in response to environmental variables including nutrients, temperature and light. Only the influence of

147

light was investigated in this study. It is known that some of these cellular components,

particularly marker pigments and CHLa change with these environmental variables, but

tend to co-vary in the same way (Goericke and Montoya 1988; Schulter et al., 2000). The relationships that co-vary and are hence most stable are best for use in the mathematical applications used for estimating algal biomass in terms of CHLa. However, converting

CHLa to a more useful currency (unit) of biomass for ecological modeling is still difficult.

In this study attempts were made to determine ratios for the relationships between chlorophyll CHLa and proteins and CHLa and two functional classes of carbohydrates, so that these more useful units of algal biomass could be determined from chlorophyll

CHLa. The study showed that robust relationships exist between CHLa and these other biomass units and prove these relationships to be useful in estimating algal biomass in more useful currencies. CHLa is easily measured or derived from remote sensing. This study showed that algal biomass cannot only be confined to one biomass parameter, but to several – in our case, pigments, proteins and carbohydrates. Fats and total organic carbon could also be considered. The relations determined here can thus find application in the current mathematical methods used for estimating algal biomass, providing more studies are done to assess these relationships in more algal species under different conditions of light and possibly variant nutrient regimes.

148

VI. CHARACTERIZATION OF NOVEL PIGMENT

The ‘scytoneman’ skeleton

The presence of scytonemin in cyanobacteria has been observed in more than 300

species, where sheaths covered with yellow to brown pigments are described (Edwards et

al., 2000). Scytonemin absorbs strongly in the UVA spectral region (315-400 nm),

however, there is absorbance in the violet and blue region as well as in the UVB (280-

320 nm) and UVC (190-280 nm) regions. Until 2004, scytonemin has been the only sunscreen pigment identified from the series, with the characteristic indolic and phenolic subunits – termed the ‘scytoneman’ skeleton (shown in Figure 39). Additionally, some cyanobacteria contain a red to purple pigment called gloeocapsin, instead of scytonemin

(Garcia-Pichel et al., 1993; Garcia-Pichel and Castenholz, 1993). The structure of gloeocapsin is still not known. The production of these molecules is believed to be related to those of other suncreens such as mycosporin-like amino acids in phytoplankton and fungi (Sinha et al., 1998).

Three new pigments, related to the scytonemin skeleton, were isolated and structurally identified in a study aimed at investigating plant succession in the Mitaraka

Inselberg in French Guyana (Butel-Ponce et al., 2004). These molecules are believed to be derived from condensation of tryptophanyl- and tyrosyl-derived subunits with a

149

linkage between the units unique among natural compounds. These new pigments have

been termed tetramethoxyscytonemin, dimethoxyscytonemin and scytonine. They are

related to the scytoneman skeleton and the structures and select spectroscopic

characteristics are presented in Table 10. All three of these, as well as scytonemin exhibit

even molecular masses. By the “nitrogen-rule” they have an even number of N atoms, a

point to be brought out below.

8 O 5 OH 4a 8a 3' 8b 1' 11' 4 N N 3a 1 11 3 12 O 9 5' HO 10 15

Figure 40: The scytoneman skeleton (oxidized form of scytonemin is shown)

We report here the isolation and putative structural elucidation of a novel

sunscreen pigment, isolated form lab grown cultures of Scytonema hofmanii grown at

high light intensities (300-1800 µmol photons·m-2·s-1; Grant and Louda, 2010) as well as

from samples collected in areas of the Florida Everglades (Figure 40). We believe this pigment to possess the scytoneman skeleton. It has similar spectroscopic properties to scytonemin, though with enhanced absorbance maximas in visible region of the electromagnetic spectrum.

150

Figure 41: Red Rock aerial - areas where samples, scraped off rocks, contain the visible light sunscreen pigment.

At this time we are also postulating that the ecological significance of this

pigment in the photosynthetic unit is for the protection of CHLa and cytochrome soret

bands, as well as the α and β bands of cytochromes (e.g. cyt-c562). This is detailed at the end of this chapter

The oxidized form of scytonemin was also isolated from the same sources as the unknown pigment. Spectroscopic results were used to compare the isolated scytonemin with that reported by Proteau and coworkers (Proteau et al, 1993). The 1H and 13C NMR

data presented in Table 11 shows that our data compares very well with that of Proteau

and coworkers. We are therefore sufficiently satisfied that our extraction and purification

method gave scytonemin in good purity.

151

Table 10: Three new pigments isolated form Scytonema sp. collected on Mitaraka Inselberg, French Guyana (Butel-Ponce et al., 2004).

Derivative Name; color m/z [M+H]+ UV/Vis (nm)

OH Tetramethoxy 671 212; 562 scytonemin; O MeO OMe HN NH Purple, amorphous

MeO O OMe solid

OH Dimethoxy 609 215; 316; 422 Scytonemin; O MeO

OMe

HN Dark red, NH amorphous O solid

H3CO Scytonine; 519 207; 225; 270

O HN Brown, amorphous

O O solid NH H3CO

152

Table 11: Scytonemin- a comparison of literature values with those that we obtained (* Proteau et al., 1993, ** FAU – OGG lab work, X signal not observed) no. Scytonemin* 1 Scytonemin 1 13 H * 13 H ** C ** C 1 129.83 x 2 194.17 195.22 3 118.67 118.46 3a 174.30 175.15 4a 163.94 164.48 5 122.08 7.76d (7.7) 122.09 7.51 d (6.0) 6 135.14 7.49 ddd (7.7, 7.6, 135.89 7.54 t (12, 1.1) 9.0) 7 126.64 7.22 dd (7.6, 7.2) 126.59 7.18 t (6.0) 8 129.67 7.89 d (7.2) 129.72 7.64 d (6.0) 8a 125.61 x 8b 158.63 x 9 139.42 8.00 s 139.36 7.58 s 10 126.36 125.84 15,11 136.86 9.00 d (8.7) 136.86 8.64 d (6.0) 14,12 117.08 7.34 d (8.7) 116.77 6.94 d (6.0) 13 163.55 163.46 HO 10.34 bs 10.59 bs

The new highly polar pigment is red-to-mahogany colored in solution and also exhibits spectroscopic properties, somewhat similar to those reported for scytonemin and scytonemin derivatives. The pigment is more much polar than any of the pigments that we usually encounter, eluting from our reversed phase HPLC system before five minutes, as shown in Figure 41. This red compound gave absorption maximas in both the UV and

Visible regions (237 nm, 366 nm, 437 nm, and 564 nm) of the electromagnetic spectrum, as given in Figure 42. When compared to the spectra of scytonemin (Figure 43), the increased intensity of the absorbance from the new pigment in the visible region of the spectrum suggests the presence of an altered chromophore. 153

New pigment

Oxidised scytonemin Solvent front

Reduced scytonemin scytonemin Reduced

Figure 42 HPLC of observed scytonemin and new pigment

Figure 43: UV/VIS absorption spectra of the new pigment

154

Scytonemin – oxidized

New pigment

Figure 44: Scytonemin oxidized and new pigment – overlay

New pigment – putative structure elucidation: (NMR spectra provided in Appendix X)

1 The H NMR spectrum of this red compound (C39H27N3O4 : ESI-TOF-MS)

indicated a methyl group resonating as a singlet at δ 1.91 and two geminal protons resonating as doublets at δ3.20 and 3.67. A primary ketimine signal was observed at

δ175.83 and a methylene signal at δ54.76. The HMBC correlations between these proton and carbon signals were used to build this first part of the molecule and to establish an attachment to the quaternary carbons at positions 3 and 3a and 4a as shown in Figure 44 and Table 12. At this time, the δ175.83 chemical shift observed for the primary imine is not known from the literature. However, we validated the proposed structure of this novel pigment, using mass spectroscopy and other chemical tests.

155

Figure 45: Molecular structure of new pigment from 1H, HSQC, HMBC (black arrows) and COSY (double headed arrows) NMR spectroscopic analyses in DMSO- d6 solution. Chemical name: 3, 3’-Bis-(4-hydroxy-benzylidine)-3a-(2-imino-propyl)-3a,4- dihydro-3H, 3’H-[1,1’]bi[cyclopenta[b]indolyl]-2,2’-dione

Two protons were observed at δ10.13 and 10.20. These are likely the hydroxyl protons of the para-substituted phenols. Two carbonyl signals were observed at δ 193.84 and 197.02, quaternary carbons at δ 105.99 and 63.52 , two aromatic quaternary carbons

(possibly bearing the phenol functions), and several other quaternary carbons. A broad

156

1 13 Table 12: H and C NMR data for putative structure of pigment in DMSO-d6 (13C assignment achieved from HSQC and HMBC experiments) No δ1H (m, JHz) δ13C * 1 129.19 2 193.84 3 105.99 3a 63.52 4 11.91 (broad) 4a 135.12 5 7.49 (d, 7.8, 1H) 126.18 6 6.48 (t, 7.4,7.98 1H) 110.58 7 6.5 (t, 7.98 7.4, 1H) 118.91 8 6.88(d, 8.64, 1H) 128.56 8a 116.21 8b 150.63 9 7.36 (s, 1H) 136.24 10 136.17 11 8.26 (d, 8.26, 1H) 135.83 12 6.86 (d, 8.64, 1H) 116.22 13 160.91 14 6.86 (d, 8.64, 1H) 116.2 15 8.26 (d, 8.26, 1H) 135.83 16 10.20 (s, 1H) 17 3.67 (d, 18.0, 1H) 54.76 17a 3.2 (d, 18.6, 1H) 54.76 18 175.83 19 1.91 (s, 3H) 19.91 20 11.91 (broad) 1' 129.21 2' 197.02 3' 132.44 3'a 173.05 4'a 144.64 5' 7.21(d, 8.22, 1H) 127.23 6' 7.12 (t, 7.68, 7.44, 1H) 121.87 7' 7.24 (t, 7.56, 7.68, 1H) 125.46 8' 7.51 (d, 7.8, 1H) 113.88 8'a 119.12 8'b 150.49 9' 7.34 (s, 1H) 129.54 10' 126.04 11' 7.72 (m, 1H) 131.81 12' 6.95 (d, 1H) 116.9 13' 160.56 14' 6.95 (d, 1H) 116.9 15' 7.72 (m, 1H) 131.87 16' 10.30 (s, 1H)

157

weak signal at δ11.91 was observed in the 1H spectrum, and we are postulating that this represents the N-H protons at positions 4 and 20. The amine signals of compounds having the scytoneman skeleton have reported 1H chemical shifts between 11 ppm and

12ppm (Butel-Ponce et al., 2004; Proteau et al., 1993). Signals H-4 and H-20 were not

observed in any of our COSY, NOSY, HMBC and NH HMBC experiments.

A second part of the molecule was built, starting from the H-11/H-15 and H-

12/H-14 signals, as these protons were observed in COSY correlations. HMBC

correlations were observed with H-11/H-15 protons and the quaternary carbon (δ136.17)

at position 10 and the methine carbon at position 9. The H-9 signal had correlations with

the carbonyl and the quaternary carbons at positions 2 and 3 and 10, thus showing the

link between the first and second parts of the molecule. The indole ring was established

by COSY and HMBC correlations of the signals H-5 to H-8 and long range correlations

observed for H-8 to positions 8a and 8b, and H-6 to 4a. The H-17 signals showed long

range correlations with position 4a and 3a, and H-5 showing long range correlation to 3a,

established the link between the indole and parts one and two of the molecule.

The remaining half of the molecule (labeled prime) was built similarly to that of

scytonemin. Starting with positions H-11’/H-15’ and H-12’/H-14’, the para- substituted

phenol as well the vinyl proton at position 9 and the quaternary carbons at positions 3’

and 2’ were established. The indole ring was also established by HMBC and COSY

correlations of the signals H-5’ to H-8’ and long range correlations of H-8’ to positions

8’b and H-1’. The quaternary carbons at positions C-1 and C-1’ provided the connection

between the two halves of the molecule.

158

Only small amounts of this pigment could be isolated in good purity (≤ 3 mg). As a result, spectral data is limited: For example, clean 13C NMR signals could not be

obtained. All NMR spectra are shown in Appendix X. Mass spectroscopy and several

analytical tests were used to give further verification of the compound’s structure.

Mass interpretation

The proposed structure for this unknown pigment, from mass analyses, has a mass

of 601.2074 Da, calculated for proposed formula C39H27N3O4. The high resolution ESI-

TOF- MS gave m/z 602 [M+H] + and m/z 624 [M+Na] +, as is shown in Figures 45. Our

proposed structure is consistent with this molecular weight and chemical formula. The odd number molecular weight of this compound would suggest that it has an odd number of nitrogens, per the nitrogen rule for organic compounds (McLafferty, 1980). Since the scytoneman skeleton contains two nitrogens, a third nitrogen would have to be present as part of a substituent on the scytoneman core to support the observed odd number molecular weight. Thus, our reasoning for the presence of a ketimine functionality. The imine functionality would also justify the polarity of this pigment, as shown in the chromatogram (Figure 41).

Initially we had proposed acetate or methyl ketone functionalities, but these groups would create more discrepancies: A carbon signal of an acetate group would give chemical shifts between δ160 and δ180, which is consistent with our observed chemical shift. However, an acetate group would give a molecular weight of 618 Da, which is not consistent with our confirmed molecular weight. The carbon signal of a methyl ketone would not give a chemical shift as low as δ 175 and the molecular weight would not be 159

consistent with our confirmed molecular weight. Additionally, acetate and methyl ketone functional groups would give the compound an even number molecular weight.

Figure 46: HR ESI-TOF MS of new pigment – m/z 602 [M+H]+, m/z 624 [M+Na]+

160

Mass analysis was obtained with LC-MS and MALDI-TOF instrumentation at

Florida Atlantic University. Additional verification of the molecular weight was obtained from high resolution ESI TOF-MS instrumentation at the University of Florida,

Gainesville. Fragmentation patterns (MSn) were also obtained from LC-MS instrumentation at the University of Florida. The positive mode ESI- MS/MS of the m/z

602 [M+H] + ion produced prominently m/z 545 via loss of 56 Da. As is shown in Figures

46, this mass is consistent with the loss of the CH2 C (NH) CH3 functionality that

O O OH OH + + H3C H2N HN N N HN O O

m/z 602 [M+H]+ m/z 545 OH OH

Molecular Formula = C39H27N3O4 Monoisotopic Mass = 601.200156 Da

Figure 47: Initial dissociation of m/z 602 [M+H] + ion

we are proposing the structure to have. The m/z 545 is highly aromatic and is thus difficult to dissociate or interpret any observed dissociation. Figure 47 shows the fragmentation of this ion. Though resistant to dissociation, this ion produced m/z 528,

517/518 and 489/490 ions. These ions are consistent with those observed from the atmospheric pressure chemical ionization (APCI) liquid chromatography/MSn of scytonemin (Squier et al., 2004). According to these authors, the ions at m/z 528 represents the loss of 17 Da and is likely due to elimination of a hydroxyl radical from 161

one of the phenol groups. The m/z 517/518 ions represent a loss of 28 Da, and have been assigned to expulsion of CO from one of the cyclopentyl carbonyl groups. The ion at m/z

489 corresponds to loss two molecules of CO. Alternately, the CO losses could occur from the ketone tautomer of the phenol substituents (McLafferty, 1980), as shown in

Figure 48.

SEQ-17095-02 8/16/2011 2:03:56 PM CIDYA; 5 uL Hypurity C8;0.25;95:5(5)>5:95(45-60)/254 nm/(+)ESI

SEQ-17095-02 #1316-1328 RT : 38.32-38.50 AV: 2 NL: 7.07E5 T: + c ESI sid=1.00 Full ms [ 125.00-1000.00] 602.2 100

50 603.3

40 546.3

Relative Abundance Relative 30

20 545.3 604.2 547.3 10 264.2 149.1 186.9 202.7 218.8 284.8 413.4 624.1 239.2 300.8 332.7 340.9 391.2 428.8 464.7 505.0 543.2 548.5 601.3 660.3 674.4 707.0 738.3 750.2 782.0 794.7816.8866.7 877.0 902.5 937.4 958.1 986.6 0 150 200 250 300 350 400 450 500 550 600 650 700 750 800 850 900 950 1000 m/z

SEQ-17095-02 #1317 RT : 38.34 AV: 1 NL: 1.00E6 T: + c sid=1.00 d Full ms2 [email protected] [ 185.00-1215.00] 545.4 100 90 80 70 546.3 60

50 40 Abundance Relative 30 20 547.3 556.3 10 557.4 329.4 496.4 510.4 574.5 584.4 0 140 160 180 200 220 240 260 280 300 320 340 360 380 400 420 440 460 480 500 520 540 560 580 600 m/z SEQ-17095-02 #1317-1329 RT : 38.37-38.55 AV: 2 NL: 2.32E5 T: + c sid=1.00 d Full ms3 [email protected] [email protected] [ 140.00-1100.00] 517.4 100 489.4 90 80 528.3 70

60 50 40 490.5 Relative Abundance Relative 30 516.5

20 529.3 546.3 10 425.4 515.5 400.4 501.4 547.3 344.4 373.4423.4 426.4 451.4472.8 488.6 530.3 0 140 160 180 200 220 240 260 280 300 320 340 360 380 400 420 440 460 480 500 520 540 560 580 600 m/z

Figure 48: (+) ESI- MS/MS dissociation of m/z 602 produced m/z545, which further dissociates to give m/z 528,518, 517 and 489 ions.

162

O N -OH N

C36H20N2O3 Exact Mass: 528.16 OH Mol. Wt.: 528.56

-CO

N N O Alternate -

C35H22N2O3 Exact mass: 518.16 OH Mol. Wt.: 518.56

Figure 49: Fragmentation patterns of the highly aromatic portion of the new pigment (consistent with that of scytonemin).

An additional molecular weight of 727 Da was also observed during this analysis. The

HPLC data is provided (Figure 49), but no mass interpretation is given, as we feel that

163

this may be an impurity that is associated with our pigment. NMR spectra showed that the sample was not fully pure. All mass spectra and associated HPLC data are given in

Appendix XI.

SEQ-17095-02 8/16/2011 2:03:56 PM CIDYA; 5 uL Hypurity C8;0.25;95:5(5)>5:95(45-60)/254 nm/(+)ESI RT: 0.00 - 60.02 39.00 NL: 5.26E6 602.3 Base Peak F: + c 100 ESI sid=1.00 Full ms [ 80 125.00-1000.00] MS SEQ-17095-02

50.80 51.63 40 386.7 684.2 49.12 54.32 56.33 38.50 237.1 832.1 59.19 Relative Abundance Relative 48.27 906.0 20 602.2 257.2 980.1

RT: 39.00 NL: 5.26E6 BP: 602.3 m/z= 601.7-602.7 F: 100 + c ESI sid=1.00 Full m s [ 80 125.00-1000.00] MS SEQ-17095-02

40 RT: 38.50 Relative Abundance Relative 20 BP: 602.2

0 40.90 NL: 3.42E5 728.1 m/z= 727.6-728.6 F: 100 + c ESI sid=1.00 Full m s [ 80 40.55 125.00-1000.00] 728.1 MS SEQ-17095-02 60

40 59.02 54.32 58.51 728.3 53.49 728.4 728.1 Relative Abundance Relative 20 39.84 46.40 51.63 727.8 728.1 727.7 727.6

0 RT: 39.03 NL: 7.08E-2 BP: 0.0 UV Analog 0.070 SEQ-17095-02

0.065 1.87 0.0

0.060 RT: 38.36 46.18 46.95 49.07 49.96 52.24 53.93 56.27 57.25

Intensity 41.91 43.59 60.00 30.72 31.81 34.20 35.97 BP: 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 19.19 20.38 22.70 24.90 25.69 29.57 0.0 0.0 0.0 1.11 2.70 4.69 7.00 9.20 9.48 9.86 12.54 15.10 18.36 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.055 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0

0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50 52 54 56 58 60 Time (min)

Figure 50: HPLC/UV. The 601 Da compounds are shaded; The 727 Da compound is likely an impurity associated with the pigment

164

Deuterium exchange reactions were used to confirm the presence of the two

alcohol substituents, as well as the NH protons: The pigment was dissolved in deuterated

1 methanol (CD3OD) and H NMR analysis was done (spectrum shown in Appendix X).

The disappearance of the two phenol protons, due to exchange with the heavier hydrogen,

confirmed their presence. The disappearance of the postulated N-H signals (position 4

and 20 in Figure 44) also confirmed their presence.

Further, acetylation reactions were carried out to confirm the 601 molecular weight as

well as the presence of the NH and OH functional groups. As is shown in Figure 50, the

increased mass of m/z 686 (M+H) + mass units verify that two functional groups on the

compound was acetylated. A second signal at m/z 728 (M+H) + mass units is likely the

same impurity that was observed in the LC/MSn analysis done at the University of

Florida.

Figure 51: mass analysis of pigment after acetylation- m/z 686 [M+H] + confirms acetylation of the two phenol OH groups

165

The pigment was further acetylated, in an attempt to acetylate all of the OH and

NH groups on the compound. HPLC analysis shows that multiple actylations took place.

However, LC-MS of this acetylated product failed to give any further information. One

distinct peak was seen in the LC analysis, but a mass was not obtained for this peak, due

to the fact that if all the NH and OH protons became acetylated, then no H would be

available for ionization in ESI mode.

IR analysis

The IR spectrum of the unknown pigment is shown in Figure 51. The broad signal

at 3411.8 cm-1 likely corresponds to the phenol functionality. Alcohol/phenol compounds

are known to have characteristic broad O-H stretching vibrations between 3200 cm-1 and

3552 cm-1. Signals for N-H functional groups of amines and imines are typically

observed between 3300 cm-1 and 3500 cm-1, thus it is likely that the broad O-H phenol

signal has overlapped these signals. The signals at 2962cm-1 and 2922cm-1 are typical of

C-H stretching vibrations of alkanes. Alkanes give C-H stretching signals between 2850

cm-1 and 3000 cm-1.Therefore, these signals are likely due to the aliphatic portion of the

molecule. However, C-H stretching vibration modes of aromatics also resonate in this

area of the spectrum, so these signals may also reflect vibrations from the aromatic

portion of the molecule as well. Signals for C=O of ketones, C=C of aromatics and

aliphatics, C=N of imines and N-H of primary amines are typically observed at 1665 cm-

1- 1710 cm-1, 1640 cm-1- 1680 cm-1, 1500 cm-1- 1700 cm-1, 1620 cm-1- 1690 cm-1 and,

1580 cm-1- 1650 cm-1 respectively. Since all of these functional groups are present in our

proposed structure, the signals at 1716 cm-1 and 1655 cm-1 are the overlapping vibrations 166

from these groups. The signal at 1451 cm-1 may be due to C-H bending vibrations, which

are typically observed between 1450 cm-1- 1470 cm-1. The C-O stretching vibration

modes are generally observed at 1000 cm-1- 1320 cm-1. Thus, the medium signal at

1376cm-1 is probably from this functional group, as it is present in our structure.

Figure 52: IR spectra of new pigment

Ecological significance of the new pigment

This pigment was isolated from samples of Scytonemin hofmanii grown at light intensities between 300 and 1800 µmol photons·m-2·s-1(Grant and Louda, 2010). It was not observed in any of the cultures grown at 100 or 180 µmol photons·m-2·s-1. The

pigment was also isolated from samples collected in areas of the Florida Everglades that

are subject to intense light conditions (> 1500 µmol photons·m-2·s-1). For this reason, and

according to the spectral overlay and absorption emission spectrum shown in Figures 52 167

and 53, we believe that the role of this pigment is to protect the chlorophyll a and

cytochrome soret bands from the excesses of visible light radiation, around 430 - 440 nm.

Figure 53: New pigment and chlorophyll a – overlay spectra

Figure 54: Excitation, emission spectral overlay of new pigment

It likely functions to reduce the amount of excitation energy reaching the chlorophyll a molecules in PS II. In doing this, electron transport chains and the critical D-1 protein are not damaged, and photosynthesis can continue without a reduction in efficiency.

Chlorophyll a soret absorption maxima occur at 430 nm, while one of the absorption maximas for this pigment is at 434 nm. The concentration per cell of the new pigment is

168

typically higher than that of chlorophyll a at the light intensities where it observed (Grant

and Louda, 2010). That is, assuming εmm of the new pigment is of similar magnitude to

that of scytonemin in the violet/UVa region.

Electrons that are generated as a result of excitation of the specialized chlorophyll

a molecules of PSII are carried to PS I via an assembly of membrane proteins known as

cytochrome b563 and c552. These cytochromes have α band absorbance maxima at 563 nm

and 552 nm respectively. Studies have shown that in addition to shuttling electrons to

PSI, the cytochromes also play a protective role during the photoactivation role of PSII

(Schweitzer and Brudvig 1995). That is, they act as a possibly cyclic ‘molecular switch’, redirecting electron flow within PS II by changing from a high to a low potential form

(Falkowski et al., 1986; Prasil et al., 1996; Mor et al., 1997). This molecular switch would thus shuttle excess electrons away from photosynthetic electron transport chain. It would therefore seem critical for these cytochromes to not become activated by light, but only by resonance energy transfer from the components in the PS II and PS I pathway.

Excitation of the cytochromes by light would only serve to introduce more electrons

(excited states) that could ultimately directly or indirectly (reactive oxygen species generation) damage the reaction centers of the photosystems. Since our new pigment has absorbance maxima at 562-564 nm (solvent effect), we speculate at this time that it may also be protecting the cytochrome α and β absorbance bands as well as the Soret from light activation. Future work is obviously needed to verify its protective role.

169

VII. APPENDICES

170

I- Pigment calculation and data handling

Data Handling: The entire method, from purchase of the samples to the generation of ratios, taxonomic division and concentration per cell will be illustrated here with a

Amphidinium carterae sample.

The dinoflagellate Amphidinium carterae was purchased from the Carolina Biological

Supply Company. Growth media was prepared and the sample was grown and analyzed as per materials and methods section.

Pigment Calculation: The absorbance (µV·s) data was next entered into an in-house

(Florida Atlantic Organic Geochemistry group) generated Excel® spreadsheet called

“PIGCALC”. This spreadsheet contains standardized equations and specific absorption coefficients and was used to calculate the quantity of each pigment, sums and ratios of pigments and then converted that information into taxonomic divisions of algae. The

PIGCALC spreadsheet presented here only contains the pigments that are specific to

Amphidinium carterae. Otherwise, a generic PIGCALC, containing all of the pigments associated with cyanobacteria and eukaryotic microalgae was used.

Pigment calculation was carried out as follows: The sample name and weight or volume was entered in the appropriate cells. The number of dilutions (from UV/Vis aliquot) was entered in cell I2. The UV/Vis spectra of the chlorophylls and carotenoids of this species was below 1, so no dilution was needed. The corrected weight of the sample is calculated in cell J1 by dividing the original weight by the dilution factor. The corrected weight is given in column K also (0.051).

171

The Internal Standard UV/Vis value of the day’s extractant solvent (A394 – A750) was entered at cell G4. The software calculates the IS added in I4 and H36. The Internal

Standard recovered from HPLC was quantified from the peak at 394nm using ε mM = 305, this is the extinction per millimolar solution per 1cm light path. The extinction coefficient of the Internal Standard is different from that used for pigments. The pigment extinction

1% coefficients are E 1cm: 1% values. That is, the absorption of a solution over a 1cm light

path, as given in the literature (see Davies, 1965).

The absorbance (µV·s) value of the Internal Standard peak at 23.60 minutes at the

394nm integration (0.004819617) was entered in cell O36. The molecular weight was

entered in cell C36. The µV·s value, is divided by the molecular weight and the result is

multiplied 0.001 to give the number of moles of Internal Standard (cell F36) found in the

sample (IS found). The molecular weight divided by the number of moles (C36/F36) gives

the weight of Internal Standard found in the sample. The correction factor in J36 was IS

added/IS found (H36/F36). The correction factor formed was 1.57x. All other pigment

corrections are then by a factor of 1.57x.

The absorbance (µV·s) values for each identified peak was entered in column O,

1% the molecular weights in column C and the extinction coefficients (E 1cm) in column B.

The weight of each pigment (column E) was calculated by dividing the µV·s values by

the extinction coefficient, multiplied by 100. The number of moles of each pigment

(column F) was determined by dividing the corresponding values in column E by those in column C. The moles of total chlorophyll a were summed and are shown in G7:G18 and

D19 (1.38775 E-10). Total chlorophyll represents the sum of all chlorophyll a derivatives

172

and the degradation products. The number of moles of the pigments from F7: F18 are

then each expressed as a percentage of total chlorophyll a (column H).

The ion pairing solution was added before injection (0.125 mL IP to 1.00 mL

extract). The final volume of the prepared injectate was 1.125 mL, of which 0.100 mL was injected. Thus, the extract actually injected was 88.89 µL. Correction to original 3.0 mL was 3.0000/0.08889 =33.75. The concentration per mL of the pigments (column J) was determined by multiplying the weight of each pigment by 33.75 and then dividing that by the corrected weight of the sample.

Chlorophyll b and chlorophyll c are never included in the total for chlorophyll a, since they are not found in all algae. In this case, chlorophylls c1/c2 are present in

Amphidinium carterae, and the calculations for this pigment is shown in row 32.

The carotenoid pigments found in Amphidinium carterae, as well as those associated with

cyanobacteria, phytoplankton and plants are shown in column A, lines 41 to 69, with two

spaces for unknown carotenoids also included. In this illustration, the absorbance (µV·s)

data for Peridinin (retention times 15.28 and 16.45) are entered in cell O44. The Peridinin

weight (E44) and number of moles (F44) are calculated as stated above. The number of

moles of all the carotenoids was summed and is shown in G41: G69 (same number). Each

pigment is then expressed as a percentage of total number of moles of carotenoids

(column H, lines 41:69). In this example, Peridinin made up 64.89% of the total

carotenoids. Molar ratios of each pigment to total CHLa (D19) is given in column I

(lines 41:69) and the inverse, total CHLa to each pigment is given in column M (lines

41:69). The molar ratio of peridinin to total CHLa was found to be 0.82 (I44) and the

inverse, total CHLa to PERI was calculated to be 1.22 (M44).

173

Ratios of CHLa to specific carotenoids were then calculated by summing the

moles of chlorophyll a (i.e. CHLa allomer, CHLa and CHLa epimer) and dividing that by

the moles of the carotenoid i.e. (SUM F11: F13) /F44. In this example the chlorophyll

a/Peridinin ratio is 1.18 (B75). These are the ratios of interest, and they are shown in the

section titled ‘Other Ratios’.

For the Divisional Estimate, the only divisional estimator in this example is

Peridinin for the Dinoflagellates. The number of moles of this marker pigment is entered in cell H85. The chlorophyll a estimate per division marker is calculated in cell H86 by multiplying H85 by 1.5. This 1.5 represents what we have determined the appropriate ratio of total chlorophyll to the marker pigment for this species should be. It is based on ratios derived from extracting algae in our Organic Geochemistry group’s laboratory and on previously reported ratios, all of which were not related to light studies. The calculated total CHLa / PERI ratio for this sample is 1.22 (M44), potentially revealing a need for adjusting the ratios for changes in light intensity. A histogram plot is then made to show

the percentage contribution that each division makes to the sample. In this case, the

sample is obviously 100% dinoflagellate.

The chlorophyll a concentration was calculated as follows: The mass of ‘live’

chlorophyll a was summed in cell C108: ((SUM(E7:E8) + SUM(E12:E14)). This total is

then multiplied by the IS correction factor to give the corrected mass in C109. The

corrected mass is multiplied by 33.75 to give the total mass of live chlorophyll a (C110).

The mass in grams is converted to micrograms in C111. Finally, C111 is divided by the corrected volume of the sample (H107) to give the concentration of chlorophyll a in

174

μg/mL of sample. The same method was used to calculate the concentration of

chlorophylls c and phaeophytins in the sample.

The percent CHLa estimate from HPLC was calculated as follows: The CHLa

estimate per division (E117) was divided by the sum of live chlorophyll a (C88/(F7+F8)

+SUM (F11:F13))) and multiplied by 100. Summing the number of moles of

pheopigments and dividing by the molar sum of chlorophyll a, then multiplying by 100

calculated the percentage of pheo pigments. Taking the sum of the number of moles of

the derivatives and dividing by the molar sum of CHLa calculated the percentage of

CHLa derivatives.

Pigment per cell was calculated as follows: The number of cells in 1 mL of this

Amphidinium carterae grown at 70-75 μmol photons·m-2·s-1 was determined to be

208638 (0.209 x 106). The concentration per mL of each pigment in the sample was

determined from PIGCALC (column J). This data was exported to another spreadsheet and pigment concentration was divided by the number of cells counted, to give the concentration of that pigment per cell.

The protein concentration, colloidal carbohydrate concentration and storage concentrations were determined as described in chapter two. The concentrations were

exported to spreadsheets where they were compared to chlorophyll a concentration for

that sample, per milliliter and per cell.

175

PERI CHLa CHLs c1/c2

DIAD

o ll a BETA a HL DINO DINO DIAT C CHLa` P468 P457

HPLC Chromatogram (λ = 440 nm) for Amphidinium carterae

Chromatogram peak areas, measured at 440 nm

176

Pigment calculation (PIGCALC)

177

Pigment calculation (PIGCALC), continued

178

II. Select photoprotectorant and accessory pigments

Selected carotenoids: photosynthetic accessory and photoprotectorant pigments

β- Carotene α- Carotene

OH O OH

Diatoxanthin Diadinoxanthin

OH OH

OH AcO

Dinoxanthin Lutein

OH HO

OH OH HO O

Neoxanthin Zeaxanthin

O HO

Antheraxanthin Violaxanthin

179

Canthaxanthin Echinenone

OH C6H11O4-O

OH HO OH HO

Myxoxanthophyll (Myxo) Myxoxanthophyll (Myxol)

HO HO OH HO

Fucoxanthinol Astaxanthin

N N N N Mg Mg N N N N

H O

H COOCH3 O COOH COOCH3 COOH Chlorophyll c1 Chlorophyll c2

180

III- Spectroradiometric output

Spectroradiometer output for the three main light levels – obtained using HR4000 Spectrometer (Ocean Optics Inc.), coupled to OOI base 32 software using a Dell PC

436

476

405 635 655

708 811 365

High Light Experiment – Inside empty growth chamber

436

476 635 655

405 708 811

365

High Light Experiment – Inside (center) Thalassiosira weisflogii culture

181

542

581 604 436

Medium Light Experiments – Inside empty growth chamber

581

542 604

436

Medium Light Experiments – Inside (center) Thalassiosira weissflogii culture

182

612 546 586

487 709

404

Low Light Experiments – Inside empty growth chamber

546

586 612 487 709

404

Low Light Experiments – Inside (center) Dunaliella tertiolecta culture

183

UVC-B

Near-IR UVA PAR

Approximately 90% of light is transmitted through the growth containers

184

IV- Calibration curves and equations

Bovine Serum Albumin calibration curve and equation for microbiuret assay

Glucose calibration curve and equation used for phenol sulfuric acid assay

185

TOC standard curve 1.2

1 0.8

0.6 y = 0.4401x

Absorbance 0.4 R² = 0.9984 0.2 0

00.511.522.5 TOC (mg/mL)

Potassium hydrogen phthalate standard curve for Walkley-Black assay of total organic carbon

186

V- Retention times and UV-Vis maximas

Retention times and UV/Vis (PDA) spectral data for Chlorophylls, Chlorophyll derivatives, carotenoids and scytonemin for C-18 column on Waters® 996 HPLC-PDA system. PIGMENT______TIME(min.)____UV/VIS(nm)______Solvent Front ~4.5 N/A Scytonemin-like 5.10 372, 440, 562 Bacteriochlorophyllides-d unkn 412, 428, 616, 658 “P468” 5.12 472 “P457” 5.14 460 Scytonemin (Reduced) 7.103 Chlorophyllinde-b unkn 464, 654 Chlorophyllide-a 9.31 426, 582, 616, 660 Chlorin-e6 free acid unkn 414, 514, 554, 606, 660 Chlorophylls-c1/-c2 5.95 446, 582, 628 Scytonemin (oxidized form) 11.11 388 Fucoxanthinol 12.24 452 Cu-chlorophyllin unkn 406,508, (575), 628 Pyro-Chlorophyllide-a* 11.57 426, 582, 616, 660 Peridinin 13.98 474 Pyropheophorbide-b 14.61 438, 530, 600, 656 Vaucheriaxanthin (19-hydroxy-neoxanthin)unkn (422), 440, 476 19’-butanoyloxyfucoxanthin 15.59 446, 470 Siphonoxanthin* unkn 448, (468) Pheophorbide-a 15.95 408, 506, 534, 610, 668 Fucoxanthin 16.26 452 Neoxanthin 16.94 414, 438, 466 Bacteriopheophorbides-d unkn 408, 426, 614, 656 “Polar” MYXO (= aphanizophyll ?) unkn (448), 476, 508 19’-hexanoyloxyfucoxanthin 15.68 446, 468 Pyropheophorbide-a 16.95 412, 510, 540, 608, 666 Violaxanthin 18.60 418, 442, 470 Prasinoxanthin 18.27 454 Pheophorbide-b ME unkn 436, 526, 598, 654 Pheophorbide-b’ ME unkn 436, 526, 598, 654 Myxoxanthophyll (MYXO) 18.76 (448), 476, 508 Astaxanthin unkn 480 Cu-Pheophorbide-a-ME unkn 408, 500,540, (590), 642 Dinoxanthin 19.58 418, 442, 470 cis-Fucoxanthin 20.28 320 , 440, (462) Diadinoxanthin 20.61 (426), 448, 476 Cu-Mesopyropheophorbide-a-ME unkn 418, 544, 592, 636

187

PIGMENT______TIME(min_UV/VIS(nm)______Bacteriochlorophyll-d(1) 26.11 408, 428, 614, 656 Antheraxanthin* 21.95 446, 473 Cu-Chlorine-e6-TME unkn 406, 500, 634 Pyropheophorbide-b ME unkn 436, 526, 598, 654 Bacteriochlorophyll-d(2) unkn 408, 428, 614, 656 BCHL-c3(7%:4n-Pr, 5Et, 2S)* unkn 434, 630, 666 Pheophorbide-a ME 19.40 410, 508, 538, 608, 666 Cu-Chlorin-p6-TME unkn 406, 500, 538, 640 Phoenicoxanthin unkn 480 Alloxanthin 22.45 (426), 454, 482 BCHL-c5 (71%: 4Et, 5Et, 2R)* unkn 434, 630, 666 Pheophorbide-a’ ME 21.26 410, 508, 538, 608, 666 Diatoxanthin 23.58 (426), 454, 484 BCHL-c4 (17%: 4n-Pr, 5Et, 2R)* unkn 434, 630, 666 BCHL-c1 (5%: 4iBu, 5Et, 2S)* unkn 434, 630, 666 Monadoxanthin unkn (422), 448, 476 Bacteriochlorophyll-d(3) unkn 408, 428, 614, 656 Pyropheophorbide-a ME 21 410, 508, 538, 608, 666 Cu-Purpurin-18-ME unkn 416, 504, 544, 622, 670 Phoenicoxanthin unkn 476 Lutein (,-carotene-3,3’-diol) 24.27 (422), 446, 476 Isozeaxanthin (,-carotene-4,4’-diol) unkn (424), 454, 480 Zeaxanthin (,-carotene-3,3’-diol) 24.61 (424), 454, 480 Bacteriochlorophyll-d(4) unkn 408, 428, 614, 656 4’-hydroxy-echinenone* 26.11 462 (7-?) cis-zeaxanthin* unkn 336, (426), 448, 474 Siphonein* unkn 334, 452, (478) Bacteriochlorophyll-d(5) unkn 408, 428, 614, 656 Bacteriochlorophyll-agg unkn 360, 580, 770 Canthaxanthin (,-carotene-4,4’-dione) 27.45 472 Cu Mesoporphyrin-IX DME (Int.Std.) 27.78 394, 524, 558 Gyryoxanthin Diester 26.61 (422), 448, 470 Bacteriopheophytin-c3* ( 7%) unkn 412, 518, 550, 614, 668 Monodemethylated spirilloxanthin* unkn 468, 494, 530 Rhodovibrin* unkn 458, 484, 518 Bacteriopheophytin-d(1) unkn 424, 520, 612, 652 Bacteriopheophytin-c5* ( 62%) unkn 412, 518, 550, 614, 668 Bacteriopheophytin-c4* ( 27%) unkn 412, 516, 552, 614, 668 Bacteriopheophytin-d(2) unkn 424, 520, 612, 652 Bacteriochlorophyll-ap 29.97 358, 580, 772 Chlorophyll-b 30.78 458, 596, 646 Cyclopyropheophorbide-a-enol unk 360,426, 156, 628, 686 3,4-Didehydrorhodopin* unkn (458), 486, 520

188

PIGMENT______TIME(min_UV/VIS(nm)______Crocoxanthin unkn (422), 448, 476 Bacteriopheophytin-c1* ( 4%) unkn 412, 518, 550, 614, 668 Rhodopin* unkn 474, (505) Spirilloxanthin unkn 470, 496, 530 Chlorophyll-b’ (epimer) 31.61 458, 596, 646 1 - 13 -oxydeoxo-Chlorophyll-a (prep:BH4 .) unkn 416, 514, 562, 606, 654 Chlorophyll-a-allomer (“132-OH-Chl-a”) 32.11 430, 582, 616, 662 Cryptoxanthin unkn (428), 456, 480 Isocryptoxanthin unkn (428), 456, 480 Chlorophyll-a 32.76 430, 582, 616, 662 Echinenone (,-caroten-4-one) unk 462 Chlorophyll-a’ (epimer) 33.61 430, 582, 616, 662 Anhydrorhodovibrin* unkn 460, 482, 518 Pheophytin-b-allomer (“132-OH-PP-b”) unkn 436, 528, 598, 656 Bacteriopheophytin-agg unkn 358, 526, 750 Bacteriopheophytin-ap 34.41 358, 526, 750 Pheophytin-b 35.28 436, 528, 598, 656 Bacteriopheophytin-ap'(epimer) unkn 358, 526, 750 Pheophytin-b’ (epimer) unkn 436, 528, 598, 656 Astaxanthin esters (Panulirus argus) unkn 478 Lycopene unkn 448, 474, 506 Pheophytin-a-allomer (“132-OH-PP-a”) 36.28 410, 502, 536, 610, 666 Pyrobacteriopheophytin-ap unkn 358, 526, 750 Pyropheophytin-b 36.78 436, 528, 598, 656 Pheophytin-a 37.13 410, 502, 536, 610, 666 -Carotene 37.82 440, 465, 495 Pheophytin-a’ (epimer) 37.45 410, 502, 536, 610, 666 -Carotene 39.18 (422), 448, 476 -Carotene (all-trans, all-E) 39.26 (428), 456, 482 cis--Carotene (15-Z, tent.) 39.56 338, (424), 448, 476 Purpurin-18-phytyl Ester* unkn 360, 408, 546, 696 Pyropheophytin-a 40.19 410, 502, 536, 610, 666 Pheophorbide-a-steryl ester(s) 41.27 410, 502, 536, 610, 666 Pyropheophorbide-a-steryl esters* 41.94 410, 502, 536, 610, 666

189

Amphidinium carterae - CHLa/PERI ratios

Descriptives CHLa/PERI 95% Confidence Interval for Mean

N Mean Std. Deviation Std. Error Lower Bound Upper Bound Minimum Maximum 1.00 5 1.0320 0.11032 0.04934 0.8950 1.1690 0.92 1.15 2.00 6 1.2983 0.14091 0.05753 1.1505 1.4462 1.12 1.49 3.00 6 1.0967 0.25944 0.10591 0.8244 1.3689 0.80 1.48 Total 17 1.1488 0.20964 0.05084 1.0410 1.2566 0.80 1.49 VI-ANOVA tables

190 Test of Homogeneity of Variances CHLa/PERI Levene Statistic df1 df2 Sig. 2.982 2 14 0.083

ANOVA CHLa/PERI Sum of Squares df Mean Square F Sig. Between Groups .219 2 0.109 3.159 0.074 Within Groups .484 14 0.035 Total .703 16

Robust Tests of Equality of Means CHLa/PERI Statistica df1 df2 Sig. Brown-Forsythe 3.365 2 9.804 0.077 a. Asymptotically F distributed.

Post Hoc tests Multiple Comparisons Dependent Variable: CHLa/peridinin

(I) amphidinium (J) amphidinium Mean 95% Confidence Interval Difference (I-J) Std. Error Sig. Lower Bound Upper Bound 191 Tukey HSD 1.00 2.00 -0.26633 0.11265 0.079 -0.5612 0.0285

dimension3 3.00 -0.06467 0.11265 0.836 -0.3595 0.2302 2.00 1.00 0.26633 0.11265 0.079 -0.0285 0.5612

dimension2 dimension3 3.00 0.20167 0.10740 0.182 -0.0794 0.4828 3.00 1.00 0.06467 0.11265 0.836 -0.2302 0.3595

dimension3 2.00 -0.20167 0.10740 0.182 -0.4828 0.0794 Games-Howell 1.00 2.00 -0.26633* 0.07579 0.016 -0.4780 -0.0547

dimension3 3.00 -0.06467 0.11684 0.848 -0.4088 0.2795 2.00 1.00 0.26633* 0.07579 0.016 0.0547 0.4780

dimension2 dimension3 3.00 0.20167 0.12053 0.275 -0.1454 0.5488 3.00 1.00 0.06467 0.11684 0.848 -0.2795 0.4088

dimension3 2.00 -0.20167 0.12053 0.275 -0.5488 0.1454

Multiple Comparisons Dependent Variable: CHLa/peridinin

(I) amphidinium (J) amphidinium Mean 95% Confidence Interval Difference (I-J) Std. Error Sig. Lower Bound Upper Bound Tukey HSD 1.00 2.00 -0.26633 0.11265 0.079 -0.5612 0.0285

dimension3 3.00 -0.06467 0.11265 0.836 -0.3595 0.2302 2.00 1.00 0.26633 0.11265 0.079 -0.0285 0.5612

dimension2 dimension3 3.00 0.20167 0.10740 0.182 -0.0794 0.4828 3.00 1.00 0.06467 0.11265 0.836 -0.2302 0.3595

dimension3 2.00 -0.20167 0.10740 0.182 -0.4828 0.0794 Games-Howell 1.00 2.00 -0.26633* 0.07579 0.016 -0.4780 -0.0547

dimension3

192 3.00 -0.06467 0.11684 0.848 -0.4088 0.2795 2.00 1.00 0.26633* 0.07579 0.016 0.0547 0.4780

dimension2 dimension3 3.00 0.20167 0.12053 0.275 -0.1454 0.5488 3.00 1.00 0.06467 0.11684 0.848 -0.2795 0.4088

dimension3 2.00 -0.20167 0.12053 0.275 -0.5488 0.1454 *. The mean difference is significant at the 0.05 level.

Amphidinium carterae – Protein/CHLa relationships

Descriptives Protein/CHLa 95% Confidence Interval for Mean

N Mean Std. Deviation Std. Error Lower Bound Upper Bound Minimum Maximum 1.00 5 2.3300 0.08520 0.03810 2.2242 2.4358 2.25 2.43 2.00 6 2.6762 0.11764 0.04802 2.5528 2.7997 2.46 2.80 3.00 6 2.6944 0.06483 0.02647 2.6264 2.7625 2.61 2.77 Total 17 2.5808 0.18805 0.04561 2.4841 2.6775 2.25 2.80

193

Test of Homogeneity of Variances Protein/CHLa Levene Statistic df1 df2 Sig. 0.315 2 14 0.735

ANOVA Protein/CHLa Sum of Squares df Mean Square F Sig. Between Groups 0.447 2 0.223 26.215 0.000 Within Groups 0.119 14 0.009 Total 0.566 16

Post Hoc Tests

Multiple Comparisons Dependent Variable: Protein/CHLa

(I) Amphidinium (J) Amphidinium Mean 95% Confidence Interval Difference (I-J) Std. Error Sig. Lower Bound Upper Bound Tukey HSD 1.00 2.00 -0.34620* 0.05588 0.000 -0.4925 -0.1999

dimension3 3.00 -0.36443* 0.05588 0.000 -0.5107 -0.2182 2.00 1.00 0.34620* 0.05588 0.000 0.1999 0.4925

dimension2 dimension3 3.00 -0.01823 0.05328 0.938 -0.1577 0.1212 3.00 1.00 0.36443* 0.05588 0.000 0.2182 0.5107

dimension3 2.00 0.01823 0.05328 0.938 -0.1212 0.1577 194 Games-Howell 1.00 2.00 -0.34620* 0.06130 0.001 -0.5178 -0.1746

dimension3 3.00 -0.36443* 0.04639 0.000 -0.4992 -0.2296 2.00 1.00 0.34620* 0.06130 0.001 0.1746 0.5178

dimension2 dimension3 3.00 -0.01823 0.05484 0.941 -0.1759 0.1394 3.00 1.00 0.36443* 0.04639 0.000 0.2296 0.4992

dimension3 2.00 0.01823 0.05484 0.941 -0.1394 0.1759 *. The mean difference is significant at the 0.05 level.

Amphidinium carterae – colloidal CHO/CHLa relationships

Descriptives Colloidal CHO/CHLa 95% Confidence Interval for Mean

N Mean Std. Deviation Std. Error Lower Bound Upper Bound Minimum Maximum 1.00 5 1.6238 0.12242 0.05475 1.4718 1.7758 1.47 1.78 2.00 6 1.7930 0.06895 0.02815 1.7206 1.8653 1.71 1.92 3.00 6 2.0163 0.03540 0.01445 1.9792 2.0535 1.97 2.06 Total 17 1.8221 0.17993 0.04364 1.7295 1.9146 1.47 2.06

195 Test of Homogeneity of Variances Colloidal CHO/CHLa Levene Statistic df1 df2 Sig. 3.200 2 14 0.072

ANOVA Colloidal CHO/CHLa Sum of Squares df Mean Square F Sig. Between Groups 0.428 2 0.214 33.295 0.000 Within Groups 0.090 14 0.006 Total 0.518 16

Post Hoc Tests

Multiple Comparisons Dependent Variable: Colloidal CHO/CHLa

(I) Amphidinium (J) Amphidinium Mean 95% Confidence Interval Difference (I-J) Std. Error Sig. Lower Bound Upper Bound Tukey HSD 1.00 2.00 -0.16911* 0.04855 0.010 -0.2962 -0.0421

dimension3 3.00 -0.39249* 0.04855 0.000 -0.5196 -0.2654 2.00 1.00 0.16911* 0.04855 0.010 0.0421 0.2962

dimension2 dimension3 3.00 -0.22338* 0.04629 0.001 -0.3445 -0.1022 3.00 1.00 0.39249* 0.04855 0.000 0.2654 0.5196

dimension3 2.00 0.22338* 0.04629 0.001 0.1022 0.3445 196 Games-Howell 1.00 2.00 -0.16911 0.06156 0.074 -0.3575 0.0193

dimension3 3.00 -0.39249* 0.05662 0.003 -0.5833 -0.2017 2.00 1.00 0.16911 0.06156 0.074 -0.0193 0.3575

dimension2 dimension3 3.00 -0.22338* 0.03164 0.000 -0.3152 -0.1316 3.00 1.00 0.39249* 0.05662 0.003 0.2017 0.5833

dimension3 2.00 0.22338* 0.03164 0.000 0.1316 0.3152 *. The mean difference is significant at the 0.05 level.

Amphidinium carterae – Storage carbohydrate/ CHLa relationships

Descriptives Storage/CHLa 95% Confidence Interval for Mean

N Mean Std. Deviation Std. Error Lower Bound Upper Bound Minimum Maximum 1.00 5 2.2627 0.04569 0.02043 2.2060 2.3195 2.20 2.31 2.00 6 2.3528 0.04679 0.01910 2.3037 2.4019 2.29 2.41 3.00 6 2.5879 0.08262 0.03373 2.5012 2.6746 2.50 2.69 Total 17 2.4093 0.15236 0.03695 2.3310 2.4876 2.20 2.69

197

Test of Homogeneity of Variances Storage/CHLa Levene Statistic df1 df2 Sig. 2.815 2 14 0.094

ANOVA Storage Sum of Squares df Mean Square F Sig. Between Groups 0.318 2 0.159 41.663 0.000 Within Groups 0.053 14 0.004 Total 0.371 16

Post Hoc Tests

Multiple Comparisons Dependent Variable:Storage

(I) Amphidinium (J) Amphidinium Mean 95% Confidence Interval Difference (I-J) Std. Error Sig. Lower Bound Upper Bound Tukey HSD 1.00 2.00 -0.09004 0.03741 0.073 -0.1879 0.0079

dimension3 3.00 -0.32518* 0.03741 0.000 -0.4231 -0.2273

2.00 1.00 0.09004 0.03741 0.073 -0.0079 0.1879

dimension2 dimension3 3.00 -0.23513* 0.03567 0.000 -0.3285 -0.1418 3.00 1.00 0.32518* 0.03741 0.000 0.2273 0.4231

dimension3 *

198 2.00 0.23513 0.03567 0.000 0.1418 0.3285 Games-Howell 1.00 2.00 -0.09004* 0.02797 0.027 -0.1686 -0.0115

dimension3 3.00 -0.32518* 0.03944 0.000 -0.4379 -0.2125 2.00 1.00 0.09004* 0.02797 0.027 0.0115 0.1686

dimension2 dimension3 3.00 -0.23513* 0.03876 0.001 -0.3462 -0.1241 3.00 1.00 0.32518* 0.03944 0.000 0.2125 0.4379

dimension3 2.00 0.23513* 0.03876 0.001 0.1241 0.3462 *. The mean difference is significant at the 0.05 level.

Cyclotella meneghiniana – Chlorophyll a: marker pigment relationship

Descriptives CHLa/FUCO 95% Confidence Interval for Mean

N Mean Std. Deviation Std. Error Lower Bound Upper Bound Minimum Maximum 1.00 5 1.1820 0.21730 0.09718 0.9122 1.4518 1.02 1.56 2.00 6 1.1200 0.04858 0.01983 1.0690 1.1710 1.07 1.20 3.00 6 1.1083 0.09766 0.03987 1.0058 1.2108 0.99 1.28 Total 17 1.1341 0.12870 0.03121 1.0679 1.2003 0.99 1.56

199 Test of Homogeneity of Variances CHLa/FUCO Levene Statistic df1 df2 Sig. 2.640 2 14 0.106

ANOVA CHLa/FUCO Sum of Squares df Mean Square F Sig. Between Groups 0.017 2 0.008 0.469 0.635 Within Groups 0.248 14 0.018 Total 0.265 16

Post Hoc Tests

Multiple Comparisons Dependent Variable:Chla /FUCO

(I) Cyclotella (J) Cyclotella Mean 95% Confidence Interval Difference (I-J) Std. Error Sig. Lower Bound Upper Bound Tukey HSD 1.00 2.00 0.06200 0.08065 0.728 -0.1491 0.2731

dimension3 3.00 0.07367 0.08065 0.641 -0.1374 0.2848 2.00 1.00 -0.06200 0.08065 0.728 -0.2731 0.1491

dimension2 dimension3 3.00 0.01167 0.07690 0.987 -0.1896 0.2129 3.00 1.00 -0.07367 0.08065 0.641 -0.2848 0.1374

dimension3 2.00 -0.01167 0.07690 0.987 -0.2129 0.1896 200 Games-Howell 1.00 2.00 0.06200 0.09918 0.815 -0.2792 0.4032

dimension3 3.00 0.07367 0.10504 0.773 -0.2605 0.4078 2.00 1.00 -0.06200 0.09918 0.815 -0.4032 0.2792

dimension2 dimension3 3.00 0.01167 0.04453 0.963 -0.1180 0.1414 3.00 1.00 -0.07367 0.10504 0.773 -0.4078 0.2605

dimension3 2.00 -0.01167 0.04453 0.963 -0.1414 0.1180

Cyclotella meneghiniana – Protein/Chlorophyll a relationships

Descriptives Protein/CHLa 95% Confidence Interval for Mean

N Mean Std. Deviation Std. Error Lower Bound Upper Bound Minimum Maximum 1.00 5 1.9958 0.03677 0.01644 1.9501 2.0414 1.96 2.04 2.00 7 2.0780 0.04607 0.01741 2.0354 2.1206 1.99 2.14 3.00 6 2.1670 0.03579 0.01461 2.1294 2.2046 2.12 2.21 Total 18 2.0848 0.07861 0.01853 2.0457 2.1239 1.96 2.21

201

Test of Homogeneity of Variances Protein/CHLa Levene Statistic df1 df2 Sig. 0.049 2 15 0.952

ANOVA Protein/CHLa Sum of Squares df Mean Square F Sig. Between Groups 0.081 2 0.040 24.596 0.000 Within Groups 0.025 15 0.002 Total 0.105 17

Post Hoc Tests Multiple Comparisons Dependent Variable: Protein/CHLa

(I) Cyclotella (J) Cyclotella Mean 95% Confidence Interval Difference (I-J) Std. Error Sig. Lower Bound Upper Bound

Tukey HSD 1.00 2.00 -0.08228* 0.02369 0.009 -0.1438 -0.0208

dimension3 3.00 -0.17124* 0.02450 0.000 -0.2349 -0.1076 2.00 1.00 0.08228* 0.02369 0.009 0.0208 0.1438

dimension2 dimension3 3.00 -0.08896* 0.02251 0.003 -0.1474 -0.0305 3.00 1.00 0.17124* 0.02450 0.000 0.1076 0.2349

dimension3 2.00 0.08896* 0.02251 0.003 0.0305 0.1474 *

202 Games-Howell 1.00 2.00 -0.08228 0.02395 0.016 -0.1482 -0.0164

dimension3 3.00 -0.17124* 0.02200 0.000 -0.2333 -0.1092 2.00 1.00 0.08228* 0.02395 0.016 0.0164 0.1482

dimension2 dimension3 3.00 -0.08896* 0.02273 0.006 -0.1504 -0.0275 3.00 1.00 0.17124* 0.02200 0.000 0.1092 0.2333

dimension3 2.00 0.08896* 0.02273 0.006 0.0275 0.1504 *. The mean difference is significant at the 0.05 level.

Cyclotella meneghiniana – Colloidal carbohydrate/ Chlorophyll a relationships

Descriptives Colloidal/CHLa 95% Confidence Interval for Mean

N Mean Std. Deviation Std. Error Lower Bound Upper Bound Minimum Maximum 1.00 5 0.9166 0.04747 0.02123 0.8577 0.9755 0.85 0.98 2.00 7 1.0773 0.07010 0.02649 1.0125 1.1422 1.02 1.21 3.00 6 1.3819 0.06586 0.02689 1.3128 1.4510 1.30 1.45 Total 18 1.1342 0.20114 0.04741 1.0342 1.2342 0.85 1.45

203 Test of Homogeneity of Variances Colloidal/CHLa Levene Statistic df1 df2 Sig. 0.520 2 15 0.605

ANOVA Colloidal CHO/CHLa Sum of Squares df Mean Square F Sig. Between Groups 0.628 2 0.314 78.199 0.000 Within Groups 0.060 15 0.004 Total 0.688 17

Post Hoc Tests

Dependent Variable: Colloidal CHO/CHLa

(I) Cyclotella (J) Cyclotella Mean 95% Confidence Interval Difference (I-J) Std. Error Sig. Lower Bound Upper Bound

Tukey HSD 1.00 2.00 -0.16074* 0.03709 0.002 -0.2571 -0.0644

dimension3 3.00 -0.46532* 0.03836 0.000 -0.5649 -0.3657 2.00 1.00 0.16074* 0.03709 0.002 0.0644 0.2571

dimension2 dimension3 3.00 -0.30457* 0.03524 0.000 -0.3961 -0.2130 3.00 1.00 0.46532* 0.03836 0.000 0.3657 0.5649

dimension3 2.00 0.30457* 0.03524 0.000 0.2130 0.3961 *

204 Games-Howell 1.00 2.00 -0.16074 0.03395 0.002 -0.2538 -0.0677

dimension3 3.00 -0.46532* 0.03426 0.000 -0.5612 -0.3694 2.00 1.00 0.16074* 0.03395 0.002 0.0677 0.2538

dimension2 dimension3 3.00 -0.30457* 0.03775 0.000 -0.4067 -0.2024 3.00 1.00 0.46532* 0.03426 0.000 0.3694 0.5612

dimension3 2.00 0.30457* 0.03775 0.000 0.2024 0.4067 *. The mean difference is significant at the 0.05 level.

Cyclotella meneghiniana – Storage carbohydrate/ Chlorophyll a relationships

Descriptives Storage/CHLa 95% Confidence Interval for Mean

N Mean Std. Deviation Std. Error Lower Bound Upper Bound Minimum Maximum 1.00 5 1.7463 0.13582 0.06074 1.5777 1.9150 1.58 1.90 2.00 7 2.1402 0.08153 0.03082 2.0648 2.2156 2.05 2.26 3.00 6 2.3777 0.05055 0.02064 2.3246 2.4307 2.30 2.44 Total 18 2.1100 0.26834 0.06325 1.9765 2.2434 1.58 2.44

205 Test of Homogeneity of Variances Storage CHO/CHLa Levene Statistic df1 df2 Sig. 3.009 2 15 0.080

ANOVA Storage CHO/CHLa Sum of Squares df Mean Square F Sig. Between Groups 1.098 2 0.549 65.101 0.000 Within Groups 0.126 15 0.008 Total 1.224 17

Post Hoc Tests Multiple Comparisons Dependent Variable: Storage CHO/CHLa

(I) Cyclotella (J) Cyclotella Mean 95% Confidence Interval Difference (I-J) Std. Error Sig. Lower Bound Upper Bound

Tukey HSD 1.00 2.00 -0.39392* 0.05376 0.000 -0.5336 -0.2543

dimension3 3.00 -0.63136* 0.05560 0.000 -0.7758 -0.4869 2.00 1.00 0.39392* 0.05376 0.000 0.2543 0.5336

dimension2 dimension3 3.00 -0.23744* 0.05108 0.001 -0.3701 -0.1048 3.00 1.00 0.63136* 0.05560 0.000 0.4869 0.7758

dimension3 2.00 0.23744* 0.05108 0.001 0.1048 0.3701 *

206 Games-Howell 1.00 2.00 -0.39392 0.06811 0.003 -0.6023 -0.1855

dimension3 3.00 -0.63136* 0.06415 0.000 -0.8413 -0.4215 2.00 1.00 0.39392* 0.06811 0.003 0.1855 0.6023

dimension2 dimension3 3.00 -0.23744* 0.03709 0.000 -0.3389 -0.1360 3.00 1.00 0.63136* 0.06415 0.000 0.4215 0.8413

dimension3 2.00 0.23744* 0.03709 0.000 0.1360 0.3389 *. The mean difference is significant at the 0.05 level.

Thalassiosira weissflogii – Chlorophyll a: marker pigment relationship

Descriptives CHLa/FUCO 95% Confidence Interval for Mean

N Mean Std. Deviation Std. Error Lower Bound Upper Bound Minimum Maximum 1.00 5 1.1420 0.03114 0.01393 1.1033 1.1807 1.11 1.19 2.00 6 1.1717 0.02137 0.00872 1.1492 1.1941 1.14 1.20 3.00 6 1.1517 0.02787 0.01138 1.1224 1.1809 1.12 1.19 Total 17 1.1559 0.02808 0.00681 1.1414 1.1703 1.11 1.20

207 Test of Homogeneity of Variances CHLa/FUCO Levene Statistic df1 df2 Sig. 0.396 2 14 0.680

ANOVA CHLa/FUCO Sum of Squares df Mean Square F Sig. Between Groups 0.003 2 0.001 1.787 0.204 Within Groups 0.010 14 0.001 Total 0.013 16

Post Hoc Tests Multiple Comparisons Dependent Variable: CHLa/FUCO

(I) Thalassiosira (J) Thalassiosira Mean 95% Confidence Interval Difference (I-J) Std. Error Sig. Lower Bound Upper Bound

Tukey HSD 1.00 2.00 -0.02967 0.01622 0.196 -0.0721 0.0128

dimension3 3.00 -0.00967 0.01622 0.825 -0.0521 0.0328 2.00 1.00 0.02967 0.01622 0.196 -0.0128 0.0721

dimension2 dimension3 3.00 0.02000 0.01547 0.422 -0.0205 0.0605 3.00 1.00 0.00967 0.01622 0.825 -0.0328 0.0521

dimension3 2.00 -0.02000 0.01547 0.422 -0.0605 0.0205

208 Games-Howell 1.00 2.00 -0.02967 0.01644 0.237 -0.0782 0.0189

dimension3 3.00 -0.00967 0.01798 0.855 -0.0608 0.0415 2.00 1.00 0.02967 0.01644 0.237 -0.0189 0.0782

dimension2 dimension3 3.00 0.02000 0.01434 0.382 -0.0197 0.0597 3.00 1.00 0.00967 0.01798 0.855 -0.0415 0.0608

dimension3 2.00 -0.02000 0.01434 0.382 -0.0597 0.0197

Thalassiosira weissflogii – protein/ chlorophyll a relationships

Descriptives Protein/CHLa 95% Confidence Interval for Mean

N Mean Std. Deviation Std. Error Lower Bound Upper Bound Minimum Maximum 1.00 5 2.2081 0.05003 0.02237 2.1460 2.2702 2.14 2.27 2.00 5 2.2540 0.02572 0.01150 2.2221 2.2860 2.22 2.29 3.00 6 2.3265 0.03702 0.01511 2.2876 2.3653 2.28 2.38 Total 16 2.2668 0.06266 0.01567 2.2334 2.3002 2.14 2.38

209 Test of Homogeneity of Variances Protein/CHLa Levene Statistic df1 df2 Sig. 1.231 2 13 0.324

ANOVA Protein/CHLa Sum of Squares df Mean Square F Sig. Between Groups 0.039 2 0.020 13.124 0.001 Within Groups 0.020 13 0.002 Total 0.059 15

Post Hoc Tests Multiple Comparisons Dependent Variable: Protein/ CHLa

(I) Thalassiosira (J) Thalassiosira Mean 95% Confidence Interval Difference (I-J) Std. Error Sig. Lower Bound Upper Bound

Tukey HSD 1.00 2.00 -0.04594 0.02450 0.185 -0.1106 0.0188

dimension3 3.00 -.011835* 0.02346 0.001 -0.1803 -0.0564 2.00 1.00 0.04594 0.02450 0.185 -0.0188 0.1106

dimension2 dimension3 3.00 -0.07241* 0.02346 0.022 -0.1343 -0.0105 3.00 1.00 0.11835* 0.02346 0.001 0.0564 0.1803

dimension3 2.00 0.07241* 0.02346 0.022 0.0105 0.1343

210 Games-Howell 1.00 2.00 -0.04594 0.02516 0.240 -0.1232 0.0313

dimension3 3.00 -0.11835* 0.02700 0.007 -0.1971 -0.0396 2.00 1.00 0.04594 0.02516 0.240 -0.0313 0.1232

dimension2 dimension3 3.00 -0.07241* 0.01899 0.011 -0.1257 -0.0191 3.00 1.00 0.11835* 0.02700 0.007 0.0396 0.1971

dimension3 2.00 0.07241* 0.01899 0.011 0.0191 0.1257 *. The mean difference is significant at the 0.05 level.

Thalassiosira weissflogii – colloidal carbohydrate/chlorophyll a relationships

Descriptives Colloidal CHO/CHLa 95% Confidence Interval for Mean

N Mean Std. Deviation Std. Error Lower Bound Upper Bound Minimum Maximum 1.00 5 0.8918 0.08721 0.03900 0.7835 1.0001 0.80 1.02 2.00 5 1.0720 0.01649 0.00738 1.0515 1.0925 1.05 1.09 3.00 6 1.2020 0.03848 0.01571 1.1617 1.2424 1.15 1.25 Total 16 1.0645 0.14185 0.03546 0.9889 1.1400 0.80 1.25

Test of Homogeneity of Variances 211 Colloidal CHO/CHLa Levene Statistic df1 df2 Sig. 5.685 2 13 0.017

ANOVA Colloidal CHO/CHLa Sum of Squares df Mean Square F Sig. Between Groups 0.263 2 0.131 43.920 0.000 Within Groups 0.039 13 0.003 Total .302 15

Post Hoc Tests Multiple Comparisons Dependent Variable: Colloidal CHO/CHLa

(I) Thalassiosira (J) Thalassiosira Mean 95% Confidence Interval Difference (I-J) Std. Error Sig. Lower Bound Upper Bound

Tukey HSD 1.00 2.00 -0.18020* 0.03460 0.000 -0.2716 -0.0888

dimension3 3.00 -0.31025* 0.03313 0.000 -0.3977 -0.2228 2.00 1.00 0.18020* 0.03460 0.000 0.0888 0.2716

dimension2 dimension3 3.00 -0.13005* 0.03313 0.005 -0.2175 -0.0426 3.00 1.00 0.31025* 0.03313 0.000 0.2228 0.3977

dimension3 2.00 0.13005* 0.03313 0.005 0.0426 0.2175 *

212 Games-Howell 1.00 2.00 -0.18020 0.03969 0.020 -0.3174 -0.0430

dimension3 3.00 -0.31025* 0.04205 0.001 -0.4444 -0.1761 2.00 1.00 0.18020* 0.03969 0.020 0.0430 0.3174

dimension2 dimension3 3.00 -0.13005* 0.01735 0.000 -0.1811 -0.0790 3.00 1.00 0.31025* 0.04205 0.001 0.1761 0.4444

dimension3 2.00 0.13005* 0.01735 0.000 0.0790 0.1811 *. The mean difference is significant at the 0.05 level.

Thalassiosira weissflogii – Storage Carbohydrate/Chlorophyll a relationships

Descriptives Storage CHO/CHLa 95% Confidence Interval for Mean

N Mean Std. Deviation Std. Error Lower Bound Upper Bound Minimum Maximum 1.00 5 2.0324 0.05363 0.02398 1.9658 2.0990 1.95 2.08 2.00 5 2.1568 0.06144 0.02748 2.0805 2.2331 2.10 2.23 3.00 6 2.2875 0.06205 0.02533 2.2224 2.3527 2.19 2.37 Total 16 2.1670 0.12224 0.03056 2.1018 2.2321 1.95 2.37

Test of Homogeneity of Variances 213 Storage CHO/CHLa Levene Statistic df1 df2 Sig. 0.358 2 13 0.706

ANOVA Storage CHO/CHLa Sum of Squares df Mean Square F Sig. Between Groups 0.178 2 0.089 25.271 0.000 Within Groups 0.046 13 0.004 Total 0.224 15

Post Hoc Tests Multiple Comparisons Dependent Variable: Storage CHO/CHLa

(I) Thalassiosira (J) Thalassiosira Mean 95% Confidence Interval Difference (I-J) Std. Error Sig. Lower Bound Upper Bound

Tukey HSD 1.00 2.00 -0.12442* 0.03756 0.014 -0.2236 -0.0252

dimension3 3.00 -0.25513* 0.03596 0.000 -0.3501 -0.1602 2.00 1.00 0.12442* 0.03756 0.014 0.0252 0.2236

dimension2 dimension3 3.00 -0.13071* 0.03596 0.008 -0.2257 -0.0358 3.00 1.00 0.25513* 0.03596 0.000 0.1602 0.3501

dimension3 2.00 0.13071* 0.03596 0.008 0.0358 0.2257 *

214 Games-Howell 1.00 2.00 -0.12442 0.03647 0.023 -0.2290 -0.0198

dimension3 3.00 -0.25513* 0.03489 0.000 -0.3526 -0.1577 2.00 1.00 0.12442* 0.03647 0.023 0.0198 0.2290

dimension2 dimension3 3.00 -0.13071* 0.03737 0.018 -0.2358 -0.0257 3.00 1.00 0.25513* 0.03489 0.000 0.1577 0.3526

dimension3 2.00 0.13071* 0.03737 0.018 0.0257 0.2358 *. The mean difference is significant at the 0.05 level.

Dunaliella tertiolecta – Chlorophyll a: marker pigment relationships

Descriptives CHLa/CHLb 95% Confidence Interval for Mean

N Mean Std. Deviation Std. Error Lower Bound Upper Bound Minimum Maximum 1.00 4 2.4125 0.13401 0.06700 2.1993 2.6257 2.26 2.58 2.00 5 2.5380 0.25636 0.11465 2.2197 2.8563 2.24 2.92 3.00 6 2.1467 0.23019 0.09397 1.9051 2.3882 1.77 2.41 Total 15 2.3480 0.27019 0.06976 2.1984 2.4976 1.77 2.92

215 Test of Homogeneity of Variances CHLa/CHLb Levene Statistic df1 df2 Sig. 0.622 2 12 0.553

ANOVA CHLa/CHLb Sum of Squares df Mean Square F Sig. Between Groups 0.440 2 0.220 4.542 0.034 Within Groups 0.582 12 0.048 Total 1.022 14

Post Hoc Tests

Multiple Comparisons Dependent Variable: CHLa/CHLb

(I) Dunaliella (J) Dunaliella Mean 95% Confidence Interval Difference (I-J) Std. Error Sig. Lower Bound Upper Bound Tukey HSD 1.00 2.00 -0.12550 0.14769 0.681 -0.5195 0.2685

dimension3 3.00 0.26583 0.14212 0.189 -0.1133 0.6450 2.00 1.00 0.12550 0.14769 0.681 -0.2685 0.5195

dimension2 dimension3 3.00 0.39133* 0.13332 0.031 0.0357 0.7470 3.00 1.00 -0.26583 0.14212 0.189 -0.6450 0.1133

dimension3 2.00 -0.39133* 0.13332 0.031 -0.7470 -0.0357 216 Games-Howell 1.00 2.00 -0.12550 0.13279 0.634 -0.5286 0.2776

dimension3 3.00 0.26583 0.11542 0.113 -0.0644 0.5961 2.00 1.00 0.12550 0.13279 0.634 -0.2776 0.5286

dimension2 dimension3 3.00 0.39133 0.14824 0.068 -0.0299 0.8126 3.00 1.00 -0.26583 0.11542 0.113 -0.5961 0.0644

dimension3 2.00 -0.39133 0.14824 0.068 -0.8126 0.0299 *. The mean difference is significant at the 0.05 level.

Dunaliella tertiolecta – Protein/ Chlorophyll a relationships

Descriptives Protein/CHLa 95% Confidence Interval for Mean

N Mean Std. Deviation Std. Error Lower Bound Upper Bound Minimum Maximum 1.00 4 1.8492 0.01749 0.00875 1.8213 1.8770 1.83 1.86 2.00 5 2.0387 0.11846 0.05298 1.8916 2.1858 1.87 2.12 3.00 6 2.2383 0.04013 0.01638 2.1962 2.2804 2.18 2.27 Total 15 2.0680 0.17625 0.04551 1.9704 2.1656 1.83 2.27

217

Test of Homogeneity of Variances Protein/CHLa Levene Statistic df1 df2 Sig. 14.618 2 12 0.001

ANOVA Protein/CHLa Sum of Squares df Mean Square F Sig. Between Groups 0.370 2 0.185 34.083 0.000 Within Groups 0.065 12 0.005 Total 0.435 14

Robust Tests of Equality of Means Protein/CHLa Statistica df1 df2 Sig. Brown-Forsythe 35.066 2 5.036 0.001 a. Asymptotically F distributed.

Post Hoc Tests Multiple Comparisons Dependent Variable: Protein/CHLa

(I) Dunaliella (J) Dunaliella Mean 95% Confidence Interval

218 Difference (I-J) Std. Error Sig. Lower Bound Upper Bound Tukey HSD 1.00 2.00 -0.18953* 0.04941 0.006 -0.3213 -0.0577

dimension3 3.00 -0.38911* 0.04754 0.000 -0.5160 -0.2623 2.00 1.00 0.18953* 0.04941 0.006 0.0577 0.3213

dimension2 dimension3 3.00 -0.19958* 0.04460 0.002 -0.3186 -0.0806 3.00 1.00 0.38911* 0.04754 0.000 0.2623 0.5160

dimension3 2.00 0.19958* 0.04460 0.002 0.0806 0.3186 Games-Howell 1.00 2.00 -0.18953* 0.05370 0.048 -0.3764 -0.0026

dimension3 3.00 -0.38911* 0.01857 0.000 -0.4433 -0.3349 2.00 1.00 0.18953* 0.05370 0.048 0.0026 0.3764

dimension2 dimension3 3.00 -0.19958* 0.05545 0.037 -0.3832 -0.0159 3.00 1.00 0.38911* 0.01857 0.000 0.3349 0.4433

dimension3 2.00 0.19958* 0.05545 0.037 0.0159 0.3832

Multiple Comparisons Dependent Variable: Protein/CHLa

(I) Dunaliella (J) Dunaliella Mean 95% Confidence Interval Difference (I-J) Std. Error Sig. Lower Bound Upper Bound Tukey HSD 1.00 2.00 -0.18953* 0.04941 0.006 -0.3213 -0.0577

dimension3 3.00 -0.38911* 0.04754 0.000 -0.5160 -0.2623 2.00 1.00 0.18953* 0.04941 0.006 0.0577 0.3213

dimension2 dimension3 3.00 -0.19958* 0.04460 0.002 -0.3186 -0.0806 3.00 1.00 0.38911* 0.04754 0.000 0.2623 0.5160

dimension3 2.00 0.19958* 0.04460 0.002 0.0806 0.3186 Games-Howell 1.00 2.00 -0.18953* 0.05370 0.048 -0.3764 -0.0026

dimension3 * 219 3.00 -0.38911 0.01857 0.000 -0.4433 -0.3349 2.00 1.00 0.18953* 0.05370 0.048 0.0026 0.3764

dimension2 dimension3 3.00 -0.19958* 0.05545 0.037 -0.3832 -0.0159 3.00 1.00 0.38911* 0.01857 0.000 0.3349 0.4433

dimension3 2.00 0.19958* 0.05545 0.037 0.0159 0.3832 *. The mean difference is significant at the 0.05 level.

Dunaliella tertiolecta – Colloidal carbohydrate/ Chlorophyll a relationships

Descriptives Colloidal CHO/CHLa 95% Confidence Interval for Mean

N Mean Std. Deviation Std. Error Lower Bound Upper Bound Minimum Maximum 1.00 4 0.9315 0.03977 0.01988 0.8682 0.9948 0.88 0.97 2.00 5 0.6514 0.11837 0.05293 0.5044 0.7984 0.51 0.77 3.00 6 0.7442 0.10814 0.04415 0.6307 0.8577 0.60 0.85 Total 15 0.7632 0.14569 0.03762 0.6825 0.8439 0.51 0.97

220

Test of Homogeneity of Variances Colloidal CHO/CHLa Levene Statistic df1 df2 Sig. 3.748 2 12 0.054

ANOVA Colloidal CHO/CHLa Sum of Squares df Mean Square F Sig. Between Groups 0.178 2 0.089 8.952 0.004 Within Groups 0.119 12 0.010 Total 0.297 14

Post Hoc Tests

Multiple Comparisons Dependent Variable: Colloidal CHO/CHLa

(I) Dunaliella (J) Dunaliella Mean 95% Confidence Interval Difference (I-J) Std. Error Sig. Lower Bound Upper Bound

Tukey HSD 1.00 2.00 0.28007* 0.06687 0.003 0.1017 0.4585

dimension3 3.00 0.18726* 0.06435 0.033 0.0156 0.3589

2.00 1.00 -0.28007* 0.06687 0.003 -0.4585 -0.1017

dimension2 dimension3 3.00 -0.09282 0.06036 0.309 -0.2539 0.0682 3.00 1.00 -0.18726* 0.06435 0.033 -0.3589 -0.0156

dimension3

221 2.00 0.09282 0.06036 0.309 -0.0682 0.2539 Games-Howell 1.00 2.00 0.28007* 0.05655 0.009 0.0970 0.4631

dimension3 3.00 0.18726* 0.04842 0.016 0.0435 0.3310 2.00 1.00 -0.28007* 0.05655 0.009 -0.4631 -0.0970

dimension2 dimension3 3.00 -0.09282 0.06893 0.410 -0.2883 0.1027 3.00 1.00 -0.18726* 0.04842 0.016 -0.3310 -0.0435

dimension3 2.00 0.09282 0.06893 0.410 -0.1027 0.2883 *. The mean difference is significant at the 0.05 level.

Dunaliella tertiolecta – storage carbohydrate/ Chlorophyll a relationships

Descriptives Storage CHO/CHLa 95% Confidence Interval for Mean

N Mean Std. Deviation Std. Error Lower Bound Upper Bound Minimum Maximum 1.00 4 1.7770 0.02382 0.01191 1.7390 1.8149 1.76 1.81 2.00 5 1.9321 0.05264 0.02354 1.8667 1.9974 1.85 1.99 3.00 6 2.0357 0.03575 0.01460 1.9982 2.0732 1.97 2.06 Total 15 1.9322 0.11335 0.02927 1.8694 1.9949 1.76 2.06

222 Test of Homogeneity of Variances Storage CHO/CHLa Levene Statistic df1 df2 Sig. 1.080 2 12 0.370

ANOVA Storage CHO/CHLa Sum of Squares df Mean Square F Sig. Between Groups 0.161 2 0.080 50.279 0.000 Within Groups 0.019 12 0.002 Total 0.180 14

Post Hoc Tests Multiple Comparisons Dependent Variable: Storage CHO/CHLa

(I) Dunaliella (J) Dunaliella Mean 95% Confidence Interval Difference (I-J) Std. Error Sig. Lower Bound Upper Bound

Tukey HSD 1.00 2.00 -0.15513* 0.02682 0.000 -0.2267 -0.0836

dimension3 3.00 -0.25877* 0.02580 0.000 -0.3276 -0.1899 2.00 1.00 0.15513* 0.02682 0.000 0.0836 0.2267

dimension2 dimension3 3.00 -0.10364* 0.02421 0.003 -0.1682 -0.0391 3.00 1.00 0.25877* 0.02580 0.000 0.1899 0.3276

dimension3 2.00 0.10364* 0.02421 0.003 0.0391 0.1682 *

223 Games-Howell 1.00 2.00 -0.15513 0.02638 0.003 -0.2369 -0.0734

dimension3 3.00 -0.25877* 0.01884 0.000 -0.3126 -0.2049 2.00 1.00 0.15513* 0.02638 0.003 0.0734 0.2369

dimension2 dimension3 3.00 -0.10364* 0.02770 0.018 -0.1856 -0.0216 3.00 1.00 0.25877* 0.01884 0.000 0.2049 0.3126

dimension3 2.00 0.10364* 0.02770 0.018 0.0216 0.1856 *. The mean difference is significant at the 0.05 level.

Scenedesmus quadricauda – Chlorophyll a: marker pigment relationships

Descriptives Pigment/CHLa 95% Confidence Interval for Mean

N Mean Std. Deviation Std. Error Lower Bound Upper Bound Minimum Maximum 1.00 4 2.7425 0.15924 .07962 2.4891 2.9959 2.52 2.88 2.00 7 2.6929 0.29004 .10963 2.4246 2.9611 2.07 2.92 3.00 6 2.5700 0.46143 .18838 2.0858 3.0542 1.98 3.18 Total 17 2.6612 0.32871 .07972 2.4922 2.8302 1.98 3.18

Test of Homogeneity of Variances 224 Pigment/CHLa Levene Statistic df1 df2 Sig. 3.647 2 14 0.053

ANOVA Pigment/CHLa Sum of Squares df Mean Square F Sig. Between Groups 0.083 2 0.042 0.355 0.708 Within Groups 1.645 14 0.118 Total 1.729 16

Post Hoc Tests

Multiple Comparisons Dependent Variable: Pigment/CHLa

(I) Scenedesmus (J) Scenedesmus Mean 95% Confidence Interval Difference (I-J) Std. Error Sig. Lower Bound Upper Bound Tukey HSD 1.00 2.00 0.04964 0.21488 0.971 -0.5128 0.6120

dimension3 3.00 0.17250 0.22129 0.721 -0.4067 0.7517

2.00 1.00 -0.04964 0.21488 0.971 -0.6120 0.5128

dimension2 dimension3 3.00 0.12286 0.19073 0.799 -0.3763 0.6221 3.00 1.00 -0.17250 0.22129 0.721 -0.7517 0.4067

dimension3

225 2.00 -0.12286 0.19073 0.799 -0.6221 0.3763 Games-Howell 1.00 2.00 0.04964 0.13549 0.929 -0.3287 0.4280

dimension3 3.00 0.17250 0.20451 0.691 -0.4389 0.7839 2.00 1.00 -0.04964 0.13549 0.929 -0.4280 0.3287

dimension2 dimension3 3.00 0.12286 0.21795 0.842 -0.4971 0.7428 3.00 1.00 -0.17250 0.20451 0.691 -0.7839 0.4389

dimension3 2.00 -0.12286 0.21795 0.842 -0.7428 0.4971

Scenedesmus quadricauda – Protein/ Chlorophyll a relationships

Descriptives Protein/CHLa 95% Confidence Interval for Mean

N Mean Std. Deviation Std. Error Lower Bound Upper Bound Minimum Maximum 1.00 5 2.7741 0.13814 0.06178 2.6026 2.9456 2.63 2.93 2.00 7 2.2583 0.05923 0.02239 2.2035 2.3131 2.22 2.37 3.00 6 2.1043 0.02094 0.00855 2.0823 2.1263 2.07 2.14 Total 18 2.3503 0.28900 0.06812 2.2065 2.4940 2.07 2.93

226 Test of Homogeneity of Variances Protein/CHLa Levene Statistic df1 df2 Sig. 15.945 2 15 0.000

ANOVA Protein/CHLa Sum of Squares df Mean Square F Sig. Between Groups 1.320 2 0.660 99.450 0.000 Within Groups 0.100 15 0.007 Total 1.420 17

Robust Tests of Equality of Means Protein/CHLa Statistica df1 df2 Sig. Brown-Forsythe 81.411 2 5.449 0.000 a. Asymptotically F distributed.

Multiple Comparisons Dependent Variable: Protein/CHLa

(I) Scenedesmus (J) Scenedesmus Mean 95% Confidence Interval Difference (I-J) Std. Error Sig. Lower Bound Upper Bound Tukey HSD 1.00 2.00 0.51579* 0.04771 0.000 0.3919 0.6397 227

dimension3 3.00 0.66978* 0.04934 0.000 0.5416 0.7979 2.00 1.00 -0.51579* 0.04771 0.000 -0.6397 -0.3919

dimension2 dimension3 3.00 0.15400* 0.04533 0.010 0.0363 0.2717 3.00 1.00 -0.66978* 0.04934 0.000 -0.7979 -0.5416

dimension3 2.00 -0.15400* 0.04533 0.010 -0.2717 -0.0363 Games-Howell 1.00 2.00 0.51579* 0.06571 0.001 0.3029 0.7287

dimension3 3.00 0.66978* 0.06237 0.001 0.4513 0.8883 2.00 1.00 -0.51579* 0.06571 0.001 -0.7287 -0.3029

dimension2 dimension3 3.00 0.15400* 0.02396 0.001 0.0849 0.2231 3.00 1.00 -0.66978* 0.06237 0.001 -0.8883 -0.4513

dimension3 2.00 -0.15400* 0.02396 0.001 -0.2231 -0.0849 *. The mean difference is significant at the 0.05 level.

Scenedesmus quadricauda – Colloidal carbohydrate/ Chlorophyll a relationships

Descriptives Colloidal/CHLa 95% Confidence Interval for Mean

N Mean Std. Deviation Std. Error Lower Bound Upper Bound Minimum Maximum 1.00 5 1.9958 0.20010 0.08949 1.7474 2.2443 1.71 2.26 2.00 7 0.8626 0.10308 0.03896 0.7673 0.9580 0.76 1.00 3.00 6 0.8675 0.02494 0.01018 0.8413 0.8937 0.83 0.90 Total 18 1.1790 0.53391 0.12584 0.9135 1.4445 0.76 2.26

228 Test of Homogeneity of Variances Colloidal/CHLa Levene Statistic df1 df2 Sig. 4.723 2 15 0.026

ANOVA Colloidal/CHLa Sum of Squares df Mean Square F Sig. Between Groups 4.619 2 2.309 152.592 0.000 Within Groups 0.227 15 0.015 Total 4.846 17

Robust Tests of Equality of Means Colloidal/CHLa Statistica df1 df2 Sig.

Brown-Forsythe 128.928 2 5.939 0.000 a. Asymptotically F distributed.

Multiple Comparisons Dependent Variable: Colloidal CHO/CHLa

(I) Scenedesmus (J) Scenedesmus Mean 95% Confidence Interval Difference (I-J) Std. Error Sig. Lower Bound Upper Bound 229 Tukey HSD 1.00 2.00 1.13321* 0.07204 0.000 0.9461 1.3203

dimension3 3.00 1.12834* 0.07449 0.000 0.9348 1.3218 2.00 1.00 -1.13321* 0.07204 0.000 -1.3203 -0.9461

dimension2 dimension3 3.00 -0.00487 0.06844 0.997 -0.1827 0.1729 3.00 1.00 -1.12834* 0.07449 0.000 -1.3218 -0.9348

dimension3 2.00 0.00487 0.06844 0.997 -0.1729 0.1827 Games-Howell 1.00 2.00 1.13321* 0.09760 0.000 0.8262 1.4402

dimension3 3.00 1.12834* 0.09006 0.000 0.8111 1.4456 2.00 1.00 -1.13321* 0.09760 0.000 -1.4402 -0.8262

dimension2 dimension3 3.00 -0.00487 0.04027 0.992 -0.1243 0.1145 3.00 1.00 -1.12834* 0.09006 0.000 -1.4456 -0.8111

dimension3 2.00 0.00487 0.04027 0.992 -0.1145 0.1243

Multiple Comparisons Dependent Variable: Colloidal CHO/CHLa

(I) Scenedesmus (J) Scenedesmus Mean 95% Confidence Interval Difference (I-J) Std. Error Sig. Lower Bound Upper Bound Tukey HSD 1.00 2.00 1.13321* 0.07204 0.000 0.9461 1.3203

dimension3 3.00 1.12834* 0.07449 0.000 0.9348 1.3218 2.00 1.00 -1.13321* 0.07204 0.000 -1.3203 -0.9461

dimension2 dimension3 3.00 -0.00487 0.06844 0.997 -0.1827 0.1729 3.00 1.00 -1.12834* 0.07449 0.000 -1.3218 -0.9348

dimension3 2.00 0.00487 0.06844 0.997 -0.1729 0.1827 Games-Howell 1.00 2.00 1.13321* 0.09760 0.000 0.8262 1.4402

dimension3 * 230 3.00 1.12834 0.09006 0.000 0.8111 1.4456 2.00 1.00 -1.13321* 0.09760 0.000 -1.4402 -0.8262

dimension2 dimension3 3.00 -0.00487 0.04027 0.992 -0.1243 0.1145 3.00 1.00 -1.12834* 0.09006 0.000 -1.4456 -0.8111

dimension3 2.00 0.00487 0.04027 0.992 -0.1145 0.1243 *. The mean difference is significant at the 0.05 level.

Scenedesmus quadricauda – Storage carbohydrate/ Chlorophyll a relationships

Descriptives Storage CHO/CHLa 95% Confidence Interval for Mean

N Mean Std. Deviation Std. Error Lower Bound Upper Bound Minimum Maximum 1.00 5 3.0290 0.22731 0.10165 2.7468 3.3112 2.78 3.27 2.00 7 1.7367 0.30617 0.11572 1.4536 2.0199 1.28 2.22 3.00 6 1.2671 0.02778 0.01134 1.2379 1.2962 1.22 1.30 Total 18 1.9391 0.75572 0.17812 1.5633 2.3150 1.22 3.27

231 Test of Homogeneity of Variances Storage CHO/CHL a Levene Statistic df1 df2 Sig. 5.098 2 15 0.020

ANOVA Storage CHO/CHLa Sum of Squares df Mean Square F Sig. Between Groups 8.936 2 4.468 86.701 0.000 Within Groups 0.773 15 0.052 Total 9.709 17

Robust Tests of Equality of Means Storage CHO/CHLa Statistica df1 df2 Sig. Brown-Forsythe 93.945 2 10.107 0.000 a. Asymptotically F distributed.

Multiple Comparisons Dependent Variable: Storage CHO/CHLa

(I) Scenedesmus (J) Scenedesmus Mean 95% Confidence Interval Difference (I-J) Std. Error Sig. Lower Bound Upper Bound Tukey HSD 1.00 2.00 1.29226* 0.13292 0.000 0.9470 1.6375

dimension3 * 232 3.00 1.76193 0.13746 0.000 1.4049 2.1190 2.00 1.00 -1.29226* 0.13292 0.000 -1.6375 -0.9470

dimension2 dimension3 3.00 0.46968* 0.12630 0.005 0.1416 0.7977 3.00 1.00 -1.76193* 0.13746 0.000 -2.1190 -1.4049

dimension3 2.00 -0.46968* 0.12630 0.005 -0.7977 -0.1416 Games-Howell 1.00 2.00 1.29226* 0.15403 0.000 0.8696 1.7149

dimension3 3.00 1.76193* 0.10229 0.000 1.4015 2.1224 2.00 1.00 -1.29226* 0.15403 0.000 -1.7149 -0.8696

dimension2 dimension3 3.00 0.46968* 0.11628 0.015 0.1148 0.8245 3.00 1.00 -1.76193* 0.10229 0.000 -2.1224 -1.4015

dimension3 2.00 -.46968* 0.11628 0.015 -0.8245 -0.1148 *. The mean difference is significant at the 0.05 level.

Synechococcus elongatus – Chlorophyll a: marker pigment relationships

Descriptives CHLa/Zea 95% Confidence Interval for Mean

N Mean Std. Deviation Std. Error Lower Bound Upper Bound Minimum Maximum 1.00 4 6.4850 0.29939 0.14969 6.0086 6.9614 6.07 6.76 2.00 7 4.7343 0.53344 0.20162 4.2409 5.2276 3.85 5.52 3.00 5 3.0580 0.20669 0.09243 2.8014 3.3146 2.86 3.40 4.00 6 0.8683 0.13045 0.05326 0.7314 1.0052 0.64 1.01 Total 22 3.6173 2.07898 0.44324 2.6955 4.5390 0.64 6.76

233

Test of Homogeneity of Variances CHLa/Zea Levene Statistic df1 df2 Sig. 2.194 3 18 0.124

ANOVA CHLa /Zea Sum of Squares df Mean Square F Sig. Between Groups 88.533 3 29.511 237.968 0.000 Within Groups 2.232 18 0.124 Total 90.766 21

Multiple Comparisons Dependent Variable: CHLa/Zea

(I) Synechococcus (J) Synechococcus Mean 95% Confidence Interval Difference (I-J) Std. Error Sig. Lower Bound Upper Bound

Tukey HSD 1.00 2.00 1.75071* 0.22072 0.000 1.1269 2.3745 3.00 3.42700* 0.23623 0.000 2.7593 4.0947 dimension3 4.00 5.61667* 0.22731 0.000 4.9742 6.2591 2.00 1.00 -1.75071* 0.22072 0.000 -2.3745 -1.1269 3.00 1.67629* 0.20620 0.000 1.0935 2.2591 dimension3 4.00 3.86595* 0.19592 0.000 3.3122 4.4197

dimension2 * 234 3.00 1.00 -3.42700 0.23623 0.000 -4.0947 -2.7593 2.00 -1.67629* 0.20620 0.000 -2.2591 -1.0935 dimension3 4.00 2.18967* 0.21324 0.000 1.5870 2.7923 4.00 1.00 -5.61667* 0.22731 0.000 -6.2591 -4.9742 2.00 -3.86595* 0.19592 0.000 -4.4197 -3.3122 dimension3 3.00 -2.18967* 0.21324 0.000 -2.7923 -1.5870 Games-Howell 1.00 2.00 1.75071* 0.25112 0.000 .9664 2.5350 3.00 3.42700* 0.17593 0.000 2.7856 4.0684 dimension3 4.00 5.61667* 0.15889 0.000 4.9499 6.2834 2.00 1.00 -1.75071* 0.25112 0.000 -2.5350 -.9664 dimension2 3.00 1.67629* 0.22180 0.000 .9708 2.3818 dimension3 4.00 3.86595* 0.20854 0.000 3.1710 4.5609 3.00 1.00 -3.42700* 0.17593 0.000 -4.0684 -2.7856 dimension3

2.00 -1.67629* 0.22180 0.000 -2.3818 -.9708 4.00 2.18967* 0.10668 0.000 1.8295 2.5498 4.00 1.00 -5.61667* .15889 0.000 -6.2834 -4.9499 2.00 -3.86595* .20854 0.000 -4.5609 -3.1710 dimension3 3.00 -2.18967* .10668 0.000 -2.5498 -1.8295 *. The mean difference is significant at the 0.05 level.

Descriptives CHLa/Echinenone 95% Confidence Interval for Mean 235 N Mean Std. Deviation Std. Error Lower Bound Upper Bound Minimum Maximum 1.00 4 58.0550 8.10271 4.05136 45.1618 70.9482 50.09 68.03 2.00 7 63.1243 4.50956 1.70445 58.9536 67.2949 57.12 68.89 3.00 5 34.1800 3.58971 1.60537 29.7228 38.6372 28.57 37.31 4.00 6 13.6233 2.09319 .85454 11.4267 15.8200 10.20 15.61 Total 22 42.1241 21.47420 4.57831 32.6030 51.6452 10.20 68.89

Test of Homogeneity of Variances CHLa/Echinenone Levene Statistic df1 df2 Sig. 3.858 3 18 0.027

ANOVA CHLa/Echinenone Sum of Squares df Mean Square F Sig.

Between Groups 9291.535 3 3097.178 142.062 0.000 Within Groups 392.430 18 21.802 Total 9683.965 21

Robust Tests of Equality of Means CHLa/Echinenone Statistica df1 df2 Sig. Brown-Forsythe 115.099 3 6.385 0.000 236 a. Asymptotically F distributed.

Multiple Comparisons Dependent Variable: CHLa/Echinenone

(I) synechococcus (J) synechococcus Mean 95% Confidence Interval Difference (I-J) Std. Error Sig. Lower Bound Upper Bound Tukey HSD 1.00 2.00 -5.06929 2.92659 0.337 -13.3407 3.2021 3.00 23.87500* 3.13221 0.000 15.0225 32.7275 dimension3 4.00 44.43167* 3.01397 0.000 35.9133 52.9500

dimension2 2.00 1.00 5.06929 2.92659 0.337 -3.2021 13.3407 3.00 28.94429* 2.73402 0.000 21.2172 36.6714 dimension3 4.00 49.50095* 2.59772 0.000 42.1590 56.8429

3.00 1.00 -23.87500* 3.13221 0.000 -32.7275 -15.0225 2.00 -28.94429* 2.73402 0.000 -36.6714 -21.2172 dimension3 4.00 20.55667* 2.82736 0.000 12.5657 28.5476 4.00 1.00 -44.43167* 3.01397 0.000 -52.9500 -35.9133 2.00 -49.50095* 2.59772 0.000 -56.8429 -42.1590 dimension3 3.00 -20.55667* 2.82736 0.000 -28.5476 -12.5657 Games-Howell 1.00 2.00 -5.06929 4.39530 0.681 -22.7630 12.6245 3.00 23.87500* 4.35783 0.019 6.0068 41.7432 dimension3 4.00 44.43167* 4.14050 0.004 25.5549 63.3084 2.00 1.00 5.06929 4.39530 0.681 -12.6245 22.7630 3.00 28.94429* 2.34145 0.000 21.7543 36.1342 dimension3

237 4.00 49.50095* 1.90667 0.000 43.5117 55.4902

dimension2 3.00 1.00 -23.87500* 4.35783 0.019 -41.7432 -6.0068 2.00 -28.94429* 2.34145 0.000 -36.1342 -21.7543 dimension3 4.00 20.55667* 1.81864 0.000 14.3218 26.7915 4.00 1.00 -44.43167* 4.14050 0.004 -63.3084 -25.5549 2.00 -49.50095* 1.90667 0.000 -55.4902 -43.5117 dimension3 3.00 -20.55667* 1.81864 0.000 -26.7915 -14.3218 *. The mean difference is significant at the 0.05 level.

Synechococcus elongatus – Protein/Chlorophyll a relationships

Descriptives Protein/CHLa 95% Confidence Interval for Mean

N Mean Std. Deviation Std. Error Lower Bound Upper Bound Minimum Maximum 1.00 4 1.7805 0.02884 0.01442 1.7346 1.8263 1.74 1.81 2.00 5 2.0862 0.03402 0.01522 2.0440 2.1284 2.04 2.13 3.00 6 2.2443 0.03420 0.01396 2.2084 2.2802 2.21 2.29 4.00 5 2.2791 0.00667 0.00298 2.2708 2.2874 2.28 2.29 Total 20 2.1207 0.19186 0.04290 2.0309 2.2105 1.74 2.29

238 Test of Homogeneity of Variances Protein/CHLa Levene Statistic df1 df2 Sig. 2.544 3 16 0.093

ANOVA Protein/CHLa Sum of Squares df Mean Square F Sig. Between Groups 0.686 3 0.229 278.265 0.000 Within Groups 0.013 16 0.001 Total 0.699 19

Multiple Comparisons Dependent Variable: Protein/CHLa

(I) Synechococcus (J) Synechococcus Mean 95% Confidence Interval Difference (I-J) Std. Error Sig. Lower Bound Upper Bound Tukey HSD 1.00 2.00 -0.30575* 0.01923 0.000 -0.3608 -0.2507 3.00 -0.46388* 0.01851 0.000 -0.5168 -0.4109 dimension3 4.00 -0.49867* 0.01923 0.000 -0.5537 -0.4436 2.00 1.00 0.30575* 0.01923 0.000 0.2507 0.3608 3.00 -0.15813* 0.01736 0.000 -0.2078 -0.1085 dimension3 4.00 -0.19292* 0.01813 0.000 -0.2448 -0.1410

dimension2 3.00 1.00 0.46388* 0.01851 0.000 0.4109 0.5168

* 239 2.00 0.15813 0.01736 0.000 0.1085 0.2078 dimension3 4.00 -0.03479 0.01736 0.228 -0.0845 0.0149 4.00 1.00 0.49867* 0.01923 0.000 0.4436 0.5537 2.00 0.19292* 0.01813 0.000 0.1410 0.2448 dimension3 3.00 0.03479 0.01736 0.228 -0.0149 0.0845 Games-Howell 1.00 2.00 -0.30575* 0.02096 0.000 -0.3753 -0.2362 3.00 -0.46388* 0.02007 0.000 -0.5294 -0.3983 dimension3 4.00 -0.49867* 0.01473 0.000 -0.5660 -0.4314 2.00 1.00 0.30575* 0.02096 0.000 0.2362 0.3753

dimension2 3.00 -0.15813* 0.02065 0.000 -0.2231 -0.0931 dimension3 4.00 -0.19292* 0.01551 0.001 -0.2538 -0.1320

3.00 1.00 0.46388* 0.02007 0.000 0.3983 0.5294

dimension3 2.00 0.15813* 0.02065 0.000 0.0931 0.2231

4.00 -0.03479 0.01428 0.178 -0.0858 0.0162 4.00 1.00 0.49867* 0.01473 0.000 0.4314 0.5660 2.00 0.19292* 0.01551 0.001 0.1320 0.2538 dimension3 3.00 0.03479 0.01428 0.178 -0.0162 0.0858 *. The mean difference is significant at the 0.05 level.

Synechococcus elongatus – Colloidal carbohydrate/ Chlorophyll a relationships

Descriptives Colloidal CHO/CHLa 95% Confidence Interval for Mean

240 N Mean Std. Deviation Std. Error Lower Bound Upper Bound Minimum Maximum 1.00 4 0.9572 0.05893 0.02947 0.8634 1.0510 0.89 1.03 2.00 5 0.9926 0.03869 0.01730 0.9446 1.0407 0.94 1.04 3.00 6 1.0173 0.02147 0.00877 0.9947 1.0398 1.00 1.05 4.00 5 1.1018 0.02877 0.01287 1.0661 1.1375 1.08 1.14 Total 20 1.0202 0.06286 0.01406 0.9908 1.0496 0.89 1.14

Test of Homogeneity of Variances Colloidal CHO/CHLa Levene Statistic df1 df2 Sig. 1.805 3 16 0.187

ANOVA Colloidal CHO/CHLa Sum of Squares df Mean Square F Sig. Between Groups 0.053 3 0.018 12.847 0.000 Within Groups 0.022 16 0.001 Total 0.075 19

Multiple Comparisons Dependent Variable: Colloidal CHO/CHLa

(I) Synechococcus (J) Synechococcus Mean 95% Confidence Interval Difference (I-J) Std. Error Sig. Lower Bound Upper Bound 241 Tukey HSD 1.00 2.00 -0.03545 0.02489 0.503 -0.1067 0.0358 3.00 -0.06011 0.02395 0.096 -0.1286 0.0084 dimension3 4.00 -0.14464* 0.02489 0.000 -0.2159 -0.0734 2.00 1.00 0.03545 0.02489 0.503 -0.0358 0.1067 3.00 -0.02466 0.02247 0.696 -0.0889 0.0396 dimension3 4.00 -0.10920* 0.02346 0.001 -0.1763 -0.0421

dimension2 3.00 1.00 0.06011 0.02395 0.096 -0.0084 0.1286 2.00 0.02466 0.02247 0.696 -0.0396 0.0889 dimension3 4.00 -0.08454* 0.02247 0.008 -0.1488 -0.0203 4.00 1.00 0.14464* 0.02489 0.000 0.0734 0.2159 2.00 0.10920* 0.02346 0.001 0.0421 0.1763 dimension3 3.00 0.08454* 0.02247 0.008 0.0203 0.1488

Games-Howell 1.00 2.00 -0.03545 0.03417 0.738 -0.1617 0.0908 3.00 -0.06011 0.03074 0.349 -0.1938 0.0735 dimension3 4.00 -0.14464* 0.03215 0.034 -0.2734 -0.0159 2.00 1.00 0.03545 0.03417 0.738 -0.0908 0.1617 3.00 -0.02466 0.01940 0.610 -0.0918 0.0425 dimension3 4.00 -0.10920* 0.02156 0.005 -0.1796 -0.0388

dimension2 3.00 1.00 0.06011 0.03074 0.349 -0.0735 0.1938 2.00 0.02466 0.01940 0.610 -0.0425 0.0918 dimension3 4.00 -0.08454* 0.01557 0.004 -0.1355 -0.0336 4.00 1.00 0.14464* 0.03215 0.034 0.0159 0.2734 2.00 0.10920* 0.02156 0.005 0.0388 0.1796 dimension3

242 3.00 0.08454* 0.01557 0.004 0.0336 0.1355 *. The mean difference is significant at the 0.05 level.

Synechococcus elongatus – Storage carbohydrate/chlorophyll a relationships

Descriptives Storage 95% Confidence Interval for Mean

N Mean Std. Deviation Std. Error Lower Bound Upper Bound Minimum Maximum 1.00 4 1.6214 0.02771 0.01386 1.5773 1.6654 1.60 1.65 2.00 5 1.7681 0.01716 0.00767 1.7468 1.7894 1.75 1.79 3.00 6 1.7351 0.02514 0.01026 1.7087 1.7615 1.70 1.77 4.00 5 1.7411 0.02362 0.01056 1.7118 1.7705 1.71 1.77 Total 20 1.7221 0.05753 0.01286 1.6952 1.7490 1.60 1.79

243 Test of Homogeneity of Variances Storage CHO/CHLa Levene Statistic df1 df2 Sig. .566 3 16 0.646

ANOVA Storage CHO/CHLa Sum of Squares df Mean Square F Sig. Between Groups 0.054 3 0.018 32.465 0.000 Within Groups 0.009 16 0.001 Total 0.063 19

Multiple Comparisons Dependent Variable:Storage CHO/CHLa

(I) Synechococcus (J) Synechococcus Mean 95% Confidence Interval Difference (I-J) Std. Error Sig. Lower Bound Upper Bound

Tukey HSD 1.00 2.00 -0.14675* 0.01580 0.000 -0.1919 -0.1016 3.00 -0.11373* 0.01520 0.000 -0.1572 -0.0702 dimension3 4.00 -0.11979* 0.01580 0.000 -0.1650 -0.0746 2.00 1.00 0.14675* 0.01580 0.000 0.1016 0.1919 3.00 0.03302 0.01426 0.136 -0.0078 0.0738 dimension3 4.00 0.02696 0.01489 0.305 -0.0156 0.0696

dimension2 * 244 3.00 1.00 0.11373 0.01520 0.000 0.0702 0.1572 2.00 -0.03302 0.01426 0.136 -0.0738 0.0078 dimension3 4.00 -0.00606 0.01426 0.973 -0.0469 0.0347 4.00 1.00 0.11979* 0.01580 0.000 0.0746 0.1650 2.00 -0.02696 0.01489 0.305 -0.0696 0.0156 dimension3 3.00 0.00606 0.01426 0.973 -0.0347 0.0469 Games-Howell 1.00 2.00 -0.14675* 0.01584 0.001 -0.2062 -0.0873 3.00 -0.11373* 0.01724 0.002 -0.1731 -0.0543 dimension3 4.00 -0.11979* 0.01742 0.002 -0.1802 -0.0594 2.00 1.00 0.14675* 0.01584 0.001 0.0873 0.2062 dimension2 3.00 0.03302 0.01281 0.115 -0.0072 0.0733 dimension3 4.00 0.02696 0.01305 0.249 -0.0158 0.0697 3.00 1.00 0.11373* 0.01724 0.002 0.0543 0.1731 dimension3

2.00 -0.03302 0.01281 0.115 -0.0733 0.0072 4.00 -0.00606 0.01473 0.975 -0.0522 0.0401 4.00 1.00 0.11979* 0.01742 0.002 0.0594 0.1802 2.00 -0.02696 0.01305 0.249 -0.0697 0.0158 dimension3 3.00 0.00606 0.01473 0.975 -0.0401 0.0522 *. The mean difference is significant at the 0.05 level.

245

Microcystis aeruginose – Chlorophyll a: marker pigment relationships

Descriptives CHLa/Zeaxanthin 95% Confidence Interval for Mean

N Mean Std. Deviation Std. Error Lower Bound Upper Bound Minimum Maximum 1.00 5 26.6680 1.57357 0.70372 24.7142 28.6218 25.14 29.12 2.00 5 19.3480 1.62603 0.72718 17.3290 21.3670 16.88 21.12 3.00 6 10.6667 1.11355 0.45460 9.4981 11.8353 9.21 12.34 Total 16 18.3800 6.98471 1.74618 14.6581 22.1019 9.21 29.12

Test of Homogeneity of Variances CHLa/ Zeaxanthin Levene Statistic df1 df2 Sig. 0.477 2 13 0.631

ANOVA CHLa/ Zeaxanthin Sum of Squares df Mean Square F Sig. Between Groups 705.113 2 352.556 171.783 0.000 Within Groups 26.680 13 2.052 Total 731.793 15 246

Multiple Comparisons Dependent Variable: CHLa/Zeaxanthin

(I) Microcystis (J) Microcystis Mean 95% Confidence Interval Difference (I-J) Std. Error Sig. Lower Bound Upper Bound Tukey HSD 1.00 2.00 7.32000* 0.90605 0.000 4.9276 9.7124

dimension3 3.00 16.00133* 0.86748 0.000 13.7108 18.2919 2.00 1.00 -7.32000* 0.90605 0.000 -9.7124 -4.9276

dimension2 dimension3 3.00 8.68133* 0.86748 0.000 6.3908 10.9719 3.00 1.00 -16.00133* 0.86748 0.000 -18.2919 -13.7108

dimension3 2.00 -8.68133* 0.86748 0.000 -10.9719 -6.3908

Games-Howell 1.00 2.00 7.32000* 1.01194 0.000 4.4278 10.2122

dimension3 3.00 16.00133* 0.83779 0.000 13.5385 18.4642 2.00 1.00 -7.32000* 1.01194 0.000 -10.2122 -4.4278

dimension2 dimension3 3.00 8.68133* 0.85759 0.000 6.1463 11.2164 3.00 1.00 -16.00133* 0.83779 0.000 -18.4642 -13.5385

dimension3 2.00 -8.68133* 0.85759 0.000 -11.2164 -6.1463 *. The mean difference is significant at the 0.05 level.

247 Microcystis aeruginose – Chlorophyll a: marker pigment relationships

Descriptives CHLa/Echinenone 95% Confidence Interval for Mean

N Mean Std. Deviation Std. Error Lower Bound Upper Bound Minimum Maximum 1 5 18.6820 2.81530 1.25904 15.1863 22.1777 15.32 22.92 2 5 22.0100 3.85852 1.72558 17.2190 26.8010 17.78 27.31 3 6 14.6017 1.06792 0.43598 13.4810 15.7224 13.26 16.23 Total 16 18.1919 4.06930 1.01732 16.0235 20.3603 13.26 27.31

Test of Homogeneity of Variances CHLa/Echinenone Levene Statistic df1 df2 Sig. 2.672 2 13 0.107

ANOVA CHLa/Echinenone Sum of Squares df Mean Square F Sig. Between Groups 151.429 2 75.715 10.152 0.002 Within Groups 96.959 13 7.458 Total 248.388 15 248

Post Hoc Tests Multiple Comparisons Dependent Variable:CHLa/ Echinenone

(I) Microcystis (J) Microcystis Mean 95% Confidence Interval Difference (I-J) Std. Error Sig. Lower Bound Upper Bound Tukey HSD 1 2 -3.32800 1.72723 0.170 -7.8887 1.2327

dimension3 3 4.08033 1.65370 0.068 -0.2862 8.4468 2 1 3.32800 1.72723 0.170 -1.2327 7.8887

dimension2 dimension3 3 7.40833* 1.65370 0.002 3.0418 11.7748 3 1 -4.08033 1.65370 0.068 -8.4468 0.2862

dimension3 2 -7.40833* 1.65370 0.002 -11.7748 -3.0418

Games-Howell 1 2 -3.32800 2.13607 0.321 -9.5527 2.8967

dimension3 3 4.08033 1.33239 0.062 -0.2677 8.4284 2 1 3.32800 2.13607 0.321 -2.8967 9.5527

dimension2 dimension3 3 7.40833* 1.77981 0.024 1.3871 13.4296 3 1 -4.08033 1.33239 0.062 -8.4284 0.2677

dimension3 2 -7.40833* 1.77981 0.024 -13.4296 -1.3871 *. The mean difference is significant at the 0.05 level.

249 Microcystis aeruginose – Protein/Chlorophyll a relationships

Descriptives Protein/CHLa 95% Confidence Interval for Mean

N Mean Std. Deviation Std. Error Lower Bound Upper Bound Minimum Maximum 1.00 5 1.7317 0.03736 0.01671 1.6853 1.7781 1.69 1.78 2.00 6 1.8037 0.03795 0.01549 1.7639 1.8435 1.75 1.85 3.00 6 1.9112 0.04213 0.01720 1.8670 1.9554 1.86 1.98 Total 17 1.8205 0.08372 0.02031 1.7774 1.8635 1.69 1.98

Test of Homogeneity of Variances Protein/CHLa Levene Statistic df1 df2 Sig. 0.005 2 14 0.995

ANOVA Protein/CHLa Sum of Squares df Mean Square F Sig. Between Groups 0.090 2 0.045 29.249 0.000 Within Groups 0.022 14 0.002 250 Total 0.112 16

Post Hoc Tests Multiple Comparisons Dependent Variable: Protein/CHLa

(I) Microcyctis (J) Microcyctis Mean 95% Confidence Interval Difference (I-J) Std. Error Sig. Lower Bound Upper Bound Tukey HSD 1.00 2.00 -0.07198* 0.02382 0.023 -0.1343 -0.0096

dimension3 3.00 -0.17951* 0.02382 0.000 -0.2418 -0.1172 2.00 1.00 0.07198* 0.02382 0.023 0.0096 0.1343 dimension2

dimension3 3.00 -0.10753* 0.02271 0.001 -0.1670 -0.0481

3.00 1.00 0.17951* 0.02382 0.000 0.1172 0.2418 dimension3

2.00 0.10753* 0.02271 0.001 0.0481 0.1670 Games-Howell 1.00 2.00 -0.07198* 0.02279 0.029 -0.1360 -0.0080

dimension3 3.00 -0.17951* 0.02398 0.000 -0.2465 -0.1125 2.00 1.00 0.07198* 0.02279 0.029 0.0080 0.1360

dimension2 dimension3 3.00 -0.10753* 0.02315 0.002 -0.1711 -0.0440 3.00 1.00 0.17951* 0.02398 0.000 0.1125 0.2465

dimension3 2.00 0.10753* 0.02315 0.002 0.0440 0.1711 *. The mean difference is significant at the 0.05 level.

251 Microcystis aeruginosa – Colloidal carbohydrate/Chlorophyll a relationships

Descriptives Colloidal CHO/CHLa 95% Confidence Interval for Mean

N Mean Std. Deviation Std. Error Lower Bound Upper Bound Minimum Maximum 1.00 5 0.7596 0.04382 0.01960 0.7052 0.8140 0.71 0.82 2.00 6 0.8372 0.05806 0.02370 0.7762 0.8981 0.76 0.92 3.00 6 0.8410 0.04089 0.01669 0.7981 0.8839 0.76 0.87 Total 17 0.8157 0.05874 0.01425 0.7855 0.8459 0.71 0.92

Test of Homogeneity of Variances Colloidal CHO/CHLa Levene Statistic df1 df2 Sig. 0.596 2 14 0.564

ANOVA Colloidal CHO/CHLa Sum of Squares df Mean Square F Sig. Between Groups 0.022 2 0.011 4.750 0.027 Within Groups 0.033 14 0.002 Total 0.055 16 252

Post Hoc Tests Multiple Comparisons Dependent Variable: Colloidal CHO/CHLa

(I) Microcystis (J) Microcystis Mean 95% Confidence Interval Difference (I-J) Std. Error Sig. Lower Bound Upper Bound Tukey HSD 1.00 2.00 -0.07753* 0.02935 0.048 -0.1543 -0.0007

dimension3 3.00 -0.08136* 0.02935 0.037 -0.1582 -0.0045 2.00 1.00 0.07753* 0.02935 0.048 0.0007 0.1543

dimension2 dimension3 3.00 -0.00383 0.02798 0.990 -0.0771 0.0694 3.00 1.00 0.08136* 0.02935 0.037 0.0045 0.1582

dimension3 2.00 0.00383 0.02798 0.990 -0.0694 0.0771

Games-Howell 1.00 2.00 -0.07753 0.03075 0.076 -0.1635 0.0084

dimension3 3.00 -0.08136* 0.02574 0.030 -0.1542 -0.0085 2.00 1.00 0.07753 0.03075 0.076 -0.0084 0.1635

dimension2 dimension3 3.00 -0.00383 0.02899 0.990 -0.0848 0.0771 3.00 1.00 0.08136* 0.02574 0.030 0.0085 0.1542

dimension3 2.00 0.00383 0.02899 0.990 -0.0771 0.0848 *. The mean difference is significant at the 0.05 level.

253 Microcystis aeruginose – Storage carbohydrate/Chlorophyll a relationships

Descriptives Storage/CHLa 95% Confidence Interval for Mean

N Mean Std. Deviation Std. Error Lower Bound Upper Bound Minimum Maximum 1.00 5 1.9186 0.03368 0.01506 1.8768 1.9604 1.88 1.95 2.00 6 1.9207 0.07289 0.02976 1.8442 1.9972 1.80 2.00 3.00 6 2.0264 0.03296 0.01345 1.9918 2.0610 1.98 2.06 Total 17 1.9574 0.07101 0.01722 1.9209 1.9939 1.80 2.06

Test of Homogeneity of Variances Storage CHO/CHLa Levene Statistic df1 df2 Sig. 1.999 2 14 0.172

ANOVA Storage CHO/CHLa Sum of Squares df Mean Square F Sig. Between Groups 0.044 2 0.022 8.459 0.004 254 Within Groups 0.037 14 0.003 Total 0.081 16

Post Hoc Tests Multiple Comparisons Dependent Variable: Storage/CHLa

(I) Microcystis (J) Microcystis Mean 95% Confidence Interval Difference (I-J) Std. Error Sig. Lower Bound Upper Bound Tukey HSD 1.00 2.00 -0.00213 0.03093 0.997 -0.0831 0.0788

dimension3 3.00 -0.10778* 0.03093 0.010 -0.1887 -0.0268 2.00 1.00 0.00213 0.03093 0.997 -0.0788 0.0831 dimension2

dimension3 3.00 -0.10565* 0.02949 0.008 -0.1828 -0.0285 3.00 1.00 0.10778* 0.03093 0.010 0.0268 0.1887 dimension3

2.00 0.10565* 0.02949 0.008 0.0285 0.1828 Games-Howell 1.00 2.00 -0.00213 0.03335 0.998 -0.0994 0.0951

dimension3 3.00 -0.10778* 0.02019 0.001 -0.1647 -0.0509 2.00 1.00 0.00213 0.03335 0.998 -0.0951 0.0994

dimension2 dimension3 3.00 -0.10565* 0.03266 0.034 -0.2020 -0.0093 3.00 1.00 0.10778* 0.02019 0.001 0.0509 0.1647

dimension3 2.00 0.10565* 0.03266 0.034 0.0093 0.2020 *. The mean difference is significant at the 0.05 level.

255

Rhodomonas salina – Chlorophyll a: marker pigment relationships

Descriptives CHLa/Alloxanthin 95% Confidence Interval for Mean

N Mean Std. Deviation Std. Error Lower Bound Upper Bound Minimum Maximum 1.00 5 2.5880 0.25927 0.11595 2.2661 2.9099 2.37 3.03 2.00 6 2.6133 0.27818 0.11357 2.3214 2.9053 2.26 3.03 3.00 6 2.5400 0.10450 0.04266 2.4303 2.6497 2.42 2.67 Total 17 2.5800 0.21316 0.05170 2.4704 2.6896 2.26 3.03

Test of Homogeneity of Variances CHLa/Alloxanthin Levene Statistic df1 df2 Sig. 1.455 2 14 0.267

ANOVA CHLa/Alloxanthin Sum of Squares df Mean Square F Sig. Between Groups 0.017 2 0.008 0.163 0.851 256 Within Groups 0.710 14 0.051 Total 0.727 16

Post Hoc Tests Multiple Comparisons Dependent Variable: CHLa/Alloxanthin

(I) Rhodomonas (J) Rhodomonas Mean 95% Confidence Interval Difference (I-J) Std. Error Sig. Lower Bound Upper Bound Tukey HSD 1.00 2.00 -0.02533 0.13640 0.981 -0.3823 0.3317

dimension3 3.00 0.04800 0.13640 0.934 -0.3090 0.4050 dimension2 2.00 1.00 0.02533 0.13640 0.981 -0.3317 0.3823 dimension3

3.00 0.07333 0.13006 0.841 -0.2671 0.4137 3.00 1.00 -0.04800 0.13640 0.934 -0.4050 0.3090

dimension3 2.00 -0.07333 0.13006 0.841 -0.4137 0.2671 Games-Howell 1.00 2.00 -0.02533 0.16230 0.987 -0.4800 0.4293

dimension3 3.00 0.04800 0.12355 0.921 -0.3517 0.4477 2.00 1.00 0.02533 0.16230 0.987 -0.4293 0.4800

dimension2 dimension3 3.00 0.07333 0.12132 0.823 -0.2925 0.4392 3.00 1.00 -0.04800 0.12355 0.921 -0.4477 0.3517

dimension3 2.00 -0.07333 0.12132 0.823 -0.4392 0.2925

257

Rhodomonas salina – Protein/Chlorophyll a relationships

Descriptives Protein 95% Confidence Interval for Mean

N Mean Std. Deviation Std. Error Lower Bound Upper Bound Minimum Maximum 1.00 5 2.4459 0.02405 0.01075 2.4160 2.4757 2.42 2.48 2.00 5 2.4363 0.02884 0.01290 2.4004 2.4721 2.42 2.49 3.00 6 2.5543 0.04969 0.02029 2.5022 2.6065 2.49 2.62 Total 16 2.4835 0.06649 0.01662 2.4481 2.5190 2.42 2.62

Test of Homogeneity of Variances Protein/CHLa Levene Statistic df1 df2 Sig. 1.955 2 13 0.181

ANOVA Protein/CHLa Sum of Squares df Mean Square F Sig. Between Groups 0.048 2 0.024 17.464 0.000

258 Within Groups 0.018 13 0.001 Total 0.066 15

Post Hoc Tests Multiple Comparisons Dependent Variable :Protein/CHLa

(I) Rhodomonas (J) Rhodomonas Mean 95% Confidence Interval Difference (I-J) Std. Error Sig. Lower Bound Upper Bound Tukey HSD 1.00 2.00 0.00962 0.02353 0.913 -0.0525 0.0717

dimension3 3.00 -0.10844* 0.02252 0.001 -0.1679 -0.0490

dimension2 2.00 1.00 -0.00962 0.02353 0.913 -.0717 0.0525

dimension3 3.00 -0.11806* 0.02252 0.000 -0.1775 -0.0586

3.00 1.00 0.10844* 0.02252 0.001 0.0490 0.1679

dimension3 2.00 0.11806* 0.02252 0.000 0.0586 0.1775 Games-Howell 1.00 2.00 0.00962 0.01679 0.838 -0.0387 0.0579

dimension3 3.00 -0.10844* 0.02296 0.004 -0.1750 -0.0418 2.00 1.00 -0.00962 0.01679 0.838 -0.0579 .0387

dimension2 dimension3 3.00 -0.11806* 0.02404 0.003 -0.1864 -0.0497 3.00 1.00 0.10844* 0.02296 0.004 0.0418 0.1750

dimension3 2.00 0.11806* 0.02404 0.003 0.0497 0.1864 *. The mean difference is significant at the 0.05 level.

259

Rhodomonas salina – Colloidal carbohydrate/Chlorophyll a relationships

Descriptives Colloidal/CHLa 95% Confidence Interval for Mean

N Mean Std. Deviation Std. Error Lower Bound Upper Bound Minimum Maximum 1.00 5 1.4569 0.07929 0.03546 1.3584 1.5553 1.37 1.56 2.00 5 1.3765 0.04288 0.01917 1.3233 1.4297 1.32 1.42 3.00 6 1.2948 0.04519 0.01845 1.2474 1.3423 1.24 1.37 Total 16 1.3710 0.08738 0.02185 1.3244 1.4175 1.24 1.56

Test of Homogeneity of Variances Colloidal/CHLa Levene Statistic df1 df2 Sig. 1.815 2 13 0.202

ANOVA Colloidal/CHLa Sum of Squares df Mean Square F Sig. Between Groups 0.072 2 0.036 10.929 0.002

260 Within Groups 0.043 13 0.003 Total 0.115 15

Post Hoc Tests

Multiple Comparisons Dependent Variable: Colloidal/CHLa

(I) Rhodomonas (J) Rhodomonas Mean 95% Confidence Interval Difference (I-J) Std. Error Sig. Lower Bound Upper Bound Tukey HSD 1.00 2.00 0.08036 0.03625 0.105 -0.0154 0.1761

dimension3 3.00 0.16203* 0.03471 0.001 0.0704 0.2537 dimension2 2.00 1.00 -0.08036 0.03625 0.105 -0.1761 0.0154 dimension3

3.00 0.08167 0.03471 0.083 -0.0100 0.1733 3.00 1.00 -0.16203* 0.03471 0.001 -0.2537 -0.0704

dimension3 2.00 -0.08167 0.03471 0.083 -0.1733 0.0100 Games-Howell 1.00 2.00 0.08036 0.04031 0.193 -0.0424 0.2031

dimension3 3.00 0.16203* 0.03997 0.015 0.0400 0.2841 2.00 1.00 -0.08036 0.04031 0.193 -0.2031 0.0424

dimension2 dimension3 3.00 0.08167* 0.02661 0.033 0.0071 0.1563 3.00 1.00 -0.16203* 0.03997 0.015 -0.2841 -0.0400

dimension3 2.00 -0.08167* 0.02661 0.033 -0.1563 -0.0071 *. The mean difference is significant at the 0.05 level.

261

Rhodomonas salina – Storage carbohydrate/Chlorophylll a relationship

Descriptives Storage/CHLa 95% Confidence Interval for Mean

N Mean Std. Deviation Std. Error Lower Bound Upper Bound Minimum Maximum 1.00 5 2.0041 0.08606 0.03849 1.8972 2.1109 1.92 2.15 2.00 5 1.9560 0.12717 0.05687 1.7981 2.1139 1.85 2.15 3.00 6 2.2331 0.02436 0.00995 2.2076 2.2587 2.20 2.27 Total 16 2.0749 0.15128 0.03782 1.9943 2.1556 1.85 2.27

Test of Homogeneity of Variances Storage/CHLa Levene Statistic df1 df2 Sig. 4.292 2 13 0.037

ANOVA Storage/CHLa Sum of Squares df Mean Square F Sig. Between Groups .246 2 0.123 16.436 0.000 Within Groups .097 13 0.007 Total .343 15 262

Robust Tests of Equality of Means Storage/CHLa Statistica df1 df2 Sig. Brown-Forsythe 14.835 2 7.349 0.003 a. Asymptotically F distributed.

Multiple Comparisons Dependent Variable :Storage/CHLa

(I) Rhodomonas (J) Rhodomonas Mean 95% Confidence Interval Difference (I-J) Std. Error Sig. Lower Bound Upper Bound

Tukey HSD 1.00 2.00 0.04806 0.05471 0.663 -0.0964 0.1925

dimension3 3.00 -0.22907* 0.05238 0.002 -0.3674 -0.0908 2.00 1.00 -0.04806 0.05471 0.663 -0.1925 0.0964

dimension2 dimension3 3.00 -0.27713* 0.05238 0.000 -0.4154 -0.1388

3.00 1.00 0.22907* 0.05238 0.002 0.0908 0.3674

dimension3 2.00 0.27713* 0.05238 0.000 0.1388 0.4154 Games-Howell 1.00 2.00 0.04806 0.06867 0.771 -0.1540 0.2501

dimension3 * 263 3.00 -0.22907 0.03975 0.007 -0.3633 -0.0949 2.00 1.00 -0.04806 0.06867 0.771 -0.2501 0.1540

dimension2 dimension3 3.00 -0.27713* 0.05774 0.016 -0.4775 -0.0768 3.00 1.00 0.22907* 0.03975 0.007 0.0949 0.3633

dimension3 2.00 0.27713* 0.05774 0.016 0.0768 0.4775 *. The mean difference is significant at the 0.05 level.

VII- Cellular concentration of CHLa and photosynthates Cellular concentration of chlorophyll a and products of photosynthesis, in relation to biovolume

Genus Lt R 1 R 2 R 3 R 4 R 5 R 6 Biovol* Item/ biovol pg cell- pg cell- pg cell- pg cell- pg cell- pg cell -¹ fg (item) ¹ ¹ ¹ ¹ ¹ cell-¹ (µm³) µm³·cell A. carterae 432 CHLa L 0.44 0.26 0.31 0.37 0.29 0.1901 M 0.33 0.54 0.24 0.17 0.35 0.21 0.2333 H 0.33 0.27 0.33 0.4 0.35 0.1728 protein L 91.85 45.98 84.02 65.75 73.11 39.6792 M 95.07 243.1 130.4 111.2 169.9 110.6 73.4227 H 179.2 124.3 196.4 216.28 139.59 93.4330 colloidal L 13.04 93.87 12.72 22.25 14.48 40.5518 CHO M 16.85 31.87 14.08 11.1 21.37 17.64 13.7678 H 36.2 31.28 30.67 38.26 35.87 15.6384 storage CHO L 76.42 49.9 49.09 74.87 55.98 33.0134 M 65.03 108.62 5.25 43.54 78.04 54.91 46.9238 H 108.25 93.49 158.64 195.07 137.59 84.2702 C. 2720 meneghiniana CHLa L 0.17 0.99 0.14 0.07 0.37 2.6928 M 0.62 0.2 1.09 0.15 0.09 0.19 2.9648 H 1.17 0.44 1.32 2.15 2.45 1.16 6.6640 protein L 15.28 91.54 14.76 72.31 32.86 248.9888 M 85.21 24.59 155.76 17.55 9.23 23.15 423.6672 H 155.12 73.34 197.46 290.73 380.01 167.61 1033.6272 colloidal L 1.61 1.77 1.18 0.63 2.6 4.8144 CHO M 7.34 3.27 0.16 1.63 0.91 1.97 19.9648 H 31.8 11.5 32.95 43.3 48.86 32.82 132.8992 storage CHO L 6.43 0.16 0.11 4.11 0.16 17.4896 M 0.89 0.36 204.34 0.18 0.12 0.21 555.8048 H 235.17 118.87 368.16 496.26 575.01 268.58 1564.0272 T.weissflogii 2813 CHLa L 0.182 0.139 0.727 0.62 0.747 2.1013 M 0.164 0.173 0.636 0.109 0.205 1.7891 H 2.55 30.4 88.2 1.36 58.9 47.1 248.1066 protein L 31.42 25.89 99.8 99.25 115.96 326.1955 M 29.39 31.49 124.46 18.09 35.85 350.1060 H 489.62 60.5 210.01 272.04 133.66 103.31 1377.3011 colloidal L 1.15 1.47 6.22 4.4 5.34 17.4969 CHO M 1.86 2.13 7.53 1.23 2.5 21.1819 H 41.77 5.37 12.58 19.81 9.2 8.15 117.4990 storage CHO L 21.19 12.24 79.6 65.3 90.84 255.5329 M 20.62 22.84 107.78 17.94 26.79 303.1851 264

H 602 61.1 176 238 121 73.4 1693.4260 D. tertiolecta 43.42 CHLa L 0.69 0.86 0.78 0.79 0.0373 M 0.52 0.87 0.64 0.45 0.41 n/a 0.0378 H 0.93 1.25 0.7 0.85 0.91 0.87 0.0543 protein L 48.19 62.32 56.94 53.26 2.7059 M 47.64 114.78 84.91 33.07 54.15 n/a 3.6868 H 141.73 226.76 109.84 151.46 170.79 163.04 7.4157 colloidal L 6.11 6.42 6.72 7.38 0.3204 CHO M 2.39 5.09 3.72 1.69 1.32 n/a 0.2210 H 3.82 7.78 2.81 5.62 5.62 6.13 0.3378 storage CHO L 44.68 50.45 45.69 45.46 2.1905 M 43.5 84.28 53.94 32.1 38.38 3.6594 H 96.83 116.56 79.01 96.69 105.34 99.8 5.0610 S. 45 quadricauda CHLa L 0.028 0.055 0.096 0.109 0.126 n/a 0.0057 M 0.401 0.3 0.689 0.774 1.12 0.875 0.0394 H 10.5 4.19 4.45 6.79 4.41 3.83 0.3056 protein L 13.48 28.68 78.3 93.34 53.33 n/a 0.3321 M 94.7 50.5 114 135 189 144 8.5050 H 1239 575 567 871 551 489 55.7550 colloidal L 3.37 2.85 9.31 19.65 11.02 n/a 0.8843 CHO M 3.61 1.88 6.94 4.75 6.86 4.98 0.3123 H 74.55 30.77 30.06 52.96 32.63 30.12 3.3548 storage CHO L 13.62 29.85 51.83 70.22 71.51 n/a 3.2180 M 19.12 11.79 31.85 22.79 30.27 23.84 0.0057 H 1929.25 69.46 81.91 1250.72 88.54 73.74 56.2824 S.elongatus 4.2 CHLa DL 0.55 0.94 0.23 0.7 0.0039 L 0.19 2.64 0.76 1.9 5.73 0.0241 M 0.36 0.79 1.77 0.2 0.4 2.39 0.0100 H 9.25 7.88 5.8 3.18 0.1 0.0389 protein DL 35.2 59.46 12.99 41.72 0.1752 L 0.26 291.82 0.96 229.7 674.68 2.8337 M 0.58 137.02 346.61 0.31 0.75 430.16 1.8067 H 1706.69 1539.28 1095.9 599.76 1986.11 8.3417 colloidal DL 4.79 4.26 0.22 7.48 0.0314 CHO L 1.82 0.23 8.31 0.19 0.6 0.0349 M 3.59 8.61 0.18 1.97 1.54 0.24 0.0362 H 114.26 0.94 0.77 0.44 125.51 0.5271 storage CHO DL 0.24 0.37 0.93 0.32 0.0039 L 0.11 149.67 0.48 112.74 336.79 1.4145 M 0.2 0.4 0.93 0.1 0.22 142.21 0.5973 H 519.69 462.17 313.28 161.2 590.26 2.4791 M. 65 aeruginosa CHLa L 0.147 0.086 0.1 0.059 0.246 0.0160

265

M 0.0509 0.089 0.107 0.102 0.248 0.177 0.0161 H 0.826 0.896 0.207 0.216 0.533 0.899 0.0584 protein L 7.06 5.22 5.89 3.36 12.1 0.7865 M 3.65 5.25 7.33 5.79 16 11.1 0.7215 H 70.41 85.9 16.9 16.8 41.7 5.5835 colloidal L 0.862 0.438 0.577 0.39 0.132 0.0560 CHO M 0.389 0.589 0.62 0.849 0.155 0.122 0.0552 H 6.15 6.34 1.19 1.52 3.86 6.49 0.4219 storage CHO L 10.85 6.51 8.6 5.3 2.17 0.7053 M 4.5 8.92 9.16 9.47 15.54 13.4 0.1411 H 95.62 91.06 23.86 23.68 51.28 90.78 6.2153 R. salina 141 CHLa L 0.024 0.044 0.047 0.035 0.069 0.0097 M 0.095 0.097 0.152 0.14 0.151 0.0213 H 0.297 0.503 0.449 0.602 0.493 0.509 0.0849 protein L 7.05 11.61 12.88 9.41 20.55 2.8976 M 24.76 25.87 46.62 32.66 39.93 6.5734 H 122.87 195.66 168.54 185.01 158.27 180.8 27.5881 colloidal L 0.777 1.015 1.191 0.984 2.503 0.3529 CHO M 2.23 2.57 3.97 3.82 3.67 0.5598 H 5.71 9.14 8.9 12.38 11.6 8.92 1.7456 storage CHO L 4.14 6.08 5.27 5.73 14.4 2.0304 M 9.61 9.43 21.27 13.19 12.5 1.8598 H 47.57 69.62 68.42 89.54 70.79 78.21 12.6251 * Olenina et al., 2006

(R= run; blank cells = run not performed; VL = 10 µmol photons·m-2·s-1; L = 37 µmol photons·m-2·s-1; ML = 70-75 µmol photons·m-2·s-1; HL = 200 µmol photons·m-2·s-1; 1µm3 = 1x10-9 µL).

266

VIII- Typical chromatograms of species studied

Typical chromatograms of the species studied (see Appendix V for pigment abbreviation codes)

LUT CHLa

VIOLA VIOLA O CHLb E N

de hlli C

BETA

o r Chllide a ANTH y P

Scenedesmus quadricauda chromatogram showing characteristic pigments

CHL a ALLO

CHLs c1/c2

ALPH Chllide a CHL a allo

Rhodomonas salina chromatogram showing characteristic pigments

267

CHLa

MYXOL CANTH ZEA CHLa`+ ECHIN

ALPHA MYXO

Microcystis aeruginosa chromatogram showing characteristic pigments

FUCO

DIAD

CHL a CHL a 2 /c 1

CHLs c

Chllide a O cis FUCO ro Chllide a DIAT CHL a allo y CHL a`+ ECHIN

P BETA

Thalassiosira weissflogii chromatogram showing characteristic pigments

268

CHLa

LUT

CHL b

O I

ALPHA CHLa` CHLa`

NEO

ANTH Chllide a

Dunaliella tertiolecta chromatogram showing characteristic pigments

CHLa BETA

ZEA MYXOL CHLa+ ECHIN Chllide a

Synechococcus elongatus chromatogram, showing characteristic pigments

269

CHLs c1/c2 PERI CHLa

DIAD DIAD

o ll

a BETA DINO DINO

HL DIAT DIAT C CHLa` P468 P457

Amphidinium carterae chromatogram showing characteristic pigments

FUCO CHL a

CHLs c1/c2

CHL a allo

DIAD

19` Hex BETA DIAT DIAT CHLa`

FUCOL Phytylated - c a CHllide

Cyclotella meneghiniana chromatogram showing characteristic pigments

270

IX- Specific growth rate (µ) curves

Specific growth rate constants (μ) calculated from the slopes of the semilog plots of growth versus time for the 8 species at each light level.

Amphidinium carterae- specific growth rates LL 16.00 ML 14.00 12.00 HL 10.00 8.00 y1 = 0.075x + 10.844 y3 = 0.1764x + 10.982 R² = 0.8327 6.00 R² = 0.8608

density) ln(cell 4.00 y2 = 0.1493x + 10.873 2.00 R² = 0.8931 0.00 0 5 10 15 20 25 30

Time (day)

LL (y1: μ = 0.075 day-1); ML (y2: μ = 0.1493 day-1); HL (y3: μ = 0.1764 day-1)

Cyclotella meneghiniana - specific growth rate LL 18.00 16.00 ML

14.00 HL 12.00 10.00

8.00 y1 = 0.1707x + 10.606 y3 = 0.2408x + 11.616 6.00 R² = 0.9524 R² = 0.9712 density) (cell ln 4.00 2.00 y2 = 0.238x + 11.766 R² = 0.9363 0.00 0 5 10 15 20 25 Time (day)

LL (y1: μ = 0.1707 day-1); ML (y2: μ = 0.238 day-1); HL (y3: μ = 0.2408day-1)

271

Thalassiosira weissflogii - specific growth rate 20.00 LL 18.00

16.00 ML 14.00 HL 12.00 10.00 y1 = 0.132x + 11.007 8.00 y3 = 0.436x + 11.12 R² = 0.9834 R² = 0.9512 density) (cell ln 6.00 4.00 y2 = 0.2262x + 11.611 2.00 R² = 0.8233 0.00 0 5 10 15 20 25 30 Time (day)

LL (y1: μ = 0.132 day-1); ML (y2: μ = 0.2262 day-1); HL (y3: μ = 0.436 day-1)

Dunaliellla tertiolecta - specific growth rate 18.00 LL 16.00 ML 14.00 12.00 HL 10.00 8.00 y1 = 0.1589x + 11.309 y3 = 0.26x + 12.025 6.00 R² = 0.8913 R² = 0.7852 ln (cell density) (cell ln 4.00 y2 = 0.1652x + 11.228 2.00 R² = 0.9337 0.00 0 5 10 15 20 25 Time (day)

LL (y1: μ = 0.1589 day-1); ML (y2: μ = 0.1652day-1); HL (y3: μ = 0.26 day-1)

272

Scenedesmus quadricauda - specific growth rate 16.00

14.00 LL 12.00 ML 10.00 HL

8.00 y1 = 0.1008x + 10.585 y3 = 0.1712x + 10.541 6.00 R² = 0.9692 R² = 0.998 ln (cell density) (cell ln 4.00 y2 = 0.1671x + 10.793 2.00 R² = 0.9293

0.00 0 5 10 15 20 Time (day)

LL (y1: μ = 0.1008 day-1); ML (y2: μ = 0.1671 day-1); HL (y3: μ = 0.1712 day-1)

Microcystis aeruginose - specific growth rate 18.00 16.00 LL

14.00 ML 12.00 HL 10.00

8.00 y1 = 0.2074x + 12.971 y3 = 0.2284x + 13.158 R² = 0.9712 R² = 0.8983

density) (cell ln 6.00 4.00 y2 = 0.228x + 12.496 2.00 R² = 0.852 0.00 0 5 10 15 20 Time (day)

LL (y1: μ = 0.2074 day-1); ML (y2: μ = 0.228 day-1); HL (y3: μ = 0.2284 day-1)

273

Synechococcus elongatus - specific growth rate 16.00 VL 14.00 LL 12.00 ML

10.00 HL 8.00 y1 = 0.0485x + 11.865 6.00 y3 = 0.2275x + 10.298 R² = 0.8806

ln (cell density) (cell ln R² = 0.9721 4.00 2.00 y2 = 0.171x + 10.571 y4 = 0.2549x + 10.328 R² = 0.9785 R² = 0.9597 0.00 0 5 10 15 20 25 30 35 Time (day)

DL (y1: μ = 0.0485 day-1); LL (y2: μ = 0.171 day-1); ML (y3: μ = 0.2275 day-1); HL (y4: μ = 0.2549 day-1)

Rhodomonas salina - specific growth rate 16.00 14.00 LL 12.00 ML 10.00 HL 8.00 y1 = 0.0833x + 10.856 y3 = 0.2431x + 10.5 6.00 R² = 0.6205 R² = 0.9558 ln (cell density) (cell ln 4.00 y2 = 0.1209x + 11.342 2.00 R² = 0.7885 0.00 0 5 10 15 20 25 Time (day)

LL (y1: μ = 0.0833 day-1); ML (y2: μ = 0.1209 day-1); HL (y3: μ = 0.2431 day-1)

274

X – NMR SPECTRA

1H NMR spectra of the new pigment (with integration)

1H NMR spectra of the new pigment (possible N-H and OH signals expanded)

275

1H NMR spectra – new pigment, aromatic region (expanded)

HSQC – unknown pigment

1H NMR spectra of new pigment (aliphatic region expanded)

276

HSQC – new pigment

HMBC – new pigment

277

Selective HMBC – new pigment

COSY - new pigment 278

NOESY – new pigment

NH – HMBC – new pigment

279

1 H NMR – new pigment (in CD3OD) – Deuterium exchange experiment

280

1H NMR - Scytonemin Oxidized

HSQC – Scytonemin oxidized

281

HMBC – Scytonemin oxidized

Selective HMBC – Scytonemin oxidized

282

COSY – Scytonemin oxidized

283

XI – Mass Spectra

MASS SPECTRA (experiments used to characterize the new pigment)

HR MS-ESI-TOF: Unknown pigment m/z 602.2074[M+H]+; m/z 624.1899 [M+Na]+

284

HR –MS-ESI: Unknown pigment m/z 1226.3617 [2M+Na] +

285

HR-MS-ESI formula results for the new pigment.

286

LC-MS for the new pigment m/z 602 [M+H] +

287

MALDI –TOF MS for the new pigment m/z 602 [M+H] +

288

LC-MS of acetylated new pigment: m/z 686 [M+H] + (likely from the acetylation of the two OH moieties of the phenols) and 728 [M+H] + (possible contaminant).

289

SEQ-17095-01 8/16/2011 12:27:32 PM CIDYA; 5 uL Hypurity C8;0.25;95:5(5)>5:95(45-60)/254 nm/(-)ESI RT: 2.23 - 60.16 RT: 38.99 NL: 4.82E6 BP: 600.5 Base Peak F: - c 100 ESI sid=1.00 Full MW 601-B ms [ 80 125.00-700.00] MS SEQ-17095-01 60

40 MW 601-A

20 R elative Abundance R elative

40.79 NL: 1.22E6 726.3 Base Peak F: - c 100 ESI sid=3.00 Full ms [ 80 690.00-1800.00] MS SEQ-17095-01 60 MW 727 isomers 40

20 Relative Abundance Relative

RT: 38.99 NL: 4.82E6 BP: 600.5 m/z= 600.0-601.0 F: 100 - c ESI sid=1.00 Full ms [ 80 125.00-700.00] MS SEQ-17095-01 60

40 RT: 38.50 20 BP: 600.5 Relative Abundance Relative

0 40.79 NL: 1.22E6 726.3 m/z= 725.8-726.8 F: 100 - c ESI sid=3.00 Full ms [ 80 690.00-1800.00] MS SEQ-17095-01 60

20 Relative Abundance Relative

0 RT: 38.99 NL: 7.27E-2 BP: 0.0 UV Analog SEQ-17095-01 38.91 0.070 0.0

0.065

In te n s ity RT: 38.34 48.81 49.67 0.060 41.86 44.02 44.71 47.33 51.93 54.78 59.07 59.37 30.61 32.07 34.31 35.43 BP: 0.0 0.0 0.0 20.61 21.83 24.02 25.89 27.05 29.68 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 4.24 7.09 8.47 9.70 11.37 13.34 14.15 16.49 19.39 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.055 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0

4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50 52 54 56 58 60 Tim e (min)

C8 HPLC/UV/(-)ESI-MSn. There were at least two MW 601 isomers as indicated by the labels on the shaded peaks. There were also at least two MW 727 isomers. (not shaded).

290

SEQ-17095-01 8/16/2011 12:27:32 PM CIDYA; 5 uL Hypurity C8;0.25;95:5(5)>5:95(45-60)/254 nm/(-)ESI

SEQ-17095-01 #1441-1450 RT: 38.30-38.50 AV: 2 NL: 7.35E5 T: - c ESI sid=1.00 Full ms [ 125.00-700.00] 600.5 100

60 601.5

Relative Abundance Relative 30

20 602.4

10 141.0 603.5 622.5 682.1 0 140 160 180 200 220 240 260 280 300 320 340 360 380 400 420 440 460 480 500 520 540 560 580 600 620 640 660 680 700 m/z

SEQ-17095-01 #1441-1450 RT: 38.31-38.52 AV: 2 NL: 7.36E5 T: - c sid=1.00 d Full ms2 [email protected] [ 185.00-1215.00] 544.5 100

40 545.5 Relative Abundance Relative 30 601.5 20 602.4 10 600.5 286.6 453.3 462.8546.4 572.5 0 140 160 180 200 220 240 260 280 300 320 340 360 380 400 420 440 460 480 500 520 540 560 580 600 m/z

SEQ-17095-01 #1439-1449 RT: 38.21-38.35 AV: 2 NL: 1.94E5 T: - c sid=1.00 d Full ms3 [email protected] [email protected] [ 135.00-1095.00] 544.4 100

90 545.4 80

40 515.7 543.6

Relative Abundance Relative 30 499.7 20 516.6

10 500.7 546.4 407.8 423.6 487.7 395.7 451.9 517.6 541.7 0 140 160 180 200 220 240 260 280 300 320 340 360 380 400 420 440 460 480 500 520 540 560 580 600 m/z

MW 601-A: The MW 601-A produced an m/z 600 [M-H]- ion (top) which was dissociated to form m/z 544 via loss of 56 u (middle). The m/z 544 was relatively resistance to dissociation but did produce some m/z 515/516 and 499/500 product ions (bottom)

291

50.80 51.63 40 386.7 49.12 684.2 54.32 56.33 38.50 237.1 832.1 59.19 R elative Abundance R elative 48.27 906.0 20 602.2 257.2 980.1

RT: 39.00 NL: 5.26E6 BP: 602.3 m/z= 601.7-602.7 F: 100 + c ESI sid=1.00 Full ms [ 80 125.00-1000.00] MS SEQ-17095-02

40 RT: 38.50 Relative Abundance Relative 20 BP: 602.2

0 40.90 NL: 3.42E5 728.1 m/z= 727.6-728.6 F: 100 + c ESI sid=1.00 Full ms [ 80 40.55 125.00-1000.00] 728.1 MS SEQ-17095-02 60

40 59.02 54.32 58.51 728.3 53.49 728.4 728.1 Relative Abundance Relative 39.84 20 46.40 51.63 727.8 728.1 727.7 727.6

0 RT: 39.03 NL: 7.08E-2 BP: 0.0 UV Analog 0.070 SEQ-17095-02

0.065 1.87 0.0

0.060 RT: 38.36 46.18 46.95 49.07 49.96 52.24 53.93 56.27 57.25

In te n s ity 41.91 43.59 60.00 30.72 31.81 34.20 35.97 BP: 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 19.19 20.38 22.70 24.90 25.69 29.57 0.0 0.0 0.0 1.11 2.70 4.69 7.00 9.20 9.48 9.86 12.54 15.10 18.36 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.055 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0

0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50 52 54 56 58 60 Time (min)

HPLC/UV/(+)ESI-MSn analysis. The MW 601 compounds are shaded

292

SEQ-17095-02 8/16/2011 2:03:56 PM CIDYA; 5 uL Hypurity C8;0.25;95:5(5)>5:95(45-60)/254 nm/(+)ESI

SEQ-17095-02 #1337-1357 RT: 38.85-39.31 AV: 4 NL: 3.26E6 T: + c ESI sid=1.00 Full ms [ 125.00-1000.00] 602.2 100

40 603.3 546.3

Relative Abundance Relative 30

20 545.3 604.2 10 547.3 149.2 159.0 186.9218.7 239.1 264.2 300.8324.7 365.1 375.0 413.4 426.2 455.2 473.0 494.6 517.4 548.4 601.3 624.3 646.4 660.1 702.6 722.3 748.9 782.6 813.1 828.9 857.5 892.4 928.9 958.1 994.1 0 150 200 250 300 350 400 450 500 550 600 650 700 750 800 850 900 950 1000 m/z

SEQ-17095-02 #1336-1354 RT: 38.86-39.16 AV: 3 NL: 5.64E6 T: + c sid=1.00 d Full ms2 [email protected] [ 185.00-1215.00] 545.6 100

50 546.4

Relative Abundance Relative 30

10 556.4 545.0 574.3 585.3 0 140 160 180 200 220 240 260 280 300 320 340 360 380 400 420 440 460 480 500 520 540 560 580 600 m/z

SEQ-17095-02 #1336-1354 RT: 38.90-39.19 AV: 3 NL: 1.76E6 T: + c sid=1.00 d Full ms3 [email protected] [email protected] [ 140.00-1100.00] 517.5 100

70 518.6 528.3 60

50 489.8

Relative Abundance Relative 30 490.6 529.4 20 491.3 546.3 543.5 10 547.2 425.7 493.6 515.9 397.1 400.8 411.7451.7 472.9 488.7 530.5 0 140 160 180 200 220 240 260 280 300 320 340 360 380 400 420 440 460 480 500 520 540 560 580 600 m/z

(+)ESI-MS produced m/z 602 [M+H]+, m/z 624 [M+Na]+ and m/z 546 [M+H-56 u] fragment ion (top). (+)ESI-MS/MS dissociation of m/z 602 produced m/z 545 and 546 (middle) which were further dissociated t produce m/z 528, 518, 517 and 489

293

SEQ-17095-02 8/16/2011 2:03:56 PM CIDYA; 5 uL Hypurity C8;0.25;95:5(5)>5:95(45-60)/254 nm/(+)ESI

SEQ-17095-02 # 1397 RT: 40.57 AV: 1 NL: 3.38E5 T: + c sid=1.00 d Full ms2 [email protected] [ 230.00-1470.00] 671.1 10 0

60 672.2 50

R Abundance elative 30

20 682.2 10 440.9 4 55.1 465.4 4 9 5.5 555.7 572 .1 636.0 687.0 700.1 710.6 0 240 260 280 300 320 340 360 380 400 420 440 460 480 50 0 52 0 54 0 56 0 58 0 600 620 640 660 680 70 0 72 0 740 m/ z

SEQ-17095-02 # 1396 RT: 40.60 AV: 1 NL: 9.79E4 T: + c sid=1.00 d Full ms3 [email protected] [email protected] [ 170.00-1350.00] 643.2 10 0

70 644.2

50 544.2

R Abundance elative 30 615.2 516.5 642.3 654.1 20 616.3 488.4 515.6 54 5.2 10 543.4 626.2 6 55.2 423.3 451.7 474.3 499.4 526.4 551.3 577.7 597.0614.6 641.2 669.0 671.4 314.4 344.2383.4 388.0 399.2 409.3 426.6 454.3 514.4 562.3 580.2 0 240 260 280 300 320 340 360 380 400 420 440 460 480 50 0 52 0 54 0 56 0 58 0 600 620 640 660 680 70 0 72 0 740 m/ z

SEQ-17095-02 # 1399 RT: 40.57 AV: 1 NL: 3.38E5 T: + c sid=1.00 d Full ms2 [email protected] [ 230.00-1470.00] 671.1 10 0

60 672.2 50

R Abundance elative 30

20 682.2 10 440.9 4 55.1 465.4 4 9 5.5 555.7 572 .1 636.0 687.0 700.1 710.6 0 240 260 280 300 320 340 360 380 400 420 440 460 480 50 0 52 0 54 0 56 0 58 0 600 620 640 660 680 70 0 72 0 740 m/ z

SEQ-17095-02 # 1409 RT: 40.96 AV: 1 NL: 9.31E4 T: + c sid=1.00 d Full ms3 [email protected] [email protected] [ 170.00-1350.00] 643.3 10 0 615.2 90

70 644.2 654.0 60

50 642.3 40

R Abundance elative 30 655.1 616.2 20 544.2 614.4 10 515.8 626.4 488.5 526.0 669.0 371.1 397.3 407.5 4 11. 5 425.2 440.3 451.4471.5 499.9 536.8 544.9 551.1 576 .9 579 .1 596.7 625.2 641.0 675.0 0 240 260 280 300 320 340 360 380 400 420 440 460 480 50 0 52 0 54 0 56 0 58 0 600 620 640 660 680 70 0 72 0 740 m/ z

MW 727: (+)ESI-MS/MS (top and 3rd) and –MS/MS/MS (2nd and bottom) of the MW 727 isomers appeared very similar but (+)ESI-MS/MS/MS mass chromatograms show some differences (not shown)

294

VIII. REFERENCES

Alderkamp, A., Nejstgaard, J. C., Verity, P. G., Zirbel, L. M. J., Sazhin, A. F.,& van Rijssel, M. (2006). Dynamics in carbohydrate composition of Phaeocystis pouchetti colonies during spring blooms in mesocosms. Journal of Sea Research, 55, 169–181.

Alldredge, A. L., Cole, J. J., & Caron, D. A. (1986). Production of heterotrophic bacteria inhabiting macroscopic organic aggregates (marine snow) from surface waters. Limnology and Oceanography, 31, 68–78.

American Public Health Association (APHA) 1998: Standard Methods for the Examination of Water and Waste Water, method 10300 C, D. 20th ed. Washington DC.

American Public Health Association (APHA) 1991: Standard Methods for the examination of Water and Waste Water, 17th ed. Washington DC; APHA.

Aminot, A. 2000. Standard procedure for the determination of chlorophyll a by spectroscopic methods. International council for the exploration of the sea. Denmark. ISSN 0903-2606.

Andersen, R. A., Bidigare, R. R., Keller, M. D., & Latasa, M. (1996). A comparison of HPLC pigment signatures and electron microscopic observations for oligotrophic waters of North Atlantic and Pacific Ocean. Deep Sea Research II, 43, 517–537.

Anderson, D. M., & Lobel, P. S. (1987). The continuing enigma of ciguatera. Biology Bulletin, 172, 89–107.

Ansotegui, A., Trigueros, J. M., & Orive, E. (2001). The use of pigment signatures to assess phytoplankton assemblage and structure in estuarine waters. Estuarine, Coastal and Shelf Science, 52, 689–703.

Antoine, D., Andre`, J. M., & Morel, A. (1996). Oceanic primary production. 2. Estimation at global scale from satellite (coastal zone color scanner) chlorophyll. Global Biogeochemical Cycles, 10, 57–59.

Archibald, A. R., Hirst, E. L., Manners, D. J., & Ryley, J. F. (1960). Studies on the metabolism of the protozoa. VII. The molecular structure of a starch-type 295

polysaccharide from Chilomonas paramecium. Journal of Chemical Society, 556– 560.

Aspinall, G. O., 1983. Polysaccharides. Academic Press, New York.

Barlow, R. G., Mantura, R. F. C., Peinert, R. D., Miller, A. E. J., & Fileman, T. W. (1995). Distribution, sedimentation and fate of pigment biomarkers, following thermal stratification in western Alboran Sea. Marine Ecology Progress Series, 70, 173–198.

Beardall, J., Roberts, S., & Millhouse, J. (1991). Effects of nitrogen limitation on uptake of inorganic carbon and specific activity of ribulose-1,5-bisphosphate carboxylase/oxygenase in green microalgae. Canadian Journal of Botany, 69, 1146–1150.

Berges, J. A., Fisher, A. E., & Harrison, P. J. (1993). A comparison of Lowry, Bradford and Smith protein assay using different protein standards and protein isolated from the marine diatom Thalassiosira pseudonana. Marine Biology, 115, 187– 193.

Bigham, D. L., Hoyer, M. V., & Canfield Jr., D. E. (2009). Survey of toxic algal (microcystin) distribution in Florida Lakes. Lake and Reservoir Management, 25, 264–275.

Borsheim, K. Y., Vadstein, O., Myklestad, S. M., Reinertsen, H., Kirkvold, S., & Olsen, Y. (2005). Photosynthetic algal production, accumulation and release of phytoplankton storage carbohydrates and bacterial production in a gradient in a daily supply. Journal of Plankton Research, 27, 743–755.

Bourne, C. E. M., Palmer, J. D., & Stoermer, E. F. (1992). Organization of the chloroplast genome of the freshwater centric diatom Cyclotella meneghiniana. Journal of Phycology, 28, 347–355.

Boyce, D., Lewis, M., & Worm, B. (2010). Global phytoplankton decline over the past century. Nature, 466 (7306), 591–596.

Bradford, M. (1976). A rapid and sensitive method for quantitation of microgram quantities of protein utilizing the principle of protein dye-binding. Analytical Biochemistry, 72, 248–254.

Bratbak, G. (1985). Bacterial biovolume and biomass estimation. Applied Environmental Microbiology, 49, 1488–1493.

296

Britton, G. (1995). Structure and properties of carotenoids in relation to function. Journal of the Federation of American Societies for Experimental Biology (FASEB), 9, 1551–1558.

Browne, J. L. (2010). Comparison of chemotaxonomic methods for the determination of periphyton community composition. M.S. Thesis, Florida Atlantic University.

Brown, M. R., & Jeffrey, S. W. (1992). Biochemical composition of microalgae from the green algal classes Chlorophyceae and Prasinophyceae. 1. Amino acids, sugars, and pigments. Journal of Experimental Marine Biology and Ecology, 161, 91– 113.

Burdloff, D., Etcheber, H., & Buscail, R. (2001). Improved procedures for the extraction of water extractable carbohydrates from particulate organic matter. Oceanologica Acta, 24, 343–347.

Bursa, A. S. (1968). Starch in the Oceans. Journal of Fisheries and Research. Bd. Canada, 25, 1269–1284.

Butel-Ponce`, V., Felix-Theodose, F., Sarthou, C., Ponge, J. F., & Bodo, B. (2004). New Pigments from the terrestrial cyanobacterium Scytonema sp. collected on the Mitaraka Inselberg, French Guyana. Journal of Natural Products, 67, 678–681.

Carpentier, C. J., Ketelaars, H. A. M., Wagenvoort, A. J., & Pikoor-Schoonen, K. P. R. (1999). Rapid and versatile measurements of phytoplankton biovolume with BACCHUS. Journal of Phytoplankton Research, 21, 1877–1889.

Chapman, D. J. (1966). Three new carotenoids isolated from algae. Phytochemistry, 5, 1331–1333.

Charles, M. J. & Simmons, M. S. (1986). Methods for the determination of carbon in soils and sediments: A review. Analyst, 111, 385–390.

Chiovitti, A., Molino, P., Crawford, S. A., Ten, R., Spurek, T., & Wetherbee, R. (2004). The glucans extracted with warm water are mainly derived from intracellular chrysolaminaran and not extracellular polysaccharides. European Journal of Phycology, 39, 117–128.

Clayton, J.R., Jr., Dortch, Q., Thorensen, S., & Ahmed, S. I. (1988). Evaluation of methods for the separation and analysis proteins and free amino acids in phytoplankton samples. Journal of Plankton Research, 10, 341–358. Craige, J. S. (1974). In: Algal Physiology and Biochemistry – Botanical Monographs. Chapter 7, Volume 10. Stewart, W. D. P. (ed). University of California press.

297

Cuhel, R. L., Ortner, P. B., & Lean, D. R. S. (1984). Night synthesis of protein by algae. Limnology and Oceanography, 29, 731–744.

Cullen, J. J. (1982). The deep chlorophyll maximum: Comparing vertical profiles of chlorophyll a. Canadian Journal of Fisheries and Aquatic Science, 39, 791–803.

Daley, R. J., & Hobbie, J. E. (1975). Direct counts of aquatic bacteria by a modified epifluorecsence technique. Limnology and Oceanography, 20, 875–882.

De Phillipis, R., Margheri, M. C., Sili, C., & Vincenzini, M. (1995). Cyanobacteria: a promising group of exopolysaccharide producers. Proceedings of the 2nd European Workshop ‘Biotechnology of microalgae’ IGV Institut fur Getreidevererbeitung GmbH, Bergholz-rehbrucke, 78–81.

Decho, A. W. (1990). Microbial exopolymer secretions in ocean environments: their role(s) in food webs and marine processes. Oceanographic Marine Biology, Annual Review 28, 73–153.

Demmig-Adams, B. (1990). Carotenoids and photoprotection in plants: A role for xanthophyll zeaxanthin. Biochimica et Biophysica Acta (BBA) – Bioenergetics, 1020, 1–24.

Dillon, J. G., & Castenholz, R. W. (1999). Scytonemin, a cyanobacterial sheath pigment, protects against UVC radiation: implications for early photosynthetic life. Journal of Phycology, 35, 673–681.

Dortch, Q. (1982). Effects of growth conditions on accumulation of internal pools of nitrate, ammonium, amino acids and protein in three marine diatoms. Journal of Experimental Marine Biology and Ecology, 61, 243–264.

Dortch, Q., Clayton, J.R., Jr., Thorensen, S. S., & Ahmed, S. I. (1984). Species differences in accumulation of nitrogen pools in phytoplankton. Marine Biology, 81, 237–250.

Dubois, M., Gilles, K. A., Hamilton, J. K., P. A., & Rebers., F. Smith. (1956). Colorimetric method for determination of sugars and related substances. Analytical Chemistry, 28, 350–356.

Eby, G. N. (2004). Principles of Environmental Geochemistry. Thompson Brooks-Cole, Pacific Grove, California pp 514.

Edwards, M., Beaugrand, G., Reid, P. C., Rowden, A. A., Jones, & M. B. (2002). Ocean climate anomalies and the ecology of the North Sea. Marine Ecology Progress Series, 239, 1–10.

298

Fabiano, M.,& Danovaro, R. (1994). Composition of organic matter in sediments facing a river estuary (Tyrrhenian Sea): relationships with bacteria and microphytobenthic biomass. Hydrobiologia, 277, 71–84.

Falkowski, P. G. (1994). The role of phytoplankton photosynthesis in global biogeochemical cycles. Photosynthesis Research, 39, 235–258.

Falkowski, P. G., Fujita, Y., Ley, A., & Mauzerall, D. (1986). Evidence for cyclic electron flow around photosyntem II in Chlorella pyrenoidosa. Plant Physiology, 81, 310–312.

Falkowski, P. G., Oweds, T. G., Arthur, C. L., & Mauzerall, D. C. (1981). Effects of growth irradiance levels on the ratio of reaction centers in two species of marine phytoplankton. Plant Physiology, 68, 969–973.

Falkowski, P. G., & Owens, T. G. (1980). Light-shade adaptation: Two strategies in marine phytoplankton. Plant Physiology, 66, 592–595.

Fernandez, E., Serret, P., Demadariaga, I., Harbour, D. S., & Davies, A. G. (1992). Photosynthetic carbon metabolism and biochemical composition of spring phytoplankton assemblages enclosed in mesocosms: The diatom Phaeocystis sp. succession. Marine Ecology Progress Series, 90, 89–102.

Finlay, B. J., Monaghan, E. B., & Maberly, S. C. (2002). Hypothesis: the rate and scale of dispersal of freshwater diatoms species is a function of their global abundance. Protist, 153, 261–273.

Fleming, E. D., & Castenholz, R. W. (2007). Effects of periodic dessication on the synthesis of the UV-screening compound, scytonemin, in cyanobacteria. Environmental Microbiology, 9(6), 1448–1455.

Fogg, G. E. (1952). The production of extracellular nitrogenous substances by a blue- green a blue-green alga. Proceedings of the Royal Society: Biological Sciences 139, 372–397.

Furhop, J. H., & Smith, K. M. (1975). Laboratory Methods. In: Smith, K. M. (ed.), Porphyrins and Metalloporphyrins. Elsevier, Amsterdam pp. 757–789.

Garcia-Pichel, F., & Castenholz, R. W. 1991. Characterization and biological implications of scytonemin, a cyanobacterial sheath pigment. Journal of Phycology, 27, 395–409.

299

Garcia-Pichel, F., & Castenholz, R. W. (1993). Occurrence of UV-absorbing Mycosporine-like compounds among cyanobacterial isolates and an estimate of their screening capacity. Applied Environmental Microbiology, 59, 163–169.

Garcia-Pichel, F., Sherry, N. D.,& Castenholz, R. W. (1992). Evidence for an ultraviolet sunscreen role of extracellular pigment scytonemin in the terrestrial cyanobacterium Chlorogloeopsis sp. Photochemical Photobiology, 56, 17–23.

Garcia-Pichel, F., Wingard, C. E., & Castenholz, R. W. 1993. Evidence regarding the UV sunscreen role of a mycosporine-like compound in the cyanobacterium Gloeocapsa sp. Applied Environmental Microbiology, 59, 170–176.

Geesey, G. G. (1982). Microbial exoploymers: ecological considerations. Applied Society of Microbiology. 48, 9–14.

Geider, R. J. (1987). Light and Temperature dependence of the carbon to chlorophyll a ratio in microalgae and cyanobacteria: Implications for physiology and growth of phytoplankton. New Phytology, 106, 1–34.

Geider, R. J. (1984). Light and nutrient effects on microbial physiology. Ph. D. Dissertation, Dalhousie University. Halifax, Nova Scotia.

Geider, R. J., LaRoche, J., Greene, R. M., & Olaizola, M. (1993). Response of photosynthetic apparatus of Phaeodactylum tricornutum (Bacillariophyceae) to nitrate, phosphate, or iron starvation. Journal of Phycology, 29, 755–766.

Gieskes, W. W. C., & Kraay, G. W. (1983). Dominance of Cryptophyceae during the phytoplankton spring bloom in the central North Sea by HPLC analysis of pigments. Marine Biology, 75, 179–185.

Gieskes, W. W. C., Kraay, G. W., Nontji, A., Septiapermana, D., & Sutomo. (1988). Monsoonal alteration of a mixed and layered structure in the phytoplankton of the euphotic zone of the Banda Sea (Indonesia): A mathematical analysis of algal pigment fingerprints. Netherland Journal of Sea Research, 22, 123–137.

Goericke, R., & Montoya, J. P. (1998). Estimating the contribution of microalgal taxa to chlorophyll a in the field – variations of pigment ratios under nutrient-and-light- limited growth. Marine Ecology Progress Series, 169, 97–112.

Goericke, R., & Welschmeyer, N. A. (1992). Pigment turnover in the diatom Thalassiosira weissflogii: II. The CO2-labelling kinetics of carotenoids. Journal of Phycology, 28, 507–517.

300

Goto, N., Mitamura, O., & Terai, H. (2001). Biodegradation of photosynthetically produced extracellular organic carbon from intertidal benthic algae. Journal of Experimental Marine Biology and Ecology, 257, 73–86.

Gouveia, L & Oliveira, A. (2009). Microalgae as raw material for biofuel production. Journal of Industrial Microbiology and Biotechnology, 36, 269–274.

Govindjee and Braun, B. Z. (1974). Botanical Monographs. Stewart, W. D. (ed). In: Algal Physiology and Biochemistry .Chapter 12, Volume 10. University of California press.

Grant, C. S., & Louda, J. W. (2010). Microalgal pigment ratios in relation to light intensity: implications for chemotaxonomy. Aquatic Biology, 11, 127–138.

Granum, E. & Myklestad, S. M. (2001). Mobilization of β-1-3-glucan and biosynthesis of amino acids induced by NH4+ addition to N-limited cells of the marine diatom Skeletonema costatum (Bacillariophyceae). Journal of Phycology, 37, 772–782.

Granum, E., & Myklestad, S.M., (2002). A simple combined method for determination of β-1,3 glucan and cell wall polysaccharides in diatoms. Hydrobiologia, 477, 155– 161.

Granum, E., Kirkvold, S., & Myklestad, S. M. (2002). Cellular and extracellular production of the carbohydrates and amino acids by the marine diatom Skeletonema costatum: diel variations and effects of N depletion. Marine Ecology Progress Series, 242, 83–94.

Guillard, R. R.L. (1975). Culture of phytoplankton for feeding marine invertebrates. In: Smith, W.L., and Chantley, M. H. (eds). Culture of Marine Invertebrate Animals. Plenum Press, New York, pp 26-60.

Hach, C. C., Bowden, B. K., Kopelove, A. B., & Brayton, S. T. (1987) More powerful peroxide Kjeldahl digestion method. Journal Association of Analytical Chemistry, 70, 783–787.

Hagar, A. (1980). The reversible, light-induced conversions of xanthophylls in the chloroplast. In: Czygan F. C. (ed.) Pigments in Plants, 2nd Edition. Gustav-Fisher, Stuttgart. pp. 57-80.

Hagerthey, S. E., Louda, J. W., Mongkhonsri, P. (2006). Evaluation of pigment extraction methods and recommended protocol for periphyton chlorophyll a determination and chemotaxonomic assessment. Journal of Applied Phycology, 42, 1125-1136.

301

Hallegraeff, G. M. (1993). A review of harmful algal blooms and their apparent global increase. Phycologia, 32, 79–99.

Hallegraeff, G. M. (2010). Ocean climate change, phytoplankton community responses and harmful algal blooms: a formidable challenge. Journal of Phycology, 2, 220– 235.

Hambrook Berkman, J.A., & Canova, M.G. (2007). Algal biomass indicators (ver. 1.0): U.S. Geological Survey Techniques of Water-Resources Investigations, book 9, chap. A7, section 7.4, August, available online only from http://pubs.water.usgs.gov/twri9A/.

Handa, N. (1969). Carbohydrate metabolism in marine diatom Skeletonema costatum. Marine Biology, 4, 208–214.

Hansell, D. A., & Carlson, C. A. (2001). Marine Dissolved Organic Matter and the Carbon Cycle. Oceanography, 14, 41–49.

Harrison, P. J., & Berges, J. A. (2005). Marine Culture Media. In: Andersen, R. A. (ed). Algal Culturing Techniques. Chapter 3. Elsevier Academic Press.

Havskum, H., Schulter, L., Scharek, R., Berdalet, E., & Jacquet, S. (2004). Routine quantification of phytoplankton groups – microscopy or pigment analyses? Marine Ecology Progress Series, 273, 31–42.

Hedges, J. I., Cowie, G. L., Richey, J. E., Quay, P. D., Benner, R., Strom, M., & Forsberg, B. R. (1994). Origins and processing of organic matter in the Amazon River as indicated by carbohydrates and amino acids. Limnology and Oceanography, 39, 743–761.

Heldt, H. W., Chon, C. J., Maronade, D., Herald, A., Stankovic, Z. S., Walker, D. A., Kraminer, A., Kirk, M. R., & Heber, U. (1977). Role of orthophosphate and other factors in the regulation of starch formation in leaves and isolated chloroplasts. Plant Physiology, 89, 1146–1155.

Higgins, H. W., & Mackey, D. J. (2000). Algal class abundance, estimated from chlorophyll and carotenoid pigments in the western Equatorial Pacific under El Nino and non- El Nino conditions. Deep- Sea Research,I, 47, 1461–1483.

Hillebrand, H., Durselen, C. D., Kirschtel, D., Pollingher, D. & Zohary, T. (1999). Biovolume calculation for pelagic and benthic microalgae. Journal of Phycology, 35, 403–424

302

Hou, J., Huang, B., Cao, Z., Chen, J., & Hong, H. (2007). Effects of nutrient limitation on pigments in Thalassiosira weissfligii and Prorocentrum donghaiense. Journal of Integrative Plant Biology, 49, 686–697.

Hough, L., Jones, J. K. M., & Wadman, W. H. (1952). An investigation of the polysaccharide components of certain fresh-water algae. Journal of the Chemical Society, 3393–3399.

Hulbert, E. M. (1957). The taxonomy of unarmoured dinophyceae of shallow embayments on Cape Cod, Massachusetts. Biology Bulletin, 112, 196–219. Itzhaki, R. F., & Gill, P. M. (1964). A microbiuret method for estimating proteins. Analytical Biochemistry, 9, 401–410.

Janse, I., van Rijssel, M., van Hall, P. J., Gerwig, G. J., Gottschal, J. C., & Prins, R. A. (1996b) The storage glucan of Phaeocystis globosa (prymnesiophyceae) cells. Journal of Phycology, 32, 382–387.

Jayappriyan, K. R., Rajkumar, R., Sheeja, L., Nagaraj, S., Divya, S., & Rengasamy, R. (2010). Discrimination between the morphological and molecular identification in the genus Dunaliella. International Journal of Current Research, 8, 73–78.

Jeffrey, S. W., & Humphrey, G. R. (1975). New Spectrophotometric equations for determining chlorophylls a, b, c1 and c2 in higher plants, algae, and natural phytoplankton. Biochemical Physiology, Pflanzen Bd, 167,191–194.

Jeffrey, S. W., & Vesk, M. (1997). Introduction to marine phytoplankton and their pigment signatures. In: Jeffrey, S. W., Mantoura, R. F. C., and Wright, S. W. (eds). Phytoplankon Pigments in Oceanography. UNESCO publishing, pp 74-75.

Jeffrey, S. W., & Vesk, M., (1997). Introduction to marine phytoplankton and their pigment signatures, In: Jeffrey, S.W., Mantoura, R.F.C. & Wright, S.W. (eds). Phytoplankton Pigments in Oceanography SCOR-UNESCO, Paris, pp 34-87.

Johansen, J. R., & Theriot, E. (1987). The relationship between valve diameter and number of central fultoportulae in Thalassiosira weissflogii (Bacillariophyceae). Journal of Phycology, 23, 663–665.

Johnsen, G. N., & Sakshaug. (1993). Bio-optical characteristics and photoadaptive responses in the toxic and bloom forming dinoflagellates Gymnodinium aureolum, Gymnodnium galatheanum and two strains of Prorocentrum minimum. Journal of Phycology, 29, 627–642.

303

Johnson, P. W., & Sieburth, J. (1982) In-situ morphology and occurrence of eukaryotic phototrophs of bacterial size in the picoplankton of estuarine oceanic waters. Journal of Phycology, 18, 318–327.

Klein, M. P., Sauer, K., & Yachandra, Y. K. (1993). Perspectives on the structure of the photosynthetic oxygen evolving manganese complex and its relation to the Kok’s cycle. Photosynthetic Research, 38, 265–277.

Kok, B., & Businger, J. A., (1956). Kinetics of Photosynthesis and Photoinhibition. Nature, 177, 135-136.

Kok, B., Forbush, B., & McGloin, M. (1970). Cooperation of charges in photosynthetic oxygen evolution – I. A linear four step mechanism. Photochemistry and Photobiology, 11(6), 457–475.

Krauus, R. W., & Thomas, W. H. (1954). The growth and inorganic nutrition of Scenedesmus obliquus in mass culture. Plant Physiology, 29(3), 205–214.

Krinsky, N. I. (1971) In: Isler, O. (ed). Carotenoids. Birkhauser-Verlag, Basel pp. 669- 716.

Lancelot, C., & Mathot, S. (1985). Biochemical fractionation of primary production by phytoplankton in Belgian coastal waters during short-and-long term incubations with 14C-bicarbonate. II Phaeocystis pouchetti colonial population. Marine Biology, 86, 227–232.

Lentner M. (1993). Experimental Design and Analysis. Valley Book Company, Blacksburg, VA.

Letelier, R. M., Bridigare, R. R., Hebel, D. V., Ondrusek, M., Winn, C. D., & Karl, D. M. 1993. Temporal variability of phytoplankton community structure based on pigment analysis. Limnology and Oceanography, 38, 1420–1437.

Lewin, J. C. (1955). The capsule of the diatom Navicula pelliculosa. Journal of General Microbiology, 13, 162–169.

Lewin, R.A. (1978). Biochemistry and Physiology of Algae: taxonomic and phylogenetic considerations. In: Jackson, D. F. (ed). Algae, Man and the Environment, pp 15- 26. Syracause University Press.

Lewin, R. A. (1956). Extracellular polysaccharides of green algae. Canadian Journal of Microbiology, 2, 665-672.

304

Llewelyn, C. A., & Gibb, S. W. (2000). Intra-class variability in carbon, pigment and biomineral content of prymnesiophytes and diatoms. Marine Ecology Progress Series, 193, 33–44.

Louda, J. W. (2008). Pigment-based chemotaxonomy of Florida Bay phytoplankton; Development and difficulties. Journal of Liquid Chromatography and Related Technologies, 31, 295–323.

Louda, J. W., & Mongkhonsri, P. (2004). Comparison of spectrophotometric and HPLC estimations of chlorophylls –a, -b, -c and phaeopigments in Florida Bay Seston. Florida Scientist, 67 (4), 281–292.

Loftus, M. E., & Carpenter, J. H. (1971). A fluorometric method for determining chlorophylls a, b, c. Journal of Marine Research, 29, 319–338.

Lourenco, S.O., Barbarino, E., Lanfer Marquez, U. M., & Aidar, E. (1998). Distribution of intracellular nitrogen in marine microalgae: basis for the calculation of specific nitrogen-to-protein conversion factors. Journal of Phycology, 34, 798–811.

Lorenzen, C. J. (1967). Determination of chlorophyll and phaeo-pigments: Spectrophotometric equations. Limnology and Oceanography, 12, 343–346.

Lowry, O. H., Rosebrough, N. J., Farr, A. L., & Randall, R. L. (1951). Protein measurement with the folin phenol reagent. Journal Biological Chemistry, 193, 265–275.

MacIntyre, H. L & Cullen, J.J. (2005). Using cultures to investigate the physiological ecology of microalgae. In: Anderson, R.A. (ed). Algal culturing techniques. Chapter 19. Elsevier Academic Press.

MacIntyre, H. L., Kans, T. M., Anning, T., & Geider, R. J. (2002). Photoacclimation of photosynthesis irradiance response curves and photosynthetic pigments in microalgae and cyanobacteria. Journal of Phycology, 38, 17–38.

Mackey, D. J., Higgins, H. W., Mackey, M. D., & Holdsworth, D. (1998). Algal class abundances in the western equatorial pacific: estimation from HPLC measurements of chloroplast pigments using CHEMTAX. Deep Sea Research, 45, 1441–1468.

Mackey, M. D., Mackey, D. J., Higgins, H. W., & Wright, S. W. (1996). CHEMTAX – A program for estimating class abundances from chemical markers: application to HPLC measurements of phytoplankton. Marine Ecology Progress Series, 144, 265–283.

305

Mantoura, R. F. C., Lewelyn, C. A. 1983. The rapid determination of algal chlorophyll and carotenoid pigments and their breakdown products in natural waters by reversed phase High Performance Liquid Chromatography. Analytica Chimica Acta, 151, 297–314.

Marinov, I., Doney, S. C., & Lima, I. D. (2010). Response of ocean phytoplankton community structure to climate change over the 21st century: partitioning the effects of nutrients, temperature and light. Biogeosciences Discussions, 7, 4565– 4606.

McCormick, P. V., Newman, S., Miao, S., Gawlik, D. E., Marley, D., Reddy, K. R., & Fontaine, T. P. (2001). Effects of anthropogenic phosphorus inputs on the Everglades. In: Porter, J. W., and Porter, K. W. (eds). The Everglades, Florida Bay, and Coral Reefs of the Florida Keys. An Ecosystem sourcebook. Boca Raton, FL. pp 83-126.

McLafferty, F. W. (1980). Interpretation of Mass Spectra. In: Turro, N. J. (ed) Organic Chemistry Series. University Science Books. Mill Valley California.

McLaughlin, J. J. A., Zahl, P. A., Novak, A., Marchisotta, J., & Prager, J. (1960). Mass cultivation of some phytoplankton. Annals of the New York Academy of Science, 90, 856–865.

Mebius, L.J. (1960). A rapid method for the determination of organic carbon in soil. Analytical Chimica Acta, 22, 120–124.

Meeuse, B. J. D., & Smith, B. N. 1962. A note on the amylolytic breakdown of some raw algal starches. Planta, 57, 624–635.

Menzel, D. W., & Corwin, N. (1967). The measurement of total phosphorus in sea water based on the liberation of organically bound fractions by persulfate oxidation. Limnology and Oceanography, 10, 280–282.

Meyers, J., & Kratz, W. A. (1955). Relations between pigment content and photosynthetic characteristics in a blue-green alga. Journal of General Physiology, 39, 11–12.

Miller, J. N., & Miller, J. C. 2005. Statistics and Chemometrics for Analytical Chemistry. Pearson Education Limited. pp 39–73.

Millie, D. F., Pearl, H. W., & Hurley, J. P. (1993). Microalgal pigment assessments using High Performance Liquid Chromatography: A synopsis of organismal and ecological applications. Canadian Journal of Fisheries and Aquatic Science, 50, 2513–2527.

306

Mitrovic, S. M., Hitchcock, J. N., Davie, A. W., & Ryan, D. A. (2010). Growth responses of Cyclotella meneghiniana (Bacillariophyceae) to various temperatures. Journal of Plankton Research, 32, 1217–1221.

Moline, M. A., & Prezlin, B. B. (1996). Long-term monitoring and analyses of physical factors regulating variability in coastal Antarctic phytoplankton biomass, in situ productivity and taxonomic composition over subseasonal, seasonal and interannual time scales. Marine Ecology Progress Series, 145, 143–160.

Montagnes, D. J. S, Berges, J. A., & Harrison, P. J. (1994). Estimating carbon, nitrogen, protein, and chlorophyll a from volume in marine phytoplankton. Limnology and Oceanography, 39, 1044–1060.

Mor, T. S., Hundal, T., Ohad, I., & Andersson, B. (1997). The fate of cytochrome b559 during anaerobic photoinhibition and its recovery processes. Photosynthetic Research, 53, 205–213.

Moore, B. G., & Tischer, R. G. (1965). Biosynthesis of extracellular polysaccharides by the blue-green alga Anabaena flos-aquae. Canadian Journal of Microbiology, 11, 877–885.

Morris, I., Glover, H., & Yentsch, C. S. (1974). Products of photosynthesis by marine phytoplankton: The effect of environmental factors on relative rates of protein synthesis. Marine Biology, 27, 1–9.

Mortain-Bertrand, A., Bennet, J., & Falkowski, P. G. (1990). Photoregulation of the light- harvesting chlorophyll protein complex associated with photosystem II in Dunaliella tertiolecta. Plant Physiology, 94, 304–311.

Muscatine, L., & Marian, R. E. (1982). Dissolved inorganic nitrogen flux in symbiotic and non-symbiotic Medusae. Limnology and Oceanography, 27 (5), 910–917.

Myklestad, S. Holm-Hansen, O., Varum, M. & Volcani, B. E. (1989). Rates of release of extracellular aminoacids and carbohydrates from marine diatom Chaetoceros affinis. Journal of Plankton Research, 11, 763–773.

Myklestad, S. M. (1974). Production of carbohydrates by marine planktonic diatoms. I. Comparison of nine different species in culture. Journal of Experimental Marine Biology and Ecology, 15, 261–274.

Nichols, B. W. (1973). Lipid composition and metabolism. In: Carr, N. G., Whitton, B. A. (eds), The Biology of Blue-green Algae, Blackwells, Oxford, pp. 144-161.

307

Niyogi, K. K., Bjorkman, O., & Grossman, A.R. (1997). The roles of specific xanthophylls in photoprotection. Plant Biology, 94, 14162–14167.

Olenina, I., Hajdu, S., Edler, L., Andersson, A., Wasmund, N., Busch, S., Göbel, J., Gromisz, S., Huseby, S., Huttunen, M.,Jaanus, A., Kokkonen, P., Ledaine, I, & Niemkiewicz, E. (2006). Biovolumes and size-classes of phytoplankton in the Baltic Sea. HELCOM Balt. Sea Environ Proc No 106. 144 pp.

Olson, R. J., Vaulot, D., & Chisholm, S. W. (1985). Marine phytoplankton distributions measured using shipboard flow cytometry. Deep-sea Research, 32, 1273–1280.

Osborne, B. A., & Geider, R. J. (1986). Effects of nitrate limitation on photosynthesis of the diatom Phaeodactylum tricornutum (Bacillarophyceae). Plant, Cell and Environment, 9, 617–625.

O’Reilly, J. E., Maritorena, S., Mitchell, B. G., Siegel, D. A., Corder, K. L., Garver, S. A., Kahru, M., & McClain, C. (1998). Ocean color chlorophyll algorithms for SeaWIFS. Journal of Geophysical Research, 103, 24937–24953.

O’Reilly, J. E., Maritorena, S., Siegel, D. A., O’Brien, M. C., Toole, P., Mitchell, B. G., Kahru, M., Chavez, F. P., Strutton, P., Cota, G. F., Hooker, S. B., McClain, C. R., Carder, K. L., Muller-Karger, F., Harding, L., Magnuson, A., Phinney, D., Moore, G. F., Aiken, J., Arrigo, K. R., Letelier, R., & Culver, M. (2001). Ocean color chlorophyll a algorithms for seaWIFS, OC2 and OC4: version 4. In: S. B. Hooker and E. R. Firestone (eds), seaWIFS postlaunch calibration and validation analyses: Part 3. NASA tech memo. 2000 – 206892, 11 (page 9-23). Greenbelt, MD: NASA Goddard Space Flight Center.

Paasche, E. (1960). On the relationship between primary production and standing stock of phytoplankton. Journal Conseil, Conseil Penn, Intern. Exploration Mer, 26, 33–48.

Paerl, H. W. (1997). Coastal eutrophication and harmful algal blooms: Importance of atmospheric deposition and groundwater as “new” nitrogen and other nutrient sources. Limnology and Oceanography, 42, 1154–1165.

Paerl, H. W., Fulton, R. S., Moisandander, P. H., & Dyble, J. (2001). Harmful freshwater algal blooms, with an emphasis on cyanobacteria. Science World I, 76–113.

Palmer, C. M. (1962). “Algae in Water Supplies.” US Department of Health, Education and Welfare, Division of Water Supply and Pollution Control, Washington DC.

Peat, S., Whelan, W. J., & Lawley, H. G. (1958). The structure of Laminarin. Part I. The main polymeric linkage. Journal of the Chemical Society, 724–728.

308

Philips, E. J., Zeman, C., & Hansen, P. (1989). Growth, photosynthesis, nitrogen fixation and carbohydrate production by a unicellular cyanobacterium, Synechococcus sp. (Cyanophyta). Journal of Applied Phycology, 1, 137–145.

Phlips, E. J., Badylak, S., & Lynch., T. C. (1999). Blooms of picoplanktonic cyanobacterium Synechococcus in Florida Bay, a sub-tropical inner-shelf lagoon. Limnology and Oceanography, (4) 44, 1166–1175.

Phlips, E. J., Bledsoe, E., Badylak, S., & Frost, J. (2002). The distribution of potentially toxic cyanobacteria in Florida. Proceedings of the health effects of exposure to cyanobacterial toxins: State of the Science Conference, August 13-14. www.doh.state.fl.us.

Piorreck, M., & Pohl, P. (1984). Formation of biomass, total protein, chlorophylls, lipids and fatty acids in green and blue-green algae during one growth phase. Photochemistry, 23, 217–223.

Powles, S.B. (1984). Photoinhibition of photosynthesis induced by visible light, Annual Review. Plant Physiology, 35, 15–44.

Prasad, A. K. S. K., Nienow, J. A., & Livingston, R. J. (1990). The genus Cyclotella (Bacillariophyta) in Choctawatchee Bay, Florida, with special reference to C. striata and C. choctawatcheeana sp. Phycologia, 29, 418–436.

Prasad, A. K. S. K. & Nienow, J. A. (2006). The centric diatom genus Cyclotella, (Stephanodiscaceae: Bacillariophyta) from Florida Bay, USA, with special reference to Cyclotella choctawhatcheeana and Cyclotella desikacharyi, a new marine species related to the Cyclotella striata complex. Phycologia, (2):45, 127– 140.

Prasil, O., Adir, N., & Ohad, I. (1992). Dynamics of Photosystem II: Mechanism of photoinhibition and recovery processes. In: Barber, J. (ed). The Photosystems: Structure, Function and Molecular Biology. 11: 295-348. Elsevier, Amsterdam.

Prezlin, B. B., & Alberte, R. S. (1978). Photosynthetic characteristics and organization of chlorophyll in marine dinoflagellates. Proceedings of the National Academy of Sciences, USA. 75, 1801–1804.

Proteau, P. J., Gerwick, W. H., Garcia-Pichel, F., and Castenholz, R. W. (1993). The structure of scytonemin, an ultraviolet sunscreen pigment from the sheaths of cyanobacteria. Experientia, 49, 825–829.

309

Pybus, C. (1996). The planktonic diatoms of Galway Bay: seasonal variations during 1974/75: biology and environment. Proceedings of the Royal Irish Academy Section B, 96, 169–176.

Raimboult, P., Diaz, F., Pouvesle, W., & Boudjellal, B. (1999). Simultaneous determination of particulate organic carbon, nitrogen and phosphorus collected on filters, using a semi-automatic wet-oxidation method. Marine Ecology Progress Series, 180, 289–295.

Rausch, T. (1981). The estimation of micro-algal protein content and its meaning to the evaluation of algal biomass I. Comparison of methods for extracting protein. Hydrobiologia, 78(3), 237–251.

Richards, F. A., & Thompson, T. F. (1952). The estimation and characterization of plankton populations by pigment analyses. II. A spectrophotometric method for the estimation of plankton pigments. Journal of Marine Research, 11, 156–172.

Richardson, D. H. S., Hill, D. J., & Smith, D. C. (1968). Lichen physiology. XI. The role of the alga in determining the pattern of carbohydrate movement between lichen symbionts. New Phytology, 67, 469–486.

Riemann, F. (1989). Gelatinous phytoplankton detritus aggregates on the Atlantic deep- sea bed. Marine Biology, 100, 533–539.

Rodriguez, F., Chauton, M., Johnsen, G., Andresen, L. M., & Zapata, M. (2006). Photoacclimation in phytoplankton: implications for biomass estimates, pigment functionality and chemotaxonomy. Marine Biology, 148, 963–967.

Ross, C., Santiago-vazquez, L., & Paul, V. (2006). Toxin release in response to oxidative stress and programmed cell death in the cyanobacterium Microcystis aeruginosa. Aquatic Toxiciology, 78, 66–73.

Rowan, K. S. (1989). Photosynthetic Pigments of Algae. Cambridge University Press. Cambridge.

Ruivo, M., Amorim, A., & Cartaxana, P. (2011). The effects of growth phase and irradiance on phytoplankton pigment ratios: implications for chemotaxonomy in coastal waters. Journal of Plankton Research, 33, 1012–1022.

Sakshaug, E., Bricaud, A., Dandonneau, Y., Falkowski, P., Kiefer, D., Legendre, L., Morel, A., Parslow, J., & Takahashi, M. (1997). Parameters of photosynthesis: definitions, theory and interpretation of results. Journal of Plankton Research, 19, 1637–1670.

310

Sapozhnikov, D. J. (1972). In: Carotenoids other than vitamin A-III, Proceedings of the Third International Symposium on Carotenoids, Butterworths, London.

Sieferermann-Harms, D. (1987). The light –harvesting and protective functions of carotenoids in photosynthetic membranes. Plant Physiology, 69, 561–568.

Schulter, L., Mohlenberg, F., Havskum, H., & Larsen, S. (2000). The use of phytoplankton pigments for identifying and quantifying phytoplankton groups in coastal areas: testing the influence of light and nutrients on pigment/chlorophyll a ratios. Marine Ecology Progress Series, 192, 49–63.

Shulter, L., & Havskum, H. (1997). Phytoplankton pigments in relation to carbon content in phytoplankton communities. Marine Ecology Progress Series, 155, 55–65.

Scott, J. M. (1980). Effects of growth rate of the food algae on the growth/ingestion efficiency of a marine herbivore. Journal of Marine Biology Association UK, 60, 681–702.

Scweiter, R. H., & Brudvig, G. W., (1995). Parallel low-temperature fluorescence and EPR measurements of Mn-depleted photosystem II. In: Matthis, P. (ed). Photosynthesis: From Light to Biosphere. 1: 807-810. Kluwer Academic Publishers, Dordrecht, the Netherlands.

Senger, H., & , P. H. (1987). Adaptation of the photosynthetic apparatus of Scenedesmus obliquus to strong and weak light. I. Differences in pigments, photosynthetic capacity, quantum yield and dark reactions. Physiologia Plantarum, 43, 35–42.

SFWMD. Ecological effects of Phosphorus enrichment in The Everglades. In: Garth Redfield (ed). (2001). Everglades consolidated report South Florida Water Management District, West Palm Beach. Florida.

Shapiro, L. P., Haugen, E. M., & Keller, D. M. (1989). Taxonomic affinities of marine coccoid ultroplankton: a comparison of immunochemical surface antigen cross- reactions and HPLC chloroplast pigment signatures. Journal of Phycology, 24, 794–797.

Sieburth, J. McN. (1969). Studies on algal substances in the sea III. The production of extracellular organic matter by littoral marine algae. Journal of Experimental Marine Biology and Ecology, 3, 290–309.

Sinha, R. P., Klisch, M., Groniger, A., & Hader, D. P., (1998). Ultraviolet- absorbing/screening substances in cyanobacteria, phytoplankton and macroalgae. Journal of Photochemical Photobiology, 47, 83–94.

311

Slate, J. E., & Stevenson, R. J. (2000). Recent abrupt environmental change in The Florida Everglades indicated from silicious microfossils. Wetlands, 20, 346–356.

Smayda, T. J. (1978). From phytoplankters to biomass. In: Sournia, A. (ed), Phytoplankton manual. Unesco Paris. 273–279.

Smith, G. M. (1961). A monograph of the algal genus Scenedesmus based upon pure culture studies. Transactions of the Wisconsin Academy of Sciences, Arts, and Letters, 18, 422–530.

Solte, W., Kraay, G. W., Noordeloos, A. M., & Riegerman, R. (2000). Genetic and physiological variation in pigment composition of Emiliania huxleyi (prymnesiophyceae) and the potential use of its pigment ratios as a quantitative physiological marker. Journal of Phycology, 36, 529–539.

Sournia, A. (ed) (1978). Phytoplankton Manual. Monographs on Oceanographic Methodology 6. UNESCO, Paris.

Spaulding, S., & Edlund, M. (2009). Thalassiosira. In: Diatoms of the United States. (http://westerndiatoms.colorado.edu/taxa/genus/Thalassiosira).

Squier, A. H., Airs, R. L., Hodgson, D. A., & Keely, B.J. (2004). Atmospheric pressure chemical ionization liquid chromatography/mass spectrometry of the ultraviolet screening pigment scytonemin: characteristic fragmentations. Rapid communications in Mass Spectrometry, 18, 2934–2938.

Stauber, J. L., & Jeffrey, S. W. (1988). Photosynthetic pigments in fifty-one species of marine diatoms. Journal of Phycology, 24, 158–172.

Steinman, A. D. & Lamberti, G. A. (1996). Biomass and pigments of benthic algae. Hauer, F. R. and Lamberti, G. A. (eds) In: Methods in Stream Ecology. Academic Press, San Diego, California.

Stevenson, R. J., Singer, R., Roberts, D. A., & Boyelyn, C. W. (1985). Patterns of benthic algae abundance with depth, trophic states, and acidity in poorly buffered New Hampshire Lakes. Canadian Journal of Fisheries and Aquatic Science, 42, 1501– 1512.

Stewart, W. D. (ed). (1974). Algal physiology and Biochemistry. Botanical Monographs Volume 10 . University of California Press. Berkley and Los Angles.

Takaichi, S. (2000). Characterization of carotenes in a combination of a C-18 HPLC column with isocratic elution and absorption spectra with a photodiode-array detector. Photosynthesis Research, 65, 93–99.

312

Tedrow, O., Julius, M. L., & Schoenfuss, H. L. (2002). The effects of biogenically active compounds on Cyclotella meneghiniana (Bacillariophyta). Journal of Phycology, 38, 34–35.

Tester, P. A., Geesey, M. E., Guo, C., Paerl, H. W., & Millie, D. F. (1995). Evaluating phytoplankton dynamics in the Newport river estuary (North Carolina) by HPLC- derived pigment profiles. Marine Ecology Progress Series, 124, 237–245.

Thompson, E. W., & Preston, R. D. 1967. Proteins in the cell walls of some green algae. Nature, London, 213, 684–685.

Thompson, P. A., Harrison, P. J., & Parslow, J. S. (1991). Influence of irradiance on cell volume and carbon and carbon quota for ten species of marine phytoplankton. Journal of Phycology, 27, 351–360.

Thornton, D. C. O. (2001). Diatom aggregation in the sea: mechanisms and ecological implications. European Journal of Phycology, 37, 149–161.

Underwood, G. J. C., & Smith, D. J. (1998). Predicting diatom exopolymer concentrations in intertidal sediments from sediment chlorophyll a. Microbial Ecology, 35, 116–125.

Underwood, G. J. C., Paterson, D. M. & Parkes, R. J. (1995). The measurement of microbial carbohydrate exopolymers from intertidal sediments. Limnology and Oceanography, 40, 1243–1253.

Underwood, G. J. C., Boulcott, M., & Rains. C. A. (2004). Environmental effects on exopolymer production by marine benthic diatoms: dynamics, changes in composition and pathways of production. Journal of Phycology, 40, 293–304.

UNESCO. (1996). Monograph on oceanographic methodology. I. Determination of photosynthetic pigments in sea water. United Nations Education, Science and Cultural Organization, Paris.

U.S. Environmental Protection Agency (U.S. EPA). (1995a). Generic quality assurance project plan guidance for programs using community- level biological assessment in streams and wadeable rivers. U.S. Environmental Protection agency, Office of Water, Washington D. C. EPA 841-B-95-004. van den Meersche, K., & Soetaert. (2009). BCE. Bayesian composition estimator: estimating sample (taxonomic) composition from biomarker data. R. package version 1.4. http://CRAN.R project.org/package=BCE.

313

van den Meersche, K., Soetaert, K., & Middleburg, J. J. (2008). A Bayesian Compositonal estimator for microbial taxonomy based on biomarkers. Limnology and Oceanography, Methods 6, 190–199.

van Grondelle, R., & Amesz, J. (1986). Excitation energy transfer in photosynthetic systems. In: Govingjee, Amesz, J., and Fork, D. C. (eds) Light Emission by Plants and Bacteria. Academic Press, New York. 191–224.

van Rijssel, M., Janse, I., Noordkamp, D. J. B., & Gieskes, W. W. C. (2000). An inventory of factors that affect polysaccharide production by Phaeocystis globosa. Journal of Sea research, 43, 297–306.

Walkley, A. & Black, I.A. (1934). An examination of the Degtjareff method for determining soil organic matter and proposed modification of the chromic acid titration method. Soil Science, 37, 29–38.

Watanabe, M. M. (2005). Freshwater Culture Media. In: Andersen, R. A. (ed) Algal Culturing Techniques. Chapter 2. Elsevier academic press.

Weber, C. I., Fay, L. A., Collins, G. B., Rathke, D. E., & Tobin, J. (1986). A review of methods for the analysis of chlorophyll in periphyton and plankton of marine and freshwater systems. Ohio State University Sea Grant Program Technical Bulletin. OHSU-TB-15.

Wehr, J.D., & Sheath, R. G. (2003). (eds): Freshwater Algae of North America: Ecology and Classification, Elsevier Science USA, pp. 255–258.

Wilhelm, C., & Manns, L. (2000). Changes in pigmentation of phytoplankton species during growth and stationary phase – consequences of reliability of pigment- based methods of biomass determination. Journal of Applied Phycology, 3, 305– 310.

Wilhelm, C., Rudolph, I., & Renner, W. (1991). A quantitative method based on HPLC- aided pigment analysis to monitor structure and dynamics of the phytoplankton assemblages – a study from Lake Meerfelder Maar (Eifel, Germany). Archives of Hydrobiology, 123, 21–35.

Wright, S. W., & Jeffrey, S. W. (2005). Pigment markers for phytoplankton production. Environmental Chemistry, 2, 71–104.

Wright, S. W., Jeffrey, S. W., Mantoura, R. F. C., Llewelyn, C. A., Bjornland, T., Repeta, D., & Welschmeyer, N. (1991). Improved HPLC method of analysis of chlorophylls and carotenoids from marine phytoplankton. Marine Ecology Progress Series, 77, 183–196.

314

Wright, S. W., Thomas, D.P., Marchant, H. J., Higgins, H. W., Mackey, M. D., & Mackey, D. J. (1996). Analysis of phytoplankton of the Australian sector of the Southern Ocean: comparisons of microscopy and size frequency data with interpretations of pigment HPLC data using the ‘CHEMTAX’ matrix factorization program. Marine Ecological Progress Series, 144, 285–298.

Yasumoto, N., Seino, Y., Murakami & Murata, M. (1987). Toxins produced by benthic dinoflagellates. Biology Bulletin, 172, 128–131.

Yentsch, C. S., & Menzel, R. W. (1963). A method for the determination of phytoplankton chlorophyll by fluorescence. Deep Sea Research, 10, 221–231.

Zak, E., Norling, B., Maitra, R., Huang, F., Andersson, B., & Pakrasi, H. B. (2001). The initial steps of biogenesis of cyanobacterial photosystems occur in plasma membranes. Proceedings of the National Academy of Sciences, 98, 13443–13448.

315