PHOTOACCLIMATION RATE OF THE DIATOM ASTERIONELLA FORMOSA

A clone of the diatom Asterionella formosa was studied to determine the ability of the species to photoacclimate as they were passed through a light gradient at varying rates. Columnar incubators 4 m in height, held at 4 � C, with a light gradient of 250 � 10 mmol photons m^-2 sec^-1 were used to simulate vertical mixing as found in Lake Michigan.

Asterionella formosa used both increases in the number and size of photosynthetic units to acclimate to lower irradiances. These increases occurred within 24 hours of being introduced into a new light climate, though clear trends in the response of photosynthetic parameters and pigments to light history was tenuous in two of three experiments. In general, the photoacclimation response included synthesis of photosynthetic pigments that appeared to be proportional to each other across all light histories and maximal intensities. This proportionality in pigmentation included samples taken from the darker and bluer BOT treatments, where the potential for chromatic adaptation to alter pigment ratios was highest. However, after one week the cells in the BOT treatment became so light limited that they were incapable of synthesizing pigment and the cell density began to decline. This indicates that cells require certain total daily or weekly light dose in order to successfully photoacclimate to reduced light conditions.

Contrary to published trends, the TOP samples in all experiments increased or maintained chlorophyll content despite being in light at non-limiting intensities. It appeared that the cells needed to add chlorophyll until they approached a maximum of about 2.5 pg cell^-1, where cells in experiment 0205 seemed to maintain their content. The final experiment yielded the most consistent evidence of periodic photoacclimation that was correlated with daily light history. The fact that this periodic acclimation only became apparent after ten days under the mixed light regime suggests that acclimation to a non-diel, cyclic light cycle may be occurring.

List of figures

List of tables

Introduction

Photoacclimation

Pelagic phytoplankton communities in the oceans and in large, deep lakes, such the Laurentian Great Lakes, are strongly influenced by vertical mixing processes. This turbulent physical process controls the availability of nutrients (Brooks and Edgington 1994), the light to which the cells are exposed (Falkowski and Raven 1997), and distribution of the organisms themselves (Reynolds 1998). The interaction of location-imposed resource constraints forces phytoplankton to constantly optimize life processes. Photosynthesis is a critical life process for photoautotrophic organisms and is modulated, within a single cell, in a changing light environment via photoacclimation.

One major effect that vertical water movement has on phytoplankton is to force the cells through a light gradient. As attenuation of light increases, the gradient becomes steeper through shading from living and non-living particles and substances (Grobbelaar et al. 1992).

The rate of vertical transport and the direction of cell movement through the water column of lakes and oceans dictates the rate and specific physiological response of phytoplankton to a changing light environment. If the mixing time is greater than the time it takes a community to acclimate to the light at that depth, a vertical gradient of some photoacclimation parameter will form. On the other hand, if the time of mixing is less, the parameter will be uniform across the region of transport. No clearly stated values for mixing velocities in the euphotic zone of Lake Michigan could be found. However, for perspective, results from the Marine Boundary Layer Experiment off the coast of Monterey, California describe measured mixing velocities associated with Langmuir circulations (Farmer et al. 1998). They found maximal mixing rates of 10 cm sec^-1 and averages of 4-8 cm sec^-1 at 15 m depth in response to average wind velocities of 15.9 m sec^-1 (30.9 knots) for two days. At the mean mixing rates, the time for a parcel of water to traverse the 15 m would be approximately 3 to 6 minutes.

As the depth of mixing increases, the average light intensity that the cells are exposed to will decrease. Therefore, as a community, the rate of photosynthesis integrated over the depth of the water column will decrease. However, the rate of respiration of the phytoplankton population is largely independent of mixing depth (Li and Garrett 1998). Therefore, there exists a critical depth below which the consumption of carbon by respiration is greater than that gained by photosynthesis and a net loss of community biomass occurs (see (Platt et al. 1991; Sverdrup 1953). The ability of a phytoplankton community to photoacclimate extends the critical depth and ultimately enhances net biomass production. A phytoplankton community that is light limited should benefit greatly from a small degree of photoacclimation, in terms of annual primary production.

The processes of photoacclimation have been studied both experimentally in marine environments, (Falkowski 1983; Gallegos and Platt 1982; Lewis et al. 1984a; Lewis et al. 1984b; Yoder and Bishop 1985) and theoretically (Cullen and Lewis 1988; Dusenberry 2000; Eilers and Peeters 1988; Falkowski and Wirick 1981). The ability to photoacclimate is ubiquitous throughout photoautotrophic plankton. However, the mechanisms and rates can vary between taxa. The following description presents some of the photoacclimation responses that are common in many taxa (Falkowski and La Roche 1991). In response to changes in photon flux density (light intensity) and spectral quality, morphological photoacclimation can be manifested by changes in cell volume, the number and density of thylakoids membranes (Berner et al. 1989; Post et al. 1984) the size of storage bodies (Sukenik et al. 1987) and changes in the number or size of plastids per cell. Physiological changes can include changes in lipid and pigment content as well as composition (Falkowski and Owens 1980; Perry et al. 1981; Prezelin and Alberte 1978). These physiological changes occur to minimize quantum requirements for photosynthetic oxygen evolution (Dubinsky et al. 1986), respiration (Falkowski et al. 1985; Geider et al. 1986; Langdon 1987) and growth rate (Falkowski et al. 1985; Laws and Bannister 1980; Post et al. 1984). Though they can be subtle and difficult to quantify, cellular modifications occur over a cell�s generation (Cullen and Lewis 1988) and should not be confused with adaptation, which occurs over multiple generations (Falkowski and La Roche 1991).

Phytoplankton physiological responses to changes in light intensity can also be accompanied by concurrent changes in metabolic priorities. The effect is well documented and numerous studies of cellular kinetics describe shifts in metabolic resource allocation based upon recent light histories of phytoplankton (Cuhel et al. 1984). For example, Dunaliella tertiolecta exhibits a rapid (<24 hour) diversion of lipid and carbohydrate biosynthesis to double their light harvesting chlorophyll protein complexes (LHCP) in response to a reduction in ambient light (Sukenik et al. 1990). Shortly after LHCP allocation begins, photosynthetic reaction center proteins (D1 and D2) accumulate. Once these light reaction proteins have increased, carbon metabolism shifts back to lipid production to allow for the construction of additional thylakoid membranes. These new membranes provide room for additional photosynthetic protein complexes. Each progression has been modeled with first order kinetics (Falkowski 1984; Hoffman and Senger 1988; Sukenik et al., 1990), though alternative models have been proposed (Cullen and Lewis, 1988). This adjustment to lower irradiances is not precise. For example, some compounds may be generated in excess amounts before achieving a steady state in relation to the light environment (Falkowski 1984a; Falkowski 1984b).

When cells return to higher irradiances, resources are diverted from protein production and allocated to lipids and carbohydrates (Sukenik et al. 1990). The additional carbohydrates and lipids allow the cell to rapidly increase in volume without the need for high concentrations of pigment. Cell growth, both in numbers and in volume, will dilute cellular pigment content. Of course, controlled reductions of photopigment does occur and is not always simply a result of dilution. Experiments using D. tertiolecta indicated that about half of the reduction in pigment was due to dilution by increased cell volume. The remaining reduction was likely due to direct decomposition (Falkowski 1984a). Of interest to this study, is the change in the relative amounts of specific photosynthetic pigments as the cells modify their photosynthetic apparatus.

Acclimation to changes in light intensity is overlaid with the intrinsic circadian cycles found in all natural populations of phytoplankton (Prezelin and Sweeney 1977); (Falkowski and Owens 1980; Marra 1980). In studies of acclimation it is important to differentiate between photoacclimation and physiological acclimation to diel periodicity. Post et al. (1984) found that the diatom Thalassiosira weisflogii achieved maximal chlorophyll content in the middle of the photoperiod and then declined before onset of the dark period. Also, cellular chlorophyll levels were minimal around the mid-point of the dark period and began to increase before light onset. These diel cycles were separate from reproductive cycles and remained present throughout a full range of light intensities.

In her excellent review of diel periodicity, Prez�lin (1992) suggests that the cellular developmental cycle in diatoms, �never appears to couple to a biological clock�Some diatom cell division appears to always occur throughout the light period.� However, studies of multiple species of marine diatoms indicate that synchronized, diel variation in the photosynthetic parameter P_max may be ubiquitous throughout the diatoms with peaks typically occurring at the mid-point of the light period (Harding et al. 1981). This appears to be confirmed by field studies of diatom-dominated marine systems (Harding et al. 1982; Mac Caull and Platt 1977; Prezelin and Alberte 1978; Prezelin and Sweeney 1977). (Prezelin 1992)

Most studies of photoacclimation and photoadaptation have been conducted under conditions where cells are transferred abruptly between high and low light intensities. For example, Anning et al. (Anning et al. 2000) studied photoacclimation using parameters similar to those in my study. They exposed the marine diatom Skeletonema costatum to large, abrupt changes in irradiance (50-1200 mmol photons m^-2 sec^-1) at 15 � C. Cells adjusted to 50 mmols were found to have higher levels of fucoxanthin and chlorophyll a, but lower diadinoxanthin and diatoxanthin. Cells grown at 50 mmol photons were transferred to 1200 mmols and within two days cellular fucoxanthin declined from ~ 3.0 pg cell^-1 to 0.5 and chlorophyll a from ~1.0 to 0.4 pg cell^-1. Chlorophyll c1c2 followed similar trends as fucoxanthin and chlorophyll a. After returning to 50 mmols, the cultures returned to their previous pigment contents within four days. b-carotene is a photoprotective pigment that is not coupled to the photosynthetic reaction centers, and protects the cells from photooxidative damage from the capture of excess photons in high-irradiance (Neale and Richerson 1987). b-carotene was also measured by Anning et al. (2000), but was invariant throughout the study.

In addition to b-cartotene, diatoms posses other carotenoids that play photoprotective roles, such as the xanthophylls diadinoxanthin and diatoxanthin (DD and DT). DD is thought to be a precursor to fucoxanthin and is thought to allow more rapid photoacclimation upon a cell�s return to low light (Goericke and Welschmeyer 1992; Lohr and Wilhelm 1999). Cells that were accustomed to low light would be sensitive to damage by excess excitation from high light intensities, as their reaction centers would be numerous. As expected, the photoprotective xanthophylls had inverse responses to both fucoxanthin and chlorophylls. Anning felt the lack of response by b-carotene was due to the rapid intervention of DD and DT.

Anning et al. (2000) also found that cell-specific P_max did not shift in response to changes in irradiance. However, chlorophyll-specific light saturated rates of photosynthesis increased after exposure to high light. The light-limited rates of photosynthesis, measured by a normalized to chlorophyll a (a^B) showed no temporal variability throughout the study. Cell-specific a (a^cell) did change and was greater in the shade-acclimated cells. These responses indicated that changes in a^cell were due to modulations of chlorophyll a content and that cell-specific chlorophyll a concentrations are important for controlling light-limited photosynthesis.

A handful of marine studies have considered the complex nature of photoacclimation when cells are passed through a light gradient. (Marra 1978) in a field setting using mixing and light gradients focused upon optimal photosynthetic rates in response to varied light histories. The study described in this thesis is one of a few to consider photoacclimation to a smoothly changing light gradient in a laboratory setting (Flameling and Kromkamp 1997; Ibelings et al. 1994; Kromkamp and Limbeek 1993; Kroon et al. 1992a; Kroon et al. 1992b).

A study very similar to mine was published in 1997 using the chlorophyte Scenedesmus protuberans (Flameling and Kromkamp 1997). They exposed cells, at 20 � C, to oscillating light over a range of 200 to ~10 mmol photons m^-2 sec^-1 at a periodicity of 1, 4 and 8 oscillations every ten hours. In each of these trials, the total daily light dose (TDLD) remained constant, but the time spent at peak intensity was modulated. They found that, when corrected for cell size, chlorophyll content, Chl a/b ratio, the chlorophyll-specific absorption cross-section or the carotenoid-Chl a in vivo absorbance ratio did not change throughout the study. They proposed that algae that do not successfully photoacclimate to an oscillating light environment would exhibit a reduced rate of daily-integrated photosynthesis and would lose biomass.

For S. protuberans, like Skeletonema costatum and Phaeocystis globosa, they determined that it was not the total daily light dose (TDLD) that determined P^B_max, but rather the daily maximal irradiance experienced by the algae. This is in contrast to their findings that light-saturated photosynthesis normalized to chlorophyll a (P^B_max) in the diatoms Thalassiosira weissflogii and Phaeodactylum tricornutum and were not influenced by the daily maximal irradiance. Also, they found that cells held in a fluctuating light environment did not exhibit declines in a^B compared with those held in a high light environment. This supports Anning et al. (2000) finding that light-limited photosynthesis is controlled by changes in chlorophyll a content per cell.

Though there are many rates and strategies that a given algal group may employ to photoacclimate, in general when light levels are reduced, pigments and proteins for light harvesting chlorophyll protein complexes (LHCPs) are produced. As light increases, maintenance of high concentrations of photopigments is unnecessary and cells will reduce their pigmentation.

As a common member of freshwater pelagic communities, Asterionella possesses critical preadaptations to allow it to be successful in a deep-mixed water column (Reynolds 1998). Physiological features such as the rate and degree of chlorophyll a production, both absolutely and in relation to the accessory pigments allows genera such as Planktothrix (cyanobacterium), Aulacoseira (diatom) and Asterionella to be successful in pelagic systems. For example, the chlorophyll:carotenoid ratio is a longstanding index that can be used to determine the photosynthetic health of phytoplankton (Margalef, 1958 as cited by Reynolds 1998). The adjustment of pigment concentration/pigment ratios is a strategy that must be employed by all phytoplankton when their light environment has shifted for a sufficient time. However, it is important to note that simply increasing chlorophyll does not linearly increase the functional photosynthetic cross-section of the photon-collecting antennas. (Margalef 1958)

A linear increase in photosynthetic rate with added chlorophyll requires that the additional chlorophyll be spread evenly throughout the cell. Failure to distribute the chlorophyll results in increased internal self-shading, an effect referred to as �package effect�. Modulating package effect is a very important strategy for photoacclimation and assays of this parameter can provide yet another descriptor as to the state of light-regulated photosynthetic efficiency (Anning et al. 2000; Falkowski and La Roche 1991; Falkowski and Raven 1997; Grobbelaar et al. 1996).

An anatomical flexibility that helps define the degree of package effect is intracellular migration of chloroplasts between a cell�s periphery and its core. Plastid relocation in response to super-saturating irradiances is a common and important photoprotective mechanism. In planktonic cells with multiple and moveable chloroplasts, cell size limits, but does not negate, the usefulness of this mechanism (Long et al., 1994). Although no images were captured, I was able to confirm that light-limited A. formosa cells positioned their chloroplasts against the cell wall when grown in light below ~30 mmols. This distribution of chlorophyll was not as apparent when viewing cells grown at ~100 mmols, where the chloroplasts tended to be located along the midline of the cell and appeared to overlap slightly.

When diatoms are exposed to prolonged doses of high light, contraction of the chromophores and chlorophyll cross-section areas occurs within minutes to hours (Neale and Richerson 1987). Under quickly changing conditions, the Asterionella genus has one or more chloroplasts per cell that can be moved to different positions within the cell cavity in response to light stress. However, the nature of entrainment of phytoplankton in the flows of a vertically mixed water column means that it is not likely that an entire population of cells would be over exposed at any time (Reynolds, 1998). In the studies mentioned above, and many others, members of the group Bacillariophyceae (diatoms) are a common focus, although no studies were found that examined Asterionella spp. in the context of photoacclimation.

Consideration of the specific strategy of A. formosa requires a knowledge of its particular pigment compliment. In general, diatoms posses chloroplasts that contain chlorophylls a, c1, c2 and fucoxanthin as the principle carotenoid. Also present is b carotene which is important as a photoprotective pigment when irradiance intensities exceed levels that the cell can safely capture and funnel to the reactions centers (Flameling and Kromkamp 1997).

Currently, it is not clear if the rate of community photoacclimation in Lake Michigan is rapid enough to allow detection; especially within the surface mixed layer under stratified conditions. It is possible that cells may simply maintain a modified photosynthetic apparatus suited to a light environment averaged through the water column they traverse.

Figures: literature examples

Figures: methods

Objectives

The primary objective of this study was to characterize photoacclimation in the diatom Asterionella formosa as it passes through a light gradient similar to those found in nature. Due to the properties of light propagation in an aquatic environment, changes in intensity are always associated with changes in quality (Kirk 1996). Though an important aspect of the physical environment, spectral changes were not considered experimentally in this project.

As part of the overall objective, I hoped to determine a rate of acclimation in relation to the rate at which the cells were �mixed� or passed through the light gradient. The methods for detecting this change focused on cellular pigment concentrations and measurements of photosynthetic efficiency and capacity. If photoacclimation was not detected for each traversal of the light gradient, a secondary objective was to determine if the cells were responding to their entrainment cycle on a longer time scale.

It was hypothesized that once transitioned from an intermediate light intensity, cells will begin to immediately photoacclimate to their new light environments. Cells exposed to higher light intensities will reduce their photopigments and shift their carbon assimilation to synthesize proteins for light harvesting complexes that allow them to use the high rate of photon flux. Cells transitioned to low light will increase their pigments and photoreceptor complexes to become more efficient at capturing the now sparse photons. These conditions were considered high-light and low-light controls for responses seen under mixed conditions. Cells that are moved through the light gradient should exhibit a moderated response that is bracketed by the response seen by the high and low light controls.

Results from previous studies using rapid and non-gradual shifts in irradiance show that a should increase and P_max should decrease in the cells as they are brought from high light to low light. This physiological shift should also occur in cells in a treatment that simulates movement across a light gradient, as in vertical water column mixing. The reverse condition where a decreases and P_max increases should occur as cells move from low light to high light.

Methods

Isolation and culturing

During March 2000 multiple monocultures of four species of diatoms were grown from isolates taken from 5m depth at Fox Point station in Lake Michigan (43� 11� 40� N, 87� 40� 11� W). Members of the genera Synedra, Fragilaria, Cyclotella and Asterionella were grown in small batch cultures. It was decided that the most versatile species upon which to focus this study would be Asterionella formosa (Figure 1). The star shaped colonies were likely to yield accurate counts from the Coulter particle counter and

previous culture experience suggested that A. formosa would be the easiest to maintain in long-term batch cultures. The clone Aster8, which grew in culture with the most vigor and reliability, was chosen for use in this study.

Maintenance cultures were grown in 150-mL, screw top Pyrex Erlenmeyer flasks using the medium DYV ((Lehman 1976) as modified by Sandgren) (Table 1). The recipe was not altered and the final concentrations of nutrients in media were intended to be non-limiting in all of the experiments.

Ingredient	Final Medium Concentration
Calcium chloride (CaCl₂)	180 mM-Ca
Magnesium sulfate (Mg(SO₄)₂ * 7H₂O)	300 uM-Mg
Sodium phosphate (Na₂HPO₄)	46.1 mM-PO₄-P
Sodium nitrate (NaNO₃)	235 mM-NO₃-N
Sodium metasilicate (Na₂SiO₃ * 9H₂O)	53.2 mM-Si
Ammonium nitrate (NH₄NO₃)	125 mM-NO₃ & 125 mM-NH₄-N
Potassium chloride (KCl)	134 mM-K
Sodium bicarbonate (NaHCO₃)	250 mM-C

Lighting in the environmental chamber that held the maintenance cultures was supplied by Cool White fluorescent bulbs at an average intensity of 80 mmol photons m^-2 sec^-1 on a 12/12-light/dark cycle. Temperature was held at 10 �C.

Sub-samples of the maintenance cultures were taken for use in the experiments. Three to six weeks were required to grow as much as 30 liters of log-phase culture and diluted to a density of about 2500 colonies ml^-1. Cultures were maintained in one-liter Nalgene polycarbonate centrifuge bottles at 15 � C and 100 mmol photons m^-2 sec^-1 supplied by Cool White High Output (HO) fluorescent bulbs.

The environmental chamber that provided sufficient space and lighting to grow 30 liters of culture was unable to maintain temperatures below 15 � C. Because the experiments were to be conducted at 4 � C simulating a homothermal Lake Michigan, I was concerned about damaging the cells with large temperature changes. Therefore, the cultures grown at 15 � C were diluted with media stored in the 10 � C chamber and mixed thoroughly in a 60-liter carboy to reduce the potential for temperature shock at the start of an experiment. The resulting mixture was then dispensed into the polycarbonate sample bottles and left in the 10 � C chamber overnight before being transferred to the 4 � C columns. In a later test, a sample bottle of media at 15 � C cooled to 4 � C in approximately three hours when placed in a 4 � C water bath.

Experimental columns

Two identical columnar incubators were used to create the light and temperature conditions necessary for this experiment (Figure 2). Each column consisted of a gray PVC pipe measuring 33 cm in internal diameter and 4.0 meters tall. Each column was wrapped with garden hose through which chilled water was circulated. The pipe and wrapped hose assembly was jacketed with insulation. Vigorous aeration was provided by an air stone at the bottom of each column that kept the water homothermal from top to bottom and throughout light and dark cycles. Testing using yo-yoing Onset Stowaway temperature loggers showed that throughout the course of an experiment, and at all depths, temperature did not deviate from the target 4.0 � C by more than � 0.7 � C.

Light was provided by Sylvania Super Metal Arc 1000-watt lamps positioned directly above each column, operating on 1 12/12 light/dark cycle. The desired

light intensity was achieved by raising or lowering the lamp height above the columns. The distance from the plane of the opening of the lamp shroud to the surface of the water was 56 cm for all experiments and both columns. The desired light intensity was achieved by raising or lowering the lamp height.

The light conditions in the columns were measured using two data logging instruments mounted on a single bracket, at the exact same level (Figure 3). One was an Onset HLI light intensity logger, which is capable of recording light intensity greater than normal room lighting. The other sensor was a Licor scalar irradiance spherical sensor. The Licor sensor records only radiation in the 400 � 700 nm region, while the Onset HLI sensor records radiation somewhat below 400 and above 1000 nm(Figure 4).Each was set to begin logging at the same time and to record light intensity at one-minute intervals. The sensors were then lowered and raised in the columns at rates ranging from 10 to 1 cm min^-1 to collect light readings continuously over depth. The Licor sensor indicated that the light intensity at 0.05 m below the water surface was 236 � 38 mmol photons m^-2 sec^-1with the air stones on. Variance in this reading was likely due to the influence of the pulsing light source, random scattering by the bubbles and noise from the sensors. The light intensity at the bottom of both columns with the air stones on was measured as 7 � 6

mmol photons m^-2 sec^-1. Aeration diffused and scattered the downwelling irradiance enough to increase attenuation of light in comparison to un-aerated conditions.

The water in the columns was chlorinated tap water that was changed at least every three weeks to minimize the accumulation of particulates and biological growth that would reduce the water clarity.

The lamps did not produce constant light intensity over short time scales. That is, a slight �pulsing� was observed both visually and with light recording devices. These pulses would build and fade with a period of less than 10 sec. and ranged as much as � 25 mmol photons m^-2 sec^-1 in surface irradiance. To detect the pulsing with an instrument, the HLI logger was set to record a measurement every five seconds without averaging. The Licor logger was left to record average values sampled over 1 minute and the intensity peaks were smoothed by this averaging.

In each experiment where sample bottles were moved through a light gradient an HLI light intensity logger was placed atop the cluster of bottles to record the actual light history of the cells. Using data from the Licor PAR scalar sensor, a conversion algorithm was created to convert intensity values from the logger to apparent scalar irradiance exposure using nonlinear fitting techniques such as the following general model (Figure 5):

where Int is the light intensity measured by the HLI light logger in lumens and PARm is the reading from the Licor spherical quantum sensor. The fourth power transformation was used to reduce the heterogeneity of variance in the readings from the HLI sensors. It is a known property of the sensors that the amplitude of the noise increases as the light intensity increases and this problem negatively influenced the modeling. Conversion of the HLI readings, and the resulting high-amplitude noise, to PAR often resulted in values that were outside of the range observed by the Licor sensor. In the experimental columns, with the bubbles and clean water, the following relationship was found.

The light intensities found in the columns approximate those found in Lake Michigan, during mid to late spring, at depths of 1.5 to 16.5 meters.

In both columns, sample bottles were attached to nylon cord with adhesive-backed 1.25 cm Velcro straps and metal clamps. The adhesive side of white Velcro was wrapped around each bottle leaving the facing side exposed. Clamps consisting of two metal strips approximately 3.2 cm long were held together using bolts and wing nuts. The clamps firmly pinched the cord to the reciprocal half of the Velcro assembly so that the bottle was securely attached to the line. The single point of attachment allowed the bottles to rotate slightly about their long axis when the cord was moved. Removing each bottle was simply a case of disengaging the Velcro straps.

The length of the metal strips, and hence the weight of the clamps, was optimized to make each bottle only slightly negatively buoyant when completely filled with medium. This was important to reduce the amount of strain on the motors used in the incubators. White Velcro, which was slightly translucent especially when wet, was used instead of black to limit the amount of shading in each bottle (Figure 6). For each experiment, 12-15 bottles were placed at the position of each treatment. Three replicate bottles were used for each sampling in all experiments. For example, in experiments where there were MIX, TOP and BOT treatments, 9 replicate bottles were sampled. The physical diameter of each bottle meant that the replicate bottles for each treatment spanned approximately one meter of column depth in the TOP and BOT treatments, and about 0.5 meters for the MIX treatments. I did not record the position of each bottle to correct for light exposure during data processing.

As stated earlier, the study used two nearly identical column incubators. One was used as a �static� column where samples were left at the top (TOP) and bottom (BOT) of the column throughout the experiment. In the second column the sample bottles were moved up and down at a fixed rate for each experiment (MIX).

The mechanism employed to move the samples through the light gradient used a stepping motor and drive (Intelligent Motion Systems Microlynx 7) with a program to control the travel rate, position at reversal and to log the positions of the samples over time (Appendix J).The stepping motor was mounted on a frame with an arm that extended over the center of the MIX column. The stepping motor was connected to an axel via a roller chain and sprocket gears. Mounted on the axle were two fixed spools that rotated in tandem. The suspending cord was wound around the spools such that a hanging loop was formed in the column. The stepping motor, rotated in decimal degree increments as specified by the program interface.

A second motor and offset pulley rotated constantly to gently jostle the bottle lines in both columns to keep the cells in suspension in the sample bottles. This technique met with limited success and the cells tended to settle after 24 hours despite the jostling. At least every 24 hours, the bottles were agitated in the columns to resuspend any settled cells. In addition, extreme care was taken to gently mix the sample bottles prior to any sampling.

Experimental Incubations

Four experimental incubations were conducted for this study (experiment numbers by date 1212, 0205, 0226, 0325). Three experiments were conducted with mixing conditions (0205, 0226, 0325). One was without a mixed sample and used only samples fixed at the top and bottom of the column (1212). In experiments 0205 and 0325 the mixing samples traversed the length of columns every 24 hours and had top and bottom fixed controls. Experiment 0226 traversed the column length in 144 hours (6 days) and had a set of bottles in the environmental chamber at 15 � C where the cultures were kept. This treatment was simply to monitor for changes in the replicates that were due to age alone. A seven-day acclimation period was used in experiment 0325. This allowed cells to become adjusted to their new light regime before sampling was started. In figure 9, the end of the acclimation period is day eight on the horizontal axis. The other experiments did not undergo acclimation in the columns prior to the sampling.

The conditions for the study were originally chosen to maximize the possibility of detecting photoacclimation. This study uses light intensities that fall within those in the literature, including static and mixed studies, where photoacclimation has been noted in less than one photoperiod (12-8 hours). Also, this study allowed the cells much longer time to acclimate to the changing light climate, with gradient traversals of 24 hours or longer. Previous studies cited in this thesis have moved cells through gradients with ranges of scalar irradiance of 40-160, 15-167 and 30-320 mmol photons m^-2 sec^-1. The same studies had traversal times of 80 minutes (Kromkamp and Limbeek 1993) to two hours (Flameling and Kromkamp 1997). The original study design called for additional experiments to further refine the nature of acclimation by A. formosa to the experimental conditions. However, mechanical failures of the stepping motor prevented further experiments.

Previous studies have cited detectable changes in many physiological parameters due to diel cycles. Lacking the time to address these changes, I opted to minimize their influence on the results and sample every 24 hours. Care was taken to not phototraumatize the cells by exposing them to light that was outside their normal diel pattern. Therefore, sampling was always conducted when the lights were off and cells were �expecting� to be in the dark based upon a 12/12-light/dark cycle. The cycle was synchronized between the column lights and chamber lights. All experiments began or samples were taken before 07:00 or after 19:00. Once sampled, the only light that the cells experienced was that of the photosynthetron. The time between sampling and when samples were placed in the photosynthetron was always less than two hours. The laboratory was dimly lit with red lighting during all analyses for photosynthetic parameters and pigmentation.

Cell counting and particle corrections

Within two hours of sampling, all replicate bottles were analyzed to determine the number of cells per milliliter of sample, or the cell density of the sample. All pigment and photosynthetic analyses were done on a per cell basis, calculated using bulk parameters normalized for cell density. Cell density was determined using a Coulter Multiziser 2 particle counter. Multiple samples from each replicate bottle were counted and the average particle density was then corrected for cell count. The correction was created by visually counting the number of cells per colony in five samples. Visual observations were made on live material at 40X magnification on a compound microscope using a plain glass slide.

The resulting histograms of the number of cells per colony were compared to the distribution of particle sizes. A typical comparison is shown in figure 7.

Through periodic visual inspections, the only particles in the cultures in the size range of 10 to 60 mm were cells and colonies of A. formosa.

Therefore, Coulter particle size frequency histograms in this range resulted exclusively from differing colony sizes. While the cells of each colony displayed some variance in size, it was reasoned that the changes in colony size must be dominated by differences in the number of cells per colony, and not differences in individual cell sizes or morphologies. This would especially be true within each replicate bottle. Under this assumption, the distribution of particle size frequencies was then broken into groups, which represented a likely number of cells per colony.

Three to four particle counts for each replicate sample bottle were made. Text files from the particle counter were imported into a spreadsheet for conversion to cell densities per bottle. The particle/colony size corrections were made for each count individually and the resulting particle densities were averaged for each bottle.

Daily specific growth rates for the duration of the sampling period were calculated as follows:

where N(T1) is the cell density at the start (T1) and N(T2) is the density at the end of the study period (T2).

Pigment extraction and analysis

After removal from the incubators, the samples were kept in darkness and exposed only to dim, red lighting prior to filtration. Filtration was performed using sintered glass filtration stands and Gelman Supor 400 0.4 micrometers filters under 12 PSI of vacuum. Each stand was rinsed once with deionized water between samples. Between experiments the stands were soaked in 5% HCl then soaked in deionized water and rinsed before drying. In this way, blockage of the sintered glass surface was kept to a minimum.

After some experimentation, the optimal volume filtered for pigment extraction was determined. Of primary consideration was to keep the maximum optical density near 1.0 in the scanning spectrophotometer. Allowing the density to exceed 1.5 yielded too much noise in the response. When the particle concentrations were on the order of 2500 � 3000 particles per milliliter (as determined in the Coulter counter within the interval of 10-60 mm diameter), 150 milliliters were filtered onto each 47mm filter. If the particles densities fell below 2000, 200 milliliters were filtered.

Two filtrations for pigments were performed for each replicate bottle. Each filter was immediately folded into quarters and stored in aluminum foil packets in the freezer until the end of the experiment. Following the experiment, the filters were extracted all at once. Each filter was placed into 14.5 milliliters of 90% acetone buffered with magnesium carbonate. The acetone was dispensed directly into glass, 25-milliliter scintillation vials from a repeating dispenser stored in the freezer. Each filter was then steeped in the freezer for one week to maximize the extraction of accessory pigments. Rowan (1989) determined that after seven days of steeping, no additional significant pigment extraction would occur. Any disparity in terms of the percent of pigment extracted from samples taken at either end of experiments should not have produced any significant differences in results. No grinding or sonication was used, as others have reported sufficient extraction without these methods (Rowan 1989), Sandgren (personal communication), Cuhel (personal communication).

After seven days the contents of each vial were poured into 15 mL, graduated, screw top centrifuge tubes and centrifuged at 2500 RPMs for 15 minutes to remove particulates. The supernatant was then drawn off and stored in clean scintillation vials in the freezer. Final volume of each extract was about 12 mL. The pigment extracts were read in a Beckman 7000 DU diode array scanning spectrophotometer with a 1nm bandwidth spectral resolution. The cuvette was a 10cm quartz microcell with a glass lid to prevent evaporation during analysis. The samples were scanned for absorbance at the 1nm resolution from 360 to 760 nm, before and after acidification with 0.25 ml 0.1 N HCl (Figure 8a, b). Each sample generated two text files that were imported into a spreadsheet for calculations.

Using the methods presented by Arar (1997), which are based upon Jeffrey and Humphrey's trichromatic equations, the concentrations of chlorophyll a / pheophytin a and c1+c2 were calculated. Chlorophyll b was not measured as it is not present in the Bacillariophyceae(Rowan 1989). The basic equations used in these calculations were:

� Abs ### = (Spectrophotometric absorbance at ###) - (spectrophotometric absorbance at 750 nm)

Using this method, the interference from particles is subtracted from the absorbance at each of the critical wavelengths before they are used in the calculations. Particle absorbance is determined at 750 nm. Therefore, in the equations in this study, Abs 664 is

equal to the spectrophotometric absorbance minus the spectrophotometric absorbance at 750. Pheophytin a was determined using Lorenzen's pheopigment-corrected Chl a and pheophytin a calculations (Arar 1997; Lorenzen 1967).

� C = concentration (mg/L) of chlorophyll a in the extract E,a solution measured,

� Abs 664 = sample absorbance at 664 nm (minus b absorbance at 750 nm) measured before acidification, and

� Abs 665 = sample absorbance at 665 nm (minus a absorbance at 750 nm) measured after acidification.

The concentration of pigment per unit volume in the whole water was then calculated using the following equations

C_S =	C_E (a,b, or c) X extract volume (L) X DF		(6)
	sample volume (L) X cell length (cm)

� C_E = concentration (mg/l) of pigment in extract E(a,b,or c) measured in the cuvette..

In addition to chlorophylls, I attempted to determine the relative contributions of the diatom accessory pigments b-carotene and fucoxanthin through spectrophotometric absorbance. Quantitative analysis of these pigments was not possible because their absorbance overlaps with the short wavelengths, or Soret bands, of chlorophyll. However, I attempted to determine if the proportion of these accessory pigments to chlorophyll a changed in response to experimental conditions. The relative contribution of the accessory pigments to the overall in vitro photopigment content was determined by summation of the changes in the absorption maxima for that pigment. Critical wavelengths for each pigment dissolved in 90:10 acetone:water were collected from the literature (Jeffrey et al. 1997; Mantoura and Llewellyn 1983; Rowan 1989). Absorbance in these regions is referred to as fucoxanthin-like absorbance and b-carotene-like absorbance (FLA and BLA). This method of numerical analysis of pigment content is not found in the literature.

In addition to measuring simple absorbance in all regions associated with fucoxanthin and b-carotene, the ratio of the absorbance due to these pigments in relation to that of chlorophyll a was also calculated. For this parameter, the primary absorbance maximum of each pigment was compared to the red absorbance peak of chlorophyll a at 664 nm.

Photosynthetic parameters

Photosynthetic parameters of the rate of carbon fixation were determined using the ¹⁴C technique of Sandgren originally derived from Lewis and Smith (1983). A 110 ml sample was taken from each experimental bottle, to which 0.25 mCi of NaH¹⁴CO₃ was added. The bulk sample was then mixed and pipetted into 20 scintillation vials. Vials were then incubated in a photosynthetron maintained at 4 � C. Cool white fluorescent tubes provided light in the incubator. A light intensity gradient spanning the range from 7 to 465 mmol photons m^-2 sec^-1 was achieved by placing neutral density screening beneath individual vials in the photosynthetron. The vials were held in the incubator for 1.5 hours and then acidified and shaken overnight in a fume hood to drive off any remaining inorganic ¹⁴C. Radioactivity in the remaining sample was determined by adding 10 ml of Universol ES LSC scintillation cocktail to each vial and counting the samples in a Beckman liquid scintillation counter. The counter was calibrated six months prior to the study by staff from the radiation safety department. Raw counts were directly output to printer and logged to a text file on an attached PC. The PC was connected to the counter via a RS-232 serial connection and the stream was captured to text file using the program HyperTerminal supplied with the Windows 95 operating system. The resulting text file, which is formatted for reading and not for data processing, was then converted to table form using a command line program written for this purpose by Joe Terranova of the computer science department. The DATPARSE program (version 0.0.3) created a single comma separated table with the values for disintegrations per minute (DPM) occurring in the 17^th column. Vials were loaded into the counter in such an order that the counts for each curve could be transferred directly into a ¹⁴C assimilation spreadsheet built by Dr. Craig Sandgren. This spreadsheet required user inputs for the following variables:

The resulting fields in the spreadsheet allowed for the calculations of photosynthetic assimilation characteristics.

The resulting fields from each sample could then be transferred to a single table with identifiers indicating sample identity.

Photosynthetic parameters were derived both on a per-cell basis and per-cellular chlorophyll a basis. One, 13-point photosynthesis versus irradiance curve was fitted for each replicate bottle. The parameters for production, corrected for cell counts in the sample, were calculated using a statistical model-fitting package and the model listed in equation (12) (Platt and Jassby 1976). The three-parameter formulation was used primarily because of its simplicity and the fact that initial trials indicated excellent fitting was achieved. A second set of photosynthetic parameters were calculated by deriving the parameters from the fit of least-squares lines, then dividing the parameters by the chlorophyll content (Fee 1998). Throughout this study, cellular chlorophyll content (pg cell^-1) was substituted for the typical method of chlorophyll contained in a unit volume of sample. The computer application developed by Fee was used for the fitting of the chlorophyll-specific data.

Where P is the rate of carbon assimilation (photosynthesis) at light intensity, I. P_max is the maximum photosynthetic rate described by the data and is represented by the maximum height of the P versus I curve. a is the slope of the initial, light-limited part of the curve where the P versus I relationship is close to linear. R is a term often referred to as the respiration term, but mathematically is simply the y-intercept of the curve (Fahnenstiel et al. 1989).

The model was run in the statistical software package SYSTAT (version 9) and the expression was fitted using the Gauss-Newton method within the non-linear regression subset of the software package. Photosynthesis versus irradiance scatter plots were constructed within SYSTAT to determine outlier data points. The software allows users to select and remove outliers with tools in the graphics editor. Command files were used to process the data in batches, though the fit of each curve was verified using the plots generated by the software. Once the software determines the least-squares best fit of the data, the parameters are presented along with pertinent statistical information. Of greatest interest is the range of uncertainty associated with the parameter estimate. The confidence range around the parameter estimates are given as Wald confidence intervals. These are defined as the parameter estimate � t * the asymptotic standard error (A.S.E.) for the t distribution with residual degrees of freedom (SYSTAT 9 manual).

The second set of photosynthetic parameters was calculated using the application PSPARMS developed by Fee (Fee 1998). The models used in the application do not include a term for respiration and the fit of the light limited portion of the curve is forced through, or near, the origin at I_k/20. The program requires chlorophyll content as an input and the mean of the cellular chlorophyll a content was substituted for the typical bulk parameter of mg chlorophyll a L^-1. The application provides the sums squares deviation for the fit of the curve to the data as an error estimate. However, the deviation was provided only for the general fit of the curve and was not suitable for determining statistical significance of the estimates.

The total daily light dose (TDLD) was calculated from the known light intensity at the top and bottom of the column for the TOP and BOT samples. TDLD for the MIX samples was calculated from the PAR values, which was converted from the HLI logger sent along with each MIX bottle cluster with each experiment. With the PAR units in mmol photons m^-2 sec^-2 and the HLI recording a sample every five minutes, TDLD had units of mol photons m^-2 24 hours^-1 and was calculated as follows:

Where TDLD is only calculated for the lighted period between lights on (t_light) and lights off (t_dark). PAR is converted from HLI readings for the MIX samples or constants for the TOP and BOT and WALK treatments. D is the duration between recordings in seconds; in this case 300 seconds. Where the noise in the HLI readings sent calculated PAR values above those known to be possible, limits of the known light range were substituted in the calculations.

Results

In experiment 0205 MIX bottles traversed the light gradient twice, once every 24 hours, and were sampled every 12 hours (Figure 9). The entire experiment lasted 48 hours. In experiment 0226 samples made the same number of gradient traversals as 0205, but it required 144 hours, or six days, for each traversal. Under these conditions, the cells were exposed to six days of increasing light intensity and had nearly an entire day at full intensity before descending through a decreasing light gradient. Experiment 0325 used the same mixing period as 0205, but added seven days of acclimation prior to the first sampling and lasted 96 hours instead of 48.

Growth rates

Experiment 0205, with the shortest duration of all the experiments, showed little change in the cell density over the 48 hours (Figure 10). Cell counts ranged from 11,945 cells mL^-1 at the start, to 13,781 cells mL^-1 at 48 hours in the TOP bottles where light was maintained at ~250 mmols throughout the experiment. In comparing the influence of increasing overall light availability, the un-mixed BOT, mixed and un-mixed TOP samples increased cell density at the daily specific growth rate of 0.02, 0.03 and 0.07, over the course of the experiment respectively (Appendix A). A regression of cell density against cumulative light exposure by treatment indicated that throughout the study the 0205 and 0325 experiments were significantly correlated with a treatment�s light exposure (Table 3, Figure 11).

Figure 9.Photosynthetically available radiation received by each treatment.

Photosynthetically available radiation received by each treatment.

Recorded light intensity for each mixed experiment (gray dots). The gray dots

also generally describe the position of the bottles in the column with lights on.

Bars along the X axis indicate 12 hour dark periods.The second vertical axis

describes the 24-hour summation of light that each replicate bottle received

during an experiment. This total daily light dose (TDLD) for the mixed

treatments is represented by large dots. Solid dots indicate TDLD's at sampling

times and open circles indicate TDLD's at times when no samples were taken.

The dashed horizontal lines represent TDLD for the control bottles. The upper

lines are for the TOP controls and the lower lines are for the BOT controls.

Experiment 0226 had a single set of controls left in the environmental chamber.

Cell densities in the second experiment (0226), which ran for 288 hours, ranged from 17,591 at the start of the experiment, to 30,010 after 288 hours in the walk-in chamber (0.04 specific growth rate). The cell density in the mixed samples was 23,377 cells mL^-1 at the conclusion of the experiment (0.02 specific growth rate). A plot of cell density against cumulative light history for the two treatments indicated that growth rate was much higher in the chamber where both cumulative light and temperatures were higher (Table 3, Figure 11).

Cells in the third experiment, having spent seven days acclimating to varied light climates showed a divergence in cell density between treatments by the first day of sampling. No cell counts were taken prior to the acclimation period. Following the acclimation period cell densities in the BOT treatment, were 11,914 cells mL^-1 and declined to 9,874 by the end of the experiment (-0.05 specific growth rate). The BOT treatment showed decreased cell density each sampling period over the entire 96-hour sampling period. Bottles in the MIX treatment increased slightly from 13,494 to 13,836 during the same four-day sampling period (0.01 specific growth rate). Cell density in the fixed TOP treatment increased the greatest from 19,274 to 25,551 cells per mL over the sampling period (0.07 specific growth rate).

Pigmentation

38.7 mg L^-1 at the start of the experiment (Figure 12 and Appendix B). After 12 hours, chlorophyll increased in all treatments. However, by 24 hours chlorophyll in the BOT and MIX treatments both declined to 37.5 and 37.0 mg L^-1 respectively. Chlorophyll in the 24-hour TOP sample declined from the 12-hour sample, but was still higher than the starting value. By 48 hours, all samples had increased whole water chlorophyll a over initial concentrations. When the whole water chlorophyll concentrations were normalized on a per cell basis, thereby providing an estimate of cellular pigment content, the trends were the same over the course of the experiment. The initial pigment content per cell was 1.9 pg cell^-1, which increased to 2.3, 1.9 and 2.4 for the BOT, MIX and TOP treatments after 48 hours.

In addition to chlorophyll a, chlorophyll c1+c2 and the carotenoids fucoxanthin and b-carotene were measured. Although the absolute concentrations of the carotenoids could not be determined spectrophotometrically, their relative importance to the in vitro pigment absorption was derived for their associated absorption maximums.

For the first experiment, fucoxanthin-like absorbance (FLA) and b-carotene-like absorbance (BLA) correlated strongly with each other and with chlorophyll a only in the MIX samples (Appendix C). TheTOP and BOT samples had strong correlation between fucoxanthin and b carotene only, and neither pigment correlated well with chlorophyll a.

Chlorophyll a in the whole samples for experiment 0226 increased in a more predictable fashion, increasing in both the mixed and stationary (WALK) treatments at every sampling period (Figure 13b). The initial sample contained 7.0 mg L^-1 that increased slightly after 24 hours to 8.4 and 10.5 mg L^-1 for the MIX and WALK treatments respectively. After 144 hours the chlorophyll a in the MIX treatment had doubled to 15.2 mg L^-1 while the chlorophyll a in the WALK samples increased by a factor of 4.5 to 31.5 mg L^-1. By 288 hours the MIX samples increased by another 8 mg L^-1 while the increases in the WALK samples slowed so that only 4 mg L^-1 was accumulated over that observed in the 144-hour samples. As in the first experiment, cellular pigment contents followed the same temporal relationships as that in the whole samples. This indicates that changes in whole sample pigment concentrations were more influenced by population growth than by cells synthesizing more pigment. The initial value was 0.4 pg cell^-1 which increased to 1.0 and 1.2 for the MIX and WALK treatments after 288 hours. In the last 144 hours, the MIX cells added 0.26 pg cell^-1 chlorophyll a, while the WALK cells added only 0.15. FLA and BLA correlated strongly with each other and with chlorophyll a in the WALK treatment. Though only the FLA:BLA correlated strongly in the mixed samples.

Chlorophyll a in experiment three (0325) followed a similar trend as cell density in the same experiment, which was generally to attain three very different concentrations by treatment before the sampling phase began. Chlorophyll a increased steadily in the TOP samples throughout the sampling period from 11.0 mg L^-1 at time zero and finishing at 32.86 mg L^-1 96 hours later. However, both the BOT and MIX treatments lost chlorophyll a over the course of the experiment. Again, the same trends were noticed in the cellular chlorophyll content as in the whole sample chlorophyll. Samples in the TOP treatment added cellular chlorophyll a (0.16 to 1.7 pg cell^-1) despite being continuously expose to irradiances of about 110 to 225 mmol throughout the bottle cluster. Samples in the MIX and low light, BOT treatments lost chlorophyll over the 96 hours. The final experiment indicated strong correlation between fucoxanthin, b carotene and chlorophyll in all treatments throughout the experiment.

Regression analysis of various pigments against TDLD was conducted to determine if particular pigment concentrations would be responsive to a cell�s light history. Cellular estimates of chlorophyll a, FLA, BLA, Pheo a and Chl c were regressed against the daily light dose received by each treatment. The relationships were significantly linear and positive for all pigments in the 0325 experiment. There were large differences in the number of replicates for each experiment in the regression analysis. This makes the comparison of significance between experiments difficult.

Photosynthetic capacity and efficiency

As stated earlier, C assimilation across a range of irradiances was parameterized for each replicate sample in two ways. The first used a statistical software package to fit the points to a model that included a term that describes the Y intercept of the fitted curve. The points used in this statistical routine were whole sample C assimilation values normalized to cell density in the sample. The parameters derived from this method are hereafter referred to as P_max and a. Scatter plots of the cell-specific photosynthesis (CELLC14) versus irradiance curves allow visual interpretation of the relationships between the treatments (Figure 13 a,b,c).

The second method fitted the same replicates using a similar model and a program developed specifically for determining photosynthesis versus irradiance parameters (Fee 1998).

The replicates of C assimilation values were then normalized to the cell density in the sample and the mean concentration of chlorophyll a per cell. The cell-specific parameters are referred to as P^cell_max and a^cell, and the cellular chlorophyll-specific parameters are referred to as P^B_max and a^B. P_max/a and P^cell_max/a^cell describe the same data, but represent two methods of calculating the parameters with P_max and a being calculated with the Platt 1976 model with R and P^cell_max and a^cell calculated with the Fee applications. P^B_max/a^B are not normalized to whole water chlorophyll a, but rather to cellular chlorophyll a in pg cell^-1 (B^cell).

For the parameters P_max and a, visual inspection and one-way ANOVA analysis indicated no difference between any two temporally consecutive parameters from the same treatment in the 48-hour experiment (0205) or the 288-hour experiment (0226) (Table 3). That is, no coherent trend was seen in figure 14 in relation to daily light exposure (dots). The uncertainty in the parameter estimates (described in the figure with bars showing the 5% confidence intervals) indicate that the parameters did not change statistically during the experiment. However, some qualitative changes in the means are discernable.

Only the 0325 experiment, which acclimated the diatom cultures to the experimental light regime prior to any photosynthetic assays, showed statistically significant change in photosynthetic parameters within a single experiment. A one-way ANOVA categorically comparing P_max and a with TDLDs as categories determined that the third and fifth measurements of Pmax in experiment 0325 were significantly distinct from the other three MIX sample periods at the 5% level (Table 2). A Tukey post hoc test confirmed the significance and also showed that sample periods 4 and 5 (72 and 96 hours) are not statistically different from each other, but were significantly different from 2 and 3 (24 and 48 hours). Samples 1,2 and 3 are not statistically separate values.

Table 2. One-way ANOVA with Tukey post hoc of 0325 MIX samples by TDLD
Tukey multiple pairwise comparisons of probability that P^cell_max in the five MIX samples are distinct values.
	1	2	3	4	5
1	1
2	0.90	1
3	0.99	1.00	1
4	0.08	0.02	0.07	1
5	0.05	0.02	0.05	1.00	1

In a scatter plot of the P_max values for the individual replicate MIX bottles versus TDLD, distinct clusters of points occur only in the 0325 experiment (Figure 16). In the previous two experiments, there was no clear trend in the relationship between 24-hour light history and maximum photosynthetic capacity. This relationship is also seen in a multiplot of the treatment means of P_max and a plotted against the TDLD (Figure 17). In this figure, the results from the practice experiment 1212 are also shown. In experiment 1212 only static TOP and BOT samples were incubated for two days at high and low light. By the second day, P_max and a had shifted very slightly from the initial reading.

Lack of estimates of variance for the parameter estimates of P^cell_max/a^cell_maxand P^B_max/a^B prevent statistical comparisons to verify whether two estimates are distinct. However, the values did reveal interesting relationships (Figures 16,17). Neither P^cell_max nor P^B_maxfor experiment 0205 varied throughout the experiment and generally agrees with the lack of variation P_max for 0205.

For experiment 0226, a^B varied little for the chamber treatment where conditions were consistent for the entire experiment. A slight decrease in a^B for the chamber during the experiment is not seen in the a^cell parameter, which indicates an increase in cellular chlorophyll with no change in a itself. In the MIX treatment, acell increases slowly throughout the experiment, though a^B declines sharply at 144 hours, along with an increase in B^cell.

Discussion

The definition of photoacclimation for this study is a clear correlation between the light history of a diatom species and the measure of a variable indicative of their photophysiological status. This study was conducted to address the hypothesis that photoacclimation occurs, and can be detected, in algae that traverse a light gradient that ranged 200 mmol over a period of hours to days.

Common parameters that can describe the physiological status of an algal population include maximum photosynthetic capacity (P_max) and efficiency (a), and chlorophyll-specific absorption cross-sections (Anning et al. 2000; Falkowski and Owens 1980; Falkowski and Raven 1997; Flameling and Kromkamp 1997; Vidussi et al. 1999; Vincent et al. 1994). The original goals of the project were to detect and quantify photoacclimation in response to a range of mixing velocities that would move algal cells through a light gradient at different rates. If possible, I hoped to determine an upper boundary of mixing rate, beyond which the rate of mixing would exceed the rate of acclimation and no difference in photosynthetic state would be discernable between cells taken at the top and bottom of the mixing cycle. Despite the slow rate of simulated mixing used throughout this study, photoacclimation was suggested in two experiments, and confirmed in only the final experiment.

Photoacclimation in pigmentation

Although state transitions and non-photochemical quenching mechanisms in the photosynthetic biochemical pathways can adjust light-absorbing cross-sections of cells on the order of minutes, photoacclimation is generally referred to as a longer-term process (Falkowski and Raven 1997). They go on to cite others when they describe the potential for chlorophyll per cell or per unit surface area to increase five- to ten-fold as irradiance decreases (Falkowski and Owens 1980; Prezelin and Matlick 1980; Ramus 1990; Richardson et al. 1983). However, there are cellular chlorophyll minima that occur at both ends of a continuum from low to high irradiances. At low irradiance, the cells become chlorotic, or lacking color, and respond by increasing cellular chlorophyll a content until a maximum quota is reached at an intermediate irradiance. At irradiances that are higher still, cells respond by decreasing chlorophyll content and production until a minimum is reached. For example, an experiment using Skeletonema costatum and Dunaliella tertiolecta found that a low cellular chlorophyll content at an irradiance of 1-10 mmol quanta m^-2 sec^-1 increased to a maximum at ~20 mmol quanta m^-2 sec^-1. With higher irradiance, a decline in cellular chlorophyll was seen until ~400 mmol quanta m^-2 sec^-1(Falkowski and Owens 1980).

Contrary to expectations chlorophyll increased more rapidly in the high-light WALK and TOP treatments than lower light treatments in each of my experiments. This suggests that cells may have been light limited and had not reached their maximum chlorophyll content. However, cellular chlorophyll in experiment 0205 did not increase significantly and began with initial replicates with cellular chlorophyll already just below 2.0 pg cell^-1. This experiment also showed the highest per cell chlorophyll concentration of the study with a single treatment having an average 2.5 pg cell^-1. This represents a cellular maximum for the conditions in this study. The remaining treatments in experiment 0205 averaged around 2.0 pg cell^-1. The TOP treatments in 0226 and 0325 experiments increased cellular pigment from about 0.4 at the start of each sampling period to a maximum of about 1.5 pg by the time each experiment ended. The continual increase in cellular chlorophyll a in the second two experiments is consistent with the assumption above that cell pigment maximums were about 2.5 pg.

The lack of chlorophyll production in the BOT replicates of experiment 0325, where light was ~10 mmol, falls within expectation based upon the previous studies on diatoms described above. Similarly, the MIX samples which were often below 50 mmol lost chlorophyll in experiment 0325 and only gained chlorophyll after 24 hours in experiment 0226 when the cells emerged from a long duration at low light.

Pigmentation did not correlate with light history in any consistent way except that cells with exposure to high light approached cellular chlorophyll a maximums for this study. At the same time, the light limited bottom treatments appeared to be unable to synthesize more pigment. Variability in the pigmentation in each treatment was greatest in the top samples where the light gradient due to the position of the bottles in relation to the attenuation of light was steepest. Simply due to the attenuation of light in the columns, bottles at the top of the TOP cluster experienced average irradiance of about 225 mmol, while those at the bottom of this cluster received slightly over 110 mmol. Shading by the bottles themselves may have further increased the differential.

The amount of chlorophyll a and the estimate of accessory pigments did not change relative to each other throughout the study. That is, changes in pigmentation seemed to occur for all photosynthetic pigments in concert. Throughout the study, chlorophyll generally correlated with FLA and BLA. However, in some cases the relationships were inexplicably negative or in some cases not correlated at all. No clear trend was seen in regards to treatment of light exposure, which would have also hinted at a chromatic response (Appendix C).

The qualitative changes in the light environment within the experimental columns were not quantified, however some information was gained through simple observation. Light reaching the bottom of the columns was visible when clear plates of Plexiglass on the column bottoms were uncovered. By holding a white piece of paper underneath the plates, the white light entering the columns at the top was blue at the bottom. It was clear that the longer visible wavelengths had been filtered by the water in the columns. The scattering and refracting influence of the bubbles in the columns appeared to mute this selective filtering and the light became slightly more white when bubbles were present. Without a spectroradiometer there was no way of knowing the exact spectral quality that reached the bottom of the columns. However, despite the bluing of the light, Morel et al. 1987 determined that light intensity was of much greater importance to photoacclimation than light quality in the diatom Chaetoceros protuberans. (Morel et al. 1987)

The changes in light color with depth in the columns were apparent. According to theories of chromatic acclimation, A. Formosa should have modified its pigment ratios to optimize its photosynthetic action spectra for the bluer light deep in the columns. Kirk (1996) summarizes numerous studies and taxa-specific responses when he says that accessory pigments will often increase more rapidly than chlorophyll a in response to lowered light or shifted spectra. With fucoxanthin as the primary, photosynthetically coupled accessory pigment, any chromatic acclimation should have been detected through a ratio change with chlorophyll a. However, the lack of consistent shift in the concentration of pigments in relation to each other might have indicated that the light gradient and the change in quality was not great enough for the cells to undergo chromatic acclimation. Although shifts in pigment ratios due to chromatic acclimation are common across algal groups, further studies cited by Kirk (1996) seem to indicate that fucoxanthin is particularly stable when faced with shifts in light quality (Brown and Richardson 1968; Shimura and Fujita 1975). This lack of plasticity in fucoxanthin:chlorophyll ratios agreed with results described previously in this study.

Photoacclimation in photosynthesis

Detection of a coherent pattern of acclimation in this study was difficult. The results from the cellular pigmentation and photosynthetic parameters indicate that changes in the cells held under control conditions were occurring despite being held in an unchanging environment. Comparison of the photosynthetic parameters P^cell_max and a^cell describe cell-specific acclimation regardless of the cellular chlorophyll content. These parameters can describe the particular strategy of acclimation by the plankton. Photosynthesis versus irradiance relationships will tend to change with changes in size or number of photosynthetic units (PSUs).

Increasing number or size of PSUs each represents a different strategy for photoacclimation and was thought to be taxa specific. However, conflicting reports have prevented simple taxonomic-based classifications (Falkowski and La Roche 1991). Recently, some authors have proposed that the nature of the photosynthetic response is more due to the peak intensity received by the cells or the total light dose (Ibelings et al. 1994; Kromkamp and Limbeek 1993; Kroon et al. 1992b). In a simple visualization of P_max versus TDLD for the MIX treatment replicates, only the 0325 experiment showed a positive relationship between the maximal photosynthetic rate and the overall light exposure (Figure 16).

If the size of the PSU is increased, a^cell will increase due to an increase in the number of light harvesting pigment molecules. However, P^cell_maxwill not change because additional reactions centers are not added. If P^cell_max increases, it is due to the synthesis of more PSUs or reactions centers. The addition of reaction centers and PSUs will also raise a^cellbecause they contain chlorophyll a (Falkowski and La Roche 1991; Lewis et al. 1984a).

Use of chlorophyll as the normalization factor describes the efficiency of photosynthesis, whereas simply normalizing photosynthesis per cell provides a measure of how effective a cell is at assimilating carbon.

It is important to remember that the TOP and BOT controls in the first experiment provide secondary results that describe a situation as found in studies such Anning et al. (2000). Those experiments were conducted using abrupt changes in light intensity to study acclimation. The same changes were included in this experiment as well as the oscillating MIX treatment. Because the cells were initially grown at ~100 mmols (INIT) and placed directly into the BOT treatment at ~10 mmols or the TOP treatment at ~250 mmols, these treatments may be used as time series test to determine when a steady state is reached after a shift to a new light intensity.

In the BOT treatment, a^cell continued to climb at a rate of about 2 mg C (20,000 cells)^-1 mmol photons^�1 m^-2 per day over the entire experiment and never came to a complete steady state (Figure 15 D). The increase in photosynthetic efficiency in the BOT cells may have been reflected in the B^cell values as they increased from 1.9 pg chl a cell^-1 in the INIT replicates to 2.0 at 24 hours then to 2.3 at 48 hours (Appendix F). The sample taken after 12 hours showed a value of 2.2, but this may have been due to diel effects and cell cycle differences between the other two samples. The influence of diel effects within each experiment were removed because each sample was taken at 24 hour intervals, except for 0205 where samples were taken at 12 hours intervals as well. Because a^cell increased steadily and P^cell_max decline steadily, I_k^cell also declined over the experiment in a response cited in most static photoacclimation studies. Unfortunately, the TOP and MIX treatments in 0205 did not have a consistent response to their new light climate.

In a study of Skeletonema costatum, Kromkamp and Limbeek (1993) determined that during acclimation to a fluctuating light climate (PAR oscillations every two hours at 17 � C), the diatom S. costatum decreased photosynthetic unit (PSU) size and increased the number of PSUs. They also found that when the peak PAR exposure throughout the cycle was 167 mmol m^-2 sec^-2 (at ~30 mmol, TDLD=0.86 mol day^-1), the photoacclimation response was muted when compared to similar treatments which oscillated from 100 �320 mmol m^-2 sec^-1 (TDLD=1.72 mol day^-1). Because this study did not consider chlorophyll-specific photosynthetic cross-sections, I cannot describe with certainty the strategy of photoacclimation employed by A. formosa. However, cell-specific photosynthesis parameters, modeled using the Fee application, suggest that A. formosa both increased the number and size of PSUs when adjusting to low light in all experiments (Figure 15).

Influences of temperature and diel periodcity on photosynthesis

Diatoms as a group are tolerant of cool to cold temperatures within the range of those used in this study. As evidence of this, measurements taken in Antarctic freshwaters at 0 to 4 � C reveal estimates of photosynthetic rates similar to, or greater than, temperate or tropical waters (Hawes 1990). Butterwick et al. studied the implications of temperature on growth rates for several species of freshwater phytoplankton. Under a constant light of 100 mmol, Asterionella formosa and seven other species were grown at temperatures from 2 � to 35 � C. A. formosa had a peak growth rate of 1.68 divisions day ^�1 at 20 � C and was the most tolerant of low temperatures with a growth rate of 0.61 divisions day ^�1 at 5 � C (Butterwick et al. 1987). Kozitskaya (1992) conducted a similar study with four marine species including the diatoms Phaeodactylum tricornutum and Navicula atomus at temperatures of 10 �, 20 � and 30 � C. She found that at 10 � C both species had average growth rates over the course of 20-25 days, of 0.059 and 0.110 divisions day ^�1 for P. tricornutum and N. atomus respectively. She also found that as a rule, diatoms in her study preferred temperature at or below 15 � C.

Cold temperatures appear to have the effect of prematurely inducing photoinhibition and excessive photoexcitation in cells as they are brought into high light from low light (Falkowski and Raven 1997; Platt et al. 1980). The primary cause for this is that the rate of photon capture is largely temperature independent, as indicated by the lack of sensitivity of light-limited photosynthetic efficiency (a) to changes in temperature. However, the rate of enzyme-mediated electron transport is temperature sensitive and the maximum rate of photosynthesis (P_max) will decline with decreasing temperature. Therefore, reduced temperatures allow for the same rate of photon capture, but a reduced ability to process potentially damaging excessive photo excitation. This effect is further aggravated when cells are shade-adapted and their photon-capturing �antennae� are fully exposed.

For these reasons, the potential for a photoprotective response by the cells seemed possible. However, there were no unusual increases b-carotene throughout the study that would have indicated a photoprotective response. However, I was unable to measure the other xanthophyll photoprotectants, which could have increased when the low-light acclimated cells were brought into the high light present particularly at the top of the columns.

Thompson et al. (1992) in a study of the effects of temperature on eight species of marine phytoplankton found that chlorophyll a per cell was always lower at 10 � C than at 25 � C, for the same light intensities. Similarly, the carbon:chlorophyll a ratios increased with increasing temperatures for all species, suggesting a temperature-mediated partitioning of resources. The cold conditions likely hampered the cells� abilities to photoacclimate under all light environments. Cells that are limited by temperature and light had the least ability to photoacclimate, despite the fact that the media was replete with nutrients. At the end of experiment 0205 dissolved phosphorus was still above 8.2 mmol L^-1 after beginning the experiment at 8.5 mmol L^-1. Phosphorus does not become limiting until well below 1 mmol L^-1. Dissolved silica, the next likely limiting nutrient, was not measured (Kilham and Tilman 1976).

Most prior studies have considered photoacclimation under conditions that were 15 � C or warmer (Anning et al. 2000; Cullen and Lewis 1988; Falkowski 1980; Falkowski 1984a; Geider et al. 1986; Grobbelaar et al. 1995; Grobbelaar et al. 1996; Ibelings et al. 1994; Marra 1978). There is strong experimental evidence that Pmax strongly correlates with temperature, though a does not consistently do so (Coles and Jones 2000; Jones 1998; Keller 1988). This appears to be true in this study where P^cell_max remained constant in the replicates that were left in the environmental chamber (WALK), while bottles that were placed in the columns that were 11 � C colder showed depressed, but largely invariant P^cell_maxs over the experiment. In the same experiment, a^cell actually increased in both treatments, which was expected for the MIX bottles, but not those left in the constant conditions of the chamber.

Only after a ten day entrainment in the oscillating irradiance did photoacclimation become significance enough to confirm. These results suggest the possibility that under poorly lit and cold conditions, diatom photoacclimation occurs more slowly than indicated by previous studies conducted under warmer and brighter conditions. The reduction in available light energy and suppression of biosynthesis by sub-optimal temperatures appeared to reduce the ability of A. Formosa tooptimize its photophysiology to match its slow traversal of a light gradient.

Others have suggested the potential for phytoplankton to become accustomed to regular light fluctuations overlaid with normal diel patterns (Prezelin and Matlick 1980; Prezelin and Sweeney 1977). That ability would explain the results found in the final experiment. Such a complex adaptability to a fluid environment that is known for being highly turbulent and irregular seems unlikely. Turbulence at the scale of Langmuir circulations or smaller likely present highly variable oscillations of the light environment to the entrained plankton.

Implications for in situ primary production

Although complete darkness was not achieved at the bottom of the columns, the cells were unable to gather enough light to synthesize significant quantities of pigment, improve their photosynthetic efficiency or increase in number. This light level, which proved insufficient to support cell growth, approximates that observed at 16.5 meters at the Fox Point station during the spring (Figure 18).

A. formosas� ability to photosynthesize at low irradiances seemed to be impacted by their low chlorophyll, as P^cell_maxand a^cell were slow to increase during experiment 0325 and followed cell chlorophyll content. Low chlorophyll and the requirement for sufficient photon capturing ability to synthesize more represent a negative compounding effect. For example, the BOT samples in this study were shown to be unable to synthesize photopigments, perhaps due to light limitations. This suggests that photoacclimation at depths near 16.5m, or deeper, would be hampered or even cease if cells remained at such low intensities for a week or longer. That is, the compensation depth would be shallower than 16.5 m.

Experiment 0325 indicates that cells would have the ability to acclimate if they were entrained in turbulence circulation near the surface for long periods. However, due to their severe light stress, cells that spent long periods in darker conditions would suffer a lag in their ability to rapidly acclimate to higher light. This may account for descriptions of hystersis in the literature (Cullen and Lewis 1988). It also might mean that in profiles of algal cells and chlorophyll in Lake Michigan, cells that were taken from consistently dark depths (<10 mmols) and are found to be deficient in chlorophyll a, likely have spent seven days or more at or below that depth. My study suggests that cells that are mixed to greater depths must get to the surface often enough to capture sufficient light to conduct biosynthesis and hence photoacclimate.

This study did not detect strong photoacclimation in samples that were immediately introduced into an oscillating light regime after an extended period under static lighting. Some results indicated that photoacclimation was occurring in the samples that were passed through a very slowly oscillating light environment. However, that acclimation did not result in the modulation of cellular pigment content. Instead, changes in the �package effect� of the photosynthetic apparatus in the cells likely accounted for the adjustments in photosynthetic capacity.

The relationship between I_k^cell and the total daily light dose is a classical one that agrees with expectations. The offset of the data points from the WALK treatment in figure 19 may be because they were much warmer than the rest of the treatments at 4 � C. Unfortunately, the trend should have raise I_k values above the colder treatments, not decreased it. As the cells are exposed to greater and greater amounts of light, their maximal photosynthetic rates increase to make use of the added light. As the depth of mixing shoals and seasonal light intensity increases in the spring, the community photosynthetic status should move from left to right in figure 19, and degrade in the fall.

Conclusion

Asterionella formosa used both increases in the number and size of photosynthetic units to acclimate to lower irradiances. These increases occurred within 24 hours of being introduced into a new light climate, though clear trends in the response of photosynthetic parameters and pigments to light history was tenuous in two of three experiments. In general, the photoacclimation response included synthesis of photosynthetic pigments that appeared to be proportional to each other across all light histories and maximal intensities. This proportionality in pigmentation included samples taken from the darker and bluer BOT treatments, where the potential for chromatic adaptation to alter pigment ratios was highest. However, after one week the cells in the BOT treatment became so light limited that they were incapable of synthesizing pigment and the cell density began to decline. This indicates that cells require certain total light dose in order to successfully photoacclimate to dark conditions.

Contrary to published trends, the TOP samples in all experiments increased or maintained chlorophyll content despite being in light that should not be limiting. Unless the light at the top was indeed sub-saturating to the cells. It appeared that the cells needed to add chlorophyll until they approached a maximum of about 2.5 pg cell^-1, where cells in experiment 0205 seemed to maintain their content. The final experiment yielded the most consistent evidence of periodic photoacclimation that was correlated with daily light history. The fact that this periodic acclimation only became apparent after ten days under the mixed light regime suggests that acclimation to a non-diel, cyclic light cycle may be occurring.

Bibliography

Anning, T., H. L. MacIntyre, S. M. Pratt, P. J. Sammes, S. Gibb and R. J. Geider (2000). �Photoacclimation in the marine diatom Skeletonema costatum.� Limnology and Oceanography 45(8): 1807-1817.

Arar, E. J. (1997). Method 446.0: In vitro determination of chlorophylls a, b, c1 + c2 and pheopigments in marine and freshwater algae by visible spectrophotometry. Cincinnati, OH 45268, National Exposure Research Laboratory, Office of Research and Development, U.S. Environmental Protection Agency: 1-26.

Berner, T., Z. Dubinsky, K. Wyman and P. G. Falkowski (1989). �Photoadaptation and the "package effect" in Dunaliella tertiolecta (Chlorophyceae).� J. Phycol. 25(70-78).

Brooks, A. S. and D. N. Edgington (1994). �Biogeochemical Control of Phosphorus Cycling and Primary Production in Lake-Michigan.� Limnology and Oceanography 39(4): 961-968.

Brown, T. E. and F. L. Richardson (1968). �The effect of growth environment on the physiology of algae: light intensity.� J. Phycol. 4: 38-54.

Butterwick, C., S. I. Heaney and J. F. Talling (1987). �The Influence of Temperature on Algal Growth.� British Phycological Journal 22(3): 301-301.

Coles, J. F. and R. C. Jones (2000). �Effect of temperature on photosynthesis-light response and growth of four phytoplankton species isolated from a tidal freshwater river.� Journal of Phycology 36(1): 7-16.

Cuhel, R. L., D. Ortner and R. S. Lean (1984). �Night synthesis of protein by algae.� Limnol. Oceanogr. 29(4): 731-744.

Cullen, J. J. and M. R. Lewis (1988). �The kinetics of algal photoadaptation in the context of vertical mixing.� J. Plankton Res. 10(5): 1039-1063.

Dubinsky, Z., P. G. Falkowski and K. Wyman (1986). �Light harvesting and utilization by phytoplankton.� Plant Cell Physiology 27: 1335-1349.

Dusenberry, J. A. (2000). �Steady-state single cell model simulations of photoacclimation in a vertically mixed layer: implications for biological tracer studies and primary productivity.� Journal of Marine Systems 24(3-4): 201-220.

Eilers, P. H. C. and J. C. H. Peeters (1988). �A Model for the Relationship between Light-Intensity and the Rate of Photosynthesis in Phytoplankton.� Ecological Modelling 42(3-4): 199-215.

Fahnenstiel, G. L., J. F. Chandler, H. J. Carrick and D. Scavia (1989). �Photosynthetic characteristics of phytoplankton communities in Lakes Huron and Michigan: P-I parameters and end-products.� Journal of Great Lakes Research 15(3): 394-407.

Falkowski, P. G. (1980). Light shade adaptation in marine phytoplankton. New York, NY, Plenum Press.

Falkowski, P. G. (1983). �Light-shade adaptation and vertical mixing of marine phytoplankton: A comparative field study.� J. Marine Res. 41: 215-237.

Falkowski, P. G. (1984a). �Kinetics of adaption to irradiance in Dunaliella tertiolecta.� Photosynthetica 18: 62-68.

Falkowski, P. G. (1984b). �Physiological responses of phytoplankton to natural light regimes.� J. Plankton Res. 6(6): 295-307.

Falkowski, P. G., Z. Dubinsky and K. Wyman (1985). �Growth-irradiance relationships in marine phytoplankton.� Limnol. Oceanogr. 30(2): 311-321.

Falkowski, P. G. and J. La Roche (1991). �Acclimation to spectral irradiance in algae.� J. Phycol.

Falkowski, P. G. and T. G. Owens (1980). �Light-shade adaptation: two strategies in marine phytoplankton.� Plant Physiology 66: 592-595.

Falkowski, P. G. and J. Raven (1997). Aquatic photosynthesis. Malden, MA, Blackwell Science, Inc.

Falkowski, P. G. and C. D. Wirick (1981). �A simulation model of the effects of vertcal mixing on primary productivity.� Mar. Biol. 65: 69-75.

Farmer, D., J. Gemmrich and V. Polonichko (1998). Velocity, temperature and spatial structure of Langmuir circulation. Physical Processes in Lakes and Oceans. J. Imberger. Washington, D. C., American Geophysical Union. 54: 87-100.

Fee, E. J. (1998). �Revision of: Computer programs for calculating in situ phytoplankton photosynthesis.� Can. Tech. Rept. Fisheries and Aquatic Sci. 1740.

Flameling, I. A. and J. Kromkamp (1997). �Photoacclimation of Scenedesmus protuberans (Chlorophyceae) to fluctuating irradiances simulating vertical mixing.� Journal of Plankton Research 19(8): 1011-1024.

Gallegos, C. L. and T. Platt (1982). �Phytoplankton production and water motion in surface mixed layers.� Deep-Sea Research 29(1a): 65-76.

Geider, R. J., B. A. Osborne and J. A. Raven (1986). �Growth photosynthesis and maintenance metabolic cost in the diatom Phaeodactylum tricornutum at very low levels.� J. Phycol. 22: 39-48.

Goericke, R. and N. A. Welschmeyer (1992). �Pigment turnover in Thalassiosira weisflogii. II. The 14CO2 labelling kinetics of carotenoids in a marine diatom.� J. Phycol. 28: 507-517.

Grobbelaar, J. U., B. M. A. Kroon, T. Burgerwiersma and L. R. Mur (1992). �Influence of Medium Frequency Light Dark Cycles of Equal Duration on the Photosynthesis and Respiration of Chlorella- Pyrenoidosa.� Hydrobiologia 238: 53-62.

Grobbelaar, J. U., L. Nedbal, L. Tichy and I. Setlik (1995). �Variation in Some Photosynthetic Characteristics of Microalgae Cultured in Outdoor Thin-Layered Sloping Reactors.� Journal of Applied Phycology 7(2): 175-184.

Grobbelaar, J. U., L. Nedbal and V. Tichy (1996). �Influence of high frequency light/dark fluctuations on photosynthetic characteristics of microalgae photoacclimated to different light intensities and implications for mass algal cultivation.� Journal of Applied Phycology 8(4-5): 335-343.

Harding, L. W. J., B. W. Meeson, B. B. Prezelin and B. M. Sweeney (1981). �Diel periodicity of photosynthsis in marine phytoplankton.� Mar. Biol. 61: 95-105.

Harding, L. W. J., B. B. Prezelin, B. M. Sweeney and J. L. Cox (1982). �Primary production as influenced by diel periodicity of phytoplankton photosynthesis.� Mar. Biol. 67: 179-186.

Hawes, I. (1990). �The effects of light and temperature on photosynthate in Antarctic freshwater phytoplankton.� Journal of plankton research 12(3): 513.

Ibelings, B. W., B. M. A. Kroon and L. R. Mur (1994). �Acclimation of photosystem II in a cyanobacterium and an eukaryotic green-alga to high and fluctuating photosynthetic photon flux densities, simulating light regimes induced by mixing in lakes.� New Phytol. 128: 407-424.

Jeffrey, S. W., R. F. C. Mantoura and S. W. Wright (1997). Phytoplankton pigments in oceanography. Paris, United Nations Educational, Scientific and Cultural Organization.

Jones, R. C. (1998). �Seasonal and spatial patterns in phytoplankton photosynthetic parameters in a tidal freshwater river.� Hydrobiologia 364: 199-208.

Keller, A. A. (1988). �Estimating Phytoplankton Productivity from Light Availability and Biomass in the Merl Mesocosms and Narragansett Bay.� Marine Ecology-Progress Series 45(1-2): 159-168.

Kilham, P. and D. Tilman (1976). �Some biological effects of atmospheric inputs to lakes: Nutrient ratios and competitive interactions between phytoplankton.� J. Great Lakes Res. 2(Suppli. 1): 187-191.

Kirk, J. T. O. (1996). Light and Photosynthesis in Aquatic Ecosystems. New York, Cambridge Univ. Press.

Kromkamp, J. and M. Limbeek (1993). �Effect of short-term variation in irradiance on light harvesting and photosynthesis of the marine diatom Skeletonema costatum: a laboratory study simulating vertical mixing.� J. Gen. Microbiol. 139: 2277-2284.

Kroon, B. M. A., M. Latasa, B. W. Ibelings and L. R. Mur (1992a). �An algal cyclostat with computer controlled light regime.� Hydrobiologia 238: 63-71.

Kroon, B. M. A., M. Latasa, B. W. Ibelings and L. R. Mur (1992b). �The effect of dynamic light regimes on Chlorella. I. Pigments and cross-sections.� Hydrobiologia 238: 71-79.

Langdon, C. (1987). �On the causes of interspecific differences in the growth-irradiance relationship for phytoplankton. Part I. A comparative study of the growth-irradiance relationship of three marine phytoplankton species: Skeletonema costatum, Olithodiscus leuteus and Gonyaulax tamarensis.� J. Plankton Res. 9: 459-482.

Laws, E. a. and T. T. Bannister (1980). �Nutrient and light limited growth of Thalassiosira fluviatilis in continuous culture, with implications for phytoplankton growth in the ocean.� Limnol. Oceanogr. 25: 457-73.

Lehman, J. T. (1976). �Ecological and nutritional studies on Dinobryon Ehrenb.: seasonal periodicity and the phosphate toxicity problem.� Limnol. Oceanogr. 21: 646-658.

Lewis, M. R., J. J. Cullen and T. Platt (1984a). �Relationships between vertical mixing and photoadaptation of phytoplankton: similarity criteria.� Mar. Ecol. Prog. Ser. 15: 141-149.

Lewis, M. R., E. P. W. Horne, J. J. Cullen, N. S. Oakey and T. Platt (1984b). �Turbulent motions may control phytoplankton photosynthesis in the upper ocean.� Nature 311(5981): 49-50.

Lewis, M. R. and J. C. Smith (1983). �A small volume, short incubation time method for the measurement of photosynthesis as a function of incident irradiance.� Mar. Ecol. Prog. Ser. 13: 99-102.

Li, M. and C. Garrett (1998). Large eddies in the surface mixed layer and their effects on mixing, dispersion and biological cycling. Physical Processes in Oceans and Lakes. J. Imberger. Washington, D.C., American Geophysical Union. 54: 61-86.

Lohr, M. and C. Wilhelm (1999). �Algae displaying the diadinoxanthin cycle also possess the violaxanthin cycle.� Proc. Natl. Acad. Sci. USA. 96: 8784-8789.

Lorenzen, C. J. (1967). �Determination of chlorophyll and pheo-pigments: spectrophotometric equations.� Limnol. Oceanogr. 12: 343-346.

Mac Caull, W. A. and T. Platt (1977). �Diel variations in the photosynthetic parameters of coastal marine phytoplankton.� Limnol. Oceanogr. 22: 723-731.

Mantoura, R. F. C. and C. A. Llewellyn (1983). �The rapid determination of algal chlorophyll and carotenoid pigments and their breakdown products in natural waters by reverse-phase high performance liquid chromatography.� Anal. Chim. Acta. 151: 297-314.

Margalef, R. (1958). Temporal succession and spatial heterogeneity in phytoplankton. Perspectives in Marine Biology. Berkeley, CA, University of California Press: 323-349.

Marra, J. (1978). �Phytoplankton photosynthetic response to vertical movement in a mixed layer.� Mar. Biol. 46: 203-208.

Marra, J. (1980). Vertical mixing and primary production. In: Primary productivity in the sea, P. G. Falkowski. New York, Plenum Press: 531.

Morel, A., L. Lazzara and J. Gostan (1987). �Growth rate and quantum yield time response for a diatom to changing irradiances (energy and color).� Limnol. Oceanogr. 32(5): 1066-1084.

Neale, P. J. and P. J. Richerson (1987). �Photoinhibition and the Diurnal-Variation of Phytoplankton Photosynthesis .1. Development of a Photosynthesis-Irradiance Model from Studies of Insitu Responses.� Journal of Plankton Research 9(1): 167-193.

Perry, M., M. Talbot and R. S. Alberte (1981). �Photoadaptation in marine phytoplankton: response of the photosynthetic unit.� Mar. Biol. 62: 91-101.

Platt, T., D. F. Bird and S. Sathyendranath (1991). �Critical Depth and Marine Primary Production.� Proceedings of the Royal Society of London Series B-Biological Sciences 246(1317): 205-217.

Platt, T., C. L. Gallegos and W. G. Harrison (1980). �Photoinhibition of photosynthesis in natural assemblages of marine phytoplankton.� Journal of Marine Research 38: 687-701.

Platt, T. and A. D. Jassby (1976). �The relationship between photosynthesis and light for natural assemblages of coastal marine phytoplankton.� J. Phycol. 12: 421-430.

Post, A. F., Z. Dubinsky, K. Wyman and P. G. Falkowski (1984). �Kinetics of light-intensity adaptation in a marine planktonic diatom.� Mar. Biol. 83(231-238).

Prezelin, B. B. (1992). �Diel Periodicity in Phytoplankton Productivity.� Hydrobiologia 238: 1-35.

Prezelin, B. B. and R. S. Alberte (1978). �Photosynthetic charateristics and organization of chlorophyll in marine dinoflagellates.� Proc. Nat. Acad. Sci. 75: 1801-1804.

Prezelin, B. B. and H. A. Matlick (1980). �Time-course of photoadaptation in the photosynthesis-irradiance relationship of a dinoflagellate exhibiting photosynthetic periodicity.� Marine Biology 58: 85-96.

Prezelin, B. B. and B. M. Sweeney (1977). �Characterization of photosynthetic rhythms in marine dinoflagellates. II. Photosynthesis-irradiance curves and in vivo chlorophyll a fluorescence.� Plant Physiology 60: 388-392.

Ramus, J. (1990). �A Form-Function Analysis of Photon Capture for Seaweeds.� Hydrobiologia 204: 65-71.

Reynolds, C. S. (1998). Plants in motion: Physical - biological interaction in the plankton. Physical Processes in Lakes and Oceans. J. Imberger. Washington, D. C., American Geophysical Union. 54: 535-560.

Richardson, K., J. Beardall and J. A. Raven (1983). �Adaptation of unicellular algae to irradiance: an analysis of strategies.� New Phytol. 93: 157-171.

Rowan, K. S. (1989). Photosynthetic Pigments of Algae. New York, NY, Press Syndicate - University of Cambridge.

Shimura, K. and Y. Fujita (1975). �Changes in the activity of fucoxanthin-excited photosynthesis in the marine diatom Phaeodactylum tricornutum grown under different culture conditons.� Mar. Biol. 33: 185-194.

Sukenik, A., J. Bennett and P. G. Falkowski (1987). �The Molecular-Basis of Photoadaptation in the Marine Chlorophyte Dunaliella-Tertiolecta.� Israel Journal of Botany 36(1): 50-50.

Sukenik, A., J. Bennett, A. Mortainbertrand and P. G. Falkowski (1990). �Adaptation of the Photosynthetic Apparatus to Irradiance in Dunaliella-Tertiolecta - a Kinetic-Study.� Plant Physiology 92(4): 891-898.

Sverdrup, H. U. (1953). �On conditions for vernal blooming of phytoplankton.� J. Cons. Perm. int. Explor. Mer. 18: 287-295.

Vidussi, F., J. C. Marty and J. Chiaverini (1999). �Phytoplankton pigment variations during the transition from spring bloom to oligotrophy in the northwestern Mediterranean sea - separation of chlorophyll a from divinyl-chlorophyll a and zeaxanthin from lutein.� Deep Sea Research Part I: Oceanographic Research Papers 47(3): 423-445(23).

Vincent, W. F., N. Bertrand and J. J. Frenette (1994). �Photoadaptation to Intermittent Light across to St-Lawrence Estuary Fresh-Water-Saltwater Transition Zone.� Marine Ecology-Progress Series 110(2-3): 283-292.

Yoder, J. A. and S. S. Bishop (1985). �Effects of mixing-induced irradiance fluctuations on photosynthesis of natural assemblages of coastal phytoplankton.� Mar. Biol. 90: 87-93.

Appendix

Appendix A. Cell counts and growth rates

Appendix B. Pigments

Appendix B, cont.

Appendix C. Pigment correlations

Appendix D. Pigment TDLD regression results

Appendix E. Photosynthetic parameters modeled with Platt, 1976

Appendix F. Photosynthetic parameters modeled with Fee application

Appendix G.DYV Freshwater Phytoplankton Medium

For the WC working solution, add the WC metals �A� to 10 ml of super stock �B� & then dilute to 1000 ml.

The vitamin working stock solution is a 1:100 dilution of the following super stock

To make 2 L of media stock suitable for diluting to 40 liters media. Dissolve the MOPS buffer into 985 mL DDH2O. Correct to pH 6.8 with about 10 pellets of NaOH and fine tune with 1 N HCl. Weigh each reagent from Table 1 and add, in the order listed, to the buffer solution while stirring continuously. Add the working solutions and dilute to 2 L. Store in refrigerator; stable indefinitely. To make the media, add 50 mL of the media stock to 950 mL DDH2O for each liter of media. Autoclave or filter sterilize prior to use.

Appendix H. Photosynthetron SOP

Radioactive stock: NaH¹⁴CO₃ (100 mCi/mL) found in Radioactive Jeff�s refrigerator

�Samples� describes below will yield one PvI curve of nine points; one block can hold three samples.

Pre measure sample (110 mL for #4) which will leave ~10 mL when fully dispensed.

a. Verify that all lights in the block are working. Light measurements should be up to date for each well.

c. Set working temperatures on thermostat and wait for blocks to cool to desired temperature.

1. For each sample, label three vial caps with �Spike� and two caps with �T₀�.

5. Add 0.25 mL of isotope stock and mix with repeater (resulting activity = 0.0022 mCi�vial^-1)

6. Working quickly, swirl the sample and pipette 1 mL volumes into vials labeled �Spike� and add 10 mL Universol

8. Swirling the sample, add 5 mL to two vials labeled �T₀�, acidify with 1 mL of 1N HCl. Cap and return vials to the flat

9. Swirling the sample, fill 9 vials with 5 mL along the light gradient in the block. Also fill two vials in the dark wells. The darks give dark excretion of carbon (RD)

12. In each block there should be: 2 pairs of dark controls and two gradients filled with 5 mL samples.

15. Fill out bench sheet for sample run, return to check incubator every 15 minutes

-Put everything away. Add <50 mL rinse-acid to glass beaker, rinse into rad waste

-Perform wipe test of designated areas using moistened (2 drops of solution) circles

-With forceps, stuff circle into vial, pour in Cytoscint cocktail. Mark room/station on vial cap

16. At 90 minutes, turn off the lights and quickly add 1mL 1N HCl to each vial. Cap the vials in the blocks.

18. Load the vials into flats for counting, ordered as they were in the incubator

1. Exhaust the samples on a shaker table for 24 hours in a radiation qualified exhausted hood.

3. Read in the counter. Be sure to include wipe test vials and report CPM counts to radiation safety officer.

Appendix I. Background for DATPARSE Beckman scintillation parsing program

DATPARSE is a command line program that is intended to parse the highly-formatted text generated by the serial capture of data output to the RS-232 connection from the Beckman scintillation counter currently found on the third floor, north wing of Lapham Hall.

The output from this program is a continuous, comma-delimited, table, suitable for use in a spreadsheet program, with a row for each vial read in the counter. There is currently no limitation on the number of samples it will process. Note that due to the highly formatted nature of the output from the device, datparse has strict requirements regarding the format of the data that it will successfully parse.

The data files are captured using a terminal communication application, such as HyperTerminal which is included with the Windows operating system. The connection settings can be found in the device manual for RS-232 output on the Beckman. Currently, the device is connected to a Pentium 90 Mhz pc running Windows95. Hardcopy and serial capture are both currently enabled in the device.

Here is an example of the captured text that will be used as input for the program:

..that's my full name (yes, Cosmo really IS my middle name). About Datparse, I didn't really keep track of version numbers. I did do it in Borland, but since it's a console app (as opposed to windowed), it can be used with gcc or any other compiler with little or no modification. Feel free to include the source, it should be fine.

Here is an example of the output. The first field is a record ID counter. The second field describes the rack number and the sample number within the rack. The 17th field shows disintegrations per minute, which is most useful for carbon assimilation studies.

1,1-1,1,349056.00,0.48,0.50,0.87,96.0,0,0.00,0,0,1F,105,C2,90.98,383646.3,FF,100,BD

2,1-2,1,356497.75,0.50,0.45,1.91,97.0,0,0.00,0,0,2E,105,C2,90.93,392037.0,F1,100,BD

3,1-3,1,381084.44,0.48,0.45,2.94,96.0,0,0.00,0,0,2C,105,C2,90.98,418848.7,06,100,BD

Appendix J. Background on the stepping motor

Program code used to drive the mixing simulation motor is listed below. This motor and drive consisted of an Intelligent Motion Systems Microlynx 7 controller/drive, an IMS power supply and an IMS 34-Frame stepping motor (3437-2) with a holding torque of 420 ounce-inches. (www.imshome.com). The unit was purchased through the control systems reseller All Control Enterprises (www.allcontrol.com).

Corporate Office:
1644 Cambridge Drive
Elgin, Il 60123
Phone: 847- 488-9200
Fax: 847- 488-9400

Rockford:
6834 Forest Hills Road,
Loves Park, IL. 61111
Phone: 815-637-4000
Fax: 815-637-4044

Wisconsin:
N35 W21140 Capitol Dr.,
Suite 3,
Pewaukee, WI. 53072
Phone: 262-781-6789
Fax: 262-781-6415

A Tripplite powerline conditioner in tandem with a high-quality surge suppressor was used to prevent power fluctuations on the supply side from damaging the unit. I did not use a line filter between the motor and controller. This device was prohibitively expensive, but would have prevented reversing the current into the controller from slippage of the motor. This motor and controller, though of the highest quality and feature set, presented no end of problems. If others in the future wish to recreate this arrangement, please consider the following recommendations. The holding torque of the motor itself appeared to be sufficient for this purpose. However, technicians from the supplier suggested using a more expensive type motor, a Yaskawa servo motor/controller, which allows for closed-loop diagnostics between the computer and the motor. Also, this type of motor/controller allows greater control and ease of programming. Inclusion of both a control-side feedback filter and a braking mechanism would prevent destruction of the delicate controller electronics from having the weight of attached samples rotate the unpowered motor, thereby turning it into a destructive generator.

Documentation for the following code can be found in PDF files available from www.imshome.com.

' THE 'TRAVELT' FUNCTION IS LIKELY NOT NEEDED WHEN VM AND DELAY ARE BOTH ABOVE 1000

TRAVELT = degs/SPEED * 1000' A CALC TO CORRECT THE TIME OUTPUT FOR SLOW MOVEMENTS

BR MOVEDOWN, POS < 16000 'NUM EXPERIMENTALLY DETERMINED TO BE POS AT BOTTOM OF COLUMN

Abstract

ABSTRACT

Table of contents

List of figures

List of tables

Introduction