ABSTRACT In situ instruments report a daily periodicity in fluorescence (f) and beam attenuation (c) in the ocean. The daily changes are hypothesized to result from changes in phytoplankton physiology. To determine the relation between f,c, chlorophyll (chl), and photosynthetically active radiation (PAR), time series measurements were made on cultures of phytoplankton which were exposed to different light conditions. F normalized to chl (f*) varied inversely to PAR. c* displayed daily changes that did not correlate to PAR. The f response is well above experimental noise and its amplitude is two times as strong as the in situ signal. The c* response is not significantly outside of the experimental error, and it's amplitude is within the range of the in situ signal. Absorption (a) was also measured. A* did not correlate to PAR or time, but steadily decreased as cultures grew older. conclude c' varies due to b* (scattering per chl.) with a circadian rhythm, and f' varies in response to PAR. This is consistent with in situ data. Further, I present what is believed to be the first reported empirical method to measure chl from both f and PAR. INTRODUCTION Stimulated fluorescence and beam attenuation may be powerful tools to measure in situ chlorophyll and biomass in the open ocean. If perfected, these methods would provide oceanographers with several orders of magnitude more data to study primary productivity and algal ecology. However, these signals are not yet reliable tools for ecological parameters. This is due in part to diel oscillations of the signals which are hypothesized to result from changes in phytoplankton physiology (Dickey 1992; Stegman et al 1992; Stramska and Dickey 1992). Understanding and predicting these diel cycles is an essential step in the determination of accurate algal biomass from shipboard and moored fluorometers (Prezelin and Ley 1980; Mauzerall 1972; Abbott 1982). Stimulated fluorescence in the ocean has been correlated to phytoplankton populations (Prezelin and Ley 1980; Keifer 1973a; Abbot et al. 1982). Recently, a diel pattern of this signal has been observed from drogues, ships, and moorings in the Sargasso Sea, Monterey Bay, and the equatorial Pacific. (Dickey 1992; Chavez pers. comm.; Abbot et al. 1990) The daily changes in fluorescence may be due to a combination of phytoplankton physiology, growth, grazing, mixing, demography, and dissolved organic sources of variability. A model of the diel patter due to physiology would decouple signals due to physiology from signals due to actual changes in biomass. In this paper, cultures of phytoplankton exposed to sunlight consistantly had a daily minimum in fluorescence equal to 50% of their daily maximum (a factor of 2 difference between night and day). Beam attenuation decreased to 80% of its maximum (a factor of 1.25 between night and day). These data are consistent with in situ observations, and provide the basis for an empirical determination of chlorophyll from strobe-stimulated fluorescence and PAR. Theory: Fluorescece is a measure of photons which were absorbed by pigments but whose energy was not shunted into photosynthesis. When plants absorb light, there are three sinks for the energy: vibrations, fluorescence and redox reactions. Vibrations (heat) are insignificant, and redox reactions are synonymous with photosynthesis (Arnon 1968). In the steady state, conservation of energy requires: Eabsorbed - Efluoresced + Ephotosynthesis Thus, changes in fluorescence must be correlated with changes in absorption and photosynthesis (Gentry et al 1989; Kolber et. al. 1990; Falkowski et al 1986, 1988; Mauzerall 1972; Bannister and Weidmann 1984) The emitted photon has a discrete wavelength which depends on the available quantum states of the pigment molecules. The most intense fluorescence emission from phytoplankton is at 685 nm, corresponding to the Stokes shifted decay of an electron excited by a 670 nm photon - right in the middle of a peak in the absorption spectra of Chl. Beam attenuation is a measure of how many photons starting out toward a target do not make it there. Attenuated light is either absorbed or scattered. Thus, beam attenuation is the sum of absorption and scattering (Table 1) c =a+b If one could collect all the scattered and unattenuated light that passed through sample, one could measure of absorption. Similarly, if one excluded all scattered light, one can measure beam attenuation. The spectrophotometer measures all of a and part of b, because it collects some but not all of the scattered light. Measured absorption = a + x*b where Oxx1. If all of the scattered light is collected, x = 1; if none of the scattered light is collected,x - 0. x depends on the volume scattering cross section and the geometry of the detector (Bricaud 1983; Roach 1974). The Shibada technique for measureing absorption with a spectrophotometer uses a diffusing plate to redirect forward scattered light. (Mitchell and Kiefer 1987; Shibada 1954; ) Approximate attenuation can be measured by positioning the sample far from the detector so that the apperture eclipses most of the scattered light. If a and c are measured, b is known as well. (Keifer pers. comm.; Chavez pers. comm.) a and b depend on phytoplankton physiology: Changes in absorption indicate changes in pigment concentration, quantum state, and package effect. Changes in scattering evidence changes in particle size, density, and index of refraction. Materials and Methods: Dunalliela , Phaeodactylum , and Amphidinium were grown in sterile enriched sea water. Cultures in their exponential phase were subjected to four different lighting conditions (Table 2). To assess the relationship between phytoplankton populations and optical phenomena, five properties were measured: Chlorophyll a (hereafter referd to simply as chlorophyll). fluorescence, attenuation, absorption and PAR. Chlorophyll concentration was quantified using acetone extractions and a Turner fluorometer (Holm Hansen et al. 1965). Absorption was measured with a 10 cm cuvette and a diffusing plate close to the detector of a diode array spectrophotometer (Shibada 1954; Mitchell 1988). Attenuation was measured using a 1 cm cuvette far from the detector (Bricaud 1983). Fluorescence was measured using a Sea Tech strobe stimulated fluorometer. PAR was measured with a photodiode callibrated to detect quanta of light. Over a 48 hour time series, fluorescence was measured 22 times, chlorophyll was measured 11 times, and spectra were measured 5 times. The sampeling rate was not constant. Fluorescence, absorbtion and attenuation were normalized to chlorophyll and analyzed on a UNIX workstation. The spectra were smoothed and integrated. A Sea Tech fluorometer and a Sea Tech transmissometer have been recording data at 15 minute intervals from a mooring in Moneterey Bay since January 16, 1992. The fluorometer is the same instrument as used with the cultures. The transmissometer opperates at 660 nm and returns a voltage. Voltage is converted into beam c by the following equation: C =-4*log(VDC/5) The mooring also has a PAR sensor which samples every 10 minutes. However, there is no instrument to measure chlorophyll on the mooring. Results: The mooring fluorometer shows long term (3- 15 day) events which are hypothesized to represent spring phytoplankton blooms. (Figure 1) There is a short term (1 day) signal superimposed on this signal. (Figure 2). The diel cycle has a maximum at night (usually either one hour after sun set or one hour before sunrise) and a minimum near noon. The amplitude of the diel cycle correlates with the mean fluorescence value; when the mean signal is high, the variations (the difference between daily maximum and minimum) are large e.g. the value of the mean correlates with the envelope of the raw - mean timeseries (Figure 3). The minimum value is consistently half the maximum value for each day (+- 10%). This is true even when the daily oscillations grow larger during the phytoplankton bloomes e.g. the envelope for the raw/mean time series is nearly constant (Figure 3) The raw/mean fluorescence timeseries has an inverse functional relationship with PAR both within days and between days (Figure 4). The raw/mean timeseries is hypothesized to represent f* in situ, because mean fluorescence has been shown to correlate with chlorophyll (Chavez 1991). This relationship is apparent in cloudy days as the fluorescence is higher than on sunny days. Sometimes the fluorescence cycle is perturbed for one or more days after a cloudy day regardless of PAR. (Figure 2) The beam attenuation increased during the spring bloom, yet it was also high in January and February. The daily cycles in care less consistent than the fcycles, c oscillates over the course of the day, and during the bloom the daily variations are larger. yet they are not as clearly related to PAR as fis. (Figure 2) Ten different time series were made with cultures: Dunalliela in the dark (DD). incubator (DI, windowsill (DW), sunlight (DS); Phaeodactylum: (PD), (PW), (PD. and (PS); and Amphidinium: (AD), and (Al) (Table 3) (Figures 5 to 11). The cultures grew at different rates under each condition. In general, chlorophyll concentration grew at an exponential rate when cultures were exposed to light, and not at all when they were kept dark. The rate of growth was not proportional to light intensity The best environment for Dunalliela was the window sill, and Phaeodactylum grew quickest in the incubator. Amphidinium died in the direct sunlight, and grew best in thi incubator. These rates may be related to the change in fluorescence on the days after a cloudy day Normalized fluorescence (f*) was constant for DD, PD, AD, DI, AI, There is a small change in f* (inverse to PAR) in PI, PW, and DW. There is a clear inverse relationship between f“ and PAR for DS and PS. Raw f correlated well with chlorophyl for specemins in the dark and in the incubator, but not for samples in the sun (Figure 12). f* for Dunalliela = .013 volts/ (ug chl/1) when it is dark adapted (i.e. when the f has adapted to dark conditions). f* for phaodactylum = 035 volts when it is dark adapted. It took 20 minutes for Dunalliela and 30 minutes for Phaodactylum to adapt to changes in PAR (Figure 13). The PS f* timeseries is more of a square shape than DS. This observation, and the fact that the PI f“ timeseries changed with PAR, while the DI timeseries did not. suggest different species have different sensitivities to PAR and that the fluorescence response to PAR can be saturated The normalized absorbtion spectra for each species had charactaristic shapes. The spectrum for Dunalliela is similar to Phaeodactylum and Amphidinium except that it is higher at the longer wavelenths. This offset appears in the c'spectrum as well. This is consistent with the observation that Dunalliela appears green to the human eve and Amphidinium and Phaeodactylum appear red-brown. Absorbtion spectra from all species in all different light conditions decreased with time irrespective of PAR or time of day (Figure 14 a). The exceptions to this decreasing trend are in PW and PS. Measurements ofc do not have a decreasing trend over time as the absorption spectra do. Instead, on average, c changes over the time of day inversely with respect to expected daylight; it appears to be correlate with time of day, not PAR (Figure 14 b) (Table 4). For combined timeseries of Dunalliela this trend is larger than two standard deviations. For Phaeodactylum the trend is not outside one standard deviation. (Figure 15). Discussion A similar diel cycle in fluorescence is observed in cultures and in the field. In the cultures chlorophyll does not oscillate over the day as fluorescence does. This supports the hypothesis that diel variations are primarily due to changes in phytoplankton physiology, not growth, grazing or mixing. Normalized fluorescence varies inversely to PAR, regardless of time of day. This suggests the cycle in fluorescence is related to PAR, not circadian rhythms. The observation that f“ takes an average of 20 minutes to adapt to changes in PAR, means a timeseries of f* lags approximately 20 minutes behind a timeseries of PAR. Thus, the correlation between f* and PAR can be made tighter by taking a running average of PAR over 20 minutes and shifting PAR ahead in time by 20 minutes. Qualitatively correcting for this lag time makes a small but noticable improvement in the uncertainty of f* as a function of PAR (Figure 4) These data provide the basis for an empirical way to measure mean fluorescence from instantaneous fluorescence corrected f = (.0005)*(shifted and smoothed PAR)"(mean f) + f This is mathematically equivalent to adding a constant fraction of the irradiance peaks to the fluorescence troughs in a time series of f*. The coefficient .0005 comes from the slope of f* vs PAR in situ (Figure 4). This coefficient was only qualitatively determined. A statistically derrived higher order fit to Figure 4 may provide a better correction algorithm. The mean f must be included in the correction because the magnitued of the in situ diel oscillations varies directly with the mean (Figure 3). It should be noted that this is just a step in the direction of a real-time empirical correction, because this model requires the time averaged f. Yet this correction tightens the relationship between f and mean f by a factor of 2 (Figure 16). This is useful because mean f correlates well to chlorophyll concentration (Chavez 1991) and chlorophyll concentration correlate to rates of primary productivity (Balch et. al 1992). Primary productivity rates are the key element in the models of the biological carbon budget in the ocean. Moreover, primary productivity in the ocean may be the major sink for atmospheric carbon dioxide. Thus, instantaneous fluorescence measurements may provide useful data for current issues in global oceanography if correction algorithms such as the one presented here are made more consistant. Despite the many unknowns, this correction algorithm already correlates instantaneous fluorescence with carbon budgets by a factor of 2 more certainty. Specifically, this correction algorithm is useful for transects made with a fluorometer, because one cannot observe a full daily cycle in each parcel of water from a moving ship. Even if an accurate mean is not known, the emperical relationship to PAR discussed thus far will indicate the direction of error due to physiological changes in phytoplankton during in situ observations of f. The observed perturbation of fluorescence one and two days after a cloudy day may represent real changes in chlorophyll concentration. This is likely because all phytoplankton cultures chlorophyll concentrations increased faster when they were not in direct sun. Thus, on a cloudy day, the increased fluorescence signal is likely to be a physiological response to reduced PAR, while high f values on the next sunny day are hypothesized to represent actual increases in chlorophyll. Thus far, I have discussed ways to correct for the diel cycle in fluorescence. Yet this cycle may contain useful data. For example, changes in the diel fluorescence cycles may identify species. The culture results support this hypothesis. Fluorescence changes in Phaeodactylum were to be more sensitive to PAR than in Dunalliella in the incubator and unlike Dunalliela the fluorecence response for Phaeodactylum in the direct sun did not change significantly for PAR above 1000 mE/m2. These species-specific responses to light may explain some of the more fine points of the in situ results. For example, the in situ f* (f divided by mean f) does not quite have a constant amplitude: it is largest at the end of the first of the spring blooms. This may reflect a higher population of more sensitive phytoplankton, such as Phaodactylum , at that time in the Monterey bay. Yet, this is a difficult problem, because many changes: sensitivity, threshholds, and growth rates are all superimposed on each other. Thus, it is difficult at this point to make significant claims about in situ species distributions with these remote bio-optical data. High attenuation in the bay during January and Feburary is hypothesized to result from suspended sediment load, not phytoplankton, because it does not correlate with the increase in mean fluorescence, the increase in amplitude of diel fluorescence changes, or the increase in amplitude of diel transmissivity changes. This is plausable because Winter is the stormy season in Monetery Bay, and more turbulent waters suspend more sediment. However, the lack of large diel cycles in attenuation does not necessacerily indicate lack of phytoplankton because, in the cultures Dunalliela exhibited diel changes in attenuation while Phaodactylum did not. In the cultures a* did not change with light adaptation. Thus, E(absorbed) for a constant amount of PAR (e.g. the fluorometer strobe light) did not change. But E(fluoresced) did change with light adaptation. Thus, by concervation of energy, a change in the efficiency of photosynthesis (photons used for photosynthesis/ photons absorbed) must account for the changes in fluorescence. The changes in a* were two orders of magnitude smaller than the changes in c*. c* did change with a circadian rhythm. c* varies inversely to expected PAR (not actual PAR) in Dunalliela but not significantly in Phaeodactylum. By inference (because c= a + b) b* also changes inversely to expected PAR. Thus, cell swelling, change in refractive index or cell division (change from fewer larger cells to more smaller cells) might explain the change in beam attenuation, because these are the factors that affect b“. Further, pigment density, quantum state of the pigments, and optical depth of the phytoplankton cannot be the factors that affect the diel cycle in c*, because thier net affect would have affected a* on a diel pattern. However, a* did change over time, though not with a diel cycle. a* decreased; i.e.: the observed percent change in a is less than the percent change in chl at each successive measurement in the experiment regardless of PAR or time of day. (al - a2)a (cl-c2)c Three working hypotheses may explain how a* decreases over time. 1) some absorption may be due to things other than chlorophyll in the sample which are not in the blank. This is possible because the blank was enriched filtered sea water, not filtered culture. Thus, there may be biogenic solutes or particles in the cultures which do not contain chlorophyll but do absorb light. 2) The relationship of a to chl may be non linear, This may be true, because when the cultures are thicker there is more scattered light, and scattered light travels through a longer path length in the cuvette, therefore, it has more opportunity to be absorbed. This is especially true in the 10 cm cuvette (as opposed to the 1 cm cuvette used for attenuation). 3) older plants may absorb less light per chlorophyll. This is likely, because as Dunalliela grew older, it was observed to have more chloroplasts per plant, and these chloroplasts may eclipse each other. Conclusion In situ diel variations in fluorescence are primarily due to changes in phytoplankton physiology, specifically an increased quantum yeild of photosynthesis due to increases in PAR. An empirical correction based on this hypothesis explains half of the daily variation in fluorescence Absorption decreases as cultures grow older, but it does not vary enough to affect attenuation. Attenutaion exhibits a circadian rhythm in Dunalliela but not in Phaodactylum. By inference, scattering changes inversely to expected PAR. Thus, cell swelling, change in refractive index or cell division might explain the change in beam attenuation. Chlorophyll measurements from in situ fluorescence can be more accurate by a factor of two by using the emperical equation: corrected(f) = raw(f) +.0005* (smoothed and shifted PAR)“mean(f). Further work is needed to accurately determine chlorophyl from remote bio-optical parameters, but this relationship is likely to be refined by correlating fluorescence, PAR, and attenuation on daily and instantaneous time scales. When this is done, the carbon budget for open ocean water masses may be better monitored with the aid of a fluorometer. It is also possible that attenuation and fluorescence will be able to identify species of phytoplankton by remote techniques Acknowledgements: Francisco Chavez was a superb mentor for this project. Thank you, Francisco, for letting me choose the direction of this projet, guiding me whenever I got stuck, and always being ready to pick up the pieces and look at them as a consistant whole. Keep hanging in there! Kurt Buck, the man who lives with phytoplankton, was a well-spring of knowledge regarding observations and species-specific habits. Jason Smith shared his sophistocated culture and knowledge of phytoplankton. Tim Pennington - thanks for taking me on the boat. Annie Reese - thanks for proofreading this project through to the very end. REFERENCES Abbott, M.R., Brink, K.H., Booth, C.R., Blasco, D., Codispoti, L.A., Niler, P.P., and Ramp, S. 1990. Observations of phytoplankton and nutrients from a lagrangian drifter off Northern California. J. Geophys Res 95 (C6) 9393 - 9409 Abbott, M.R., Richardson, PJ., and Powell, T.M. 1982. In situ response of phytoplankton fluorescence to rapid variations in light. Limnol. Oceanogr. 27 (2) 218- 22: Baker, E.T. and Lavelle, J.W. 1984. The effect of partical size on the light attenuation coefficient of natural suspensions. J Geophys Res. 89(C5)8197-8203. Bannister, T.T., and Weidemann, A.D. 1984. The maximum quantum yeild of phytoplankton photosynthesis in situ. J. of Plankton Res. 6(2) 275 - 28 Bricaud, A., Morel, A., and Prieur, L. 1983. Optical efficiency factors of some phytoplankters. Limnol. Oceanogr. 28 (5) 816 - 832. Chavez, F.P., Barber, R.T., Mitchael, K., Huyer, A. Ramp, S.R., Stanton, T.P. and Menliola, B.R. 1991. Horizontal Transport and the Distribution of Nutrients in the Coastal Transition Zone off North California: Effects on Primary Production Phytoplankton Biomass and Species Composition. J. Geophys Res. 96(C8), 14,833 -14,848. Cuhel, R.L., Ortner, P.B., and Lean, D.R.S. 1984. Night synthesis of protein by algae. Limnol. Oceanogr. 29(4) 731 - 744. Dickey, T. D. 1991. The emergence of concurrent high-resolution physical and bio¬ optical measurements in the upper ocean and their applications. American Geophysical Union Reviews of Geopysics 29(3) 383 - 413. Falkowski, P.G., Kolber, Z. and Fujita, Y. 1988. Effect of redox state on the dynamics of photosystem II during steady-state photosynthesis in eucaryotic algae. Biochem. Biophys. Acta 933 (1988) 432 -443. Falkowski, P.G., Wyman, K. Ley, A.C., and Mauzerall, D.C. 1986. Relationship of steady state photosynthesis to fluorescence in eucariotic algae. Biochem. Biophys. Acta 849 (1986) 183-192. Gently, B., Briantais, J.-M., and Baker, N.R. 1989. The relationship between the quantum yeild of photosynthetic electron transport and quenching of chlorophyll fluorescence. Biochem. Biophys. Acta. 990 (1989) 87 -92. Keifer, D.A. 1973. Fluorescence Properties of Natural Phytoplankton Populations. Mar. Bio. 22, 263-269. Kolber, Z., Wyman, K.D. and Falkowski, P.G., 1990. Natural variability in photosynthetic energy conversion efficiency: a field study in the Gulf of Maine. Limlnol Oceanogr. 35 (1) 72 -79 Mauzerall, D. 1972. Light-induced changes in Chorella, and the primary photoreactions for the production of oxygen. Proc. Nat. Acad. Aci. USA. 69 (6) 1358 - 1362. Mitchell, B.G., and Kiefer, D.A. 1988. Chlorophyll a specific absorption and fluorescence excitation spectra for light-limited phytoplankton. Deep-Sea Res. 35 (5. 639 -663. Prezelin, B.B, and Ley, A.C. 1980. Photosynthesis and chlorophyll a fluorescence rhythms of marine phytoplankton. Mar. Bio. 55, 295-307. Roach, David M. 1974. The determination of refractive index distribuitons for oceanic particulates. PhD Thesis, Oregon State University. Shibada, K. Benson, A.A, and Calvin, M. 1954. The absorbtion spectra of suspensions of living micro-organisms. Biochem. Biophys. Acta 15 (1954) 461 - 469. Stramska, M. and Dickey, T. 1992. Short term variations of the bio-optical properties of the ocean in response to cloud-induced irradiance fluctuations. J.G.R. in press. Stramska, M. and Dickey, T. 1992 Variability of bio-optical properties of the upper ocean associated with diel cycles in phytoplankton population. Submitted to JGR. 1992 Stegmann, P.M. and Lewis, M.R. 1992. Primary production estimates from rcordings of solar stimulated fluorescence in the equitorial pacific at 150° W. J Geophys Res. 97 (C1) 627 - 638. Figure Legends: Figure 1. Time series of surface PAR, fluorescence at 10 meters depth and transmissivity at 10 meters from the mooring in Monterey Bay from January 16 to May 20, 1992. Figure 2. Time series of surface PAR, fluorescence at 10 meters depth and transmissivity at 10 meters from the mooring in Monterey Bay from May 7 to May 14, 1992. Figure 3. Raw, mean, raw - mean, and raw/mean timeseries for fluorescence measured by the mooring in Monterey bay from March 20 to May 5, 1992. Flgure 4. raw/mean fluorescence vs. PAR measured by the mooring in Monterey bay from March 20 to May 5, 1992. Figure 5. Fluorescence vs. mean fluorescence (determined by a one day running average). [Mean fluorescence correlates well with chlorophyll (Chavez 1991).) Figure 6. Time series of fluorescence, chlorophyll, fluorescence normalized to chlorophyll and PAR for Dunalliela in the sun (DS) Figure 7. Time series of fluorescence, chlorophyll, fluorescence normalized to chlorophyll and PAR for Phaodactylum in the sun (PS) Figure 8. Time series of fluorescence, chlorophyll, fluorescence normalized to chlorophyll and PAR for Dunalliela in the incubator (DI) Figure 9. Time series of fluorescence, chlorophyll, fluorescence normalized to chlorophyll and PAR for Phaodactylum in the incubator (PI) Figure 10. Time series of fluorescence, chlorophyll, fluorescence normalized to chlorophyll and PAR for Dunalliela in darkness (DD) Figure 11. Time series of fluorescence, chlorophyll, fluorescence normalized to chlorophyll and PAR for Phaodactylum in darkness (PD) Figure 12 a. Chlorophyll concentration vs. raw fluorescence for a two day timeseries of phytoplankton cultures in direct sun Figure 12 b. Chlorophyll concentration vs. raw fluorescence for a two day timeseries of phytoplankton cultures in an incubator Figure 12 c. Chlorophyll concentration vs. raw fluorescence for a two day timeseries of phytoplankton cultures in darkness Figure 13. Time series of fluorescence and par for Phaodactylum , Dunalliela , and Amphidinium moved from darkness to direct sun at 11:00 am. Figure 14 a. Absorption spectra for Dunalliela in the incubator. Figure 14 b. Attenuation spectra for Dunalliela in the incubator. Figure 15 a. Mean and standard deviation for attenuation vs. expected light in Dunalliela. Figure 15 b. Attenuation vs. expected light for each specemin of Dunalliela. Figure 15 c. Attenuation vs. expected light for each specemin of Phaeodactylum. Figure 16 a. Mean fluorescence vs. raw fluorescence measured by the mooring in Monterey bay from March 20 to April 14. Figure 16 b. Mean fluorescence vs. emperically corrected fluorescence measured by the mooring in Monterey bay from March 20 to April 14. Table 1 SYMBOL chl PAR MEANING UNITS absorption meter -1 scattering meter - meter -1 attenuation raw fluorescence volts chlorophyll ug/l absorption / chlorophyll m2/mgchl scattering / chlorophyll m2/ mg chl attenuation / chlorophyll m2/ mgchl fluorescence / chlolophyll m2/ mg chl Photosynthetically active radiation mE/ m Table 2 LOCATION Darkness Incubator Window. Sunlight LIGHT (maximum) mEm? 35 mEm? 750 mE/m2 1800 mE/m2 Table 3 SPECIMEN ABBREVIATION DS Dunalliela in the sun Dunalliella in the incubator DD Dunalliella in the dark PS Phaeodactylum in the sun Phaeodactylum in the incubator Phaeodactylum in the dark PD Amphidinium in the incubator PD Amphidinium in the dark Table 4 a Hyothesis: c is high inversely proportional to expected light Supporting Data Refuting Data Specemin DD DW PD PW Table 4 b Hypothesis: c is directly proportional to expected light Supporting Data Specemin Refuting Data DD DW PD PW Figure 1. 2.0 x103 - 1.6 x103. 1.2 x103 - 8.0 x102 - 4.0 x102 - -0.0 x10° 4.0 - 2.0 - -.0 - 5.00 - 3.00 bhtee a 60. Julian Day in 1992 120. 2.0 x103 1.6 x103 1.2 x103 8.0 x102 40 x102 00 2100 4.0- 2.0 - -.0 - 4.50 4.00 3.50 - 128.0 Hn E 130.0 PAR at the surface Surface fluorometer Surface transmissometer 132.0 Julian Day in 1992 M h W 134.0 Figure 2. Figure 3. 6.0 raw data 9 4.0 2.0 Mhnst -.0 1 day running average 4.0 - 20 -.0 - raw data - 1 day running average 2.0 — 0 oh o 2.0 -4.0 3.00 raw data / running average 2.00 31.00 -.00 - 80. 90. 100. 110. Julian Day in 1992 120. Flgure 4. 3.00 — F3 days 82 to 126 (all) fmean(f) 2.00 4 k. .. ** ...*+ . 1.00 1.4. 4 ...: * . * : *.....* *... . . . * ...** -00 + 2.0 x103 1.0 x103 -0.0 x10° PAR Figure 5. — P 2.00 - ..** .. ... .* * 1.00 - . 2. *.* : -00 + 2.0 3.0 1.0 raw(f) 10 150. 50. 14 2102 1.0 x10-2- 6.0 X10-3 200. 100. 0. — Dunalliela in direct sun Raw Fluor * *x Chlorophyll * Normalized Fluor (f*) ** 12 PAR Fluorescence and CHlorophyll Timéserie. x Figure 6. 1.00 80. 40. 0. -L 8.0 x10-2 § 40 2102 -0.0 x10° — 200. 100. 0. - Figure 7. Phaodactylum in direct sun Raw Fluor * * Chlorophyll * * * Normalized Fluor (f*) X„ * PAR Fluorescence and CHlorophyll Timéseries ** * 10 150. 50. 4.0 x10-2 - 2.0 x10-2 -0.0 x10° 80 40 00 Dunalliela in incubator Raw Fluor r Chlorophyll Normalized Fluor (f*) „ n 5 - PAR Fluorescence and CHlorophyll Timéseries Figure 8. 1.00 - 100. 4.0 X10 § 20 2102 -0.0 x100 - .80 40 -.00 — Figure 9. Phaodactylum in incubator Raw Fluor Chlorophyll * Normalized Fluor (f* PAR Fluorescence and CHlorophyll Timéseries 4.00 2.00 400. 3 200. 0. — 2.5 x10-2- 1.5 x10-2 5.0 x10-3 - 80 - 40 -.00 - Figure 10. Dunalliela in Darkness Raw Fluor Chlorophyll Normalized Fluor (f*) + + --- --- - - PAR — ——- — — — — — Fluorescence and CHlorophyll Timéseries 4.00 2.00 — 20 5 40 210 2 2.0 x10-2— -0.0 x100 80 - 40 -.00 — Figure 11. Phaeodactylum in Darkness Raw Fluor Chlorophyll Normalized Fluor (f*) * . ——— — — — —- PAR — — - -—— — —— Fluorescence and CHlorophyll Timéserie Figure 12 a. Figure 12 b. 200 - 100.— Figure 12 c. 200. - § 100. -00 FLO vs. CHL in direct sun Dunalliela Phaeodactylum ge raw FLUORESCENCE (voltage) FLO vs. CHL in the incubator Dunalliela Amphodinium Taw FLUORESCENCE (voltage) FLO vs. CHL in the dark Amphodinium Dunalliela Phaeodactylum raw FLUORESCENCE (voltage) Figure 13. Phaodactylun 2.00 - Dunalliella X Amphidinium 00 8.00 TIME (hours) PAR 10.00 Figure 14 a. 4.0 x10-2 - -0.0 x10° - 300 Figure 14 b. 4.0 x10-3 00 r10 -4.0 x10-3 — 300 1 DI 3 DI 5 fne R DI 5 — 514 D1 3 DI 2 700 wavelength wavelength 4.0 x10-2 - 2.0 x10-2 - DARK -.00 Dunal la Expected light LIGHT 2.00 Figure 15 a. Figure 15 b. 8.0 x10-2 — 6.0 x10-2 - 5 4.0 x10-2- 2.0 x10-2 - -0.0 x10° - 8.0 x10-2 - 6.0 x10-2 - 5 4.0 x10-2 - 2.0 x10-2 - -0.0 x10° - DUNALLIELA DD DI OFF ON Expected PAR DS DW OFF ON Expected PAR Figure 15 c. 2.0 x10-3 - PD 1.6 x10-3 - 1.2 x10-3 - 8.0 x10—4 - 4.0 x10-4- -0.0 x10° - 12 - 5 06 04 - 00 OFF O Expected PAR PHAODACTYLUM PW ON OFF Expected PAR Figure 16 a. Figure 16 b. 2.00 1.00- 00 - 2.00 - 1.00 -.00 — ++ snt i ti . + n i ( 1.0 3.0 2.0 Raw F ++ 1.0 2.0 3.0 Corrected F