Ecophysiological strategies in response to UV-B radiation stress in cultures of temperate microalgae isolated from the Pacific coast of South America

Marine microalgae exposed to ultraviolet radiation (UV) have complex adaptive responses provided by a series of protection and repair mechanisms. Interspecific differences in UV sensibility could result in differential selection of the more tolerant species, having consequences for the structure of phytoplankton assemblages. The relative importance of protection and photorepair mechanisms of microalgal cells exposed to potential UV-B stress was studied in monocultures with different-taxonomic, ecological and size characteristics obtained from the Chilean coast. Differences in photosynthesis and growth rates were predicted, since the ability to effectively acclimate to UV is not universal between microalgal species. The dinoflagellate Alexandrium catenella Whedon et Kofoid Balech, the diatom Phaeodactylum tricornutum Bohlin, the chrysophyte Aureococcus sp. and the cyanobacterium Spirulina subsalsa Oersted were acclimated during exponential cell growth under PAR+ UV -A radiation (365 nm, 140-240 kJ mˉ2 d 1) and thereafter exposed 2 h d 1̄ to high and low UV-B radiation (312 nm, maximum 3.1 kJ mˉ2dˉ1) at the center of the 16 h light period. Measured parameters were growth rates (μ), in vivo spectral absorption, cellular fluorescence capacity, pigment concentration, photosynthesis and photoreactivation during three cycles in controls and treatment samples. Growth rates diminished less than 35 % in Phaeodactylum and Aureococcus compared to 80-100 % decrease in Alexandrium and Spirulina. In these two last species, a significant increase in UV absorbing substances was o b s e r v e d . probably related to the presence of mycosporine-like aminoacids (MAAs) and scytonemin, respectively, and also lower photoreactivation efficiency compared to Phaeodactylum andAureococcus. The analysis of photosynthetic performance under different PAR/UV -A ratios for Alexandtium and Phaeodactylum, could also explain the differences in μ . These results suggest that in time, species with high rates of photorepair might be more tolerant to UV-B than those species, which depend on the synthesis of UV absorbing compounds as their principal protection mechanism.


INTRODUCTION
The deleterious effect of UV-B radiation (limits 280-320 nm, Neale 2000) has been recognised as a potentially significant stress factor for populations of marine phytoplankton world-wide (Vernet 2000).At the population level, the net effect of damaging radiation is a function of the history of radiation, the radiation dose and the sensitivities of the individual species (Cullen & Neale 1993, Karentz 1994, Xiong et al. 1997).On the other hand, tolerance of phytoplankton to UV--A (320-400 nm) and UV--B radiation has received increasing attention in the past one to two decades in view of global increases in UV radiation reaching the surface of the earth as a result of the gradual depletion of stratospheric ozone (Frederick et al. 1994, El-Sayed et al. 1996).Phytoplankton species have evolved the ability to effectively acclimate to UV-B radiation in response to variations in UV levels across latitudes, seasons and depths, which could confer them, at least temporarily, some competitive advantages over species with limited tolerance to this radiation stress.
As the position of phytoplankton cells in the water column is limited by the availability of light for photosynthesis, cells are invariably exposed to the shorter, damaging wavelengths included in sunlight.Nearly all cell components can absorb UV-B radiation, yet most of the UV-B damages result from absorption of the radiation by proteins and nucleic acids.Depressions in primary production are attributed primarily, to direct damage to photosystem II (Iwanzik et al. 1983) and inhibition of the C0 2 fixing enzyme RUBISCO.Inhibition of the synthesis of chlorophyll a (Chi a) and other light harvesting pigments further limits recovery of the photosynthetic apparatus (Strid et al. 1990, Häder & Häder 1991, Fischer & Häder 1992, Molina & Montecino 1996).Damage to DNA from exposure to UV-B radiation results primarily in the production of pyrimidine dimers, which can be cumulative, resulting eventually in mutation or death (Karentz et al. 1991, Mitchell & Karentz 1993).
Our increased understanding of UV -B radiation effects on marine phytoplankton has revealed a complex and diverse array of mechanisms involved in protection and damage repair (Vernet 2000).It is possible that some of these mechanisms have a phylogenetic basis or represent pre-adapted tolerance (Reynolds 1997), whereas others have evolved in response to local conditions.Many algal cells are capable of synthesising UVabsorbing compounds that act as UV screens, thus preventing or reducing UV damage (Carreto et al. 1989, Karentz et al. 1991, Roy 2000), including Mycosporine-like amino acids (MAAs), which are ubiquitous among marine algae (Nakamura et al. 1982).Nevertheless, active synthesis ofMAAs in response to UV stress is not universal (Hannach & Sigleo 1998).The extracellular pigment scytonemin is found in several cyanobacterias especially in benthic forms (García-Pichel & Castenholz 1991, Quesada et al. 1995).Scytonemin has a broad band absorbance in the UV-A range, (maximum at 365-375 nm) and may be effective at low UV doses (Retamal 1999).Also, repair mechanisms include renewal of damaged proteins by de novo synthesis (Mate et al. 1998) and light-dependent (photorreactivation) or independent DNA repair (Karentz et al. 1991 ).
It has been suggested that the increased UV --B resistance of low latitude species is associated with highly efficient DNA repair mechanisms (Karentz 1994, Karentz & Spero 1995).
Tolerance to UV-B radiation stress is based on the efficiency of repair and protection mechanisms.The net effect reflects a balance between damage, repair and protection costs, with consequences for the survival, growth and reproductive success of the species under UV stress (Vincent & Neale 2000).
Since UV --B radiation dose decrease with latitude, differential tolerance can be expected among algae from different latitudes.In the Antarctica, phytoplankton exposed to a sudden UV --B radiation increase suffered significant growth inhibition (Jokiel & York 1984, Davidson et al. 1994, Helbling et al. 1994); whereas marked resistance and acclimation to UV --B radiation have been found at low latitudes (Helbling et al. 1992, Hazzard et al. 1997).The greater inhibition of growth at high latitudes may also be explained by the temperature dependency of biosynthetic repair (Roos & Vincent 1998).Thus, although the current increase in UV --B radiation in the southern tip of South America is considered modest compared to more equatorial latitudes (Frederick et al. 1994), it may be sufficient to elicit changes in the species composition of local phytoplankton assemblages.It is of particular importance to identify which mechanisms could be involved in the development of harmful algae blooms in Southern Chile (Mufioz & A varia 1997), based on the various ecophysiological strategies that allow species to exist in environments where chronic exposure to potentially damaging levels of UV radiation is unavoidable.Here we compare the physiological responses to UV-B radiation in four species of phytoplankton isolated along the coast of Chile.The species selected for this study are phylogenetically diverse and differ widely in habitat and cell size.
A bigger culture chamber was equipped with a combination of fluorescent lamps providing PAR (Philips TLD 18W/33), UV-A radiation (365 nm, Philips TLD 18W/08), and UV-B radiation (313 ± 12 nm, Q-Pannel).Figure lA shows the weighted emission of UV lamps.PAR was measured with a Licor quantum meter and a LI-250 cosine sensor.UV --A and UV -B radiation were measured with a VLX-3W interference filter radiometer (Cole Parmer, France).All measurements were done in air with the sensor covered by bag material and the appropriate cut-off filter.Because flat sensors were used, total flux was estimated from the sum of upward and downward measurements at specific locations within the growth chamber.
Inoculates from the stock were diluted to exponential growth densities (determined previously) and transferred in 100 ml UV --B-transparent polyethylene bags (Whirl-Pak NASCO, U.S.A.) to the big culture chamber for acclimation to PAR and UV --A radiation prior to UV --B exposure.Cells were considered acclimated when the in vivo chlorophyll a (Chi a) fluorescence per cell did not change significantly with time (3 to 4 days).Cultures were then assigned to three experimental conditions: high UV-B, low UV-B, or control (without UV-B).All bags received PAR+ UV-A radiation on a 16:8 h L:D photoperiod.UV-B exposed bags received a high or a low dose of UV-B radiation, provided daily for 2 h at the center of the 16 h light period.High and low UV-B represent unweighted doses at 3.0-3.1 and 0.7-1.4kJ mˉ² d-1 respectively, obtained by varying the distance to the emission source (Table I).Whereas control bags were wrapped in polyester film to eliminate wavelengths < 320 nm, UV --B exposed bags were maintained in cellulose acetate film envelops to cut wavelengths < 280 nm (Fig. IB ).All cut-off filters attenuated PAR by 10 %.Meanwhile, three consecutive experimental runs ( 4-5 d each) were performed under similar (A) Espectros de radiaci6n de lamparas de UV-B (313 ± 12 nm) y UV-A (365 ± 12 nm) en unidades relativas y emisiones ponderadas de ambas lamparas y (B) transmitancia de filtros bloqueadores usados en los tratamientos con UV-B (lamina de acetato de celulosa) yen las muestras controles (lamina de poliester).

Photoreactivation
In a separate experiment, photoreactivation efficiency was estimated from cell viability following exposure to UV--B radiation.Photoreactivation experiments were performed on samples removed from the PAR+ UV --A acclimated cultures described above prior to their UV --B exposure.
Two series of 20 ml samples (n = 3) were exposed to a range of UV --B doses (0-7 .5 kJ mˉ²) in the absence of PAR.Immediately after this treatment, one series of cultures was incubated for three days under full range PAR ( 400-700 nm) to allow for photorepair of UV -B damage.For the other series under PAR, lacking the blue range of the spectrum ( 450-700 nm), a cut-off yellow filter was used to block photorepair wavelengths.Cell numbers were determined using the methods described for growth rate measurements and used to calculate rates of cell survival.The percent of photoreactivation was estimated from the difference between the survival rates of the series of cells allowed photorepair by subtraction of the respective areas under the curves (Retamal1999).
The photoreactivation efficiency at each UV-B dose was estimated as the survival rate of cells not allowed to photorepair relative to the survival rate of cells allowed to photorepair.Curves were compared through the Peto-Peto test (Pike & Thomson 1986).

Photosynthesis versus irradiance
Photosynthesis versus irradiance curves were determined for UV -B exposed (high UV --B, low UV --B) and control (UV --B excluded) cultures of A. catenella and P. tricornutum, both before and after 3 and 2 days of UV-B cycles, respectively.Photosynthetic rates were measured by N aH 14 C0 3 uptake in a thermoregulated photosynthetron incubator at 12 irradiance levels ranging from 10 to 600 μmol photons m-2 sˉ¹.Aliquots of 1 ml were incubated for 1 h with 1 μCi 14 C mlˉ¹.The samples were then fixed with formaldehyde, acidified by addition of 250 μ1 6N HCI and shaked for 1 h to remove excess 14 C0 2 (Montecino et al. 1996).
Total activity was measured with phenethylamine.Data of P-E were fit to the model of Jassby & Platt (1976): where PP is the photosynthetic rate, E is PAR, p max is the light-saturated photosynthetic rate, and a is the light limited photosynthetic rate per unit of PAR.Photosynthetic rates were normalised to cell-concentration and the photosynthetic parameters were compared using the nonparametric Mann-Whitney U test (Zar 1984).

Specific in vivo absorption (a*), chlorophyll a ( Chl a) and cellular fluorescence capacity ( CFC)
In vivo absorbance spectra (280-700 nm) were recorded on 5-15 ml samples concentrated onto 25 mm GF/F glass fiber filters using a Shimadzu single beam UV --1203 spectrophotometer attached to a computer.Sample filters were extracted in hot methanol, scanned again, and the second scan subtracted from the first to obtain absorption due to pigmented substances only (Kishino et al. 1985, Bricaud & Stramski 1990).A filter saturated with culture medium was used as a blank.Spectral absorption coefficients (a (λ), mˉ¹) were obtained p from the expression: where OD(λ) is the wavelength-specific optical density, OD(750) is the optical density at 750 nm (a correction for residual scattering), is a dimensionless pathlength amplification factor, Vf is volume filtered (m 3 ), A is the clearance area of the filter (m 2 ), and 2.3 converts from log base 10 to log base e. Specific absorption coefficients (a* p(λ), m 2 mg-1 ) were obtained as ap (λ)/Chl a (Stuart et al. 1998).In order to quantify the total absorption due to UV--absorbing substances, a*p (λ) was integrated at 2 nm intervals across the range 310-340 nm.Total absorption of UV-B exposed samples was expressed relative to the control and analysed by using the Mann -Whitney U test, Kruskal-Wallis test and the Tukey a posteriori test at P = 0.05 (Zar 1984 ).
Chlorophyll a was determined spectrophotometrically using the equations of Jeffrey & Humphrey (1975) in 10-15 ml samples concentrated onto 25 mm glass fiber filters (MFS), cold extracted in 90 % acetone for 24 h and clarified by low centrifugation.
For CFC measurements, samples (4 ml, n = 3) were dark adapted for 30 min and transferred in dim light to a Turner Ill fluorometer.Fluorescence was measured over the first 5 s of exposure to the excitation beam (Fb).Thereafter an aqueous solution of 3-(3,4-dichlorophenyl)-1, 1-dimethylurea (DCMU) was injected to the samples to a final concentration of 3 x I0-5 M and fluorescence was maximised after 30 s, and this plateau value was recorded as F a. CFC was obtained according to CFC = Fa-Fb/Fa (Vincent et al. 1984) Measurements in UV --B exposed samples were expressed relative to the control.The effect of UV-B radiation on in vivo absorbance, Chl a and CFC was evaluated by the non parametric Kruskal-Wallis test (Zar 1984).

Growth rates ( )
Three replicates for each treatment were incubated during 14 days for A. catenella and counted every 72 h from 1 ml aliquots fixed with formaldehyde (5 %) in a Sedgwick-Rafter counting chamber.Pheodactylum tricornutum and Aureococcus sp.cultures (n = 3) were counted daily for 7-8 days in a Fuchs-Rosenthal chamber (Martinez et al. 2000).Samples were manipulated under sterile conditions to minimize bacterial growth.Cell densities of S. subsalsa (n = 3) were estimated daily for 10 days from the optical density at 665 nm using a Unicam UV/ /VIS spectrophotometer (Retamal 1999).Cell-specific growth rates (μ) were determined by linear regression of the natural logarithm of cell abundance versus time through the next expression: where Cf is cell concentration at final time (tr) and Ci is cell concentration at initial time (ti) during the growth exponential phase (Kain 1987).

RESULTS
Tolerance to UV--B radiation varied between taxa, and the different responses of the measured parameters are described below for each of them.uted to Mycosporine-like aminoacids (MAAs).These changes in stap* were significant for P. tricornutum with differences according to the cell density of the samples (Fig. 2A and 2B).Comparing between high and low UV-B treatments, a significant stap* increase (P = 0.006) of 70 % occurred at low UV --B radiation, and only after the first day of exposure (Tukey test, P = 0.005) in the samples that were not diluted.Therefore, another experiment was run with cultures that were maintained optically thin, by diluting 1:1 with fresh medium.In these refreshened samples, significant differences in stap * were found between treatments at both times.At low UV-B the stap* increase was 57 % after one day and 76 % after the second day of exposure when compared with the high UV --B treatment (U = 9.0, P < 0.05, Fig. 2B).The highest increase in stap*was shown by A. catenella already after one day of exposure at both low and ligh UV-B (67 and 72% respectively) when compared to time zero (P < 0.05) with no differences between these two treatments (Fig. 2C).After three days of exposure the difference in stap* decreased by 50 % with high UV --B intensity (P < 0.05), nevertheless it was 50 % higher compared to time zero (Fig. 2C).While in Aureococcus sp.changes in stap * were not significant in time or between high and Low UV-B treatments (Fig. 2D), in S. subsalsa after the second day of exposure stap * increased both under UV --B (low intensity) and when exposed to UV --A.This increase reached 74% between 310-340 nm, and 83 % between 365-375 nm after 4 days of exposure when compared to time zero (P < 0.05) (Fig. 3A and 3B).

Cellular photosynthetic capacity (CFC)
Results analysed as CFC standardized values (stCFC = CFC,reatmen/CFCcontrol), showed that in P. tricornutum after one day exposure to UV --B under non diluted and diluted conditions stCFC values always decreased, and between treatments this decrease was significantly greater (P < 0.05) with high UV-B (Fig. 4A and 4B).In the samples that were not diluted, the stCFC decreased significantly (P = 0.001) and remained depressed until day three under low and high UV-B exposure.
With high UV-B the decrease was greater and reached 90 % after the third day of exposure compared with the low UV --B treatment (Fig. 4A).
In the daily medium diluted samples of P. tricornutum, the stCFC significantly decreased with high UV --B, and contrary to the behaviour under low UV --B between days one and three (Fig. 4B) showed no recovery after the third day of exposure (P < 0.05).
A. catenella showed a variable response in its photochemical efficiency (stCFC ) in replicated experiments (codes 13a and 15).Fig. 4C shows that under low UV --B no effect was found com- pared to high UV --B, where after one day of exposure a significant 50 % of irreversible decrease occurred (Mann-Whitney test, U = 9.0, P < 0.05).
However, in the second experiment a significant increase in stCFC independent of UV --B quantities was observed, which was higher with more time of exposure to UV --B (Fig. 4D).After the first day the increase was 33 % and after the third day of exposure the increase was 54 % higher than at time zero (P < 0.05, Fig. 4D) (stCFC values higher than 1.0 may be attributed to the effect of UV --A on control samples, see below and in Discussion).Also in Aureococcus sp. the s t CFC increased after the second day of exposure to low and high UV-B respectively (Kruskal-Wallis test, P < 0.05), when compared to time zero (Fig. 4E).

Chlorophyll a
A different response was obtained between species according to UV-B cycles and with no significant Chl a differences between UV-B dose (Table 2A).Changes in the concentration of this photosynthetic pigment per cell, showed a significant decrease in A. catenella and S. subsalsa only on the third day under high UV-B (Tukey, P = 0.01).In P. tricornutum the Chl a concentration changed according to cell concentration of the non refreshened samples, and in Aureococcus sp.no differences were found between time zero and day three (Table 2B ). A. catenella with a time zero value of 18.7 ± 1.4 pgChl a cell-1 in the first experiment, showed a 12.8% decrease and start- (B) Concentration of chlorophyll a (pgChl a per cell) in A. catenella, P. tricornutum, Aureococcus sp. and S. subsalsa, arranged according to cell size, in the controls (Time = 0) and according to Table 2A.The mean pgChl a cellˉ¹ values ± standard error (n = 6) for all treatments during three UV --B cycles are shown.The range of the number of cells mlˉ¹ in control samples is given below name of each species' name.Superscripts a and b indicate differences through time (Tukey test, P < 0.05) Concentraci6n de clorofila a (pgCI a por celula) en A. catenella, P. tricornutum, Aureococcus sp.y S. subsalsa, en los controles (Tiempo = 0) y de acuerdo con Ia Tabla 2A.Se presentan los valores promedio de pgCI a celˉ¹ ±error estandar (n = 6) de todos los tratamientos durante tres ciclos de UV -B.Tam bién se indica el ran go de celulas mlˉ¹ en las muestras control bajo el nombre de cada especie.Los superfndices a y b indican diferencias en el tiempo (prueba de Tukey, P < 0,05) UV-B cycles (d-1) ing with 28.6 ± 3.6 pgChl a cellˉ¹ in the second experiment the decrease reached 31 %.In P. tricornutum Chl a diminished significantly after one UV -B cycle (from 0.66 ± 0.14 to 0.28 ± 0.01 pgChl a cell-1 ) in the not diluted cultures.This response more than an UV effect relates to PAR availability (selfshading), considering that in the diluted cultures Chl a concentrations ( -0.29 pgChl a cell-1 ) did not change significantly in time.
P versus E and photosynthetic parameters in A. catenella and P. tricornutum As a function of PAR irradiance, photosynthetic performance (P-E curves) measured through autotrophic carbon fixation (P) after A. catenella and P. tricornutum samples were exposed to UV radiation, was different between these two species and different experiments (Fig. 5).P results were normalized to cell counts because low light photosynthetic efficiency (a) values could not be compared between these two species when normalized to Chl a.This occurred as a result of the different proportions of UV A/PAR, or the accumulated amount of UV A received during the 16 hours light period in the control samples that were placed nearest to the UV --Band UV A lamps.
Consequently the amount of UVA received by the control treatment was 1.3 times higher than the treatments under high and low UV --B.This affected differentially A. catenella (Fig. SC, SD, SE and SF) compared with P. tricornutum (Fig. SA and SB).The same applies to the results in the photosynthetic parameters a and maximum carbon fixation (P max' cell-1 ), the CFC response (Fig. 4D) and also in the RUBISCO pool, that were not significantly lower than the controls (data not shown).A similar experiment (code N°15) showed the same CFC decrease in the controls, generating the observed relative increase in the UV-B treatments with a photosynthetic efficiency that was also affected (data not shown).
In P. tricornutum, the analysis of the response in a and in the light-saturated rate of photosynthesis (P max cell 1 ) in thicker growing cultures (Fig. SA), indicates no statistically significant differences between treatments, compared to those obtained when selfshading was avoided in the culture samples through daily dilution with fresh medium (Fig. SB).The comparison of both photosynthetic parameters p max cellˉ¹ and a between exposure cycles T1 and Tf in A. catenella, show a significant decrease for all treatments (Table 3A).Between experiments, these parameters in P. tricornutum under different cell concentration (non diluted and diluted) were also significantly different (Table 3B).

Photoreactivation
In A. catenella, P. tricornutum, Aureococcus sp. and S. subsalsa the highest survival percentage of the cells grown under full PAR (control) was statistically different from the cells incubated in the absence of blue light (400-450 nm) (P < 0.05).
The clearest response was shown by P. tricornutum, based on the highest difference (> 48 %) between both curves (Fig. 6A) compared to 23 % in A. catenella (Fig. 6B), 24 % in S. subsalsa (Fig. 6D) and 12 % higher survival in Aureococcus sp. (Fig. 6C).In order to further quantify the differences between species, the results are also compared through the relationship between UV --B dose versus photoreactivation efficiency (survival percentage of the control sample/ survival percentage of the treatment sample).The highest photoreactivation efficiency with increased UV --B dose was shown by P. tricornutum, evidenced by the smallest slope (-2.82,R 2 = 0.88).For S. subsalsa the slope was -3.78 (R 2 = 0.73), for A. catenella -5.97 (R 2 = 0.99) and for Aureococcus sp.-7.89 (R 2 = 0.99).By statistical comparison of these curves, the photoreactivation efficiency was the same between A. catenella and S. subsalsa.

DISCUSSION
In this study differential interespecific sensitivity to UV --B stress, was found between the four microalgal strains isolated from the Pacific coast of South America, as reflected in photosynthetic performance (Fig. 4 and 5) and growth rates (Fig. 7).This was related to the ability of the cells to acclimate, using different strategies, mainly through photoreactivation (Fig. 6) and synthesis of UV absorbing compounds (Fig. 2 and 3 ).Great variation in microalgae sensitivity (Xiong et al. 1996) is due to the presence of different mechanisms that deal with the direct effects (Siebeck 1981, Mitchell & Karentz 1993, Vincent & Roy 1993, Davidson et al. 1994 ), of which the increase in the cellular concentration of screening agents has received broad experimental support (Carreto et al. 1989, Lesser et al. 1996, Jeffrey et al. 1999).When exposed to UV, photosynthesis of the dinoflagellates Heterocapsa and Prorocentrum was less affected than Phaeodactylum (Eke lund 1994) similarly to what was found here between P. tricornutum and the dinoflagellate A. catenella.Nevertheless, the increase in screening agents in A. catenella as observed in Prorocentrum micans (Lesser 1996), did not provide complete protection against UV effect, especially in relation to the effect of UV-  SA and SB).This occurred because in the experimental design we prioritized optimal distribution of treatment replicates in the big culture chamber in relation to PAR, situating the control samples nearest to the UV --Band UV A lamps.The absence of response in Aureococcus sp. in terms of an increase in UV absorbing compounds or photoreactivation capacity is noteworthy, because its growth diminished less than 50%.The question remains if this chrysophyte is screening UV through the cell wall constituents or using other efficient repair mechanisms.Under enhanced UV-B, chrysophytes were reported to replace diatoms in a mesocosmos experiment (Mostajir et al. 1999).
The UV-B effects on the physiology and ecology of marine phytoplankton (Vernet 2000) have been categorised in four basic responses: avoidance, stress reduction, damage repair and acclimation.Moreover, photoprotection and repair processes dealing with PS II damage are known to be dependent on previous light history and photosynthetic physiology of microalgae (Roy 2000).Comparaci6n de Ia capacidad de fotorreparaci6n a traves del porcentaje de sobrevivencia medido como el área entre tratamientos despues de 3 dias de exposici6n de muestras a diferentes dosis de UV-B (kJ mˉ²), comenzando con UV-B = 0. Las muestras con troles, incubadas bajo PAR ( 400-700 nm), corresponden a las line as continuas y los tratamientos, incubados sin luz azul (450-700 nm), corresponden a las lineas punteadas, en (A) Phaeodactylum tricornutum (B) Alexandrium catenella (C) Aureococcus sp.y (D) Spirulina subsalsa.
photosynthetic reaction centre is rapidly reversible, and in natural phytoplankton communities photodamage to PS II appears to be completely repaired overnight (Vassiliev et al. 1994 ).From studies in microalgae thylakoid membranes and PS II polypeptides, it has become clear that D 1 degradation is not the immediate cause of photoinhibition but a consequence of inactivation of PS II primary photochemistry following exposure to high levels of PAR (Long et al. 1994).Furthermore, gene induction is also a defence response against UV stress (Mate et al. 1998).CFC as a crude measurement of open RC II traps (Vincent et al. 1984) showed that A. catenella was the species that recovered even under high UV -B conditions.Also their Chl a concentration per cell did not change significantly between UV-B treatments (Table 2A), and was in fact more dependent on PAR or cell density (Table 2B).The same behaviour was observed in Aureococcus sp.; nevertheless, in P. tricornutum CFC decreased under both high and low UV-B in the diluted and non diluted samples (Fig. 4A and 4B).C h l a concentration per cell changed around the limits of 0.3-0.5 pgChl a cell 1 reported for this species.P. tricornutum presented a slightly lower growth rate of 0.9 d 1 versus 1.2 d 1 (Geider et al. 1985) with the highest photoreactivation capacity.The cost of the protection mechanisms to the plants are not known (Rozema et al. 1997).Nevertheless it has to be taken into account that in small cells there is not much cytoplasm to accumulate MAAs (Garcfa-Pichel 1994) and investment in the synthesis of UV --B absorbing secondary metabolites is expensive.Photorepair is a nearly universal, blue light-dependent, and primary mechanism of DNA repair.Therefore, especially in small phytoplankton species, light correction of UV damage should be an important factor in cell survival (Karentz et al. 1991), although (Davidson et al. 1994) found no consistency in relation to size.According to our results, under high and low UV-B, photosynthesis damage was prevented in a short time scale (one day) in the larger cells through the synthesis ofUV absorbing compounds.Nevertheless, chronic UV-B exposure (3 days) decreased A. catenella's photosynthetic capacity and at a threshold of 2.9 kJ m-2 dˉ¹ of UV --B radiation, there was no investment in growth and μ reached 83 % inhibition (Martinez et al. 2000).
This would support the idea of avoidance responses, such as vertical migration in relation to the radiation source, which could be the case for this species in the natural environment.Inhibition of growth was also found despite MAAs and scytonemin synthesis inS.subsalsa (Fig. 3A and 3B ), inhabiting tidal pools and growing at low PAR intensities of 30-40 μmoles mˉ² sˉ¹.The more efficient use of low light produces a smaller dependence on irradiance of the division rate than for the carbon uptake (Rivkin & Putt 1987), concluded that benthic algae are saturated at very low PAR intensities in relation to photosynthesis and cell division.In relation to growth, short term variations in irradiance and cellular metabolism, photosynthesis and nutrient uptake is a matter of debate.Wangberg et al. (1996) described the effects of UV-B on some of these processes and suggested that biomass and composition of marine phytoplankton was modified by UV radiation.
Aside from UV absorbing compounds and increase in photoprotecti ve pigments ( carotenoids) found inS.subsalsa (Retamal 1999), the arrangement of the filaments (trichomes) can influence survival, because less damage was found in the  ] of all species, considering those experiments that received similar doses of UV-B as shown in Table 1.
larger trichomes (> 160 mm) probably related to an increased selfshading effect.Responses also varied with exposure times and in this benthic species a significant response was obtained after 2-4 cycles of UV --B (Fig. 3), suggesting adaptation to chronic UV environments.In Antarctic phytoplankton, selfshading can provide additional protection and Lesser et a!.(1996) found a decrease of 22 % in light saturated rates of photosynthesis, using cut-off filters at 375 nm, in cultures maintained at low levels of Chi a (30 μgl).
In the present study, the size of the samples was predetermined (concentration and volume) to allow for the different measurements.Therefore, in our case Chi a concentrations were up to 10 times higher, and consequently the effect of UV could have also been diminished by selfshading as it was the case in P. tricornutum (Fig. SA).Photoinhibition can also be prevented experimentally by acclimation with PAR (Neale et a!. 1994 ).In the long term a decrease in sensitivity means an increase in protection/resistance of phytoplankton (Villafane et a!. 1995).Short term inhibition and in the long term a steady state acclimation and recovery ofRUBISCO was found to ameliorate the inhibition of carbon fixation (Hazzard et a!. 1997).
Our results from monocultures would confirm that an increase of the UV component of radiation in surface waters, could modulate the structure and function of phytoplankton assemblages as observed in mixed cultures, different phytoplankton groups (Helbling eta!. 1994) and in the depthdifferential inhibition of Antarctic phytoplankton photosynthesis (Neale et a!.1998).A decrease in primary production with UV exposure was less at lower latitudes (Helbling eta!. 1993 ).Nevertheless, shifts could also be expected according to nutrient concentrations (Behrenfeld et a!.1995), because more UV --B tolerant species can have advantages during nutrient competition (Behrenfeld eta!. 1992).Other shifts can be biologically controlled, like the change from autotrophic to heterotrophic conditions reported for the brown tide forming species Aureococcus anophagefferens (Bricelj & Lonsdale 1997).At the community level, natural plankton assemblages exposed to UV -B showed an increase in the abundance of bacteria, flagellates and small phytoplankton, shifting from a herbivorous food web to a microbial food web (Mostajir et a!. 1999).
Experimental UV --B doses were lower compared with field measurements obtained with the same instrument at 50° Sin the spring of 1997 that ranged from 8.6-30.2kJ mˉ² dˉ¹.Ecological significant penetration of UV --B depends not only on phytoplankton biomass but also on the abundance of dissolved organic compounds (Kirk 1994, Goes et al. 1995).Therefore, in relation to the amount of UV-B radiation reaching the earth's surface and the extent of the Ozone Hole (Smith et al. 1992, Häder 1996, Orce & Helbling 1997, Rozema et al. 1997), the underwater light field needs also to be taken into account to predict changes in phytoplankton abundance.
Our study has extended the knowledge of microalgal response specificity and the realization of the possibility that cell acclimation can lead to a reduced effect of UV on production rates, from tropical and subtropical organisms to include organisms from lower latitudes.Species having a superior net growth performance under UV enhancement will be selected, given that species redundancy will discriminate in favour of those with superior preadapted tolerance (Reynolds 1997).In summary this study has shown that at low UV -B doses, efficient mechanisms are operating to prevent UV --B damage.The different strategies between taxa (Fig. 8) can be associated with cell size and shape or different taxonomic groups (Laurion & Vincent 1998).Extrapolations of laboratory culture results have to be taken with extreme caution because the small differences in wavelengths and in the proportions of radiation in 4 3 2 0 0 2 the experimental design, plays a major role in the time course of effects.Nevertheless, species with high rates of photorepair could have some competitive advantages over less tolerant UV-B specie, since species with lower rates of photorepair will have to expend energy for repair and acclimation mechanisms which would otherwise be used for growth.These advantages should insure that inhibition at the level of phytoplankton assemblages is hardly conceivable in the long term.

Fig. 1 :
Fig. 1: (A) Radiation spectra of the UV-B (313 ± 12 nm) and UV --A (365 ± 12 nm) lamps in relative units and weighted emission of both lamps and (B) transmittance of cut-off filters used for the UV-B treatments (cellulose acetate film) and control samples (polyester film).

Fig. 3 :Fig. 4 :
Fig. 3: Integrated specific absorption *) and standardized by each control (stap * =far* treatment fa *control) estimated from in vivo spectra at different times (in days) after exposures cycles of 2 h dayˉ¹ torlow UV-B in (A) Spirulina subsalsa fa* between 310-340 nm, and (B) Spirulina subsalsa fa* p p between 365-375 nm.Subscripts a and bon top of each bar, denote significant differences in time (Tukey groups).

Fig. 5 :
Fig. 5: Experiments of photosynthesis (normalized by cell counts, mgC cellsˉ¹ hˉ¹) versus light, in Phaeodactylum tricornutum and in Alexandrium catenella.(A) Non-daily diluted samples of Phaeodactylum tricornutum that were just previously exposed during 2 h to high and low UV --B and samples without UV --B (controls); (B) daily diluted samples of Phaeodactylum tricornutum that were just previously exposed during 2 h to high and low UV-B and samples without UV-B (controls); (C) Alexandrium catenella experiment number I (at Tl) with samples that were just previously exposed during 2 h to high and low UV-B and samples without UV-B (controls); (D) Alexandrium catenella experiment number I (at Tf) after 3 days of 2 h daily cycles of high and low UV-B and samples without UV-B (controls); (E) Alexandrium catenella experiment number 2 (at Tl) with samples that were just previously exposed during 2 h to high and low UV-B and samples without UV-B (controls); (F) Alexandrium catenella experiment number 2 (at Tf) after 3 days of 2 hours cycles of high and low UV --B and samples without UV-B (controls).Experimentos de fotosfntesis (normalizados por recuentos celulares, mgC celsˉ¹ hˉ¹) versus luz en Phaeodactylum tricornutum y Alexandrium catenella.(A) Muestras de Phaeodactylum tricornutum no diluidas diariamente y previamente expuestas a alta y baja UV-B durante 2 h y muestras sin UV-B (controles); (B) muestras de Phaeodactylum tricornutum diariamente diluidas y previamente expuestas a alta y baja UV-B durante 2 h y muestras sin UV-B (controles); (C) experimento mimero I en Alexandrium catenella (en Tl) con muestras previamente expuestas a alta y baja UV-B durante 2 h y muestras sin UV-B (controles); (D) experimento numero I en Alexandrium catenella (en Tf) despues de 3 dfas de ciclos de 2 h diarias con alta y baja UV-B y muestras sin UV-B (controles); (E) experimento numero 2 en Alexandrium catenella (en Tl) con muestras previamente expuestas a alta y baja UV -B durante 2 h y muestras sin UV -B (controles); (F) experimento numero 2 en Alexandrium catenella (en Tf) despues de 3 dfas de ciclos de 2 h diarias con alta y baja UV-B y muestras sin UV-B (controles).
Comparison of the photosynthetic parameter α (mgC cell-' h-' μmol m-2 s-1 ) and maximum Pcells (mgC cell-'h-') between experiments with Phaeodactylum tricornutum cultures after two cycles of UV--B radiation under high (selfshading) and low cell density (not selfshading) conditions.Values represents mean ± standard error our experiments, slightly different proportions of UV A/PAR affected differentially the control samples of A. catenella (Fig. SC, SD, SE and SF) compared with P. tricornutum (Fig.

Fig. 6 :
Fig. 6: Comparison of the photorepair capacity through the percentage of survival measured as the area between treatments after 3 days, from samples exposed at different UV-B doses (kJ m-2 ), starting with incident UV-B = 0. Control samples incubated under PAR ( 400-700 nm) correspond to the continuous line and treatments incubated without blue light ( 450-700 nm) correspond to the dashed lines in (A) Phaeodactylum tricornutum, (B) Alexandrium catenella, (C) Aureococcus sp. and (D) Spirulina subsalsa.

Fig. 8 :
Fig. 8: Predictive diagram for the relationships between photoreactivation, absorbing UV -B substances, and other protection mechanisms for the four microalgae.Scales are in arbitrary units.Diagrama predictivo de las relaciones entre fotorreparaci6n, sustancias absorbedoras de UV-B y otros mecanismos de protecci6n para las cuatro microalgas.Las escalas están en unidades arbitrarias.

TABLE 3
was funded by FONDECYT grant 1960875, concurrent lately to the Program FONDAP-HUMBOLDT.At Universidad de Chile, Faculty of Sciences we also acknowledge Dr. Raúl Morales for providing the UV sensor, for laboratory facilities Dr. Fernando Zambrano with the spectrophotometer and Dr. Tito Ureta for the Scintillation counter.We are deeply thankful to Gloria Collantes, Miriam Segue!, Dr. Benjamin Suarez and Dr. Nestor Lagos for providing the cultures and also Verónica Mufioz for helping during the experiments and analysis of samples.Finally, we acknowledge Walter Helbling and Alejandro Buschmann for the invitation to the Minisimposium of Photobiology at the 1999 Phycology Congress.