Epipelic diatoms are important components of microphytobenthic biofilms. Cultures of four diatom species (Amphora coffeaeformis, Cylindrotheca closterium, Navicula perminuta and Nitzschia epithemioides) and assemblages of mixed diatom species collected from an estuary were exposed to elevated levels of ultraviolet-B (UV-B) radiation. Short exposures to UV-B resulted in decreases in photosystem II (PSII) photochemistry, photosynthetic electron transport, photosynthetic carbon assimilation and changes in the pattern of allocation of assimilated carbon into soluble colloidal, extracellular polysaccharides (EPS) and glucan pools. The magnitude of the effects of the UV-B treatments varied between species and was also dependent upon the photosynthetically active photon flux density (PPFD) to which the cells were also exposed, with effects being greater at lower light levels. Both increases in non-photochemical quenching of excitation energy in the pigment antennae and photodamage to the D1 reaction centres contributed to decreases in PSII photochemistry. All species demonstrated a rapid ability to recover from perturbations of PSII photochemistry, with some species recovering during the UV-B exposure period. Some of the perturbations induced in carbon metabolism were independent of effects on PSII photochemistry and photosynthetic electron transport. Elevated UV-B can significantly inhibit photosynthetic performance, and modify carbon metabolism in epipelic diatoms. However, the ecological effects of UV-B at the community level are difficult to predict as large variations occur between species.
Many species of epipelic diatoms are motile, moving via the production of EPS from the raphe in the cell wall. Populations show endogenous rhythms of vertical migration within sediments, and must spend a portion of each day in the photic zone of the sediments to photosynthesize and grow (Underwood & Kromkamp 1999; Underwood et al. 2005). This need for light may well require a trade-off between the absorption of adequate light for photosynthesis and the increasing exposure to UV-B radiation during tidal emersion. Peletier et al. (1996) examined the effect of enhanced UV-B radiation on diatom growth rate using unialgal cultures, and found that the dose required to significantly reduce the diatom growth rate was in excess of that predicted for worse-case scenarios of ozone depletion. Roux et al. (2002) monitored ultraviolet (UV)-induced changes in carbon assimilation and assemblage composition and, again, found that UV-B radiation did not exert a major stress in intertidal systems. In planktonic and sub-tidal diatoms, UV-B radiation has been shown to alter the way in which assimilated carbon is allocated to intra- and extracellular carbon pools (Goes et al. 1996; Underwood et al. 1999; Wulff et al. 1999, 2000). The way in which UV-B affects how carbon is allocated to intra- and extracellular colloidal carbon pools has not been investigated in intertidal microphytobenthic systems. If enhanced UV-B radiation results in changes to parameters other than cell division and total primary productivity, then the stress placed on estuarine systems by increases in UV-B may be greater than that predicted by Peletier et al. (1996) and Roux et al. (2002).
The ability of epipelic diatoms to migrate may provide a means for photoprotection, either from high photosynthetically available radiation (400–700 nm) (Kromkamp, Barranguet & Peene 1998; Perkins et al. 2002), or, possibly, from damaging UV-B radiation (Underwood et al. 1999). A UV-specific migratory response has also been reported to occur in cyanobacterial mats (Bebout & Garcia-Píchel 1995). The presence of a migratory response makes it difficult to study the potential effects of environmental stress on epipelic diatom cell physiology in intact systems as cells may simply move to avoid the stress. In addition, this migration makes the control of the stress treatments impracticable. In the current study, a series of experiments was designed to test for UV-B-induced changes to photosynthetic electron transport, carbon assimilation and allocation of carbon to intra- and extracellular carbohydrate pools. Monocultures of Cylindrotheca closterium (Ehrenberg) Reimann and Lewin, and diatom assemblages that had been removed from the sediment matrix were used to examine the responses to enhanced UV-B in the absence of any vertical migration of the cells. Furthermore, experiments were performed to determine whether there were species-specific responses to UV-B radiation for four commonly occurring intertidal epipelic diatoms –Amphora coffeaeformis (Kützing) Archibold and Schoeman, C. closterium, Navicula perminuta (Grun.) in van Heurck and Nitzschia epithemioides (Grun.) in Cleve and Grunow.
MATERIALS AND METHODS
Samples of C. closterium, N. perminuta, A. coffeaeformis and N. epithemioides were obtained from the University of Essex's culture collection. Each culture was grown in a filter-sterilized f/2 medium (Guillard & Ryther 1962) at a salinity of 20 PSU (Tropic Marin, Betta Aquatics, Elmstead, UK) and placed in a controlled-environment incubator (LMS, Sevenoaks, Kent, UK) maintained at 20 °C. The logarithmic growth of each culture was maintained by subsampling into a fresh medium every 5–7 d.
The C. closterium cultures were acclimated to the experimental photosynthetically active photon flux density (PPFD) of either 150 or 300 µmol m−2 s−1 produced from sodium lamps (see inset in Fig. 1 for spectral irradiance) by placing 25 mL aliquots of a culture into 400 mL f/2 medium in 1 L flasks for 5 d before the experiment. All other species were acclimated to a PPFD of 150 µmol m−2 s−1. The acclimation, in all cases, took place in an experimental cabinet, which was constructed with Dexion (Dexion Comino Ltd, Bromsgrove, UK), screened with Mylar-D film (315 nm cut-off) (Secol Ltd, Thetford, UK). Perspex trays filled with water (30 mm deep) were placed below the sodium lamps, but above the UV-B tubes, to absorb infrared radiation and maintain the temperature at 20 ± 2 °C. Neutral density filters were placed immediately below the perspex trays to give an even PPFD of either 150 or 300 µmol m−2 s−1 to the cultures.
A day before the cultures were exposed to elevated UV-B, 300 mL of each medium was removed from above the submerged, settled biofilms. The removed media were replaced with fresh media in which the cells were then suspended. Four millilitres of the cell suspension from each culture was pipetted into eight wells of six replicated 100 × 100 mm square Petri dishes (Bibby Sterilin Ltd, Stone, UK, 25 wells per dish, with a maximum volume of 5 mL well−1).
Isolation of natural assemblages
The surface (top 5 mm) sediment scrapes were obtained from Alresford Creek, a branch of the Colne Estuary, Essex, UK (51°50.2′N, 0°59.5′E) during a low tide. The sediment was placed on trays c. 20 mm deep; sprayed with filtered, high-tide, estuary water to prevent desiccation and left overnight. The following morning, the biofilms were removed from the sediment surface by placing double layers of microscope lens tissue (Whatman, Maidstone, UK) on the sediment surface for 1 h (Eaton & Moss 1966). The upper-lens tissues were placed in 500 mL f/2 medium and agitated with a glass rod to release the diatom cells. The cell suspension was filtered through muslin and diluted with an f/2 medium to an initial concentration of 1–2 µg Chl a ml−1. Four millilitres of the cell suspension was added to the Petri dish wells as described earlier, and the dishes were left for 24 h in the experimental cabinet for the cells to settle. The resulting biofilms in the cultures of C. closterium and the mixed species were a maximum of 2–3 cells deep to minimize self-shading.
In each isolated experiment, three 500 µL samples were fixed with Lugol's iodine (2% final concentration) for the assemblage identification. The diatom assemblages were dominated by small Navicula species, including N. gregaria, N. phyllepta and N. perminuta, which accounted for ∼80% of the diatom cell numbers. Also larger diatom species such as Pleurosigma angulatum, Plagiotropis vitrea and Gyrosigma limosum were present. Euglenids accounted for less than 2% of the total cell numbers.
The dishes containing the cultures and mixed-species assemblages were placed in the experimental cabinet, kept at 20 ± 2 °C at a PPFD of 150 or 300 µmol m−2 s−1. The cabinet was divided into two sections by a sheet of Mylar-D film (315 nm cut-off). Half of the cabinet was illuminated by two UV tubes (TL20 W/12/RS, Phillips, Hamburg, Germany). The spectral output of the lamps was measured with a Macam Photometrics SR-9910-PC spectroradiometer (Livingston, UK), which was fitted with a cosine corrected sensor. The spectral irradiances for the three UV treatments and the control are shown in Fig. 1. The irradiance of 0.35 W m−2 UV-B, when given for 10 h, and weighted to the generalized plant action spectrum (normalized to 300 nm; Caldwell 1971), gave a daily cumulative exposure of 12 kJ m−2 d−1, which is approximately 2.5 times higher than the current UK maximum summer exposure (Nogués & Baker 1995; Bartlett & Webb 2000). The relatively high doses of UV-B radiation were used to provide an understanding of the potential mechanisms of UV-B damage in intertidal epipelic diatoms. The spectroradiometer scans were repeated at 10 cm intervals in a grid pattern across the cabinet, and sections were marked to ensure that the replicate dishes received the same photo-energy inputs. The doses of ultraviolet-A (UV-A) radiation (320–400 nm) received by the control and UV-B treatments were the same throughout each experiment (Fig. 1). The ultraviolet-C (UV-C) radiation was screened out with 125 µm cellulose diacetate film (290 nm cut-off) (Harrold Manufacturing, Silsby, UK), which had been pre-solarized for 6 h. A dark period of 12 h was used to simulate the night-time conditions. For the second experiment, the dishes containing C. closterium, N. perminuta, A. coffeaeformis and N. epithemioides were placed in the UV-B-irradiated section at varying distances from the UV-B tubes so that dose rates were 0.365 W m−2 (high), 0.262 W m−2 (medium) and 0.064 W m−2 (low); again, the photo-energy inputs were the same between the replicate dishes. A set of control dishes was placed under the UV-B tubes that were screened with pre-solarized cellulose diacetate and Mylar-D films to remove the UV-C and UV-B radiation. All films were replaced after an exposure to a maximum of 20 h UV radiation because of UV-induced changes in transmission properties.
PSII photochemical efficiencies
The Chl a fluorescence was measured with a Xenon-PAM Fluorometer (Heinz Walz GmbH, Effeltrich, Germany) through a 680 nm bandpass filter (Coherent, Watford, UK). A modulated (2 Hz) measuring light at ∼0.6 µmol photons m−2 s−1 was provided by a xenon flash lamp. Saturating light pulses of 500 ms (approximately 4000 µmol photons m−2 s−1) were produced from a halogen lamp. Both measuring and saturating light sources were filtered by BG39 filters (Coherent), and a heat-reflecting mirror was placed in front of the saturating light source. Fluorescence measurements were made directly from the biofilms within the dishes so that biofilm optical properties were not altered by placing a cell suspension in a cuvette. All dishes were placed on a black cotton cloth. Neither the cloth nor the dishes emitted any detectable fluorescence.
Prior to the start of the experiments, the samples were dark adapted for 1 h, after which Fo and Fm were measured. The same parameters were measured following the exposure of the samples to 10 h UV-B radiation and a 1 h dark-adaptation period. Fv/Fm = (Fm − Fo)/Fm was calculated. After 6 h of UV-B exposure, F′ and Fm′ were measured. From these measurements, the operating efficiency of PSII photochemistry [Fq′/Fm′ = (Fm′ − F ′)/Fm′] was calculated (Genty, Briantais & Baker 1989). Fq′/Fm′ was determined at 13 actinic light levels between 9 and 900 µmol m−2 s−1.
Fv/Fm was also determined after 1, 2.5, 4 and 6 h for C. closterium, N. perminuta, A. coffeaeformis and N. epithemioides exposed to the three UV-B doses and for the controls using a dark-adaptation time of 20 min, which was a compromise between times for optimal dark adaptation for the relaxation of non-photochemical quenching and to minimize the dark repair of the UV-B damage.
In an additional set of experiments, 500 µL of the f/2 medium containing streptomycin sulphate (500 µg mL−1, final concentration) (Sigma-Aldrich Co., St. Louis, MO, USA) was added to the dish wells containing 4 mL of the unialgal cultures in order to inhibit the de novo synthesis of the chloroplast-encoded D1 protein of the PSII reaction centre (Schnettger et al. 1994).
Carbon assimilation and allocation in unialgal diatom cultures and mixed-species assemblages under differing PPFD and UV-B regimes
The photosynthetic carbon assimilation and allocation of assimilated carbon into colloidal-S, EPS and glucan pools were measured at PPFDs of 150 and 300 µmol m−2 s−1 in the cultures of C. closterium and freshly isolated intertidal mixed-species assemblages, in parallel with the determinations of Fq′/Fm′.
Following 6 h of UV-B radiation (0.35 W m−2), 500 µL of NaH14CO3 (Amersham Biosciences UK Ltd, Little Chalfont, UK, 37 k Bq mL−1) radio-labelled f/2 medium was added to the UV-B-irradiated and control wells. The samples were incubated under the respective treatments for 1 h. The incubation was terminated after 1 h by the addition of glutaraldehyde (1% v/v final concentration). The controls were dark-incubated, dead cells (killed by the addition of glutaraldehyde prior to isotope addition) and unlabelled cells to correct for background counts.
The Chl a concentration was determined in 1 mL of the sample after an overnight extraction at 4 °C in 2.5 mL methanol. The absorbance was measured spectrophotometrically at 665 and 750 nm, and corrected for phaeopigment content by acidification with 90 µL 10% HCl (Jensen 1978). The Chl a concentration was calculated according to Stal, Vangemerden & Krumbein (1984).
The rates of primary productivity and allocation of assimilated carbon into colloidal-S, EPS and glucan pools were determined using the methods of Smith & Underwood (1998). A 0.5 mL subsample of the original culture was placed into 5 mL scintillation liners to determine the total primary productivity. The remaining sample was centrifuged at 3400 g for 15 min, and 1 mL of the supernatant was used to determine the colloidal-S allocation. The EPS fraction was precipitated overnight from an aliquot of the supernatant in 70% (v/v) ethanol. The glucan fraction was extracted from the remaining cellular pellet in 4 mL of 0.2 manova) was used to determine the statistically significant differences in Fv/Fm, Fo and Fm; photosynthetic carbon assimilation; carbon allocation to colloidal-S, EPS and glucan pools using incubation time and UV-B dose as factors. The fluorescence anovas used a repeated-measures design. Fv/Fm, as a percentage of the control, was calculated at each time-point and plotted against cumulative UV-B dose (in kJm−2) to examine the nature of the relationship between UV-B dose and dose rate, and reduction in Fv/Fm (Xiong 2001). Where data were significantly different, post hoc Tukey tests were used to identify the significantly different groups. The difference/effect was considered significant at the P = 0.05 level.
Effects of UV-B radiation on PSII photochemistry
Cylindrotheca closterium and the mixed assemblages were incubated at PPFDs of 150 and 300 µmol m−2 s−1 and exposed to 0.35 W m−2 of UV-B radiation. The maximum quantum efficiency of PSII photochemistry (Fv/F) was measured prior to and following 10 h ± UV-B exposure over a 2 d period (Fig. 2a). There was a progressive reduction of Fv/Fm in the UV-B-irradiated C. closterium, compared to the control cultures, in the 150 µmol m−2 s−1 PPFD experiment (F3, 32 = 32.4, P < 0.001) (Fig. 2a). At 300 µmol m−2 s−1 PPFD, the UV-B-induced depression of Fv/Fm in C. closterium also increased with an increasing duration of UV-B exposure (F3, 32 = 20.1, P < 0.001). The magnitude of UV-B-induced reduction in Fv/Fm was significantly greater at an incubation PPFD of 150 µmol m−2 s−1 than at 300 µmol m−2 s−1 (F1, 86 = 5.6, P = 0.02).
The UV-B radiation caused a significant reduction in Fv/Fm in the mixed-species assemblages at both light levels (F3, 32 = 15.9, P ≤ 0.001 and F3,32 = 22.0, P ≤ 0.001) (Fig. 2b). However, there was no significant difference in the magnitude of Fv/Fm reduction between the two light levels, which is in contrast to that found in C. closterium. After the dark periods, there was an indication of some recovery in Fv/Fm in all cases; there was a greater degree of recovery in C. closterium and the mixed assemblages that had received a higher PPFD during the UV-B irradiation.
The reductions in Fv/Fm in both C. closterium and the mixed assemblages occurred because of increases in Fo, consistent with photo-inactivation and damage of PSII reaction centres, and decreases in Fm, indicating the development of non-photochemical quenching processes. However, there were differences in the nature of the changes in Fo and Fm for the two incubation PPFD treatments (Fig. 3). When UV-B was given with 150 µmol m−2 s−1 incubation PPFD (Fig. 3a), a 30% increase in Fo occurred in C. closterium (F3, 32 = 16.7; P ≤ 0.001; Tukey test, P ≤ 0.001), and an 11% increase in Fo (Fig. 3c) was seen in the mixed assemblages at the end of the first exposure period (F3,32 = 14.0; P ≤ 0.001; Tukey test, P ≤ 0.001). Reductions in Fm were immediate in C. closterium, 14% (P ≤ 0.001), but the reduction in Fm in the mixed assemblages, at the lower PPFD, was not statistically significant. At 300 µmol m−2 s−1, Fo increased in C. closterium, by 20% (F3, 32 = 4.1, P = 0.017) and the mixed assemblages by 5% (F3, 32 = 27.1, P ≤ 0.001), but in both cases the increase in Fo occurred after the second UV-B exposure period (Tukey test, P = 0.008 and 0.012, respectively). In contrast, the reduction in Fm occurred after the first UV-B exposure period, by 15% in C. closterium (F3,32 = 7.9; P ≤ 0.001; Tukey test, P = 0.002), and by 7% in the mixed assemblages (F3, 32 = 4.8; P = 0.01; Tukey test, P = 0.043). There were also differences in how Fo and Fm changed following the dark period. Fo decreased following the dark period in C. closterium, but remained unchanged in the mixed assemblages. In contrast, Fm decreased following the dark period in C. closterium, but increased in the mixed species. This increase in Fm is consistent with the relaxation of non-photochemical quenching.
The influence of two 6 h exposures to 0.35 W m−2 UV-B radiation, at either a PPFD of 150 or 300 µmol m−2 s−1, on the operating efficiency of PSII photochemistry (Fq′/Fm′) in C. closterium and the mixed-species assemblages over a range of PPFDs from 9 to 900 µmol m−2 s−1 is shown in Figs 4 and 5. At the lower incubation PPFD, UV-B radiation caused a reduction in Fq′/Fm′, relative to the controls, throughout the light curve. This reduction was greatest during the second UV-B exposure period (Fig. 4b).
In the samples incubated at the higher PPFD, Fq′/Fm′, when measured at PPFDs between 9 and 300 µmol m−2 s−1, was lower in UV-B-irradiated C. closterium than in the controls (Fig. 4c). At PPFDs above 300 µmol m−2 s−1 in the light curve, the difference was less pronounced. Reductions in Fq′/Fm′ were greatest during the second exposure period (Fig. 4d).
UV-B radiation caused similar reductions in Fq′/Fm′ in the mixed-species assemblages as in C. closterium, although UV-B-induced reductions in these values were less pronounced. At 150 µmol m−2 s−1 incubation PPFD, UV-B radiation caused a reduction in Fq′/Fm′, and this reduction was greatest during the first exposure period (Fig. 5a). The reduction in Fq′/Fm′ was less pronounced in the mixed assemblages incubated at the higher PPFD (Fig. 5c & d).
In summary, UV-B radiation caused reductions in Fv/Fm and Fq′/Fm′ in both C. closterium and the mixed-species assemblages. These reductions were greatest in C. closterium when UV-B was given in conjunction with an incubation PPFD of 150 µmol m−2 s−1 compared with an incubation PPFD of 300 µmol m−2 s−1.
Differential sensitivity to UV-B radiation of C. closterium, N. perminuta, A. coffeaeformis and N. epithemioides
The aims of this study were to investigate whether four diatom species, which occur commonly in intertidal benthic mats, exhibit differential sensitivity to UV-B radiation, and to determine if the apparent reduced sensitivity of the mixed assemblages was because of a ‘masking’ effect from UV-B-resistant species. There were clear differences in the effect of a 6 h UV-B exposure on Fv/Fm of the four species (Fig. 6). In C. closterium, UV-B radiation caused a progressive reduction in Fv/Fm, which occurred at all three dose rates (F3,64 = 203.778, P ≤ 0.001) (Fig. 6a). In N. perminuta, UV-B radiation also caused a depression in Fv/Fm (F3,64 = 56.305, P ≤ 0.001) (Fig. 6b); however, by the end of the UV-B exposure period, only the high-dose rate values were significantly lower than those of the controls. In A. coffeaeformis, Fv/Fm was significantly lower with all three UV-B dose rates (F3,64 = 76.364, P ≤ 0.001) and changed during the period of the experiment (F3,64 = 42.208, P ≤ 0.001) (Fig. 6c). Following an initial depression with all UV-B dose rates, recovery had occurred after 6 h so that Fv/Fm was only reduced in the high-dose rate treatment (Tukey test, P ≤ 0.001), suggesting the presence of a repair mechanism in N. perminuta and A. coffeaeformis. Fv/Fm in N. epithemioides was significantly reduced with medium and high UV-B treatments (F3, 64 = 122.340; P ≤ 0.001; Tukey tests, P ≤ 0.001) (Fig. 6d). There was no recovery in Fv/Fm by the end of the exposure period in both N. epithemioides and C. closterium, suggesting the possibility of a reduced acclimatory response or repair capacity than what occurred in N. perminuta and A. coffeaeformis.
The ability of the four different species to repair UV-B-induced damage to PSII reaction centres was examined by incubating the four different species at a PPFD of 60 µmol m−2 s−1 and exposing them for 2 h to 0.262 W m−2 UV-B in the presence and absence of streptomycin, which inhibits protein synthesis by 70S ribosomes and consequently prevents de novo synthesis of the D1 reaction centre protein of the PSII (Schnettger et al. 1994). The control cultures (± streptomycin, without UV-B radiation) were also prepared, but the Fv/Fm values were not significantly different throughout the 8 h period (data not shown). For all the species, the presence of streptomycin resulted in a significant decrease in Fv/Fm during an exposure to UV-B and following a 6 h recovery period after the removal of the UV-B treatment (Table 1). In all the species, anova indicated that the inhibition of the repair of D1 had a significant effect on the rate of recovery of all four species, but this effect was greatest in A. coffeaeformis and N. perminuta. However, all species exhibited considerable recovery of Fv/Fm in the presence of streptomycin following the removal from the UV-B treatment (Table 1) indicating that factors other than D1 synthesis are important in the recovery of all of the species from UV-B perturbations.
Table 1. The effect of inhibition of synthesis of D1 protein on the changes in Fv/Fm occurring during exposure of the four diatom species to 0.262 W m−2 of UV-B radiation for 1 or 2 h and after a 6 h recovery period after the removal of the UV-B treatment
Species ± streptomycin
Time in UV-B treatment
Recovery for 6 h
The D1 protein synthesis was inhibited by the addition of streptomycin to the cultures. The data are the means of five replicates ± SE.
Effect of cumulative UV-B dose on Fv/Fm in C. closterium, N. perminuta, A. coffeaeformis and N. epithemioides
The effect of cumulative UV-B dose on PSII photochemistry was studied to determine whether the effects of UV-B on the four species were dependent upon dose rate or cumulative dose (Fig. 7). In A. coffeaeformis, C. closterium and N. perminuta a short exposure to high-dose rate UV-B radiation caused a much greater depression of Fv/Fm than a longer exposure to the same cumulative dose of UV-B radiation given at a lower-dose rate. In contrast to the other species, the depression of Fv/Fm in N. epithemioides was dependent on cumulative UV-B dose, rather than on the dosage rate; there was a consistent decline in Fv/Fm with a cumulative exposure to UV-B radiation for all three UV-B dose rates (Fig. 7d). The least squares linear regression (r2 = 0.83) showed that for every additional 1 kJ m−2 of UV-B received, Fv/Fm decreased by 2.66 ± 0.38%.
For C. closterium, N. perminuta and A. coffeaeformis, the increased cumulative UV-B dose was not accompanied by the reciprocal reduction in Fv/Fm (Fig. 7a–c).
Effects of UV-B radiation on primary productivity and carbon allocation
The effects of two 6 h exposures of 0.35 W m−2 UV-B, given on consecutive days, on the primary productivity and assimilation of carbon into colloidal-S, EPS and glucan pools were made in parallel with the measurements of PSII operating efficiency for C. closterium and the mixed-species assemblages (shown in Figs 4 & 5, respectively). At a PPFD of 150 µmol m−2 s−1, UV-B caused a reduction in total primary productivity in C. closterium following both the first (35%) and second days (55%) of UV-B exposures (F1,16 = 6.87, P = 0.001) (Fig. 8a). No significant reductions in carbon allocation into colloidal-S or EPS were found after the first UV-B exposure, but reduced allocation to EPS was observed after the second UV-B exposure (F1,16 = 7.375, P = 0.015). Significant reductions in carbon allocation into glucan pools occurred on both days of the experiment (F1,16 = 22.63, P ≤ 0.001) (Fig. 8b–d). At the higher PPFD of 300 µmol m−2 s−1, there was no significant decreases in productivity in C. closterium after the first exposure to UV-B, but after the second UV-B exposure the total primary productivity was reduced by 40% (F1,16 = 16.05, P ≤ 0.001) (Fig. 8a). However, no significant differences in the allocation of carbon to either colloidal-S, EPS or glucan pools between the control and the UV-B-irradiated C. closterium were observed on either day (Fig. 8b–d).
UV-B had no significant effect on the primary productivity of the mixed assemblages after 1 d of exposure at a PPFD of 150 µmol m−2 s−1; however, after the second UV-B exposure, there was a 40% depression in primary productivity (F1,16 = 8.34; P = 0.011; Tukey test, P ≤ 0.05) (Fig. 9a). The allocation of assimilated carbon into colloidal-S compounds was significantly lower in UV-B-irradiated samples on day 1 (F1,16 = 6.84; P = 0.019; Tukey test, ≤ 0.05), but not on day 2 (Fig. 9b). UV-B radiation caused a significant reduction in the allocation of assimilated carbon into EPS on both days (F1,16 = 17.78, P ≤ 0.001) (Fig. 9c). Exposure to UV-B had no significant effect on the allocation of assimilated carbon into glucan on either day (Fig. 9d). When kept under a PPFD of 300 µmol m−2 s−1, an exposure to UV-B radiation did not cause any significant change in primary productivity or the allocation of assimilated carbon to colloidal-S on either day (Fig. 9a & b). However, UV-B radiation did cause a reduction in carbon allocation to EPS during day 2 (F1,16 = 6.91; P = 0.018; Tukey test, P = 0.021) (Fig. 9c), and carbon allocation to glucan was significantly higher in the UV-B-irradiated samples on both days (F1,16 = 5.65, P ≤ 0.03) (Fig. 9d).
The exposure of epipelic diatoms to UV-B radiation can result in decreases in their maximum PSII photochemical efficiency, photosynthetic electron transport, photosynthetic productivity and the allocation of assimilated carbon into colloidal-S, EPS and glucan pools. However, the effects can be complex, being dependent not only on the exposure dose of UV-B, but also on the PPFD to which the cells are exposed during the UV-B treatment as well as the diatom species. Such dependencies of UV-B effects on the PPFD during exposure to UV-B have been previously reported in cyanobacteria (MacDonald et al. 2003), planktonic diatoms (Grzymski, Orrico & Schofield 2001) and higher plants (Warner & Caldwell 1983; Mirecki & Teramura 1984; Cen & Bornman 1990).
The decreases in Fv/Fm induced by UV-B exposure in C. closterium (Fig. 2a) are associated with both an increase in Fo and a decrease in Fm (Fig. 3a & b), although this was not the case in the mixed assemblages where the decreases in Fv/Fm were primarily associated with a decrease in Fm (Fig. 3c & d). The decreases in Fm are presumably because of the increased dissipation of excitation energy in the PSII antennae by non-photochemical quenching processes (Krause & Jahns 2004). The increases in Fo are indicative of a photo-inactivation, damage or loss of PSII reaction centres, which are consistent with the increased reductions in Fv/Fm observed in C. closterium when synthesis of the D1 protein is inhibited by streptomycin (Table 1).
In both C. closterium and the mixed assemblages, UV-B exposure produced significant decreases in the rate of the linear electron transport. At low PPFDs where the cells are operating at a maximum quantum yield of photosynthesis, decreases in the maximum quantum yield of PSII photochemistry (Fv/Fm) would account for the decreases in Fq′/Fm′. However, at PPFDs above c. 150 µmol m−2 s−1, the UV-B-induced decreases in electron transport are unlikely to be accounted for entirely by the decreased maximum quantum yield of PSII photochemistry because PSII will be operating at much lower efficiencies than the maximum because of the development of light-dependent non-photochemical quenching processes. In leaves (Allen et al. 1997) and marine macrophytic algae (Bischof, Hanelt & Wiencke 2000), such UV-B-induced decreases in electron transport have been shown to be associated with the loss of ribulose 1.5-bisphosphate carboxylase/oxygenase (Rubisco) content and activity. This could be the case at times in C. closterium and the mixed assemblages where the decreases in CO2 assimilation (Figs 8a & 9a) sometimes accompany the decreases in electron transport. However, there are situations where the decreases in electron transport occur but there are no significant changes in CO2 assimilation, suggesting that damage to PSII is the primary cause of the decrease in electron transport.
It is well established that PSII can be damaged by UV-B radiation (Iwanzik et al. 1983; Renger et al. 1989; Kulandaivelu, Nedunchezhian & Annamalainathan 1991; Melis, Nemson & Harrison 1992). However, it is perhaps surprising that there were such pronounced species-specific responses. The species-specific responses in other algal species are often correlated with cell size because of the effects of size on internal molecular self-shading and on the efficiency of UV-absorbing compounds (Garcia-Pichel 1994; Laurion & Vincent 1998). However, UV-absorbing compounds, such as mycosporine-like amino acids, are not found in diatoms at concentrations high enough to afford protection (Jeffrey et al. 1999; Wulff 1999). It seems likely, therefore, that differential species responses occur either as a result of an efficient repair of UV damage, or from an alternative method of energy dissipation. Navicula perminuta displayed the greatest ability to adapt to and recover from UV-B-induced depression of Fv/Fm, in spite of being the smallest diatom species used in the study. By contrast, the largest species, N. epithemioides displayed the least ability to recover. All species showed some recovery from the depressions in Fv/Fm induced by a 2 h exposure to UV-B in the presence of streptomycin for more than 6 h after removing the UV-B (Table 1), which indicates that a rapid relaxation of the UV-B-induced non-photochemical quenching processes occurs on the removal of the stress. However, Fv/Fm was still significantly lower after 6 h without UV-B in all streptomycin-treated species apart from C. closterium. This small, but significant lack of recovery indicates that, in some diatom species, the repair and replacement of damaged proteins form part of the UV-B-induced defence mechanism. However, protein repair does not appear to be as important a strategy for epipelic diatoms as it is in other algal groups (Heraud & Beardall 2000; Xiong 2001).
All four species demonstrated an ability for recovery of Fv/Fm when the UV-B radiation was removed (Table 1). As would be expected in the absence of repair processes, Fv/Fm decreased in an almost linear fashion in N. epithemioides with increasing cumulative UV-B doses (Fig. 7d). Surprisingly, this was not the case for A. coffeaeformis and N. perminuta where after an initial depression in Fv/Fm, large increases occurred with increasing cumulative UV-B doses (Fig. 7a–c), suggesting that these species have the ability to rapidly recover from more severe UV-B perturbations. The ability to rapidly recover from UV-B perturbation during the exposure period has important implications for the ability of the species to tolerate UV-B radiation and remain highly productive during periods of exposure to it. Such large differences in the response of different species to UV-B may account for the much smaller decreases in Fv/Fm and Fq′/Fm′ observed when mixed assemblages, compared to C. closterium, were exposed to UV-B (Figs 2, 4 & 5). Presumably, many of the species in the mixed assemblages have a good ability to recover from UV-B-induced perturbation during the exposure period, which is not the case for C. closterium.
An exposure to UV-B induced changes in the allocation of assimilated carbon into colloidal-S, EPS and glucan pools (Figs 8b,c & 9b,c). In some treatments, the changes in the carbon-allocation pattern occurred when there was no significant change in CO2 assimilation, suggesting that these effects on carbon allocation are independent of the effects on photosynthesis.
Diatoms show great flexibility in the patterns of allocation of carbon to intra- and extracellular carbohydrate pools under changing illumination and nutrient conditions (Underwood & Paterson 2003). A general pattern found in the experiments was in the increase in the percentage of total carbon allocated to glucan. Similar increases in storage compounds, in response to UV-B, have been seen in higher plants (He, Huang & Whitecross 1994), sub-tidal epipelic diatoms (Underwood et al. 1999; Wulff 1999) and in the green alga Chlorella (Malanga & Puntarulo 1995). The accumulation of intracellular glucan in the absence of increases in the overall productivity may be because of the reduced cell division frequency or the inhibition of glucanases (Underwood et al. 2004). Both cell division and enzymatic processes may be affected by UV-B radiation (Davidson et al. 1994; Döhler 1996; Peletier et al. 1996; Bischof et al. 2000), although the effect of UV-B on diatom glucanases specifically remains unresolved.
The exposure of diatoms to elevated UV-B levels can inhibit photosynthetic electron transport and carbon assimilation, and also modify the pattern of allocation of assimilated carbon to colloidal-S, EPS and glucan pools. Clearly, this has potential implications for the photosynthetic productivity of natural microphytobenthic mats exposed to elevated UV-B levels. However, the large variation in the effects of UV-B in different species, the differing ability of species to recover from UV-B perturbations and the effect of PPFD on the response to UV-B, coupled with a potential protective migratory response (Kromkamp et al. 1998; Underwood et al. 1999), make it extremely difficult to predict the effects on natural mats. Clearly, to determine the effects of UV-B on natural microphytobenthic mats would require in situ measurements of the photosynthetic activities and productivity of the mats. Recent studies using Chl fluorescence imaging have shown that in the natural biofilms of mixed species, the overall biofilm functioning is a consequence of varying photosynthesis–irradiance characteristics and vertical migration behaviours of the different populations of diatom species (Oxborough et al. 2000; Underwood et al. 2005). This strongly suggests that, while UV-B may in the short term appear to have little appreciable effect on the net primary production of intertidal biofilms, over longer timescale selection for UV-B tolerant species could alter the species composition of sediment communities (Wulff et al. 2000) and result in changed biofilm functioning (Underwood 2005).
J.W. was the recipient of research studentship from the UK Natural Environment Research Council. We are grateful to Mark Barnett for assistance with the fieldwork.