Direct Effects of UV-B Radiation on the Freshwater Heterotrophic Nanoflagellate Paraphysomonas sp.

Abstract

The formation of DNA photoproducts in organisms exposed to ambient levels of UV-B radiation can lead to death and/or reduced population growth in aquatic systems. Dependence on photoenzymatic repair to reverse DNA damage caused by UV-B radiation is demonstrated for Paraphysomonas sp., a member of a widely distributed genus of heterotrophic nanoflagellates. At 20°C, Paraphysomonas sp. was exposed to a range of UV-B intensities encountered in natural systems. Populations of the flagellate survived and grew in a dose-dependent manner, but only when simultaneously exposed to photorepair radiation (PRR). In contrast, flagellates exposed to UV-B at 15°C suffered 100% mortality except at the lowest UV-B level (with PRR) tested, which suggested a photorepair temperature optimum above 15°C. After acute UV-B exposures, DNA damage (measured as the formation of pyrimidine dimers) was reduced only in organisms that underwent subsequent exposure to PRR. Populations kept in the dark after UV-B exposure maintained the initial levels of pyrimidine dimers. These results are the first to demonstrate the reliance of a heterotrophic flagellate on photoenzymatic DNA repair for survival from UV-B exposure.

The importance of heterotrophic protists in aquatic ecosystem processes is well established (6, 25, 29). Protists, and especially heterotrophic nanoflagellates, are major predators of bacteria and prey of zooplankton and contribute to the recycling of essential nutrients. Consequently, factors that negatively affect biomass and alter the structure of protistan communities will influence both higher and lower trophic levels. Heterotrophic nanoflagellates in the genus Paraphysomonas are distributed worldwide in freshwater and marine systems, and Paraphysomonas vestita is the most commonly isolated species due to the range of acceptable conditions under which it grows (11). Therefore, responses of Paraphysomonas species to features of global climate change, including UV-B flux, reflect potentially important effects on aquatic microbial food webs.

UV-B radiation (UV-B; 280 to 320 nm) negatively affects many aquatic organisms primarily through damage to DNA. Cellular DNA strongly absorbs UV-B radiation, and even a relatively small exposure can result in the formation of different types of lesions between adjacent pyrimidine bases. Cyclobutane pyrimidine dimers (CPDs) account for the majority of these photoproducts, but pyrimidine [6-4] pyrimidinone dimers ([6-4] photoproducts) are also produced in measurable amounts (19). Both types of damage inhibit DNA replication and transcription and, if not repaired, can lead to inaccurate transmission of genetic information or cell death (19, 30). Repair mechanisms that have evolved to correct UVB-induced DNA damage fall into two general categories: dark repair mechanisms, of which nucleotide excision repair (NER) is the most important for UV radiation (UVR) damage in eukaryotes, and photoenzymatic repair (PER) (16, 18, 19, 30, 38). NER is a light-independent repair mechanism that cleaves the damaged DNA strand before and after the lesion, resulting in the removal of an oligonucleotide containing the dimer. DNA polymerase then fills in the gap with new nucleotides using the opposite DNA strand as a template. In PER, a photolyase enzyme binds to the UV-generated lesion and uses the energy of light to directly cleave the bond formed between pyrimidine bases (i.e., the dimer) without replacing the nucleotide. Photolyase enzymes require absorption of photoreactivating radiation (PRR) in the visible (400- to 700-nm) and UV-A (320- to 400-nm) parts of the spectrum, with a wavelength for maximum PER efficiency determined to be in the 380- to 440-nm range for some bacteria, yeast, and Euglena (16).

Long-term effects of increasing UV-B radiation on aquatic ecosystems are not known. However, most aquatic organisms, including protists, are susceptible to UV-B radiation and changes in the in situ levels of UV-B are likely to alter the ecology and community structure of aquatic systems (32). Changes in species diversity driven by differential sensitivity to UVR could modify patterns of predation and competition in the microbial food web, ultimately altering carbon and nutrient cycling. Understanding the UVR tolerance and the range of responses of individual species to UVR exposure can aid in predicting the ecological impacts of changes in UVR. The results of the present study indicate that the sensitivity of the freshwater Paraphysomonas sp. (strain GflagA) to UV-B was dependent on UV intensity and temperature. In addition, Paraphysomonas sp. relied heavily on PER to counter otherwise lethal effects of UV-B.

MATERIALS AND METHODS

Culture origin and maintenance.

Cultures of the heterotrophic nanoflagellate Paraphysomonas sp. strain GflagA used in these experiments were established by using single-cell isolations from water samples collected in June 2001 from a depth of 0.5 m within the pelagic zone of Lake Giles. Located in the Pocono Plateau of northeastern Pennsylvania, Lake Giles is an oligotrophic lake with a low dissolved organic carbon concentration (1.16 mg of C liter⁻¹) and a high optical clarity (K_dPAR 0.13; K_d320 0.32) (20). The flagellates were identified to the genus Paraphysomonas by general morphology and comparison of a partial 18S small-subunit rRNA gene sequence (650 bp) to the GenBank nucleotide database with a sequence similarity of >99% to P. vestita when a variable A/T-rich insertion (5) was removed from analysis. The sequence information for the Paraphysomonas sp. GflagA strain is available through GenBank (accession no. FJ528656).

Cultures were acclimated to temperature in light- and temperature-regulated growth chambers with a 14:10-h light/dark cycle (cool white fluorescent bulbs) at 15 or 20°C for several weeks for the two sets of experiments. The culture medium used for general maintenance was 0.1% Cerophyll containing an assemblage of coisolated, but unidentified bacteria as food. Prior to experiments, Paraphysomonas sp. was transferred to filtered, autoclaved Lake Giles water with heat-killed bacteria (HKB) as a sole food source and grown at the experimental temperatures for at least 1 week with HKB added as necessary to maintain these acclimated populations in exponential growth phase until the start of an experiment.

UV phototron experimental design.

To assess the effects of UV-B radiation on survival and growth of Paraphysomonas sp., cultures were exposed to UV-B radiation over 12 h with or without simultaneous photorepair radiation (PRR) using a UV-lamp phototron (37). This apparatus consists of a horizontal rotating wheel (5 rpm) set into an opaque box that houses photorepair lamps consisting of two 40-W cool white fluorescent bulbs (PAR) and two 40W Q-Panel 340 fluorescent bulbs (UV-A). Damaging UV-B radiation was supplied from above the box by a Spectroline XX15B UV-B lamp (Spectronics) covered with cellulose acetate to remove UV-C wavelengths. Because the energy output of individual UV-B lamps is known to vary, we followed an approximately annual calibration schedule using a custom made spectral radiometer (see reference 37 for energy spectra of the PRR and a description of the radiometer). Holes arranged in two rows around the outer edge of the wheel accommodated custom-made quartz treatment dishes (55-mm diameter, 30-ml volume) with quartz lids. Depending on the treatment, one or both of the light sources could be blocked by the placement of an opaque disk either below (interrupting the PRR) or above the dish (interrupting UV-B). The opaque disks were not used for treatments with the combination of both UV-B and PRR. A black polyvinyl chloride collar surrounded each dish to eliminate the stray light between treatments. Stainless steel mesh screens placed within the collars on the lid of individual dishes were used to modify the UV-B exposure for cumulative UV-B doses of 23, 45, or 58 kJ m⁻² over 12 h. The 12-h maximum UVR exposure, determined from the lamp's spectral output, was approximately equal to ambient surface UVR levels at 40 to 44°N latitude on a sunny day near the summer solstice (15, 37). An additional exposure level of 6 kJ of UV-B m⁻² was tested separately at 20°C and required an additional mesh screen surrounding the UV-B lamp. Each treatment and control was run in triplicate in all experiments. Experiments were performed at 15 and 20°C, temperatures within the range that Paraphysomonas sp. would naturally experience in Lake Giles USA (41°23′N, 75°06′W). Prior to UV-B exposure, the cell density of temperature-acclimated Paraphysomonas sp. was reduced to a target abundance of 5,000 flagellates ml⁻¹ by dilution with <0.22-μm-filtered Lake Giles water. HKB, a food source unaffected by UV-B, were added to a final concentration of 10⁷ cells ml⁻¹, and the flagellates with HKB were then transferred to the individual quartz dishes. This assured that the initial food and flagellate concentrations were identical in all treatments. Samples taken immediately prior to and after exposure and every 24 h for 5 days thereafter were fixed and stained with acid Lugol's solution. Paraphysomonas sp. cells were counted at ×400 magnification by using a Zeiss Axiovert 10 microscope and Phycotech settling chambers. Maximum growth rates (μ [day⁻¹]) of Paraphysomonas sp. were determined from changes in abundance over the 24-h period during which flagellate numbers increased the most in a given treatment. Rates were quantified by using the following exponential growth equation: μ = (lnN₂ − lnN₁)/(t₂ − t₁), where N₂ and N₁ are flagellate abundances at times t₂ and t₁ and (t₂ − t₁) is the time interval between sampling (1 day). This overall design allowed examination of the potential for photo-dependent recovery with a reasonable experimental endpoint based on the well-documented growth response of protozoa in batch cultures where food becomes limiting as populations grow exponentially.

In order to investigate UV-B induced DNA damage and the recovery of Paraphysomonas sp., cells were exposed to UV-B lamps and then incubated in darkness or light. For this set of acute UV exposures, the UV-lamp phototron apparatus was modified; two additional UV-B lamps were mounted above the rotating wheel, and no PRR lamps were used. The Paraphysomonas sp. culture was diluted with <0.22-μm-filtered Lake Giles water to ∼9,000 cells ml⁻¹, and cells were exposed for 5 min for a total UV-B exposure of 13 kJ m⁻² without PRR. Immediately after exposure, samples were taken for DNA damage analyses and microscopic enumeration, HKB were added, and replicate dishes were transferred to PRR or placed in the dark for 48 h. Control dishes were never exposed to UV-B and were kept in the dark for the duration of the experiment. This experiment was conducted within a temperature-controlled environmental chamber at 20°C. Additional samples to determine DNA damage and cell abundance were taken after 1.5, 24, and 48 h. For DNA damage analysis, flagellate cells were collected on sterile 3-μm filters (see below). Filters were placed in sterile 1.5-ml microcentrifuge tubes and stored at −20°C until DNA was extracted. Cell abundance samples were fixed and enumerated as previously described.

DNA dosimeter exposure and determination of dimers.

DNA dosimeters, composed of a sterile solution of salmon sperm DNA within quartz tubing, were exposed to UV-B with or without PRR in the UV-lamp phototron for 12 h in the same apparatus used to test Paraphysomonas responses to UV (in quartz dishes immersed in water from Lake Giles). The DNA dosimeters give a metric of the maximum potential DNA damage from the UV-B exposures used in the lamp phototron experiments (3).

To determine UVR-induced damage in Paraphysomonas sp., DNA was extracted from cells concentrated on sterile polycarbonate filters (3.0-μm pore size; Poretics) under low-pressure filtration (∼100 mm Hg in a vacuum). DNA was extracted from the cells in 35 μl of QuickExtract DNA Extraction Solution 1.0 (Epicenter Biotechnologies). This was just enough solution to cover individual filters in microcentrifuge tubes, which were placed in a 65°C heat block for 30 min, followed by 98°C for 15 min with mixing after 7 min. Filters were removed, and the DNA samples were stored at −20°C until analyzed for DNA damage. For both filters and dosimeters, radioimmunoassay was used to measure DNA damage as CPDs per megabase of DNA (17).

Preparation and testing of HKB.

A batch culture of the bacterium Pasteurella sp. was grown to stationary phase in 0.1% yeast extract. Bacteria were then heat killed by submerging the culture-containing flask in a 70°C water bath for 1 h (27). Using aseptic technique, the cells were washed two times via centrifugation in sterile 250-ml bottles at 8,230 × g for 20 min and resuspended in sterile-distilled water. Resuspended cells were filtered through sterile Nuclepore filters (47 mm; 5-μm pore size) to remove clumps. Sterility of the HKB solution was verified by inoculation into 0.1% yeast extract and observation for signs of bacterial growth (none detected). HKB were stored at 4°C until used, typically within 1 week.

The effect of experimental levels of UV-B on the food quality of HKB was compared to that of HKB without UV-B exposure by using the UV lamp phototron. HKB were exposed in replicate quartz dishes to UV-B (with PRR) at 58 kJ m⁻² at 16°C according to our standardized method (above). Controls were kept in the dark on the lamp phototron wheel. After UV-B exposure of the HKB, Paraphysomonas samples were added to each UV-exposed dish and dark control dish. Subsamples were preserved with Lugol's immediately after the addition of flagellates and every 24 h until population growth ceased. Initial HKB concentration in both treatments was 5.1 × 10⁶ ml⁻¹. Flagellates were enumerated as previously described.

Statistical analysis.

The significance of UV-B intensity levels and temperature effects on growth rates in treatments where growth occurred (e.g., those receiving PRR) were tested by using two-way analysis of variance (ANOVA). Prior to analyses, data were (n + 0.5)^0.5 transformed. The effect of UVR exposure on food quality of HKB (as determined by flagellate growth rate) was tested with one-way ANOVA.

RESULTS

Survival and growth responses to UV-B.

At 20°C with PRR, all populations of Paraphysomonas sp. exposed simultaneously to UV-B and PRR survived and grew. All treatments at 20°C without PRR experienced complete mortality within 24 h when exposed to greater than 6 kJ of UV-B m⁻². Maximum abundances of Paraphysomonas sp. in the dark control and PRR-only control groups were reached by 24 h (Fig. 1). Populations of flagellates exposed to 23 kJ of UV-B m⁻² in the presence of PRR had an initial lag in growth relative to controls but reached a comparable maximum abundance by 36 h after exposure. At greater UV-B exposure levels increasingly longer lag periods were observed, with flagellates exposed to the highest UV-B level having the longest interval of reduced growth postexposure (Fig. 1). In the experiment in which cells were exposed to 6 kJ of UV-B m⁻², maximum abundances comparable to dark control were reached by 12 h postexposure in the presence of PRR; cells exposed to 6 kJ of UV-B m⁻² in the absence of PRR survived and maintained a constant population with minimal growth (Fig. 2).

FIG. 1.

Abundance changes (mean ± the standard error [SE], n = 3) for Paraphysomonas sp. after a 12-h exposure to three environmentally relevant levels of UV-B (23, 45, and 58 kJ m⁻²) with (+) or without (−) simultaneous PRR at 20°C. Dark controls were not exposed to any UVR; PRR controls were exposed only to the PRR lamps.

FIG. 2.

Abundance changes (mean ± the SE, n = 3) for Paraphysomonas sp. after a 12-h low-intensity (6 kJ m⁻²) UV-B lamp exposure with (+) or without (−) simultaneous PRR at 20°C.

The highest maximum specific growth rates for Paraphysomonas sp. at 20°C were reached by cells grown in the dark or exposed to only PRR lamps; μ_max = 2.6 and 2.5 day⁻¹, respectively (Table 1). The effect of UV-B intensity on maximum specific growth rates was significant (P < 0.001), and growth rates were reduced in a dose-dependent manner in cells exposed to UV-B with simultaneous PRR. Exposure to the highest UV-B level with PRR reduced maximum specific growth rate to less than half that of controls (μ_max = 1.1 day⁻¹). In the absence of PRR, there was no growth in the UV-B exposed population except at the lowest exposure (6 kJ m⁻² μ_max = 0.73 day⁻¹).

TABLE 1.

Specific growth rates of Paraphysomonas sp. at 20 and 15°C exposed to various levels of UV-B with or without simultaneous PRR

UV treatment (kJ m−2) with or without PRRa	Mean growth rate (day−1) ± SEb
	20°C	15°C
Dark	2.6 ± 0.41	2.0 ± 0.14
PRR only	2.5 ± 0.25	1.8 ± 0.11
6
−PRR	0.73 ± 0.34	ND
+PRR	2.2 ± 0.10	ND
23
−PRR	M	M
+PRR	2.0 ± 0.09	1.4 ± 0.10
45
−PRR	M	M
+PRR	1.7 ± 0.35	M
58
−PRR	M	M
+PRR	1.1 ± 0.14	M

^a−PRR, without PRR; +PRR, with PRR.

^bMean based on three determinations. M, complete mortality; ND, no data.

The ability of Paraphysomonas sp. to recover from exposure to UV-B radiation declined in flagellates adapted to 15°C relative to the same exposures at 20°C. Flagellates that were not exposed to simultaneous PRR at 15°C experienced 100% mortality by 24 h (Fig. 3). Populations exposed to 23 kJ of UV-B m⁻² and PRR recovered most quickly and began to grow after approximately 48 h. Unlike the populations in corresponding UV-B intensities at 20°C, flagellates in all other UV-B-plus-PRR treatments failed to rebound after exposure (Fig. 3). In the 15°C PRR-only control treatment, there was a time lag before population growth started that did not occur in the dark control. These two control groups had the greatest maximum specific growth rates at 15°C (dark μ_max = 1.8 day⁻¹ and PRR μ_max = 2.0 day⁻¹) (Table 1), which were only 75% of the corresponding rates at 20°C. For treatments that did not result in complete mortality, the effects of UV-B intensity and temperature were significant (P < 0.001), and ANOVA indicated a significant interaction between intensity and temperature.

FIG. 3.

Abundance changes (mean ± the SE, n = 3) for Paraphysomonas sp. after a 12-h exposure to UV-B (kJ m⁻²) lamps with (+) or without (−) simultaneous PRR at 15°C.

Dosimeter DNA damage.

Maximum potential DNA damage (as CPDs) from exposure levels used in the lamp phototron experiments were quantified using DNA dosimeters of salmon sperm DNA in quartz tubes exposed to the UV-B lamp with or without PRR lamps. The data demonstrate dose-dependent DNA damage formation, confirming that greater exposure in the UV-lamp phototron resulted in a greater number of CPDs in the absence of DNA repair (Fig. 4). The highest UV-B level tested (58 kJ m⁻²) with or without PRR resulted in 162.9 and 158.3 CPD Mb⁻¹ DNA, respectively. Exposure to the PRR lamps showed no significant increase in CPD formation. DNA dosimeters exposed to PRR lamps alone had 4.1 CPD Mb⁻¹, while those which were kept in the dark had 3.4 CPD Mb⁻¹ DNA. Dosimeters exposed to 1 UV-B lamp for 12 h (58 kJ m⁻²) accumulated 47 times more CPD Mb⁻¹ DNA compared to dosimeters kept in the dark (Fig. 4).

FIG. 4.

CPDs per megabase of DNA in dosimeters exposed in the lamp phototron to different levels of UV-B (kJ m⁻²) with (+) or without (−) simultaneous PRR for 12 h. Dosimeters consisted of isolated salmon sperm DNA with no repair enzymes present and represent the maximum potential accumulation of dimers when no photorepair or photoprotection is present.

DNA damage and repair in Paraphysomonas.

During the acute UV-B exposure (13 kJ m⁻² over a 5-min time course), DNA damage in Paraphysomonas cells accumulated to ∼55 CPD Mb⁻¹. Cells that were transferred to dark conditions immediately after UV-B exposure showed no significant change in CPD Mb⁻¹ DNA over the following 48 h (Table 2). However, a reduction in CPD Mb⁻¹ DNA was seen in cells that were exposed to PRR after UV-B exposure, with only 36% of initial damage remaining by 48 h (Table 2). Compared to the control treatment, cells exposed to PRR after UV-B exposure demonstrated a 53% increase in population size at 24 h, while the population size for cells put into the dark decreased by 20% (Table 2).

TABLE 2.

Postexposure DNA damage and relative change in population size for Paraphysomonas sp.^a

Treatment after UV-B exposure	Mean CPD/Mb DNA ± SE at:		% CPD remaining (48 h)	Mean change in population size relative to control (24 h) ± SE
	0 h	48 h
Moved to dark	57.5 ± 3.5	58.4 ± 7.2	100	−0.20 ± 0.10
Moved to PRR	53.6 ± 5.1	20.6 ± 0.4	38	0.53 ± 0.10

^aTreatments consisted of acute 5-min UV-B exposure (13 kJ m⁻² total), followed by incubation in the dark or in PRR. The CPD Mb⁻¹ values did not change over time in the controls, which remained in darkness for the duration of the experiment.

HKB as food.

The growth rate of Paraphysomonas sp. fed with HKB exposed to UVR was not significantly different from Paraphysomonas samples fed with HKB stored in the dark. The specific growth rates of flagellates fed with HKB exposed to UVR at the highest intensity and exposure time used in our experiments versus those fed HKB kept in the dark were 1.0 ± 0.06 day⁻¹ and 1.1 ± 0.02 day⁻¹, respectively. HKB abundances changed only in the presence of flagellates, irrespective of exposure to UV-B.

DISCUSSION

Role of UV-B intensity for Paraphysomonas.

A major goal of the experiments described here was to determine whether UVR could limit the distribution of a heterotrophic flagellate in an environment that received moderate to high ambient UVR exposures. There was a clear effect of UV-B intensity on survival and growth rate of Paraphysomonas sp., with higher UV-B intensity resulting in greater inhibition and/or damage. This is a general phenomenon seen with dosimeters, as well as with organisms (e.g., Fig. 1 and 4). The role of intensity is an important component of the response of an aquatic organism because of the differential absorption of solar energy by water and dissolved substances. Organisms, including Paraphysomonas sp., that move or are mixed into deeper water are less affected by UV-B than those that remain directly at the surface, especially since attenuation is less for PRR than for UV-B (i.e., PRR penetrates deeper into the water column).

For this reason, our laboratory-based experiments used a strain of Paraphysomonas isolated from Lake Giles, which was optically clear, with levels of UVR that were ecologically relevant for that lake. The maximum calculated UV-B fluence for the lamp phototron was similar to that determined for a full sunny day in June at Lake Giles, at latitude 41.2°N (37), and in Pisa, Italy, at latitude 43.7°N (15). The levels of photoproducts produced in DNA dosimeters in our experiments were consistent with UV-B exposure in temperate latitudes. Boelen et al. (3) used similar dosimeters in marine systems and determined a conversion factor of CPDs produced in DNA dosimeters to scalar irradiance (1 CPD Mb⁻¹ of nucleotides per 3.1 J m⁻²). Using that conversion factor, the biological effective dose of our highest intensity was 496 J m⁻². This agrees well with the biological effective doses of 592 and 456 J m⁻² calculated for the ocean surface at 34°N by Boelen et al. (3) and confirms the relevance of the UV-B exposures used in our laboratory experiments.

PER in UV-B-exposed Paraphysomonas.

Depending on water clarity, organisms living in the surface waters of aquatic systems can encounter injurious levels of UVR and consequently must acquire mechanisms to minimize the negative effects. Behavioral and biochemical responses to UVR exposure, which are not mutually exclusive, are known for a broad range of species and include avoidance, various enzymatic repair pathways (e.g., PER, NER [dark]), and the presence of photoprotective compounds.

Our experiments indicated that Paraphysomonas sp. was heavily dependent on PER for survival and growth after exposure to UV-B. There was 100% mortality in the absence of PRR at all except the lowest UV-B exposure tested (Fig. 1 to 3). The reduction of CPDs in DNA from flagellate cells exposed to PRR corroborates a large role of PRR for Paraphysomonas sp. exposed to UV-B. The decrease in CPDs was not observed in the absence of PRR (Table 2). Although the PER treatment did not exclude NER from occurring, the treatments moved to the dark (no PER) after UV-B exposure gave no indication of CPD repair (Table 2). This implies little role for NER in Paraphysomonas, and a dominant role of PER, which is consistent with observations of some other heterotrophic protists, including the ciliates Paramecium tetraurelia and Glaucoma sp. (26, 31). P. tetraurelia had greater clonal longevity if it received PRR when exposed to UV-B, while populations of Glaucoma sp. accumulated lethal amounts of DNA damage in the absence of PRR but had fewer CPDs and grew in the presence of PRR (26, 31). However, PER is not ubiquitous in heterotrophic protists and can vary even within the same genus. One strain of the ciliate Cyclidium was found to have very little tolerance of UV-B, whether PRR was present during UV-B exposure or not (26), while another clone did not accumulate any measurable DNA damage when exposed to UV-B with PRR (33).

NER is a metabolically costly and relatively slow DNA repair mechanism compared to PER (30, 35), and it may be that the rapid population growth of protists selects for PER over NER. Alternatively, it is possible that PER quickly removes most dimers and NER then works over a longer time frame to remove the remaining damage. If so, our experimental design would have underestimated this potential role for NER because Paraphysomonas sp. reached peak abundances rapidly. The mechanism of a slower response by NER would partly explain the longer lag periods at higher UV-B exposure, but the lack of any dark repair of CPDs (Table 2) argues against this interpretation. The lags may instead result from a higher proportion of incompletely repaired cells at higher UV-B doses that either do not survive cell division or are unable to divide due to cell cycle arrest. Cell inactivation such as this is known for multicellular eukaryotes when CPDs are not repaired (23).

Given the rapid and total mortality observed in the absence of PRR at several dosages, it seems most likely that Paraphysomonas sp. is reliant on PER for sustained survival in UV-B-exposed populations. Although PER is widespread taxonomically (16), its importance relative to dark repair is highly variable for different organisms. PER has been shown to be a particularly important mechanism of DNA repair in several species of amphibians (2), zooplankton (7, 12, 14), protists (26, 31), and bacteria (22). Zenoff et al. (39) found several strains of aquatic bacteria with various degrees of reliance on PER and no obvious use of NER, although Cytophaga sp. had efficient CPD removal in both the dark and under PRR.

Temperature and food quality effects on the response to UVR.

Temperature has strong effects on feeding, respiration, and growth in heterotrophic protists (24), which is demonstrated here by the difference in the Paraphysomonas growth rate in the dark at 20 and 15°C (Table 2). Populations of Paraphysomonas sp. maintained at the two temperatures also showed differential abilities of recovery after identical UV-B exposures, with increased survival and growth at 20°C relative to 15°C (Fig. 1 and 3). An overall reduced metabolism at the cooler temperature could have contributed to these differences; however, it seems likely that the photolyase enzyme itself was less efficient at 15°C. The interaction of temperature with DNA repair is known for several other taxa exposed to UV-B, including a red alga, a ciliate, a rotifer, and several crustacean zooplankton species (7, 13, 21, 26, 36).

UVR also can reduce bacterial growth, so we adapted HKB as a food source to separate effects due to UV-induced changes in food from the direct consequences of UVR exposure to Paraphysomonas. The reduced population growth of Paraphysomonas sp. in the presence of UV-B in our experiments can be ascribed to direct effects on the flagellates since the quality of HKB was not affected by UV-B exposure. HKB irradiated at our highest level of UV-B prior to being used as food led to growth rates by Paraphysomonas sp. that were statistically indistinguishable from those fed nonirradiated HKB. Likewise, growth rates of Paraphysomonas sp. on live bacteria were 1.75 day⁻¹ at 15°C and 2.45 day⁻¹ at 20°C (A. L. Macaluso, unpublished data), which were similar to growth rates in the dark with HKB (Table 1). It is also unlikely that the responses of Paraphysomonas sp. at different temperatures was due to a drop in food quality over the experimental time course. The addition of HKB, whether kept at room temperature or refrigerated, both result in exponential growth of protists (R. W. Sanders, unpublished data). These data indicate that food quality was not the ultimate cause of the increased lag periods and reduced growth rates of Paraphysomonas observed at increased UV-B exposure.

In all treatments and controls where there was growth of Paraphysomonas sp., a notable decline in flagellate abundance occurred within 24 h of reaching peak abundance. This is typical of batch culture experiments and also in field experiments when predators of protists are removed, because heterotrophic protists grow rapidly and graze their prey to a threshold level that can no longer support the flagellate population at high abundance (4, 10, 28). Heterotrophic protist cultures can be resurrected from UV-B-exposed treatments (if they received PRR) by adding more food (Sanders, unpublished). This observation suggests that PER is highly effective in these organisms, and that, unless UVR causes 100% of a population to die, even a single individual could establish a new protist population via asexual reproduction.

Implications for aquatic systems.

Quantifying the limits of tolerance to UVR for individual species and the variation of sensitivity to UV-B at different temperatures are important aspects of elucidating ecological impacts of sunlight in aquatic environments. Solar radiation, including UV-B, increases rapidly in early spring when water temperature may remain below the optima for DNA repair. Thus, there is potential for aquatic communities to be shaped by species tolerance to seasonal changes that include shifts in the ratio of UV-B exposure to temperature. Temperature alone is considered to have a role in altering aquatic community composition (34), but the effect of increasing UVR levels during a period of relatively slow seasonal change in temperature is not known. Trends of increasing global temperature may alter community structure by shifting local temperature out of organisms’ tolerance ranges, thus rendering them more susceptible to UVR-induced DNA damage.

Understanding how UVR affects heterotrophic protists may be of special interest since they are recognized as having substantial effects on major ecosystem processes including nutrient recycling and transfer of carbon from bacteria and picophytoplankton through the food web (1, 9, 25). However, a direct comparison of the cumulative UVR dose from our experiments to field conditions may lead to inaccurate predictions of population change. This is because models predicting such effects typically assume reciprocity, i.e., that a cumulative dose has the same effect regardless of the dose rate. In other words, the principal of reciprocity indicates that the same DNA damage determined here for Paraphysomonas sp. exposed to 56 kJ m⁻² over a 12-h period would occur if it had received 56 kJ m⁻² over a different time period. When PER occurs, as in Paraphysomonas sp., reciprocity typically does not hold and applying the laboratory results to lakes with dissimilar radiation regimes could be misleading (8, 12). Nevertheless, these experiments demonstrate that Paraphysomonas sp., belonging to the most commonly isolated genus of heterotrophic flagellates (11), can survive and reproduce at the highest levels of UVR that it is currently likely to encounter in a temperate environment. Although higher fluences of UV-B resulting from reduced atmospheric ozone might reduce the rate of population growth, we expect indirect effects on its predators and prey are more likely to alter population growth of Paraphysomonas than direct effects of UVR.