Effects of UVB radiation on grazing of two cladocerans from high-altitude Andean lakes

Climate change and water extraction may result in increased exposition of the biota to ultraviolet-B radiation (UVB) in high-altitude Andean lakes. Although exposition to lethal doses in these lakes is unlikely, sub-lethal UVB doses may have strong impacts in key compartments such as zooplankton. Here, we aimed at determining the effect of sub-lethal UVB doses on filtration rates of two cladoceran species (Daphnia pulicaria and Ceriodaphnia dubia). We firstly estimated the Incipient Limiting Concentration (ILC) and the Gut Passage Time (GPT) for both species. Thereafter we exposed clones of each species to four increasing UVB doses (treatments): i) DUV-0 (Control), ii) DUV-1 (0.02 MJ m2), iii) DUV-2 (0.03 MJ m2) and iv) DUV-3 (0.15 MJ m2); and estimated their filtration rates using fluorescent micro-spheres. Our results suggest that increasing sub-lethal doses of UVB radiation may strongly disturb the structure and functioning of high-altitude Andean lakes. Filtration rates of D. pulicaria were not affected by the lowest dose applied (DUV-1), but decreased by 50% in treatments DUV-2 and DUV-3. Filtration rates for C. dubia were reduced by more than 80% in treatments DUV-1 and DUV-2 and 100% of mortality occurred at the highest UVB dose applied (DUV-3).


Introduction
Depletion of the ozone layer and consequent increases in ultraviolet radiation (UVR) triggered the interest on the harmful effects of UVR on organisms, such as denaturation of proteins [1] and DNA damages [2][3][4]. Although the ozone layer depletion has stabilized, the effects on UVR will remain even by the end of this century [5,6]. High-altitude Andean lakes ecosystems are particularly vulnerable, because of the extremely high levels of UVR occurring in this region [7], and the increased UVR penetration predicted for these environments due to indirect effects of climate change [7,8].
The Andean region is characterized by the presence of numerous mountain lakes, mostly of glacial origin [9]. Underground connections and soil temperatures of the Central Andes (below 15˚C but above freezing point) allow the flow of mobile forms of carbon, from the carbon-rich soils (239 to 466 TmÁha -1 ; [10]) towards the lakes [11]. According to climate change scenarios, high-altitude Andean mountain ecosystems will experience warmer and drier a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 research within the Bolivian territory. Sampling activities were approved by administrative resolution VMABCC-026/09.

Cladocerans origin and culture conditions
Zooplankton samples were collected in Totora-khocha Lake, located in The Central Andes in Bolivia (65.63˚W-17.46˚S) at 3730 m above sea level, This is one of thousands of lakes formed during the last retreat of glaciers during the Pleistocene [9]. Air temperatures vary between 3.3 and 23.6˚C during summer, and between -3.0 and 14.9˚C during winter. Water temperature varies between 10.2 and 14.6˚C. Mean and maximum depths of the lake are approximately 2.8 and 6.6 m, respectively. Secchi depth is approximately 1.6 m.
To collect the zooplankters, we took vertical and horizontal samples using a 100 μm mesh size plankton net. Captured individuals were gathered and transported alive to the laboratory in 2 L plastic bottles filled with lake water.
Approximately 100 individuals of D. pulicaria and C. dubia were isolated and acclimatized for 50-60 days to laboratory conditions (room temperature 20 ± 2˚C, 13:11 light-dark cycles). Both species were fed with cultures of Chlorella vulgaris, supplying 25 μl of a 1.51Á10 7 ± 2.01Á10 3 cells mL -1 solution every day. After the adjustment period, adult females of each species were randomly selected to establish monoclonal cultures, from which we took the individuals for the experiments.
ILC and GPT are key parameters that depend on species body size and filtration capacities [51], thus, for both species we performed preliminary tests to determine concentration of microspheres, feeding times and range of food concentrations that allow properly counting each ingested microsphere and provide gross estimations of ILC and GPT. Based on those preliminary results, we designed more precise assays.

Gut passage time
GPT of D. pulicaria and C. dubia were determined by placing adult clones individually in 5 mL of water (10 experimental units per treatment) with fluorescent microspheres at concentrations of 5Á10 4 particles mL -1 for D. pulicaria and 5Á10 3 particles mL -1 for C. dubia. Feeding solution concentrations were selected based on the ILC experiment results. Eight feeding times (1, 2, 3, 5, 10, 20, 40 and 60 min) were tested for D. pulicaria, and four feeding times (1, 3, 5 and 7 min) were tested for C. dubia.
Based on the results of the ILC and GPT assays (see Results section), we selected food concentrations lower than the ILC to prevent inhibition of ingestion rates; and feeding times shorter than the GPT to prevent excretion of particles during the experiments. Therefore, to perform filtration rate experiments we selected food concentrations of 5Á10 4 cells mL -1 and 5Á10 3 cells mL -1 for D. pulicaria and C. dubia, respectively. We established feeding times of 10 minutes for D. pulicaria and 4 minutes for C. dubia.

UVB radiation and filtration rate experiments
To determine the effect of sub-lethal doses of UVB radiation on filtration rates, adult clones of each species were exposed to four treatments with increasing doses of artificial UVB radiation: i) DUV-0 (Control, no exposure to UVB), ii) DUV-1 (0.02 MJ m 2 ), iii) DUV-2 (0.03 MJ m 2 ) and iv) DUV-3 (0.15 MJ m 2 ). Considering that daily UVB doses at the surface in the Central Andes in Bolivia range from 22 to 57 MJm 2 [52] and attenuation coefficients (Kd) in the lakes vary from 4.29 to 11.73 [40], we estimated that UVB doses used in our experiments are expected to occur in nature between 0.6-1.7 m (DUV-1), 0.5-1.6 m (DUV-2) and 0.4-1.2 m (DUV-3).
Organisms of both species (14 experimental units per treatment) were placed individually in 5 cm depth transparent polycarbonate flasks with 10 mL of water. Then, they were irradiated with a 1.2 W lamp (PL-S 9W/01/2P Philips Electrical Ltd. Sywell, Northants, England). We employed a narrowband lamp (emission range of 305-315 nm and peak emission at 311 nm) to avoid short UVB and UVC wavelengths that UVB lamps usually emit, which are not found in sunlight and can be extremely damaging [53]. Samples were incubated at laboratory conditions. UVB exposure was applied once, in a 7 h interval within the light period. Exposure levels (doses) were applied varying the distance from the UVB source to the flasks. We performed filtration rate assays subsequently to one dark period (11h) after the UVB exposure. Organisms were placed individually in flasks with 5 mL of feeding solution. D. pulicaria were fed with a 5Á10 4 particles mL -1 solution (1Á10 3 particles mL -1 of fluorescent microspheres and 4.9Á10 4 cells mL -1 of C. vulgaris) for 10 minutes. C. dubia were fed with a 5Á10 3 particles mL -1 solution (3Á10 3 cells mL -1 fluorescent microspheres and 2Á10 3 cells of C. vulgaris mL -1 ) for 4 minutes.

Laboratory procedures
At the end of all experiments, animals were anaesthetized with carbonated water to prevent microsphere excretion, and then, fixed with formaldehyde (2% final concentration). Each individual was first rinsed with deionized water to exclude microspheres that may be exteriorly adhered, placed in a glass slide and measured from the top of the helmet to the base of the tail. For the GPT assay, lengths of gut and food occupancy were measured in all individuals. For the remaining experiments, the organisms were dissolved individually with sodium hypochlorite to release the spheres from the digestive tract. All ingested microspheres were counted using epifluorescence microscopy (Olympus CX-31 microscope 400X; CX-RFL-2 UV light source; blue light).

Data and statistical analysis
Filtration rate is defined as the volume of water separated from particulate matter by the filtering apparatus of one organism in a unit of time [54]. Ingestion and filtration rates were calculated for the ILC and UVB experiments by Eq (1): where FR is the filtration rate (mL ind -1 h -1 ), IR is the ingestion rate (calculated as the total number of food particles ingested in one hour) and M is the total amount of food particles in one milliliter of feeding solution.
After verifying assumptions of normality and homoscedasticity through diagnostic plots (Normal Q-Q and Residuals vs. Fitted, respectively) and Akaike Information Criteria (AIC), dependent variables of each experiment (relative food occupancy, ingestion or filtration rates) were compared among treatments by an Analysis of covariance (ANCOVA) and, when appropriate, orthogonal contrast post-hoc tests were applied. Size of each individual was used as covariate to correct for any effect of size on each dependent variable. All statistical analyses and figures were performed in R v3.2.2 [55].

Gut passage time
Our results showed that GPT for D. pulicaria feeding in C. vulgaris solution was between 40 and 60 minutes. After 40 minutes of feeding, all individuals of D. pulicaria showed more than 76% of gut filling. After 60 minutes, 40% of the individuals of D. pulicaria showed complete gut filling. C. dubia exhibited GPT's between 5 and 7 minutes. All individuals of C. dubia showed more than 77% of gut filling after 5 minutes of feeding, whereas 20% of the individuals showed complete filling after 7 minutes of feeding (Fig 2).

Discussion
In this study, we investigated experimentally the effect of increasing doses of UVB radiation on filtration rates of D. pulicaria and C. dubia. Incipient Limiting Concentration (ILC) and Gut Passage Time (GPT) were estimated previously as key parameters to determine optimum food concentrations (algal concentrations below the ILC) and feeding times (below the GPT). Compared to former studies, we found comparable ILC values for D. pulicaria and C. dubia. ILC values for cladocerans have been reported as low as 6Á10 3 cells (Oocystis sp.) mL -1 for Moina micrura and as high as 4.2Á10 5 cells (Chlorella sp.) mL -1 for D. ambigua [50]. Our estimation of the ILC for D. pulicaria (1Á10 5 cells (C. vulgaris) mL -1) falls well within this range. The small C. dubia (mean length 961 μm) showed lower ILC (5Á10 3 cells (C. vulgaris) mL -1 ) than the larger D. pulicaria (mean length 1468 μm). Low ILC values are expected for populations adapted to oligotrophic conditions [56]. The pattern of higher feeding efficiency of larger species compared to smaller ones has been reported earlier [57,58] and is mainly explained by the increasing energy demand with organism size [51,59].
GPT values for Daphnia species vary in a wide range depending on the species (grazer and prey), method of estimation, size of the animals and food concentration [60][61][62]. GPT values between 25 and 54 minutes have been reported for different species of the genus Daphnia [49]. According to Wiedner and Vareschi [63], GPT's are shorter at higher food concentrations, and feeding times in laboratory experiments with D. pulex should not exceed 2 minutes when using food concentrations of 10 5 cells mL -1 or higher. However, GPT estimations for D. pulicaria show strong variations, ranging between 4 and 106 minutes even for the same prey species and same conditions [62,64]. Hence, our GPT estimation for D. pulicaria (40-60 minutes) is within the range previously found. To our knowledge, GPT values for C. dubia have not previously been assessed. However, food evacuation rates tend to be lower in small cladocerans than in large ones [49]. GPT values for chydorids, similar in size to C. dubia (< 1 mm), range between 5 to 7 minutes [65,66]. These results match exactly with our GPT estimate for C. dubia.
It has been demonstrated that UV radiation affects cladocerans in several ways, such as reducing the number of offspring [24, 67,68], changing swimming behavior [32, 69,70], and causing visceral [71] and DNA damage [68,72]. Some metabolic and physiological effects of sub-lethal UV doses on cladoceran species have also been noted, such as decreased/increased respiration rates [73], reduced activities of digestive enzymes, and reduced residence time of algae in the gut [41]. Our study is the first to demonstrate a negative effect of sub-lethal UVB doses on the filtration rate of cladocerans. We found that filtration rates of D. pulicaria decreased to less than 50% in individuals exposed to naturally occurring doses of UVB radiation. According to data from 14 lakes in The Central Andes of Bolivia [40], doses in treatments DUV-2 and DUV-3 are expected to occur at maximum depths of 1.6 m. Hence, it is important to highlight that diel vertical migration may play a key role in avoiding UVB radiation [28, 74,75]. Our findings suggest that if D. pulicaria remains at depths shallower than 1.6 m, filtration rates could be strongly affected. While it was expected that higher doses result in stronger effects [76], we found no further reduction in filtering rates of D. pulicaria in treatment DUV-3 (0.15 MJ m 2 d -1 ; corresponding to 1.2 m depth) compared to treatment DUV-2 (0.03 MJ m 2 d -1 ; corresponding to 1.6 m depth). Filtration rates in treatment DUV-1 (0.02 MJ m 2 d -1 ; corresponding to 1.7 m depth) were not statistically different from the control treatment DUV-0. It is unlikely that animals get exposed to higher doses under natural conditions, which means that individuals that are able to avoid the first 1.6 m of the water column during the daytime, are likely to be unaffected by UVB radiation.
The effect of UVB radiation on C. dubia was even stronger; treatment DUV-3 (0.15 MJ m 2 d -1 ) caused 100% mortality, whereas filtration rates were reduced to less than 20% in treatments DUV-1 (0.02 MJ m 2 d -1 ) and DUV-2 (0.03 MJ m 2 d -1 ). High mortality of C. dubia was also previously observed in southern Chilean wetlands when UVB doses exceeded 0.07 MJ m 2 d -1 [77]. Our results, together with field observations describing C. dubia as an unpigmented species [78] that commonly avoids the upper surface waters in transparent lakes even in absence of predators [79,80], suggest that C. dubia must remain deep in the water column to prevent lethal effects. The depth necessary for C. dubia to prevent the effects of UVB radiation will depend on the attenuation coefficient, being around 1.7 m in the region from where the organisms were collected [40].
A huge variation in tolerance to UVB radiation has been reported for cladocerans. Some studies suggest that small cladocerans are more tolerant to UVB radiation than larger ones [80]. Extremely low tolerance was reported for the large Daphnia magna (lethal doses LD 50 0.013 MJ m 2 ; [27], in comparison with smaller species such as C. dubia (LD 50 = 0.035 MJ m 2 ) and Bosmina meridionalis (LD 50 = 0.04 MJ m 2 ; [81]. This pattern may occur due to the low energy cost that repair mechanisms represent for small-sized organisms [82,83]. A similar pattern was described between individuals of the same species for Daphnia galeata and D. pulex, probably due to the transfer of photoprotective pigments from the mother [84][85][86]. However, and in accordance with results of our study, the reverse pattern has also been found: UVB tolerance of the large Daphnia pulex (LD 50 = 0.06 MJ m 2 ; [87] was found to double that of the small species C. dubia (LD 50 = 0.035 MJ m 2 ; [81]. These contrasting results suggest that factors other than body size may also play a role in UVB tolerance for cladocerans. The evidence that photoprotective pigments provide tolerance to some species [15,36,87], argue for the need of further studies to explain tolerance to UVB radiation. There is a positive correlation between cladocerans size and filtration rates [54,57,58]. Large species are more efficient grazers, and their impact on the energy flux in aquatic ecosystems is disproportionately high compared to that of small species. Top-down control on phytoplankton is (co)determined by the proportional relationship between large and small zooplankton species [58]. Even if D. pulicaria was less affected by UVB than C. dubia, a 50% reduction of filtration rates of D. pulicaria may have stronger impact on the overall zooplankton community grazing rates than an 80% reduction in the filtration rates of C. dubia. Indeed, a reduction of zooplankton grazing pressure decreases the zooplankton/phytoplankton biomass ratio [88], leading in a break up in the energy flux.
Our results indicate that UVB radiation is a strong selective force in high-altitude Andean lakes, which excludes grazers from utilizing resources in surface waters during daytime. Climate change, through an indirect increase in UVB radiation, and water extraction, through a reduction in water depth, will strongly reduce zooplankton grazing and ultimately the energy flux in these ecosystems. This could lead to increase the risk of algal blooms, even if this risk is currently low due to nutrient limitation [89] and to consequent decrease in water quality for human uses.