The avoidance of surface waters by crustacean zooplankton is a common phenomenon both in marine and freshwaters. The contemporary paradigm interprets such behavior as an antipredator strategy. However, the phenomenon has also been reported in predator-free environments, which suggests that other variables may contribute to this depth-selection behavior. We investigated the possibility that ultraviolet radiation (UVR) could influence the avoidance of surface waters in three interconnected lakes differing in their transparency. Each lake was sampled 17 or 18 times. In the two most transparent lakes, the percentage of the population occurring in the surface layer decreased during spring and summer, but this pattern was not observed in the less transparent lake. Such differences in the patterns of surface-water avoidance were unrelated to the abundance and vertical distribution of the dominant planktivores (early stages of galaxiid fishes) as well as to temperature and food availability. They were inversely related, however, to the average UVR levels within the surface layer. A UV avoidance strategy was predicted for the copepod Boeckella gracilipes, on account of its low UV tolerance and lack of photoprotective compounds. Our results support this prediction and strongly suggest that surface avoidance by freshwater Patagonian zooplankton is not just a byproduct benefit of an antipredator behavior but is a direct response to high UVR levels.

The process of ozone depletion over the Antarctic continent has resulted in the increase of incident ultraviolet-B (UVB) radiation, whose effects may be damaging to living organisms. To counteract the negative effects of ultraviolet radiation (UVR), aquatic organisms may display one or more strategies: (1) avoidance (i.e. deep distribution); (2) photoprotection through the use of ''sunscreen'' compounds, such as mycosporine-like amino acids (MAAs); and (3) enzymatic repair of the damage. The effects of UVR were assessed on four populations of the copepod Boeckella poppei from Antarctic lakes using laboratory and field experiments. The results were related to measurements of DNA enzymatic repair activity and MAA concentration. This is the first study that combines these measurements in zooplankton. Boeckella poppei was highly tolerant to UVR (LD 50 ϭ 2.2-2.78 J cm Ϫ2 ). However, measurements of photorecovery (comparison of UVB mortality in the presence and absence of photoreactivating light) and dosage of photolyase activity indicated low rates of enzymatic repair, which may be the result of low temperatures typical of Antarctic lakes. Three different MAAs were identified, both in phytoplankton and copepods: porphyra-334, mycosporine-glycine, and shinorine. The population of B. poppei from Lake Boeckella had the lowest MAA concentration, as well as the lowest tolerance to artificial and natural UVR. These findings support the idea that UV tolerance in this species is related to the accumulation of MAAs. A comparison of the strategies used to cope with potentially damaging levels of UVR by different species of Boeckella indicates a high degree of plasticity in this genus, which has probably been key for its success to colonize a wide range of UV environments.

Summary Ultraviolet radiation (UVR) damages early life stages of several fish species. Galaxias maculatus is a small catadromous fish, with landlocked forms occurring in many lakes within the Nahuel Huapi National Park (Patagonia, Argentina). In this work, the vulnerability of G. maculatus eggs exposed to both natural and artificial UVR was investigated in relation to water transparency. Field experiments were performed in two lakes differing in UVR attenuation. Galaxias maculatus eggs were exposed to in situ levels of UVR in quartz tubes incubated at various depths. For laboratory experiments, the eggs were exposed to five levels of artificial UVB radiation. Exposure to natural UVR causes various degrees of egg mortality depending on water transparency and incubation depth. In the less transparent lake (Kd320 = 3.08 m‐1), almost complete mortality was observed near the surface. At a depth of 43 cm the observed mortality was only 22%, but was still significantly different from the dark control. In the most transparent lake (Kd320 = 0.438 m‐1), almost total mortality was observed in tubes incubated at 2.56 m or shallower. A gradual decline in mortality was recorded from that depth to 3.78 m where the values approached those in the dark control treatments. A monotonic relationship between mortality and UV exposure could be observed both in field and laboratory experiments. Using the results from field incubations, a LD50 of 2.5 J cm‐2 nm‐1 was estimated. In a few mountain lakes, this value would be exceeded even if the eggs were laid at the maximum depth of the lake. Thus UVR seems a sufficient cause to explain the absence of G. maculatus populations in some mountain lakes. For most lakes, however, UVR is probably one of several important environmental factors, which together determine the habitat suitability.