In the previous section, the response to UVR of populations at one trophic level
(basal species) was considered. The interaction of UVR, however, with more than one trophic level adds substantial complexity to the possible responses, with the potential occurrence of positive and negative feedbacks (Figure 1). Both prey and predator populations might be affected by UVR, and, if so, the net effect will depend on the relative tolerance threshold of the interacting species. Yet, as soon as we consider more than one interaction between species, responses in the food web are expected to be much more complex than depicted in Figure 1, including potential changes in population size at different trophic levels. Thus, for example, in the hypothetical aquatic food web depicted in Figure 2, a potential reduction in population size of the UV-sensitive species 5 feeding on UV-sensitive basal species 7 and UV-tolerant species 8 and eaten by species 3 may increase the competitive advantage of species 8 and at the same time reduce the population size of species 3 but increase those of 2 and 1.
Both well et al.  found the first direct evidence of complex interactions in the food web during a colonization experiment with freshwater periphyton growing in artificial flumes of 1 cm depth located outdoor in British Columbia, Canada. They observed that short-term effects of UVR (mainly UV-A) caused inhibition of diatom growth and accrual rate (chlorophyll-a). However, after the third week, UVR reduced the number of algal grazers (chironomid larvae mainly Cricotopus bicinctus and Orthocladius sp.), and by the fourth week, the initial negative effect on algal biomass was reversed. This food web, however, was relatively simple with mainly one herbivore species. Other studies with periphyton in natural streams have failed to confirm the positive feedback described above. For example, in a 28-day study in Otter Creek, Nebraska, ambient levels of UVR did not affect algae or herbivores colonizing tiles submersed at a depth of 8-22 cm . The authors argued that lack of significant
differences in herbivore densities between treatments might have been due to the presence of long UV-A wavelengths in the "UV-excluded treatment", which have been suggested by Bothwell et al.  to elicit UV-avoidance responses of invertebrates. In three colonizing experiments done in the upper White Oak Creek, Tennessee, periphyton and herbivores, mainly the snail Elimia clavaefor-mis, were not significantly affected by ambient UVR levels [42, see also critics and discussions in 43 and 44]. In an experiment lasting 30 days in the Cache la Poudre River, Colorado, a negative effect of UVR on periphyton biomass accrual and abundance of invertebrates colonizing artificial substrates submersed at ~ 10 to 40 cm depth was only apparent at the end of the experiment , However, the authors concluded that it was unclear whether the effect on invertebrates was caused by UVR, by interaction with other invertebrates, or higher algal biomass in the UV-excluded treatment. They speculated that ending the experiment after 30 days and interrupting successional shifts in algal species might have avoided a positive feedback of UVR on algal biomass. McNamara and Hill  suggested that the different responses of periphyton observed in the studies mentioned above are related to the presence of more UV-resistant communities in streams at low elevations of mid-latitudes than at higher elevation or latitudes. This seems to be counterintuitive because UV-B fluxes increase with altitude (see Chapter 2). On the other hand, several authors have argued that one possible explanation for the dissimilar results obtained are the different exposure characteristics of organisms in artificial flumes and natural streams, particularly the shallow depth and exclusion of higher-level predators that may exacerbate the effects of UVR on periphyton and insect larvae [15,41,45],
A lack of indirect effects mediated by UVR has also been observed in a colonizing experiment in an alpine lake . While development of epilithon (mainly diatoms and cryptophytes) was suppressed by UV-A and UV-B radiation, zoobenthos like the sediment-dwelling Gammarus lacustris and chironomids with burrowing habits were not affected. There are only a few studies in coastal marine environments addressing this topic. During an experiment in the coast of Greece, biomass of colonizing benthic algae (mainly pennate diatoms) and species community structure were affected by ambient levels of UV-B, but invertebrate biomass was not .
Autecological studies with freshwater and marine heterotrophic nanoflagel-lates (HNF) have provided some evidence for a positive feedback between UVR and prey populations. Sommaruga et al.  reported that artificial and natural UVR (mainly UV-B but also UV-A) strongly reduced bacterivory rates of the freshwater HNF Bodo saltans. In laboratory experiments, they found that mortality rates (i.e., negative growth rates) of bacteria in the UV-B-excluded or dark treatments were higher than in the presence of UV-B. Furthermore, depending on predator density, even positive bacterial growth rates were observed in the presence of UV-B , Similar evidence was obtained by Ochs [47,48] in laboratory experiments with the marine HNF Paraphysomonas bandaiensis and P. imperforata grazing on two strains of the picocyanobacterium Synechococcus sp. Prey population size was always higher in those treatments where grazing by
HNF was more affected by UVR. On the other hand, studies with natural protist assemblages exposed to ambient and enhanced UV-B levels have provided mixed results. In a study with microbial communities from two arctic systems, although ambient UV-B levels negatively affected growth of some ciliates species, community-grazing rates were not . In a 16-day mesocosm experiment with a microbial food web (zooplankton excluded) from a UV-transparent alpine lake, negative effects of ambient UV-B radiation were observed on HNF growth and bacterivory rates. However, this did not result in higher bacterial abundance suggesting that bacteria were negatively affected as well . In an experiment with a microbial food web (organisms > 240 pm excluded) from the St. Lawrence Estuary, Canada, enhanced UV-B radiation reduced significantly the populations of large phytoplankton and ciliates after 7 days . The increase in prey abundance, mainly in HNF, was interpreted as a release from predation pressure by ciliates . Reductions in HNF bacterivory rates were not observed until the 7th day  when bacterial abundance increased . In other two mesocosm studies with estuarine communities (including zooplankton), no major effects to enhanced UV-B levels were observed except for phytoplankton in one of the studies, while positive feedbacks among different components including fish larvae were not found [55,56].
Another potential effect of UVR on the predator-prey interaction is when the prey population has a higher UV-sensitivity than the predator, potentially leading to a negative feedback (Figure 1 case 2 of predation/herbivory). Investigations on this possible scenario have been based on the observation that the UV-B-irradiated green alga Selenastrum capricornutum was ingested by Daphnia magna but digested with lower efficiency than those in the control without UV-B
. This effect was significant only after > 12 h irradiation with artificial UV-B radiation (max. at 312 nm). The authors hypothesized that changes in both mucous secretion and in thickness of the cell wall were responsible for the lower digestibility. In a later study, other species of phytoplankton (Chlamydomonas reinhardtii, Scenedesmus acutus, S. subspicatus, and Cryptomonas pyrenoidifera) cultured in the presence of a high UV-B dose were assessed for qualitative and quantitative cell changes and their effect on life-history parameters of D. pulex
. Beside reduction of algal growth rates by UV-B, important changes in the nutritional quality of the algae (e.g., total lipid concentration and fatty acids composition) were also observed. The intrinsic growth rate of D. pulex kept in the dark, however, was only significantly affected when feeding on the UV-irradiated S. subspicatus. Changes in intrinsic growth rate were mainly caused by a smaller number of offspring in the UV-B treatment. On the other hand, UV-B-irradiated C. reinhardtii and C. pyrenoidifera caused a reduction in the length of newborns in the first clutch. Changes in survival of D. pulex were not observed in all cases. Interestingly, in a similar experiment, but including another culture strain of C. reinhardtii, the survival of D. pulex neonates kept in the dark was strongly affected when feeding on UV-B-irradiated algae . Daphnia feeding on UV-B-irradiated algae also showed reduced intrinsic growth rates, clutch number and size. In this experiment, the growth rate of C. reinhardtii was only affected at the beginning but after 7 days it was similar to the control and changes in total lipids concentration were not observed. The results of these studies, although interesting, are difficult to interpret with regard to the net response of changes in population size. Considering that species of Daphnia like D. magna are known to be UV-B-sensitive [60,61], additional parallel experiments with D. magna and D. pulex concomitantly exposed to UVR would have been necessary to evaluate this interaction. On the other hand, the contrasting results obtained using the same species but different strains [58,59] stress the large biological variability found in response to the exposure to UVR.
In the only one study with freshwater periphyton, enhanced UV-B levels reduced photosynthesis and photosynthetic pigments, but algal nutritional quality (as measured by cell N and P content) and growth of juveniles of the snail Physella gyrina fed with UV-B-irradiated periphyton were not altered .
Was this article helpful?