**4. Discussion**

Autecological data on planktonic ciliate species are scarce, and almost nothing is known about their strategies to cope with potentially damaging UVR (see the review in [41]). The ciliates tested here commonly occur in subalpine lakes, and they are members of the mixotrophic assemblage found in Piburgersee [10,41,42]. Considering that, in this lake, UV-B radiation is completely attenuated after 3 m and UV-A after approximately 7 m, e ffective photoprotective strategies appear to be essential for the survival of plankton [10]. Interestingly, the ciliates tested here were obviously not as tolerant to solar radiation conditions as we assumed, and the e ffectiveness of the UV photoprotective strategies tested was unexpectedly low.

Certainly, under the experimental conditions tested here, an escape reaction away from UVR was not possible because the ciliates were "trapped" in the shallow wells. Therefore, our approach does not necessarily reflect the in situ situation but, instead, tested how di fferent species reacted to the same controlled experimental manipulation. In fact, the results from a one-year seasonal study in Piburgersee showed that *P. trachelioides, V. chlorellata*, and *S. vernalis* dwell in the uppermost meters during summer (Figure S2).

In a previous study on the overall resistance to UVR of the ciliate summer community from Piburgersee, *V. chlorellata* was the only ciliate species that survived exposure under UVR [10]. The authors concluded that, probably because of the high MAA concentrations and the presence of densely packed algal symbionts, UVR was hindered from reaching sensitive cell targets such as the nuclei [10]. In the present experiments, *V. chlorellata* again first appeared to be resistant to UVR but, finally, all individuals died. The possible sublethal e ffects were probably masked in the previous study, where the ciliates were preserved right at the end of the experiment [10]. Another factor that probably allows *V. chlorellata* to live directly at the surface of Piburgersee is its association with *B. braunii* (Figure 1c). These planktonic algae form large colonies that can even be recognized by the naked eye. As the algal colonies serve as attachment sites for *V. chlorellata*, we argue that shading from incident solar radiation by the (probably mutualistic) algal partner is important.

For *S. vernalis*, physical shading of the nuclei through algal symbionts is unlikely because they are uniquely arranged in a variable number of "packages" and because aposymbiotic individuals naturally occur (Figure 1d; [41,43]). Accordingly, the UV-screening e fficiency by self-shading is expected to be significantly lower in *S. vernalis* than in other mixotrophic species, such as *V. chlorellata* or *P. bursaria* [10,15,16]. MAA analyses of *S. vernalis* from PIB revealed that these compounds were

only detected in very low concentrations when they were there (Table 1; [14]). The synthesis of MAAs per se is energetically costly, and the concentrations required for individual photoprotection were probably too low to tolerate the UVR dose tested here. Although MAAs are considered e fficient sunscreen compounds, it remains unknown what concentration is needed to provide an e fficient protection in ciliates, although in general, the larger the cell diameter, the higher the potential sunscreen efficiency [38]. Therefore, we can only speculate that the MAA concentrations might have either been too low for e fficient UVR screening or that they may have di fferent functions other than photoprotection.

Out of the three species tested here, *P. trachelioides* was the most sensitive to UVR, especially to UV-B (Figure 1f, Figure 3a,b and Figure S1). This was unexpected because, in this large dileptid ciliate of up to >1 mm in length, we found the highest concentration of MAAs among the three tested species (Table 1) as well as the existence of numerous algal symbionts [14]. Nevertheless, the existence of a PER in *P. trachelioides* indicates the potential to counteract UVR damage (Figure 3b). Our results add to previous reports of the existence of PER in the ciliates *Glaucoma* sp. and *P. acutum* as a major strategy in UV-B tolerance [22,23].

Physical damage, including the easy loss of the prominent proboscis of *P. trachelioides* through mechanical disturbance (e.g., by a pipette), has been observed earlier [44,45], and here, we found that UVR contributed to significant length reductions (Figure 1f and Figure S1). The proboscis is necessary for feeding and swimming and can be regenerated within several hours to almost double the trunk length, especially when the ciliates remained undisturbed (Figure 1e; [44,45]). Although several studies on the regeneration of surgically amputated mouthparts in dileptids and stentorids have been done in the past, unfortunately, none of them ever mentioned a reason behind losing and regenerating a proboscis or trunk [46–51].

Algal symbiont accumulation in the posterior cell region has been reported for *P. trachelioides*, though from the scientific literature, it is unclear what caused the dislocation (Figure 1f; [45,52,53]). Some authors observed this phenomenon before cytolysis, for instance, Krainer [53] and Butkay [45] found such individuals already in their original habitats. Butkay [45] assumed that increasing water temperature and/or an escape reaction away from an irradiation source may have been the reasons. Symbiont dislocation into the posterior cell region only attains e fficient photoprotection when this cell region is simultaneously directed toward the radiation source, providing a kind of "umbrella" that actively shades the nuclei. This phenomenon has been observed in *P. bursaria*, and it is reversible depending on the level of UV exposure. Thus, when UVR is excluded, the symbionts are immediately evenly distributed inside the cell again [15,16]. Although it is unknown what exactly triggers the dislocation in *P. trachelioides*, our observations sugges<sup>t</sup> that UVR is involved. Because all three species were obviously stressed by UVR, the photoprotective role of the symbionts is called into question. However, the symbionts themselves might also have been negatively impacted by UVR, although this was not tested here.
