UV radiation increases flavonoid protection but decreased reproduction in Silene littorea

Plants respond to changes in ultraviolet (UV) radiation via morphological and physiological changes. Among the variety of plant UV-responses, the synthesis of UV-absorbing flavonoids constitutes an effective non-enzymatic mechanism to mitigate photoinhibitory and photooxidative damage caused by UV stress, either reducing the penetration of incident UV radiation or acting as quenchers of reactive oxygen species (ROS). In this study, we designed a UV-exclusion experiment to investigate the effects of UV radiation in Silene littorea. We spectrophotometrically quantified concentrations of both anthocyanins and non-anthocyanin flavonoids (flavones) in petals, calyces, leaves and stems. Furthermore, we analyzed the UV effect on the photosynthetic activity in hours of maximum solar radiation and we tested the impact of UV radiation on male and female reproductive performance. We found that anthocyanin concentrations showed a significant decrease of about 20% with UV-exclusion in petals and stems, and 30% in calyces. Flavone concentrations showed a significant reduction of approximately 25% in calyces and stems, and 12% in leaves. Photochemical efficiency of plants grown under UV stress decreased sharply at maximum light stress, but their ability for recovery after light-stress was not affected. In addition, exposure to UV radiation does not seem to affect ovule production or seed set, but decreases the total seed production per plant and pollen production by 69% and 31%, respectively. Our results demonstrate that UV radiation produced opposite effects on flavonoid accumulation and reproduction in S. littorea. UV stress increased flavonoid concentrations, suggesting a photoprotective role of flavonoids against UV radiation, but had negative consequences for reproduction. We propose that this trade-off helps this species to occupy exposed habitats with high UV radiation.

12 220 concentrations in petals and stems decreased approximately 20% in these plants, whereas in 221 calyces the decrease was of 30% and the differences were marginally significant (Fig 2, Table   222 1). Anthocyanins were nearly absent altogether in leaves (Fig 2E). Flavone concentrations in 223 plants from the UV-exclusion treatment were lower by 12%, 23%, and 25% in leaves, calyces, 224 and stems, respectively, but in petals the differences were not significant (Table 1). 232 showed. FW, fresh weight; ns, not significant; ns*, marginally significant; *, P < 0.05; **, P < 233 0.01; ***, P < 0.001. Sines and Furnas populations did not show significant differences in anthocyanin 237 concentrations in any of the sampled tissues (Table 1). Conversely, the flavone 238 concentrations were significantly higher in plants from Sines, the southern population, in all 239 tissues except for the stems (Fig 2, Table 1), and the interactions of light treatment and 240 population were not significant (i.e. the decrease of flavone concentration in UV-exclusion 241 plants was homogeneous in both populations).

242
When we analyzed each population independently, we found that the only significant  Table 2). Leaves showed significant differences in their photochemical 255 efficiency between UV-treatments and between measurement conditions (predawn or 256 afternoon), and the interaction of UV-treatments and measurement conditions was also 257 significant (Table 2). In the afternoon, leaves of the UV-exclusion treatment showed a 20.8% 258 and 57.4% reduction of Fv/Fm values in early (March) and maximum (May) stages of the 15 259 flowering period, respectively (P < 0.001 for both pairwise comparisons, Table 3; Fig 3A and 260 B). In calyces, statistical differences in their photochemical efficiency were found only 261 between measurement conditions (predawn or afternoon) both in March and May (Table 2).
262 Pairwise comparisons in calyces revealed significant lower Fv/Fm values in afternoon 263 conditions, regardless of the UV treatment or the flowering period (P < 0.032, Table 3; Fig 3C   264 and D).

278
Flower production showed statistical differences between the two experimental 279 conditions (Table 4). Plants from the UV-exclusion treatment displayed approximately five 280 times more flowers than those with UV-present (261.4 ± 30.1 and 50.7 ± 8.3, respectively; 281 mean ± SE; Fig 4A). In addition, flower production was significantly different for both 282 populations, being higher in Sines plants. Conversely, fruit set was statistically higher in the 283 UV-present treatment and in plants from Furnas population (Fig 4B, Table 4). The number of 17 284 ovules per flower and seed set was statistically similar between light treatments or 285 populations (Figs 4C and D). The total seed production per plant was approximately three 286 times higher in plants from the UV-exclusion treatment compared to the UV-present plants 287 (46.2 ± 7.8 and 14.3 ± 1.9, respectively; Fig 4E), and did not show statistical differences 288 between populations (Table 4). Pollen production decreased by ~31% in plants exposed to 289 UV radiation (2126.1 ± 99.0 and 1473.9 ± 85.8, respectively; Fig 4F), but again differences 290 between populations was not found. The interactions of UV treatment and population were 291 not significant for any of the studied reproductive outputs (Table 4).

304
When assessing the relationship between flavonoid production and male and female 305 reproductive outputs, we did not found any significant correlations at Bonferroni-corrected 306 level ( = 0.05/6 = 0.008) (S2 Table).