Climate change alters slug abundance but not herbivory in a temperate grassland

We hypothesized that seasonal trends in moisture and temperature would interact with the imposed climate treatments to alter effects of slug herbivory and influence slug abundance and slug biomass. We found that seasonal differences in slug abundance were modified by the cli- mate treatments in ways that we predicted. When moisture was limiting (in late summer), +heat treatments were hotter and drier than non-heated plots (Fig 1) and had lesser numbers of slugs (Fig 4). This result is consistent with warming-associated reductions in slug abundance at Jasper Ridge Biological Preserve [45]. At our site, added precipitation increased the abun- dance of slugs during the late summer. While we did not monitor or measure slug activity, slugs may have been more active during this time period. Slug abundance varied across the year. Slugs were present aboveground in significant numbers during the spring, early summer, early autumn, an unseasonably warm week in February, and an unseasonably cool and wet period between mid-June and early July, illustrating, as others have reported, that slugs are sensitive to the temperature and moisture levels in the field [12, 22, 23]. We additionally hypothesized that seasonal changes in temperature and moisture would interact with the climate treatments to alter plant community composition, forage quality, and biomass. We found evidence that warming significantly modified plant community dynamics and forage quality. While the abundance of C4 species and C:N ratio increased overall in all

treatments as the season progressed, +heat treatments exhibited greater increases compared to the other treatments. This is consistent with research suggesting that warming will reduce overall forage quality [46]. Heat, precipitation, and their interaction modified the seasonal effects on plant biomass. Plant biomass seasonal changes and climate modifications were inconsistent across the years, illustrating that the effects of climate change are subject to inter- annual variation, a result supported by others [47, 48]. For example, during the drought in 2012, the +precipitation treatments had greater biomass than the control plots, but during the relatively wet year of 2013, the +heat treatments had significantly greater biomass. Climate change may make inter-annual variation in harvestable biomass more variable, which is con- sistent with predictions by White et al. [46]. Table 2. Average number and estimated biomass (± S.E.) of slugs caught per plot by climate treatment across all 20 trapping events. Letters indicate significant differences across climate treatments. Treatment Number of Slugs Biomass of Slugs (mg) Control 0.92 ± 0.19 a 45.65 ± 8.39 a Precipitation 1.19 ± 0.28 a 69.14 ± 11.01 a Heat 0.66 ± 0.21 b 59.41 ± 15.36 a Heat × Precipitation 0.48 ± 0.13 b 44.15 ± 12.59 a https://doi.org/10.1371/journal.pone.0283128.t002 Table 3. ANOVA table for the average count and biomass of slugs trapped per plot for 2013 in the climate change treatments. Degrees of freedom are numerator, denominator. Bolding indicates significance. Slug Count Slug Biomass

Effect DF F P DF F P Heat 1,12 11.74 0.0050 1,62 0.27 0.6058 Precip 1,12 0.15 0.7055 1,63 0.08 0.7836 Heat × Precip 1,12 0.90 0.3603 1,63 0.59 0.4438 Time 19,304 23.52 <0.0001 14,62 0.49 0.9322 Time × Heat 19,304 3.89 <0.0001 7,61 1.83 0.0974 Time × Precip 19,304 1.67 0.0407 9,61 0.71 0.6999 Time × Heat × Precip 19,304 1.44 0.1059 5,60 0.48 0.7919 https://doi.org/10.1371/journal.pone.0283128.t003 PLOS ONE Slug herbivory & climate change PLOS ONE | https://doi.org/10.1371/journal.pone.0283128 March 14, 2023 10 / 17