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

projections for the regions—+1.4–5.8˚C mean annual temperature and +10–30% mean annual rainfall by 2100 [21]. To address this gap in knowledge, we measured slug abundance, slug bio- mass, slug herbivory, plant community composition, plant biomass production, and forage quality in a manipulative, field climate change project in a central Kentucky pasture where temperatures were increased 3˚C year-round and growing season precipitation was increased by 30% of the long-term mean. We hypothesized that slug herbivory would have a measurable negative impact on above- ground plant biomass and that climate change factors would affect slug herbivory, slug abun- dance, and slug biomass. As reported elsewhere [12, 22], we predicted that slug abundance and biomass would vary over the course of the year and that slug response to the climate treatments would be seasonally dependent. For example, warming might increase slug abundance, bio- mass, and herbivory during cool spring and autumn months but suppress them during the hot summer. We hypothesized that the combination of warming and additional precipitation would allow slugs to persist and exhibit significant herbivory throughout the summer com- pared to warmed only plots where they would be less abundant due to their intolerance of drought. However, co-occurring changes in plant community composition, both seasonally and those resulting from the climate change treatments, could modify these responses. Materials and methods Research site The experiment was conducted at the University of Kentucky Spindletop Research Farm in

Lexington, Kentucky, U.S.A. (38˚10’N, 84˚49’W). The site was established in an existing cli- mate change project as described in Brosi [32] and Slaughter, Weintraub, and McCulley [33]. After a spring application of glyphosate (Roundup Pro; Monsanto, St Louis, MO, U.S.A.) to remove existing vegetation, the research field was established in the summer of 2008 with ‘Kentucky-31’ tall fescue (Schedonorus arundinaceus (Schreb.) Durmort., seeded at a rate of 11.2 kg ha-1 pure live seed), ‘Ginger’ Kentucky bluegrass (Poa pratensis L., 7.8 kg ha-1), ‘Free- dom’ red clover (Trifolium pretense L., 6.7 kg ha-1), and ‘Patriot’ white clover (Trifolium repens L., 2.2 kg ha-1). Bermudagrass (Cynodon dactylon (L.) Pers.) was plugged into the field later in the summer at a rate of 10.76 plugs m-2. Soil at the site is deep (>1 m) Maury silt loam (fine, mixed, active, mesic Typic Paleudalfs). Long-term mean annual summer and winter tempera- tures are 23.8 and 1.6˚C, respectively [34], with average annual rainfall of 1137 mm distributed relatively evenly throughout the year. Total annual rainfall (mm) across the study period was 1656 mm (2011), 1019 mm (2012), and 1495 mm (2013) (Kentucky Mesonet; http://www. kymesonet.org/; Lexington, KY, U.S.A.). During the winter of 2008–2009, twenty 3-m diameter hexagonal plots were established and divided into five replicates of four climate treatments. The climate treatments were ambi- ent conditions (control), +heat (3˚C above ambient), +precipitation (+30% of the long-term

mean annual rainfall applied during the growing season from rainwater collected on site; +343 mm), and a combination of +heat and +precipitation. Increased temperature was maintained day and night using Salamander infrared heaters (Mor Electrical Heating Associates, Inc., Comstock Park, MI) positioned around the plot following the procedure of Kimball and Con- ley [35]. Additional rainfall was applied every two weeks throughout the growing season (April-October). Rainfall additions were April +51 mm, May +51 mm, June +62 mm, July +67 mm, August +62 mm, and September +51 mm. The climate treatments began on May 1, 2009 and ran continuously until November 11, 2013. Each plot was surrounded with aluminum flashing buried 0.5 m deep in the soil. Soil tem- perature (FW3648 shielded Type T thermocouples; TE Wire, Saddle Brook, NJ, U.S.A) and PLOS ONE Slug herbivory & climate change PLOS ONE | https://doi.org/10.1371/journal.pone.0283128 March 14, 2023 3 / 17