**Background:** Trophic cascades—indirect effects of predators on plants via herbivore suppression—are well documented, but the role of insectivorous bats in temperate forest ecosystems has been largely overlooked. Bats have high metabolic rates, exhibit functional responses to prey density, and consume many forest-defoliating insects, yet only one prior study (in tropical forests) experimentally excluded bats to measure their impact on forest defoliation. Most bird exclusion studies inadvertently excluded bats by using 24-hour exclosures, potentially conflating the effects of both groups. This study tested the hypothesis that insectivorous bats reduce insect density and defoliation in temperate deciduous forests, and whether effects differ between oak and hickory seedlings.
**Methods:** The study was conducted in Yellowwood State Forest, south-central Indiana, USA, across five thinned stands (0.4–6.8 km apart). Twenty experimental units were deployed over three summers (2018–2020; seven per year, one lost in a storm). Each unit consisted of a netted bat-excluded plot (6×7×7 m, with 4.44-cm mesh netting closed at night and opened at dawn to allow bird access) and a paired netless control plot within 15 m. Within each plot, 9–10 oak and hickory seedlings were randomly selected (total: 235 oak and 161 hickory seedlings). Leaf area was measured at the start and end of the treatment period (late May to mid-August) using either a 1-cm grid (2018) or photography with Easy Leaf Area software (2019–2020); both methods were validated and showed no significant difference (t=0.99, df=5, p=0.923). Insect abundance was surveyed on each seedling at both time points. Bat activity was monitored with acoustic detectors, and nonbat fauna with trail cameras. Reliability was high (intraclass correlation coefficient >0.99). Defoliation was analyzed using linear mixed models with treatment, tree genus, and their interaction as fixed effects, year as a covariate, and experimental unit/tree ID as random effects. Insect density was analyzed with a generalized linear mixed model (negative binomial) with treatment, time point, and their interaction as fixed effects.
**Key Results:** The full model outperformed the null model (null ΔAICc=60.83, ~100% model weight). Treatment (F₁,₃₉₅=69.64, p=0.001) and the treatment×genus interaction (F₁,₃₉₅=5.26, p=0.022) were significant predictors of defoliation. Genus alone was not significant (F₁,₃₉₅=1.61, p=0.193). Bat-excluded seedlings experienced five times more defoliation than controls (14.5% vs. 2.8%). Oaks showed nine times more defoliation when bats were excluded (16.3% vs. 1.8%), while hickories showed three times more (11.6% vs. 4.1%). Insect density did not differ between treatments at the start (z=0.90, p=0.370) but was significantly higher in bat-excluded plots at the end (z=21.93, p≤0.001). Over the season, insect density decreased by 0.4 insects/seedling in controls (z=−2.66, p=0.008) and increased by 1.2 insects/seedling in bat-excluded plots (z=3.97, p≤0.001). Six times more observations of >10 lepidopteran larvae occurred on bat-excluded seedlings, and 122 vs. 19 insect eggs were noted on bat-excluded vs. control seedlings in final surveys.
**Clinical Implications:** This study provides experimental evidence that insectivorous bats initiate strong trophic cascades in temperate deciduous forests, reducing defoliation fivefold and insect density threefold. The greater benefit to oaks (ninefold reduction) than hickories (threefold) suggests bats may influence forest composition by conferring a competitive advantage to certain tree species, with potential implications for ongoing oak declines in the eastern US. The results indicate that bat declines (e.g., from white-nose syndrome) could exacerbate defoliation and alter forest structure. The study also highlights that prior bird-exclusion studies using 24-hour exclosures likely overestimated birds' effects by inadvertently excluding bats, and recommends future studies open exclosures at night. While defoliation levels were likely sublethal (14.5% leaf area loss), even moderate defoliation can reduce seedling growth and survival under drought or other stressors. The findings underscore the need to conserve bat populations for both ecological and economic reasons, and suggest that proximity to forests may be a key predictor of bat-provided pest suppression in agricultural landscapes.