Do vertical gradients in temperate forest structure insect herbivore communities?  We tested the hypothesis that the increase in light intensity from understory to forest canopy level drives differences in leaf physical traits and budburst phenology that impact insect herbivores and thus play a role in structuring both herbivore communities and the leaf damages they cause. Twelve sugar maple (Acer saccharum) sites were monitored in southern Quebec, examining insect herbivore patterns from understory to the shaded and sun canopy over the summers of 2020, 2021, and 2022. Additionally, we recorded leaf physical traits, temperature, humidity, and sun exposure. Our findings revealed that leaf thickness increased along the vertical gradient in 2021, making leaves less favorable to herbivores in the canopy level. Accordingly, we recorded a consistent decrease in insect herbivory damage rates from the understory to the shaded canopy and sun canopy in 2020 and 2021, driven by leaf cutters, skeletonizers, stipplers, and leaf miners. These results support our hypothesis that variation in plant physical traits due to sun exposure contributes to the vertical stratification of insect damage. In 2022, the gradient of insect herbivore abundance corroborated the observed damage trends from the previous years. Moreover, we calculated an average annual herbivory rate of 9.1% of the leaf surface in our study site, suggesting limited evidence supporting a significant contribution of background herbivory to the decline of sugar maple forests. Overall, our study highlights the importance of vertical gradients in structuring insect herbivore communities and emphasizes the role of leaf traits in mediating these interactions.

Study site and plant species

We conducted our sampling at the Kenauk Nature property, located in southwestern Quebec, Canada (45°42'N; 74°53'W). The property spans two regions from the Outaouais river valley to the Laurentian Mountains. Based on the provided domains map by the Quebec Ministry of Forests it is in a maple/bitternut hickory domain of temperate deciduous forest (Quebec Ministry of Forests 2023). To consider the variation in insect herbivory throughout the growing season, three sampling sessions were conducted at the end of June, end of July, and middle of August in 2020, 2021, and 2022. During each session, we sampled the same 12 sites, each spaced more than 15 meters apart and tree species composition was relatively uniform in all sites. These sites were selected based on the similarity in tree size and safety considerations for climbing. Each site comprised one mature sugar maple tree and one sapling. We collected samples from both the sunlit and shaded canopy levels of the mature trees, as well as from the branches of saplings in the understory. Canopy branches were sampled at a height of 10-13 meters above ground on mature trees averaging around 16 meters in height, and the understory branches were defined as those on surrounding sugar maple saplings ranging from 1-3 meters in height. For visual clarity and to present all sampling events and their respective timings in one place, an infographic has been included as Figure 1.

Assessment of insect herbivore damages

Leaf herbivory damage is recognized as a good measure of the structure of insect herbivore communities and their interactions with host plants (Schowalter et al. 1986b; Schowalter 2006). Among the methods of assessing leaf herbivory damage, visual assessments are accurate enough for comparative analysis (Landsberg 1989; Johnson et al. 2016). In this study, herbivory damages were visually quantified as the percentage of leaf area damaged by different insect feeding guilds (Wint, G. R. W 1983; Houston et al. 1990; Thomas et al. 2010; Johnson et al. 2016). We averaged the damage scores for 15 fully expanded leaves per stratum per site in both 2020 and 2021. Nine insect herbivory damage types identified on sugar maples (Thomas et al. 2010) were assessed, including leaf rollers, leaf skeletonisers, leaf cutters, leaf stippling, spindle galls, erineum gall, ocellate gall, leaf miner maple leaf-cutter and maple leaf-blotch miner (Fig 1). Trained interns conducted assessments using the guidebook by Houston et al. (1990), which features clear images of insect herbivores induced damages on sugar maple leaves (Houston et al. 1990). During assessments, each leaf was visually divided into four parts, and the percentage of observed damage per part was estimated. The sum of these percentages yielded the total percentage of a particular damage type on the leaf. To ensure consistency in assessment methods, subsequent interns received supervision from the previous year's intern, and the first 20-30 leaves were assessed together. To capture both spatial and temporal effects on herbivory damage, we randomly collected 15 leaves per stratum per site at three sampling sessions (June, July, and August). In total 1620 leaves were collected (12 trees*3 strata*15 leaves*3 sampling dates)(Schowalter et al. 1986a; Zehnder et al. 2009; Turcotte et al. 2014). For understory, leaves were collected by hand from accessible branches on the ground.  Canopy leaves were accessed using single rope technique and climbing gear and were cut with a two-meter pole pruner; shaded leaves were collected from mid-center canopy branches under near-complete canopy cover receiving less light, while sunlit leaves were sampled from higher and more marginal canopy branches with less than 50% canopy cover to ensure greater sun exposure. Due to the limitations of single rope techniques, sampling was often confined to the inner section of the tree crown (Basset 1991; Thomas et al. 2010; Maguire et al. 2014). Environmental factors such as light intensity, temperature, and humidity were recorded at each sampled branch during the three sampling sessions to account for microhabitats differences between branches in sun-exposed canopies, shaded canopies, and understory. Light intensity was measured using a Reed light meter model R1930, while temperature and humidity were recorded using a Reed temperature and humidity monitor model R6000.

Measurement of leaf physical traits and phenology of leaf bud burst

In 2021, we evaluated the effect of microhabitat heterogeneity on leaf quality by collecting 36 undamaged, fully expanded leaves per stratum per sampling date (total of 108 leaves per stratum=3 leaves* 12 site* 3 sampling dates). Sampling size was determined based on previous studies that measured leaf traits in a vertical gradient of deciduous forests (Corff and Marquis 1999; Fortin and Mauffette 2002; Murakami et al. 2005; Zvereva et al. 2020). The leaf sampling was synchronized with the herbivore damage sampling. Fresh leaves were weighed on-site, and their thickness was measured using a digital caliper. Subsequently, the leaves were pressed, labeled, transported to the lab, and dried in an oven for 72 hours. Dried leaves were then weighed again, and all leaves were scanned to measure leaf area using ImageJ software. All efforts were made to ensure that dried leaves remain intact and free from any folding or breaking to maintain measurement accuracy. This allowed us to determine water content and specific leaf area (SLA) for all sampled leaves, facilitating the comparison of leaf traits variation across the three sugar maple strata.

In 2022, we conducted a comparative study on the phenology of leaf-flush between sugar maple saplings and the canopy of mature sugar maples trees. On May 10, 2022, we recorded the bud burst ranking by visually inspecting 10 buds on a single sugar maple sapling with the naked eye, and 10 buds on the canopy of mature trees adjacent to the selected sapling using binoculars. The bud burst ranking was assessed on a scale of 0 to 30; with 0 representing small buds, 10 indicating swollen buds, 20 indicating small leaves, and 30 indicating fully expanded leaves (Hannerz 1999). This recording process was repeated at 20 sites which were distinct from the 12 sites used for herbivory damages and leaf trait measurements but were located in close proximity to those sampled sites and exhibit similar tree species composition. Each sites spaced at least 10 meters apart and included one mature sugar maple with a height of 10-15 meters to assess canopy level bud burst timing and one sugar maple sapling with a height of 1-3 meters representing the understory.

Insect herbivore species identification

To establish a link between the observed herbivory damage types to the actual insect herbivore community on sugar maple trees, we employed the beat sheeting method to collect insects from all three strata at the same 12 sugar maple sites used for herbivory damage assessments. Although sampling was conducted during the summer of 2020, 2021 and 2022, technical issues resulted in the collection of fewer than 20 individual insect herbivores in the first two years, which were subsequently excluded from the analysis. Therefore, only data from 2022 was considered for analysis. Using the beat sheeting method, three branches from each stratum were struck with a one-meter length stick, while an 85 cm* 85 cm sheet with a detachable jar at the center was positioned beneath the branch. All organisms that dislodged onto the sheet were carefully collected in the jar. Subsequently, the jars containing samples from the canopy strata were lowered to the ground and filled with 70% ethanol to prevent predation on the herbivores within the container. Upon retrieval, the samples were sorted into morphotypes and identified to the lowest possible taxonomic level in the lab. Each sample was meticulously labeled, pinned, and deposited at the Lyman Entomological Museum in St-Anne-de-Bellevue, QC, Canada.