This study examines the larval mandibles of five caddisfly species, documenting their morphological and biomechanical adaptations to different feeding strategies. Three predatory species (Molanna angustata, Plectrocnemia conspersa, Rhyacophila fasciata) and two algae-grazing species (Silo nigricornis, Tinodes pallidulus) were investigated using scanning electron microscopy (SEM), energy-dispersive X-ray spectroscopy (EDX), confocal laser scanning microscopy (CLSM), and nanoindentation. SEM analysis revealed distinct mandible structures: grazing species possessed various collecting setae and sharp mandible cutting edges, while predatory species exhibited more pointed incisors. By EDX, low concentrations of Ca, Cl, Cu, Fe, K, Mg, Mn, P, S, Si, and Zn were detected with no relationship to mechanical properties, suggesting limited elemental reinforcement in larval Trichoptera mandibles. CLSM imaging, however, revealed regional material heterogeneities related to the mechanical properties, indicating that the mechanical parameters depend on the degree of cuticle tanning. We detected harder and stiffer mandible cuticle in predatory species, likely enhancing the mechanical resistance of the material during prey capture. Nanoindentation analysis identified two functional mandible types in grazers, adapted for scraping and collecting, with differential regional cuticle hardness and stiffness. Predatory species exhibited three mandible types, likely specialized for grabbing and crushing, piercing and cutting, or grabbing, piercing, and cutting. Decreased mechanical properties of condyles in predators suggested enhanced flexibility for prey handling, whereas grazers, showing higher Young’s moduli and hardness values in condyles, likely required higher pressure for scraping. In some species (Silo nigricornis and Rhyacophila fasciata), the heterogeneity between lateral and medial mandible cuticle indicated that self-sharpening mechanisms could be present. These findings provide insight into the functional morphology and material adaptations of mandibles of trichopteran larvae in relation to their feeding ecology.

Studied species

Five species were investigated and compared to the published data on Glossosoma boltoni (Spicipalpia) [42]: Plectrocnemia conspersa (Curtis, 1834) and Tinodes pallidulus McLachlan, 1878 (both Annulipalpia), Silo nigricornis (Pictet, 1834), Rhyacophila fasciata Hagen, 1859, and Molanna angustata Curtis, 1834 (all Integripalpia) (Supplementary Table 1). Two species primarily feeding through “grazing” (i.e., scraping algae from solid surfaces) (Tinodes pallidulus, Silo nigricornis) and three species primarily predatory (Plectrocnemia conspersa, Rhyacophila fasciata, Molanna angustata) were selected. Data regarding trophic specialization were obtained from the literature [36].

For this study, eight specimens of the ultimate larval instar per species were examined (Supplementary Table 2). The specimens of P. conspersa, S. nigricornis, M. angustata, and R. fasciata were obtained from the Georg Ulmer collection, curated at the Zoological Museum Hamburg (ZMH), now part of the Leibniz Institute for the Analysis of Biodiversity Change Hamburg (LIB). Most specimens were collected by August Thienemann around the turn of the 20th century in northern Germany. T. pallidulus was collected by Martin Kubiak in 2013 in Saxony-Anhalt. Specimens were preserved in 70% EtOH.

Mandibles were first documented using scanning electron microscopy (SEM) to identify specific areas affected by wear. Subsequently, elemental analysis on the mandibular cuticle was conducted using energy-dispersive X-ray spectroscopy (EDX). Nanoindentation was performed to test for mechanical property gradients within each mandible. Finally, mandibles were examined by confocal laser scanning microscopy (CLSM) to reveal material composition. Using this combination of methods, the morphologies and material properties of the mandibles were compared and analyzed in the context of feeding strategies (grazing and predatory).

Light microscopy

Initially, the specimens were documented using a Keyence Digital Microscope VHX-7000 (KEYENCE, Neu-Isenburg, Germany) equipped with automatic stacking software in dorsal, ventral, and lateral views (Supplementary Figures 1–5). Images of individuals were also used to determine the ratio of body and head length to mandibular length. Measurements of the individuals were taken by drawing a line from anterior to posterior along the larva in Adobe Photoshop, because most specimens were curled. Then, the length of the line, given in pixels, was converted into mm, taking the number of pixels of the scale bar as a conversion factor. Additionally, all extracted mandibles were documented using light microscopy before further analysis.

Scanning electron microscopy (SEM)

For documenting head morphology, seven specimens per species were cut along with the first abdominal segment, briefly cleaned in 70% EtOH using an ultrasonic bath, mounted on SEM sample holders using double-sided carbon adhesive, and air dried. The specimens were coated with a 5 nm thick layer of platinum (Polaron SC7640 Sputter Coater) and visualized with a Zeiss LEO 1525 (One Zeiss Drive, Thornwood, USA) at an acceleration voltage of 15kV. Subsequently, the specimens were rehydrated with 70% EtOH, detached from the sample holders, and the mandibles were carefully dissected from the heads. The mandibles were cleaned again in the ultrasonic bath and documented using SEM. The images were used to describe morphology and measure the length of the mandible. Afterwards, mandibles were detached from the sample holders with 70% EtOH, cleaned in a short ultrasonic bath, and used for either EDX or nanoindentation.

Energy-dispersive X-ray spectroscopy (EDX)

Two specimens per species were used for this analysis (one left mandible of one specimen and one right mandible of another specimen to compare for differences between individuals and between the right and left sides). The same localities of the countermandibles from the same specimens were later tested by nanoindentation. 

The mandibles were initially mounted on glass slides using double-sided carbon adhesive. Each sample was surrounded by a small metal ring, which was then filled with epoxy resin (Reckli Epoxi WST, RECKLI GmbH, Herne, Germany) following the protocol described by [43,44,45]. Polymerization lasted three days at room temperature. Afterwards, the glass slides and carbon adhesive were removed. Samples were polished using sandpapers of varying grit sizes until the first layers of the cuticle became visible. Subsequently, the exposed surfaces were polished with a 0.3 µm aluminum oxide polishing powder suspension (PRESI GmbH, Hagen, Germany) on a polishing machine (Minitech 233/333, PRESI GmbH, Hagen, Germany) to achieve a smooth surface. Since the various regions of interest were located at different depths within the sample, each sample was sequentially polished to expose one target region at a time. Once a target region was revealed, EDX analyses were conducted. The sample was then further polished to expose the next target region, after which the same protocol was repeated (see below) to test the subsequent regions.

After smoothing, the samples (Figure 1) were cleaned in an ultrasonic bath for five minutes, mounted onto SEM sample holders, air-dried, and coated with a 5 nm layer of platinum. The elemental composition was determined using the SEM (Zeiss LEO 1525), equipped with an Octane silicon drift detector (SDD) (TEAM microanalysis system, EDAX Inc., New Jersey, USA). Consistent settings were applied for all samples (e.g., acceleration voltage of 20 kV, working distance, lens aperture, etc.). Prior to analyzing each sample, the detector was calibrated using copper. Small areas of the cuticle sections (1–2 x 1–2 µm) were tested.

The detected elements included

aluminum (Al), carbon (C), calcium (Ca), chlorine (Cl), copper (Cu), hydrogen (H), iron (Fe), potassium (K), magnesium (Mg), manganese (Mn), nitrogen (N), sodium (Na), oxygen (O), phosphorus (P), platinum (Pt), sulfur (S), silicon (Si), and zinc (Zn).

However, some elements were excluded from discussion as they are either fundamental components of chitin and proteins (H, C, N, O), part of the coating (Pt), or derived from the polishing powder (Al, O). The remaining elements were collectively categorized as "all elements, Ae". Since Pt and P share a spectral peak and the software cannot distinguish between the two, their values were combined as (P + Pt). To refine the estimation of P content, the elemental composition of 20 areas of the epoxy resin was analyzed, yielding a mean ± SD Pt content of 0.15 ± 0.02 atomic %. The analysis confirmed that the resin contained none of the other discussed elements.

The measurements were sorted into the following mandibular zones: (A) medial cuticle of the mandible tip, (B) lateral cuticle of the mandible tip, (C) medial surface of the mandible, (D) lateral surface of the mandible, (E) dorsal condyle, and (F) ventral condyle (Figure 1). The mandible surfaces were then divided into the sections a to k (Figure 1B). Since mandibles were not of the same size, we assorted these sections by proportions. 

A total of 1006 localities from 10 mandibles were tested via EDX (Figure 1). For analysis of elemental content of the suborders, the data from G. boltoni was included from [42].

Nanoindentation

Nanoindentation was conducted on the two opposing mandibles from the same specimens analyzed by EDX. To expand the sample size, six additional mandibles from three other specimens per species were included for mechanical property testing, resulting in a total of four individuals per species studied. The preparation of samples for nanoindentation followed the same protocol as for EDX: mandibles were embedded in epoxy, polished, and smoothened [for detailed protocol, see 42].

The prepared samples were mounted onto the nanoindenter sample holder. Indentations were performed using a nanoindenter SA2 (MTS Nano Instruments, Oak Ridge, USA) equipped with a Berkovich indenter tip and a dynamic contact module (DCM) head. The mechanical properties — hardness (H) and Young’s modulus (E) — were calculated from force-displacement curves using the continuous stiffness mode. H measures a material's resistance to localized plastic deformation caused by indentation. E represents the stiffness of a solid material, describing the relationship between stress and axial strain. 

These properties were determined at penetration depths of 800–1000 nm. For each indentation site, approximately 30 measurements were recorded at varying depths and then averaged to produce one mean value for H and one for E per indent. All tests were carried out under standard room conditions (relative humidity 28–30%, temperature 22–24 °C), with each indent and its corresponding curve manually inspected. After testing, each sample was further polished to reveal the next target region for analysis.

A total of 1040 localities from 40 mandibles were tested using nanoindentation (Figure 1). For analysis of the mechanical properties of the suborders, the data from G. boltoni was included from [42].