A self-locking mechanism of the frog-legged beetle Sagra femorata.

Insect legs play a crucial role in various modes of locomotion, including walking, jumping, swimming, and other forms of movement. The flexibility of their leg joints is critical in enabling various modes of locomotion. The frog-legged leaf beetle Sagra femorata possesses remarkably enlarged hind legs, which are considered to be a critical adaptation that enables the species to withstand external pressures. When confronted with external threats, S. femorata initiates a stress response by rapidly rotating its hind legs backward and upward to a specific angle, thereby potentially intimidating potential assailants. Based on video analysis, we identified 4 distinct phases of the hind leg rotation process in S. femorata, which were determined by the range of rotation angles (0°-168.77°). Utilizing micro-computed tomography (micro-CT) technology, we performed a 3-dimensional (3D) reconstruction and conducted relative positioning and volumetric analysis of the metacoxa and metatrochanter of S. femorata. Our analysis revealed that the metacoxa-trochanter joint is a "screw-nut" structure connected by 4 muscles, which regulate the rotation of the legs. Further testing using a 3D-printed model of the metacoxa-trochanter joint demonstrated its possession of a self-locking mechanism capable of securing the legs in specific positions to prevent excessive rotation and dislocation. It can be envisioned that this self-locking mechanism holds potential for application in bio-inspired robotics.

The morphology and spectral characteristics of the compound eye of Agasicleshygrophila (Selman & Vogt, 1971) (Coleoptera, Chrysomelidae, Galerucinae, Alticini).

The first exploratory study was conducted on the compound eye morphology and spectral characteristics of Agasicleshygrophila (Selman & Vogt, 1971) to clarify its eye structure and its spectral sensitivity. Scanning electron microscopy, paraffin sectioning, and transmission electron microscopy revealed that A.hygrophila has apposition compound eyes with both eucones and open rhabdom. The micro-computed tomography (CT) results after 3D reconstruction demonstrated the precise position of the compound eyes in the insect's head and suggested that the visual range was mainly concentrated in the front and on both sides of the head. The electroretinogram (ERG) experiment showed that red, yellow, green, blue, and ultraviolet light could stimulate the compound eyes of A.hygrophila to produce electrical signals. The behavioural experiment results showed that both males and females had the strongest phototaxis to yellow light and positive phototaxis to red, green, and blue light but negative phototaxis to UV light. This study of the compound eyes of A.hygrophila will be helpful for decoding its visual mechanism in future studies.

Detection of Hindwing Landmarks Using Transfer Learning and High-Resolution Networks.

Hindwing venation is one of the most important morphological features for the functional and evolutionary analysis of beetles, as it is one of the key features used for the analysis of beetle flight performance and the design of beetle-like flapping wing micro aerial vehicles. However, manual landmark annotation for hindwing morphological analysis is a time-consuming process hindering the development of wing morphology research. In this paper, we present a novel approach for the detection of landmarks on the hindwings of leaf beetles (Coleoptera, Chrysomelidae) using a limited number of samples. The proposed method entails the transfer of a pre-existing model, trained on a large natural image dataset, to the specific domain of leaf beetle hindwings. This is achieved by using a deep high-resolution network as the backbone. The low-stage network parameters are frozen, while the high-stage parameters are re-trained to construct a leaf beetle hindwing landmark detection model. A leaf beetle hindwing landmark dataset was constructed, and the network was trained on varying numbers of randomly selected hindwing samples. The results demonstrate that the average detection normalized mean error for specific landmarks of leaf beetle hindwings (100 samples) remains below 0.02 and only reached 0.045 when using a mere three samples for training. Comparative analyses reveal that the proposed approach out-performs a prevalently used method (i.e., a deep residual network). This study showcases the practicability of employing natural images-specifically, those in ImageNet-for the purpose of pre-training leaf beetle hindwing landmark detection models in particular, providing a promising approach for insect wing venation digitization.