Mai, G.-S. | Search bioRxiv

bioRxiv2026

MorphQ: label-free quantification and visualisation of complex morphology from standardised specimen images

O_LIQuantifying complex morphology from images remains difficult because predefined descriptors capture only selected traits. Yet, supervised machine learning models for images require labels and often produce task-specific features that are hard to interpret as biological traits. C_LIO_LIWe present MorphQ, a label-free, self-supervised method that learns a quantitative morphospace from standardised specimen images. Its encoder produces feature vectors for statistical analysis, and its decoder converts analysed positions in morphospace into human-interpretable images, including hypothetical forms not represented by sampled specimens or sampled taxa. C_LIO_LIUsing 1,868 Lepidoptera species, we tested whether MorphQs label-free features were more useful for downstream analysis than features from principal component analysis (PCA) or a supervised species-classification machine learning model. As a diagnostic probe of downstream biological utility, MorphQ features supported higher low-label family-classification accuracy than comparator features, and retained stronger family-level similarity for species absent from model training, indicating better generalisation to species not seen during model training. C_LIO_LITwo case studies link MorphQ morphospaces to species-level elevation and assemblage-level functional diversity while keeping statistical patterns visually inspectable. MorphQ provides a reproducible framework for constructing interpretable morphological trait spaces when predefined descriptors are incomplete and labelled data are limited. C_LI Data/code for peer review: An anonymised repository containing the source code, trained model weights, example data, configuration files and scripts required to reproduce the analyses is available at https://anonymous.4open.science/r/MorphQ-ECD4/.

ecology↗

bioRxiv2024

Circadian Activity Predicts Breeding Phenology in the Asian Burying Beetle Nicrophorus nepalensis

Climate change continues to alter breeding phenology in a range of plant and animal species across the globe. Traditional methods for assessing when organisms reproduce often rely on time-intensive field observations or destructive sampling, creating an urgent need for efficient, non-invasive approaches to assess reproductive timing. Here, we examined three populations of the Asian burying beetle Nicrophorus nepalensis from subtropical Okinawa (500 m) and Taiwan mountains (1100-3200 m) that were reared under contrasting photoperiods in order to develop a predictive framework linking circadian activity to breeding phenology. Using automated activity monitors, we quantified adult circadian rhythms and employed machine learning to predict breeding phenology (seasonal versus year-round breeders) from behavior alone. Our model achieved 95% accuracy under long-day conditions using just three behavioural features, and notably, maintained 76% accuracy under short-day conditions when both types are reproductively active, revealing persistent behavioural differences between breeding strategies. These results demonstrate how integrating behavioural monitoring with machine learning can provide both a rapid, scalable method for tracking population responses to climate change and novel insights into species adaptive responses to shifting seasonal cues across different elevational gradients in their native range.

animal behavior and cognition↗