[Paper Review] A continental-scale dataset of ground beetles with high-resolution images and validated morphological trait measurements

S M Rayeed, Mridul Khurana+19 more|arXiv (Cornell University)|Jan 14, 2026

Environmental DNA in Biodiversity Studies0 citations

TL;DR

The paper presents a multimodal, FAIR-compliant dataset of over 13,200 NEON ground beetle specimens from 30 sites across the continental US and Hawaii, with high-resolution images and validated digital measurements of elytra length and width. Digital trait extraction is validated against manual measurements with sub-millimeter precision, enabling AI-driven analyses of carabid morphology and ecology at continental scales.

ABSTRACT

Despite the ecological significance of invertebrates, global trait databases remain heavily biased toward vertebrates and plants, limiting comprehensive ecological analyses of high-diversity groups like ground beetles. Ground beetles (Coleoptera: Carabidae) serve as critical bioindicators of ecosystem health, providing valuable insights into biodiversity shifts driven by environmental changes. While the National Ecological Observatory Network (NEON) maintains an extensive collection of carabid specimens from across the United States, these primarily exist as physical collections, restricting widespread research access and large-scale analysis. To address these gaps, we present a multimodal dataset digitizing over 13,200 NEON carabids from 30 sites spanning the continental US and Hawaii through high-resolution imaging, enabling broader access and computational analysis. The dataset includes digitally measured elytra length and width of each specimen, establishing a foundation for automated trait extraction using AI. Validated against manual measurements, our digital trait extraction achieves sub-millimeter precision, ensuring reliability for ecological and computational studies. By addressing invertebrate under-representation in trait databases, this work supports AI-driven tools for automated species identification and trait-based research, fostering advancements in biodiversity monitoring and conservation.

Motivation & Objective

• Address the under-representation of invertebrates in global trait databases by producing a publicly accessible, high-resolution image and trait dataset for ground beetles (carabids).
• Enable automated trait extraction and AI-driven analyses by providing validated digital measurements (elytral length and width) with transparent error metrics.
• Integrate morphological data with NEON environmental datasets to study trait-environment interactions across diverse ecosystems at continental scales.

Proposed method

• Standardized pitfall trap sampling across 30 NEON sites spanning 81 NEON terrestrial sites and 20 ecoclimatic domains.
• High-resolution imaging of vial and pinned specimens using site-specific protocols to ensure consistency and metadata preservation.
• Morphological trait measurements via TORAS for pinned specimens (elytral length, basal pronotum width, maximum elytral width) with scale bar calibration.
• Notes from Nature workflow used for vial specimens to measure elytral length and elytral width on group images.
• Hybrid segmentation pipeline combining Grounding DINO for initial bounding boxes and CVAT for manual refinement to isolate individual specimens.
• Conversion of polyline lengths from pixel measurements to millimeters using scale bars; inter-annotator reliability checks for vial measurements.

Figure 1 : Map of NEON terrestrial sites across the United States of America. Each circle marks a NEON terrestrial site: black fill indicates sites where we imaged carabid samples. The inset shows the Bartlett Experimental Forest BART site in detail, illustrating NEON’s spatial sampling design acros

Experimental results

Research questions

• RQ1 Can a continental-scale dataset of ground beetles with high-resolution images and validated measurements support AI-driven trait extraction and species identification?
• RQ2 How do digitally measured morphological traits (elytral length and width) compare to manual measurements in terms of precision and reliability?
• RQ3 How can integrated morphological data with environmental covariates from NEON enable cross-taxon and trait-environment analyses in carabids?
• RQ4 What is the feasibility and reliability of automated segmentation and measurement workflows for large invertebrate image datasets?

Key findings

• The dataset includes digitally measured elytra length and width for over 13,200 NEON carabids across 30 sites.
• Digital trait extraction achieves sub-millimeter precision when validated against manual measurements.
• Pinned specimens were imaged with standardized protocols using Canon EOS 7D and macro lenses, with detailed imaging parameters documented.
• Vial specimens provided two-trait measurements (elytral length and width) via Notes from Nature, with inter-annotator agreement assessed for elytral length but width variability led to exclusion of vial elytral width from final analyses.
• A hybrid segmentation workflow (Grounding DINO + CVAT) enables scalable isolation of individual specimens from complex group images.
• The dataset adheres to FAIR principles and integrates with NEON environmental data for trait-environment analyses.

Figure 2 : Workflow for standardized ground beetle collection and processing at NEON sites. The process begins with pitfall trap setup ( 69 ) and proceeds through the collection of captured organisms. Specimens are filtered to separate carabids from bycatch. If identified as a beetle, the specimen i