TaxHarmonizeR: A data-driven taxonomic harmonisation tool for ecological datasets
Abstract
Taxonomic ambiguity, arising from mixed levels of taxonomic identification within ecological datasets, can bias biodiversity estimates and ecological inference. Existing harmonisation approaches typically apply a single decision rule across all taxa, often sacrificing either taxonomic precision or the retention of abundance or occurrence data. Here, we present TaxHarmonizeR, an open-source R tool that implements a reproducible, data-driven workflow for taxonomic harmonisation. The method quantifies ambiguity within each taxonomic lineage as the ratio of occurrences or abundance between a focal taxon and all lower-levels of identification within the same lineage (focal-offspring ratio) and applies a user-defined Taxonomic Ambiguity Threshold (TAT) to determine whether taxa should be aggregated, retained, or deleted. Unlike conventional approaches, TaxHarmonizeR allows harmonisation decisions to vary among taxonomic lineages according to their observed level of ambiguity, thereby maximising taxonomic precision while ensuring that each lineage is represented only once in the harmonised dataset. We illustrate the workflow using a freshwater macroinvertebrate dataset from the Rhône River and evaluate the influence of different TAT values on taxonomic richness, occurrence / abundance retention, and α-, β-, and γ-diversity. Harmonisation had marked effects on biodiversity metrics, demonstrating that taxonomic ambiguity can substantially influence ecological interpretations and highlighting the importance of selecting an appropriate harmonisation strategy. By coupling lineage-specific harmonisation with sensitivity analyses across alternative ambiguity thresholds, TaxHarmonizeR provides a flexible and transparent framework that can be adapted to diverse ecological datasets and research objectives, while allowing users to assess the sensitivity of downstream results to the choice of TAT.
Related
- SeCAT (Sequence Consensus Amplicon Trimming): a simulation-informed Nextflow pipeline for cross-study harmonisation of 16S rRNA metabarcoding datasets
- TRIDENT (Taxonomic Resolution and IDentification using Environmental dNa Traces): An Optimized Algorithm for Vertebrate Taxonomic Assignments in eDNA Metabarcoding, Integrating Molecular, Taxonomic, and Ecological Criteria
- Ten rules for robust taxonomy in ecological datasets
- A Community-Driven Single-Cell PBMC Reference Integrating Landmark Datasets Spanning Health and Disease
- Open Datasets and AI/GenAI-Driven Computer Vision for Precision Livestock Farming: A Bidirectional Review
BibTeX
@misc{villsen2026taxharmonizer,
title = {{TaxHarmonizeR: A data-driven taxonomic harmonisation tool for ecological datasets}},
author = {Kurt Villsen and Maxence Forcellini and Johan Pansu and David Eme and Maxime Logez},
howpublished = {HAL},
year = {2026},
month = sep,
url = {https://hal.inrae.fr/hal-05766542v1},
note = {CC BY-NC-ND 4.0},
}
RIS
TY - UNPB
TI - TaxHarmonizeR: A data-driven taxonomic harmonisation tool for ecological datasets
AU - Kurt Villsen
AU - Maxence Forcellini
AU - Johan Pansu
AU - David Eme
AU - Maxime Logez
PY - 2026
DA - 2026/09/28
PB - HAL
UR - https://hal.inrae.fr/hal-05766542v1
AB - Taxonomic ambiguity, arising from mixed levels of taxonomic identification within ecological datasets, can bias biodiversity estimates and ecological inference. Existing harmonisation approaches typically apply a single decision rule across all taxa, often sacrificing either taxonomic precision or the retention of abundance or occurrence data. Here, we present TaxHarmonizeR, an open-source R tool that implements a reproducible, data-driven workflow for taxonomic harmonisation. The method quantifies ambiguity within each taxonomic lineage as the ratio of occurrences or abundance between a focal taxon and all lower-levels of identification within the same lineage (focal-offspring ratio) and applies a user-defined Taxonomic Ambiguity Threshold (TAT) to determine whether taxa should be aggregated, retained, or deleted. Unlike conventional approaches, TaxHarmonizeR allows harmonisation decisions to vary among taxonomic lineages according to their observed level of ambiguity, thereby maximising taxonomic precision while ensuring that each lineage is represented only once in the harmonised dataset. We illustrate the workflow using a freshwater macroinvertebrate dataset from the Rhône River and evaluate the influence of different TAT values on taxonomic richness, occurrence / abundance retention, and α-, β-, and γ-diversity. Harmonisation had marked effects on biodiversity metrics, demonstrating that taxonomic ambiguity can substantially influence ecological interpretations and highlighting the importance of selecting an appropriate harmonisation strategy. By coupling lineage-specific harmonisation with sensitivity analyses across alternative ambiguity thresholds, TaxHarmonizeR provides a flexible and transparent framework that can be adapted to diverse ecological datasets and research objectives, while allowing users to assess the sensitivity of downstream results to the choice of TAT.
ER -