Related Experiment Video
Updated: Feb 19, 2026
13:03
Robust DNA Isolation and High-throughput Sequencing Library Construction for Herbarium Specimens
Published on: March 8, 2018
11.1K
Widespread sampling biases in herbaria revealed from large-scale digitization
Barnabas H Daru1, Daniel S Park1, Richard B Primack2
1Department of Organismic and Evolutionary Biology, Harvard University Herbaria, Harvard University, Cambridge, MA, 02138, USA.
The New Phytologist
|October 31, 2017
Summary
Herbarium records show widespread collection biases, including proximity to roads and overrepresentation by a few collectors. These biases vary regionally but consistently affect data used in ecological and evolutionary studies.
Area of Science:
- Botany
- Ecology
- Biodiversity Science
Background:
- Herbarium records are crucial for ecological and evolutionary research.
- Nonrandom collecting practices can introduce significant sampling biases into these records.
- Digitization of regional floras provides an opportunity to analyze these biases.
Purpose of the Study:
- To assess spatial, temporal, trait, phylogenetic, and collector biases in digitized herbarium records.
- To compare biases across three major digitized floras: Australia (AU), South Africa (SA), and New England, USA (NE).
- To identify shared and unique patterns of herbarium collection bias.
Main Methods:
- Analysis of approximately 5 million herbarium records from AU, SA, and NE.
- Quantification of biases related to geography, time, species traits, phylogeny, and collectors.
- Comparison of bias patterns across the three distinct regions.
Main Results:
- Identified shared biases: proximity to roads/herbaria, peak collections in spring/summer, underrepresentation of threatened species, and similar collection numbers for close relatives.
- Documented regional differences: overrepresentation of graminoids (SA, AU) and annuals (AU), and distinct peak collection decades (NE: 1910s, SA: 1980s, AU: 1990s).
- Observed that a small number of 'mega-collectors' contributed disproportionately to collections in all regions, potentially via 'founder effects'.
Conclusions:
- Herbarium collection data exhibit pervasive and complex biases that vary regionally.
- Understanding and accounting for these sampling biases is essential for accurate interpretation of herbarium-based studies.
- Future botanical collection strategies should aim to mitigate, rather than exacerbate, existing biases.
