So you actually need to work with the Herbarium Handbook 3rd Edition. Here is how it goes down.
I run a mid-size regional herbarium, roughly 40,000 specimens. We've been running on the Herbarium Handbook 3rd Edition workflow for about four years now, and it is what it is. It's not a software product — it's a standard and a guide, published by the International Association for Plant Taxonomy and managed through BMNHP (Berlin). Think of it less as a tool and more as a rulebook that a lot of institutions quietly adopted because doing anything else creates fragmentation between databases. The third edition shifted things from a purely manual format to a more structured data model that aligns closer with Darwin Core. That matters if you are planning to push your collection data to GBIF or other aggregators. The second edition was essentially a suggestion box. The third edition is closer to a spec sheet with implementation notes.
Herbarium Handbook 3rd Edition — what it actually covers
It covers five main areas: collection acquisition and accessioning, specimen preparation and mounting, preservation and storage, cataloguing and data management, and outreach/distribution. The cataloguing section is where most people hit walls. It defines the minimum metadata set for a usable herbarium record, but the minimum set is still bigger than most lab computers can handle without a decent relational database behind it. The core fields it requires include an accession number, collector name and number, collection date, location data (with coordinate precision), habitat description, determiner, and a voucher status. That last one trips people up. "Voucher" is not a date. It's a status flag indicating whether the specimen serves as a voucher for another publication or dataset. You will see this field misused constantly in publicly available datasets.
How I actually use it day to day
We use the Handbook's framework alongside a custom PostgreSQL schema that maps to its recommended fields but adds institutional extensions. The Handbook does not prescribe software. That is the trick people miss. It gives you the ontology, not the interface. So we built a thin ETL layer that pulls from our field data sheets and normalizes them into the Handbook structure before loading into the main catalog. For accessioning new material, the Handbook recommends a three-step verification: field data check, taxonomic confirmation, and duplicate marking. I used to do all three sequentially. It took about 90 minutes per batch of twenty specimens. Now I run field data and taxonomic confirmation in parallel because the duplicate check is mostly a visual scan. That cut our intake time to roughly twenty minutes per batch. Your mileage depends on whether you have a dedicated taxonomist on staff or you are doing the determinations yourself.
Get the Full Details
The thing nobody mentions about the 3rd edition
The biggest problem I ran into was with locality descriptions. The Handbook treats "locality" as a free-text field with optional coordinate refinement. In practice, about thirty percent of our older specimens have locality data that looks useful but falls apart under scrutiny. Vague descriptors like "near the river" or "mountain slope" cannot be georeferenced reliably. The Handbook does not solve this. It acknowledges it in a footnote and basically says handle it on your own. My workaround was building a lookup table of historical place names and known misidentifications against our existing collection. I cross-referenced old collector notes with modern GIS layers and flagged anything that didn't align. This took about six weeks for our oldest holdings. But once it was done, the georeferencing pass for GBIF submission went smoothly instead of collapsing under uncertainty codes.
Where the Handbook falls short
It does not address digital imaging standards well at all. The 3rd edition references high-resolution photography as a best practice but provides no specification for resolution, color calibration, or file formats. If you are building a digital herbarium or working with partners who need image data, you will need to supplement this with something like the CCE (Center for Plant Conservation) imaging guidelines or ABIS standards. The Handbook won't help you decide between TIFF and JPEG2000 for long-term archiving. It also does not cover digitization workflows. There is no section on barcode generation, scanning throughput, or quality control for digitized records. I found myself writing internal SOPs that filled these gaps rather than adapting the Handbook. That is fine. The Handbook is a reference, not a manual. But going in expecting it to tell you how to run a digitization pipeline will disappoint you.
Practical advice if you are implementing this
Start with the metadata field definitions in Chapter 4. Do not skip ahead to the outreach section. Most institutions I talk to treat the Handbook as something they pull from when they need a citation. That is backwards. The data structure is the useful part. Get your fields right first, then worry about displays and reports. Invest in a controlled vocabulary for habitat and substrate fields. The Handbook suggests using standard terms but does not provide a list. I ended up pulling from ENVIREG and GBIF's occurrence fields and mapping them to our own terms. This mapping work took about two weeks and saved us from having to clean up inconsistent entries every time a new collector joined. Inconsistent habitat descriptions are the fastest way to make a collection unusable for ecological analysis. Don't try to retrofit the Handbook onto an existing spreadsheet-based system. It was designed for relational databases. We tried it on a shared Google Sheet once. It looked fine for a month. Then three people edited different locality fields simultaneously and the dataset became internally inconsistent. Spreadsheets can handle the field list. They cannot handle the referential integrity the Handbook expects. Use a proper database or at least a structured file like SQLite.
Where to find it
The Herbarium Handbook 3rd Edition is available through the BMNHP website and via the International Institute of Tropical Agriculture's publications catalog. It is not open access in the strictest sense — you can read it, but redistribution is restricted. For academic institutions, there is usually a licensing path through the publisher. I can point you to the direct download page if you need the exact URL. It is publicly accessible for reading purposes. For anyone actually implementing this, expect the first three months to feel like you are rebuilding your cataloging process from scratch. It is. The Handbook forces you to confront assumptions you have been making silently. Some of those assumptions are wrong. Most of them are just undocumented. Once you get through the initial friction, the structure holds up. It is tedious, not broken.
