Zhu Chaodong Research Group and other NSR co-published articles proposing six steps to establish a technical knowledge base for future taxonomic work

Species are the main units
we use to study, understand and protect the natural world.
So how we define species has knock-on effects in various fields far beyond biology
.
Despite this importance, the debate over the concept of species continues to this day and has implications for the organisms we understand, study and protect
.
Given the abundance of life on Earth, achieving a universally accepted concept of species seems impossible
.
If taxonomists wait for a universally accepted concept that delays their work describing and defining species, we risk extinction before species are
named.

Thankfully, taxonomists have been describing species
for centuries.
Therefore, a generally accepted definition of species is clearly not a prerequisite
for taxonomic research.
This is reflected in De Queiroz's (2007) Unified Concept of Species: which separates the conceptual basis of the term species from the operational methods that define them
.
As a result, taxonomists can largely escape the daunting philosophical basis of "species", but it is necessary to develop standardized definitions, optimize our operational basis and increase transparency, so that we can communicate clearly and enable them to effectively identify species
.
Funding, unpublished data, etc.
, tend to prevent multiple data types from being used, even if they are more accurate (and possibly automated)
for identification when used together.
To further standardize efforts across taxonomies and specific biological groups, we need comparable data and methodologies
.

To truly maximize the value of diverse data sources, we must make every effort to link specimens to formal species names defined by taxonomists and unify biological research, rather than rigid operations
.
Otherwise, we run the risk of starting the stove and building a parallel system of this kind
.
This is largely out of touch with centuries of research and conservation work by organizations such as the Union for Conservation of Nature (IUCN) and the Convention on International Trade in Endangered Species (CITES), which often operate
legally using species names.

Authoritative species inventories, image databases (especially type specimens), and DNA reference libraries are key
to this puzzle.
Therefore, specimens should be used as much as possible for 1) molecular data, 2) morphological data (images, etc.
) and traits, 3) geographic information, and 4) forensic history
.
Without these, we cannot determine the difference between known and unknown species, nor can we effectively build and share our knowledge
of the species described.
In addition, in order to reliably define species and assess their conservation status, we must take full account of variation
.
A single collection of each species is still insufficient for many purposes
.
To more effectively protect some of the richest taxa, active collaboration
across domains is required.

1.
Credentials and Standards
.
All research involving specimen collections, as well as cross-reference organizations for DNA, must maintain a minimum of credentials
in the collection.
The resulting specimens, as well as historically valuable specimens such as types, should be imaged and made public
.
Similarly, taxonomic revisions should include morphological, numeral, and other available evidence, including public data that is validated or flagged as problematic, to eliminate misidentification
.
Within the taxa, standards for the specific tissues to be preserved, the expected harmonized measurement and imaging parameters should be
agreed.

2.
Data standards and cross-operability.

The preservation of data on new specimens should require an evaluator field to indicate taxonomic authority as a measure of reliability (including how they are identified, which is particularly valuable for reproducibility).

Making the version and history traceable will enable multiple experts to endorse the results or mark them as problematic
.
Similarly, location-to-size data
at the appropriate scale should be mandatory.
When a new sequence is uploaded to the molecular library, it should be automatically generated as a specially tagged record on the GBIF (traceable to existing records and cross-platform versions by a unique identifier; The initial link has been established, see

3.
Single data system
.
We envisage a unified platform with a centralized system that integrates with existing databases and includes other different data types, such as relevant literature, conservation status and conservation status, applied species concepts, etc.
(Figure 1; February 2020).

Some repositories have tried this, but it's still in its infancy at best
.
At the heart of this is a unified list of species, such as the Catalogue of Life, but differences based on different species concepts should be guided by taxonomists (starting with the World List of Marine Species).

4
.
Species concept demonstration.
The description of the new species must be as comprehensive as possible, integrating all available data
.
Since these data may diverge in species descriptions (e.
g.
, DNA and morphology), we must clearly state the rationale for the naming of new species, as well as the lines of evidence used elsewhere in the revision, including how they are integrated
.
This will greatly improve the transparency and reproducibility
of taxonomy.

5.
Improved automation of the entire process
.
Universal workflows can be created by combining common sequence data and species definition methods and can be hosted online, just like BOLD's short DNA barcode BIN system
.
For many other data types, a similar common data system is still not possible
.
Potential new species can be automatically identified from this data, enabling a comprehensive taxonomic approach based on a growing library of references, utilizing validated historical specimen images and DNA sequences
.

6.
Representative sequences
of unknown populations.
As a preliminary step before full species description, a "pattern" sequence (similar to NCBI RefSeqs)
can be specified.
Multiple markers
are preferred because of the potential for problems with a single marker (pseudogenes, incomplete lineage classification).
These preliminary "species" can be used for conservation assessment and restoration until a formal species description
is available.

Most steps are premised on funding, but it is equally important to strengthen support and more relevant occupational assessments to facilitate data generation and species description
.
To support the recent push for open data, funding agencies should stipulate that relevant molecular and morphological data should be disclosed
on a fixed timeline using standard methods (not "available upon reasonable request").
The National Institutes of Health recently made this recommendation and should be adopted
more widely.
Depending on the career stage and the country, these policies may be fairer to those affected by regulations as the scale of expectations continues to expand (February 2020).

By conducting equitable research across various fields and countries, we can really optimize and reap the synergies and outcomes
of biodiversity and biodiversity researchers.

This work is the result of long-term discussion and cooperation between relevant domestic and foreign experts such as the Institute of Zoology, Chinese Academy of Sciences, Institute of Botany, Chinese Academy of Sciences, Kunming Institute of Botany, Chinese Academy of
Sciences, China Agricultural University, Nanjing Agricultural University, Capital Normal University and Imperial College London.
Dr.
Michael Orr and Dr.
Douglas Chesters received continuous support from the PIFI Program of the Chinese Academy of Sciences and the National Natural Science Foundation of China.
The taxonomic work of Zhu Chaodong's research group continues to be funded
by the Key Laboratory of Animal Evolution and Systematics of the Chinese Academy of Sciences.
At the same time, Professor Wu Zhongyi and researcher Zhai Weiwei gave important encouragement
in the process of manuscript preparation.

Article link: https://doi.
org/10.
1093/nsr/nwac284

Figure 1: Workflow
for collaborating with population sample data.
Where the data comes from, where it's stored, and what it's
used for later.
The complete workflow is given on the left, and the exemplary species-level query
for the unified data system on the right.