The Trouble with Triplets in Biodiversity Informatics: A Data-Driven Case against Current Identifier Practices

The biodiversity informatics community has discussed aspirations and approaches for assigning globally unique identifiers (GUIDs) to biocollections for nearly a decade. During that time, and despite misgivings, the de facto standard identifier has become the “Darwin Core Triplet”, which is a concatenation of values for institution code, collection code, and catalog number associated with biocollections material. Our aim is not to rehash the challenging discussions regarding which GUID system in theory best supports the biodiversity informatics use case of discovering and linking digital data across the Internet, but how well we can link those data together at this moment, utilizing the current identifier schemes that have already been deployed. We gathered Darwin Core Triplets from a subset of VertNet records, along with vertebrate records from GenBank and the Barcode of Life Data System, in order to determine how Darwin Core Triplets are deployed “in the wild”. We asked if those triplets follow the recommended structure and whether they provide an easy and unambiguous means to track from specimen records to genetic sequence records. We show that Darwin Core Triplets are often riddled with semantic and syntactic errors when deployed and curated in practice, despite specifications about how to construct them. Our results strongly suggest that Darwin Core Triplets that have not been carefully curated are not currently serving a useful role for relinking data. We briefly consider needed next steps to overcome current limitations.     

Community Next Steps for Making Globally Unique Identifiers Work for Biocollections Data

Biodiversity data is being digitized and made available online at a rapidly increasing rate but current practices typically do not preserve linkages between these data, which impedes interoperation, provenance tracking, and assembly of larger datasets. For data associated with biocollections, the biodiversity community has long recognized that an essential part of establishing and preserving linkages is to apply globally unique identifiers at the point when data are generated in the field and to persist these identifiers downstream, but this is seldom implemented in practice. There has neither been coalescence towards one single identifier solution (as in some other domains), nor even a set of recommended best practices and standards to support multiple identifier schemes sharing consistent responses. In order to further progress towards a broader community consensus, a group of biocollections and informatics experts assembled in Stockholm in October 2014 to discuss community next steps to overcome current roadblocks. The workshop participants divided into four groups focusing on: identifier practice in current field biocollections; identifier application for legacy biocollections; identifiers as applied to biodiversity data records as they are published and made available in semantically marked-up publications; and cross-cutting identifier solutions that bridge across these domains. The main outcome was consensus on key issues, including recognition of differences between legacy and new biocollections processes, the need for identifier metadata profiles that can report information on identifier persistence missions, and the unambiguous indication of the type of object associated with the identifier. Current identifier characteristics are also summarized, and an overview of available schemes and practices is provided.     

Rapid enhancement of biodiversity occurrence records using unconventional specimen data

Distributions of taxa across time and space are central to understanding biodiversity and biotic change, yet currently available occurrence data, drawn from biodiversity specimen records and observational datasets, are often insufficient to answer many driving questions. Records of “associated taxa,” taxa co-occurring with a specimen at the time and place of collection, have the potential to fill data gaps and expand the spatiotemporal scope of current occurrence records. I developed a method to extract associated taxon records from 84,328 digitized specimen records and examined the potential of these data to improve the quantity and quality of existing species occurrence data. Adding associated taxon records increased the size of the test dataset by 18.5%, spanned multiple decades (1937–2016), and potentially extended the known range of 217 taxa in Florida and up to 1500 taxa in the United States, demonstrating the capacity of these records to deepen our understanding of changes in the distributions of taxa on Earth. These results suggest that increased attention to documenting associated taxa could be a promising way to maximize the impact of every collecting event.     

The PhyloCode: a critical discussion of its theoretical foundation

The definition of taxon names as formalized by the PhyloCode is based on Kripke's thesis of "rigid designation" that applies to Millian proper names. Accepting the thesis of "rigid designation" into systematics in turn is based on the thesis that species, and taxa, are individuals. These largely semantic and metaphysical issues are here contrasted with an epistemological approach to taxonomy. It is shown that the thesis of "rigid designation" if deployed in taxonomy introduces a new essentialism into systematics, which is exactly what the PhyloCode was designed to avoid. Rigidly designating names are not supposed to change their meaning, but if the shifting constitution of a clade is thought to cause a shift of meaning of the taxon name, then the taxon name is not a "rigid designator". Phylogenetic nomenclature either fails to preserve the stability of meaning of taxon names that it propagates, or it is rendered inconsistent with its own philosophical background. The alternative explored here is to conceptualize taxa as natural kinds, and to replace the analytic definition of taxon names by their explanatory definition. Such conceptualization of taxa allows taxon names to better track the results of ongoing empirical research. The semantic as well as epistemic gain is that if taxon names are associated with natural kind terms instead of being proper names, the composition of the taxon will naturally determine the meaning of its name.
© The Willi Hennig Society 2006.     

A review of software tools for spell‐checking taxon names in vegetation databases

Correct spelling of taxon names in vegetation databases is a fundamental prerequisite for many data processing steps. However, manual detection and correction of spelling mistakes is inefficient, prone to errors and non‐reproducible, especially when scanning large databases. Here, I review six software tools that spell‐check taxon names in vegetation databases: (1) the Global Names Resolver, (2) the Interim Register of Marine and Nonmarine Genera, (3) the Taxonomic Name Resolution Service and R packages (4) Plantminer, (5) Taxonstand and (6) tpl. In particular, I test their capacity to spell‐check names across the taxonomic ranks and organism groups frequently encountered in vegetation data and challenge their ability to screen names from different geographic regions. Performance by software tools differed widely in these tests. Backed up by multiple reference lists, the Global Names Resolver emerged as the most versatile software tool. All software solutions currently suffer from some minor limitations, including an inability to spell‐check names of hybrid taxa. Furthermore, some spelling mistakes, by their nature, cannot be resolved unambiguously. Given these limitations, taxon names should be spell‐checked with software tools in a semi‐automatic rather than an automatic way.     

Ceci n'est pas une pipe: names, clades and phylogenetic nomenclature

An introduction is provided to the literature and to issues relating to phylogenetic nomenclature and the PhyloCode, together with a critique of the current Linnaean system of nomenclature. The Linnaean nomenclature fixes taxon names with types, and associates the names with ranks (genus, family, etc.). In phylogenetic nomenclature, names are instead defined with reference to cladistic relationships, and the names are not associated with ranks. We argue that taxon names under the Linnaean system are unclear in meaning and provide unstable group–name associations, notwithstanding whether or not there are agreements on relationships. Furthermore, the Linnaean rank assignments lack justification and invite unwarranted comparisons across taxa. On the contrary, the intention of taxon names in phylogenetic nomenclature is clear and stable, and the application of the names will be unambiguous under any given cladistic hypothesis. The extension of the names reflects current knowledge of relationships, and will shift as new hypotheses are forwarded. The extension of phylogenetic names is, therefore, clear but is associated to (and thus dependent upon) cladistic hypotheses. Stability in content can be maximized with carefully formulated name definitions. A phylogenetic nomenclature will shift the focus from discussions of taxon names towards the understanding of relationships. Also, we contend that species should not be recognized as taxonomic units. The term ‘species’ is ambiguous, it mixes several distinct classes of entities, and there is a large gap between most of the actual concepts and the evidence available to identify the entities. Instead, we argue that only clades should be recognized. Among these, it is useful to tag the smallest named clades, which all represent non-overlapping groups. Such taxa  – LITUs (Least Inclusive Taxonomic Units) – are distinguished from more inclusive clades by being spelled with lower-case initial letter. In contrast to species, LITUs are conceptually straightforward and are, like other clades, identified by apomorphies.     

Brassicaceae: Species checklist and database on CD-Rom

A species checklist has been prepared for the Brassicaceae (Cruciferae) family, providing the first updated list in over 70 years. The family, currently, includes 338 genera and 3709 species. The database contains approximately 14,000 taxonomic names (records). Taxon status and synonymy, taxon name, scientific authority, literature source and source verification, and the basionym are provided for each record. Electronic supplementary material to this article is available at and is accessible for authorized users.     

Towards a collaborative, global infrastructure for biodiversity assessment

Biodiversity data are rapidly becoming available over the Internet in common formats that promote sharing and exchange. Currently, these data are somewhat problematic, primarily with regard to geographic and taxonomic accuracy, for use in ecological research, natural resources management and conservation decision-making. However, web-based georeferencing tools that utilize best practices and gazetteer databases can be employed to improve geographic data. Taxonomic data quality can be improved through web-enabled valid taxon names databases and services, as well as more efficient mechanisms to return systematic research results and taxonomic misidentification rates back to the biodiversity community. Both of these are under construction. A separate but related challenge will be developing web-based visualization and analysis tools for tracking biodiversity change. Our aim was to discuss how such tools, combined with data of enhanced quality, will help transform today's portals to raw biodiversity data into nexuses of collaborative creation and sharing of biodiversity knowledge.     

Supporting taxonomic names in cell and molecular biology databases

Groups of organisms require labels or names to refer to them; however, the idea of a single static name index, although tempting for its simplicity, is both impractical and unadvisable as a basis for referring to organisms for which data has been collected and stored for analyses and sharing. The relevant issues are described and some of the challenges facing database researchers are discussed.     

Generation of a large gene/protein lexicon by morphological pattern analysis

The identification of gene/protein names in natural language text is an important problem in named entity recognition. In previous work we have processed MEDLINE documents to obtain a collection of over two million names of which we estimate that perhaps two thirds are valid gene/protein names. Our problem has been how to purify this set to obtain a high quality subset of gene/protein names. Here we describe an approach which is based on the generation of certain classes of names that are characterized by common morphological features. Within each class inductive logic programming (ILP) is applied to learn the characteristics of those names that are gene/protein names. The criteria learned in this manner are then applied to our large set of names. We generated 193 classes of names and ILP led to criteria defining a select subset of 1,240,462 names. A simple false positive filter was applied to remove 8% of this set leaving 1,145,913 names. Examination of a random sample from this gene/protein name lexicon suggests it is composed of 82% (+/-3%) complete and accurate gene/protein names, 12% names related to genes/proteins (too generic, a valid name plus additional text, part of a valid name, etc.), and 6% names unrelated to genes/proteins. The lexicon is freely available at ftp.ncbi.nlm.nih.gov/pub/tanabe/Gene.Lexicon.     

Phororhacoidea or Phorusrhacoidea? A note on the nomenclature of the “terror birds”

The Phorusrhacidae, a group of large terrestrial carnivorous birds mostly known from the Cenozoic of South America, are often placed in a superfamily, for which the taxon name Phororhacoidea Patterson, 1941 is frequently used. However, according to the International Code of Zoological Nomenclature, Phororhacoidea is not valid. The proper taxon name at the superfamily level is Phorusrhacoidea Ameghino, 1889.     

GENERIC NAMES OF NORTHERN AND SOUTHERN FUR SEALS (MAMMALIA: OTARIIDAE)

We have resolved a nomenclatural problem discovered during research on the northern fur seal that concerns the correct generic name for this taxon and for fur seals of the Southern Hemisphere. The unfortunate practice by some 19th-century authors to use names in their Latinized form but to date them from their first appearance as French common names led to the use of Arctocepbalus for southern fur seals when the name correctly applies to the northern fur seal, known today as Callorbinus ursinus . However, Arctocepbalus and Callorbinus are antedated by Otoes G. Fischer, 1817, which is the earliest available generic name for the fur seal of the northern Pacific. The earliest available generic name for southern fur seals is Halarctus Gill, 1866. To avoid the confusion that would result from replacing the currently used generic names with those required by strict adherence to the Principle of Priority, we have petitioned the International Commission on Zoological Nomenclature to preserve Arctocepbalus and Callorbinus for the southern and northern fur seals, respectively.     

Should taxon names be explicitly defined?

Overviews are provided for traditional and phylogenetic nomenclature. In traditional nomenclature, a name is provided with a type and a rank. In the rankless phylogenetic nomenclature, a taxon name is provided with an explicit phylogenetic definition, which attaches the name to a clade. Linnaeus’s approach to nomenclature is also reviewed, and it is shown that, although the current system of nomenclature does use some Linnaean conventions (e.g., certain rank-denoting terms, binary nomenclature), it is actually quite different from Linnaean nomenclature. The primary differences between traditional and phylogenetic nomenclature are reviewed. In phylogenetic nomenclature, names are provided with explicit phylogenetic definitions, whereas in traditional nomenclature names are not explicitly defined. In phylogenetic nomenclature, a name remains attached to a clade regardless of how future changes in phylogeny alter the clade’s content; in traditional nomenclature a name is not “married” to any particular clade. In traditional nomenclature, names must be assigned ranks (an admittedly arbitrary process), whereas in phylogenetic nomenclature there are no formal ranks. Therefore, in phylogenetic nomenclature, the name itself conveys no hierarchical information, and the name conveys nothing regarding set exclusivity. It is concluded that the current system is better able to handle new and unexpected changes in ideas about taxonomic relationships. This greater flexibility, coupled with the greater information content that the names themselves (i.e., when used outside the context of a given taxonomy or phytogeny) provide, makes the current system better designed for use by all users of taxon names.     

Least-inclusive taxonomic unit: a new taxonomic concept for biology

Phylogenetic taxonomy has been introduced as a replacement for the Linnaean system. It differs from traditional nomenclature in defining taxon names with reference to phylogenetic trees and in not employing ranks for supraspecific taxa. However, 'species' are currently kept distinct. Within a system of phylogenetic taxonomy we believe that taxon names should refer to monophyletic groups only and that species should not be recognized as taxa. To distinguish the smallest identified taxa, we here introduce the least-inclusive taxonomic unit (LITU), which are differentiated from more inclusive taxa by initial lower-case letters. LITUs imply nothing absolute about inclusiveness, only that subdivisions are not presently recognized.     

Cladistic information in phylogenetic definitions and designated phylogenetic contexts for the use of taxon names

A phylogenetic definition of a taxon name associates that name with a clade through its reference to a particular ancestor and all of its descendants. Depending on one's perspective, phylogenetic definitions name either clades on the one true, but unknown, phylogeny, or components on cladograms (clades on hypotheses regarding the true phylogeny). Phylogenetic definitions do not contain enough information to identify components without a reference cladogram. As a result, (1) if clades are equated with components on cladograms, a phylogenetic definition may associate a taxon name with different clades on different cladograms, and (2) the inclusiveness, diagnostic synapomorphies, and distribution in time and space of the clade with a particular name can differ markedly depending on the phylogenetic hypothesis one chooses to adopt. This potentially unacceptable lability in the clade to which a name refers can be avoided by using a taxon name in conjunction with only phylogenetic hypotheses on which specific taxa are related in a particular fashion. This designated phylogenetic context can be described in an n-taxon statement that would be appended to the phylogenetic definition. Use of the taxon name would be considered inappropriate in conjunction with cladograms on which the relationships contradict those in the n-taxon statement. Whereas phylogenetic definitions stabilize the meaning of taxon names, designated phylogenetic contexts would stabilize the usage of those names.     

Catalog of the archival collection of the Zoological Institute,Russian Academy of Sciences: Class osteichthyes (Bony Fishes), order perciformes,family Zoarcidae

The catalog contains data on the archival collection of fishes of the family Zoarcidae stored in the Laboratory of Ichthyology, Zoological Institute, Russian Academy of Sciences. For each stored item, the following information is given: valid Latin name; other names under which the species could be recorded; specified data of the original label; and category of type material. Appendices include checklists of all stored type species, as well as genera and species currently absent in the archival collection of the Zoological Institute.     

Parents Accidentally Substitute Similar Sounding Sibling Names More Often than Dissimilar Names

When parents select similar sounding names for their children, do they set themselves up for more speech errors in the future? Questionnaire data from 334 respondents suggest that they do. Respondents whose names shared initial or final sounds with a sibling’s reported that their parents accidentally called them by the sibling’s name more often than those without such name overlap. Having a sibling of the same gender, similar appearance, or similar age was also associated with more frequent name substitutions. Almost all other name substitutions by parents involved other family members and over 5% of respondents reported a parent substituting the name of a pet, which suggests a strong role for social and situational cues in retrieving personal names for direct address. To the extent that retrieval cues are shared with other people or animals, other names become available and may substitute for the intended name, particularly when names sound similar.