More than a handful of dirt: sequence-based species description and the role of the ICN (a response to Seifert) A recent Editorial in IMA Fungus (Seifert 2017) is critical of sequence-based species description (Hawksworth et al. 2016). The Editorial raises more questions than it answers, concerning the nature of discovery, the minimum evidence that should be required to describe species, and the role of the International Code of Nomenclature for algae, fungi, and plants (ICN; McNeill et al. 2012) as an enforcer of taxonomic quality. The Editorial trivializes the work of molecular ecologists and paints a bleak picture of the future of taxonomy if sequence-based species description is adopted. Here, I address some of the questions raised in the Editorial and offer a more optimistic vision for the integration of molecular ecology and taxonomy. Responses to quoted elements (in italics) are below: CORRESPONDENCE 1. As stated by Hibbett et al. (2011), “... for information about unsequenced Yes, the knowledge that someone else molecular ecology is clearly the major arena taxa, including ecology, and geographic has already detected the same DNA of contemporary species discovery...” rather distributions. This newly formed represented by the specimen in hand than conventional taxonomy. My question team might request specimens from does change the picture. First, it means is whether this process actually discovers new herbaria or culture collections to obtain that the species was already discovered. species, or simply indicates that there are new sequences to check whether the newly The taxonomist who has the specimen species to be found? discovered sequences correspond to in hand should be delighted, not species that have already been described. threatened, by this knowledge, because To “discover” something is to become At some point, they might decide that the sequence may provide insights aware that it exists. It is possible to some of the new sequences warrant that can augment information based discover new species through analyses description as new species, but under on the specimen. Most importantly, of sequences as well as specimens. the current rules of nomenclature they the sequence provides information It is a different question to ask if would be unable to validly name their concerning phylogenetic relationships sequences (or specimens) provide discoveries, because they do not have of the species. The phylogeny may make sufficient information to warrant formal a physical type specimen. However, if it possible to predict characteristics that description. the ecologist had saved a portion of are not observable in the specimen, such their handful of dirt then that could as physiological attributes. Information 2. In modern ecology, when you have a serve as type material, as demonstrated about ecology and biogeography can substrate in your hand that contains DNA by the valid publication of Piromyces be derived from metadata about the sequences of a thousand species, half of them cryptodigmaticus (Kirk 2012; but see sequence, including sampling locality unknown, have you discovered 500 new species Tripp & Lendemer 2012). and source material. For example, if or have you picked up a handful of dirt? the specimen is a mushroom growing 3. Does the act of naming a sequence provide on soil and the sequence comes from A molecular ecologist does not simply new information that is not already inherent a root tip then it may be possible to pick up a handful of dirt. They travel in the sequence itself? I would say not. make an inference about the ecology to a sampling locality, collect dirt, of the species. With an environmental record metadata, extract DNA, perform I would say yes. The act of naming sequence in hand, the picture has details shotgun metagenomics or amplicon communicates the information that that might otherwise be invisible. sequencing, feed the sequences through someone thinks they have discovered a bioinformatics pipeline that compares a new species. A DNA sequence is 5. Is there any conceptual similarity between them to millions of other sequences, just a DNA sequence, whereas a name a species based on one specimen and a species make community comparisons using embodies a taxonomic hypothesis based on a few DNA sequences? tools such as UniFrac (Loupozone (including lineage information). One & Knight 2005), and deposit their might just as well ask if the act of This is a red herring. The conceptual new sequences in publicly-accessible naming provides new information that similarity, to the extent there is one, is of databases. This is usually where the is not inherent in a morphological the most general nature, concerning the process ends, without a determination description. nature of evidence that justifies formal that any new species have been naming. This question conflates two discovered. 4. Whether it is one specimen or a hundred, different issues: (1) whether a species How might the research proceed with a specimen in hand it seems clear should be described based on a single if there was a robust feedback loop that you have made a discovery. Does the observation, and (2) whether sequences between fungal taxonomy and knowledge that someone else has detected can serve as the type material. These are molecular ecology? In that alternate the same DNA in a different handful of not equivalent questions. The answer to universe, the ecologist would employ dirt really change the picture? There are no one does not automatically inform the a bioinformatics tool to place the new characters other than nucleotides, there is no answer to the other. sequences in a phylogenetic tree to seek differential ecology or behaviour attributable their closest relatives. The ecologist to the specific unknowns, unless they can be 6. Does a double standard exist, where our might then reach out to a taxonomist inferred in some way by information inherent historical practise [sic] allows (but is now who could consult relevant monographs in the genetic sequences. actively discouraging) what some perceive as (36) IMA FUNGUS low quality species descriptions with an old description based on single collections, discovered – within the groups on CORRESPONDENCE technology, while preventing what some would although that would be consistent which they specialize (they should consider a higher quality of species description with a role as a “quality assurance already be doing this). They would then using a new technology? mechanism”. name only the species for which they In my view, the function of the think there is adequate evidence. For Yes, there is a double standard. A species ICN is to dictate the terms for valid an example, see the thoughtful (not known only from a single collection publication of names, not to assure at all banal or trivial) description of can be validly named, even if the type taxonomic accuracy. “Quality” in fungal Hawksworthiomyces sequentia by De specimen has no observable characters taxonomy is “enforced” by editors Beer et al. (2016). that differentiate it from any other and reviewers, and by the scientific species, has been reduced to dust, or community, particularly authors of 10. When should we describe species? even lost, but a species known by any monographs and other taxonomic number of independent environmental compilations. If taxonomists disapprove The answer to the question posed in the sequences, with metadata, cannot be of sequence-based species, then they title of the Editorial is that taxonomists validly named. The absurdity of this should exclude them from their should describe species whenever they double standard will become all the monographs and checklists. However, think that it is warranted based on the more evident as techniques such as the rules of nomenclature should not evidence at hand. Different taxonomists single-cell genomics become more prohibit other workers from formalizing will have different standards for widely used in ecological studies. hypotheses based on sequences as valid evidence that justifies species names with the protection of priority. description. One author might insist 7. Are the limitations of what we can By the same reasoning, the ICN on several collections and multi-gene determine about a species from a DNA should not prohibit species description phylogenies with Bayesian coalescent sequence more severe than what we can based on a single specimen (or require analyses. Another may think that a determine about a species when we have sequence data). single specimen without molecular only one specimen? If not, why are so many data is sufficient. A third may be happy journals reluctant to allow single species 9. What does it take to raise species description with multiple environmental sequences, descriptions based on morphology, but lining above banality, above trivia that could be backed up by metadata and phylogenetic up to publish controversial papers on DNA extracted by any child or by a machine? analyses, as per the proposed defines[sic] taxa that test the limits of the Do we want machine taxonomy in fungal Recommendations for sequence-based ICN? biology? From one perspective this seems typification by Hawksworth et al. like a paranoid question and, from another, (2016). Currently, the ICN would See (5) regarding single specimens vs. prescient. If DNA sequences comprise both permit the first two authors to formally sequences in species description. The the description and the type, it is a short step name their hypotheses, which would suggestion that “so many” journals to a pipeline that automatically describes then receive the protection of priority, are “lining up to publish” papers on and names the OTUs as species. The question but the third author would not have sequence-based species description is of machine-automated species description that option. This borders on scientific an exaggeration. Even the most casual is staring us in the face. Surely we should be chauvinism, and is inconsistent with the perusal of current literature shows that discussing it? principles of the ICN, which has never sequence-based species descriptions dictated the nature of evidence required are extremely rare (perhaps because Yes, this does sound paranoid.