The problem with(out) vouchers

Funk & al. • Point of View: Vouchers

TAXON — 12 Feb 2018: 3 pp.

POINT OF VIEW

The problem with(out) vouchers Vicki A. Funk,1,3 Robert Edwards1 & Sterling Keeley2 1 U.S. National Herbarium (US), Department of Botany, National Museum of Natural History, Smithsonian Institution, Washington, D.C. 20013-7012, U.S.A. 2 Joseph F. Rock Herbarium (HAW), Department of Botany, University of Hawaii, Manoa, Honolulu, Hawaii 96822, U.S.A. 3 Herbarium National Tropical Botanical Garden (PTBG), Kalaheo, Kauai, Hawaii 96741, U.S.A. Author for correspondence: Vicki A. Funk, [email protected] DOI https://doi.org/10.12705/671.1

History doesn’t repeat itself, but it does rhyme. (Mark Twain or possibly Anonymous; O’Toole, 2014)

Every few years science has a debate about Vouchers. The details are not the same but they follow a familiar pattern. An individual or a group (e.g., Minteer & al., 2014a) publishes a paper or note positing that voucher specimens are not necessary for one reason or another (e.g., driving rare organisms to extinction, making conservation difficult, takes too long, is too much trouble) that elicits responses (e.g., Krell & Wheeler, 2014; Rocha & al., 2014) and counter responses (Minteer & al., 2014b). Then things quiet down until a subsequent spark ignites a similar debate. In part, this is because in almost every new field of biology data gathering starts off without requiring vouchers but rather quickly a scientist—or group of scientists—realizes that vouchers are necessary for the science to be accurate and reproducible. In plant systematics we need only look at cytology where initially vouchers were not required for the publication of chromosome counts until 1948 when Rattenbury made a bold statement proposing “… to restrict publication of chromosome numbers to those collections which are documented by reliable vouchers …” and calling for a standardized table to be used when presenting the results. By 1960 this practice was universally adopted in cytology (e.g., Weeden & Powell, 1978 and references therein) and taking hold in other fields such as wood anatomy (Stern & Chambers, 1960) and palynology, in which vouchering environmental samples can be particularly problematic (Stafleu, 1967; Garg, 2010). Even with the explosion of sampling that accompanied the advent of molecular phylogenetics concerted attempts were made to formalize the provision of voucher tables, with journals directing editors and reviewers to enforce their inclusion amid strong community support (e.g., Funk & al., 2005, Pleijel & al., 2008). The arguments for comprehensive vouchering of experimental samples are strong: vouchers are an important reference point for much biological work and are integral to studies where knowing the identity of a sample is required (Funk & al., 2005; Pleijel & al., 2008). Accurate and checkable identifications are

needed for work in ecology, phylogenetics, taxonomy, conservation, and pharmacology to name just a few fields. Voucher citations function as the central link in a chain of evidence that stretches in two directions. They connect a tissue sample to extracted DNA, to a genetic sequence, to a morphometric matrix, to a spatial dataset and, in the other direction, they connect to a physical specimen that stands as a representative of a population or species (Astrin & al., 2013). If the link between these parts and the original voucher is broken, then the integrity of the data—and any publication based on those data—is compromised (Garbino & Nogueira, 2017). While this may seem obvious, reemphasis is particularly important in a time when interaction with whole specimens can be minimal and fragments of tissue are technically sufficient for genetic study. Indeed, in a recent survey Rabeler (2017) found that compliance with community expectations is poor. Comparison of the “guidelines to authors” from eight international journals with a strong molecular component, found that three made no mention of vouchers, four noted that vouchers had to be cited but gave no instructions, and only one (Taxon) provided a specific format. In many journals even when voucher tables are present they are often published online in the supplemental materials and likely unreviewed (Funk & al., 2005). It may be unsurprising then that a survey of voucher deposition for invertebrate studies (Turney & al., 2015) found only 35% reported doing so. And, of all papers published in 2006 in Molecular Phylogenetics and Evolution, one of the principal international journals for molecular phylogenetics, 54% had no specified vouchers (Pleijel & al., 2008). Even more concerning is the possibility that vouchers may not even exist for the samples used in many of these studies and even if they do, the reader does not know where to find them. Why are so many scientists unwilling or unable to collect vouchers? If the traditional voucher consisting of a plant or part of a plant is not available there are other parts of the plant (e.g., seeds, fruits, wood, etc.) that can be mounted on an herbarium sheet, stored in fluid, or placed in a special collection, as long as it allows identification of the species. If all else fails one can even use a series of photographs of the plant (see Funk & al., 2017 for more details), again as long as they allow identification of the species. Leaving aside neglect and naivety as reasons

Article history: Received: 22 Nov 2017 | returned for (first) revision: 20 Oct 2017 | (last) revision received: 21 Nov 2017 | accepted: 12 Oct 2017 | published: online fast track, 12 Feb 2018; in print and online issues, n/a || Associate Editor: Dirk C. Albach || © International Association for Plant Taxonomy (IAPT) 2018, all rights reserved

Version of Record (online fast track)

1


TAXON — 12 Feb 2018: 3 pp.

for failing to collect a voucher, there are circumstances where proper protocol (Funk & al., 2017) may be difficult or impossible. As touched on, species may be rare or endangered or there is insufficient material to permit the collection of a full specimen (Minteer & al., 2014a). Samples may be fragmented or destroyed during analysis (Metsger, 1999), and in some cases, especially with increased targeted collection of DNA-destined samples, full vouchers may be logistically difficult to obtain or transport. In such cases, future researchers can be faced with the problem of trying to validate the identity and provenance of a DNA sample with no physical voucher. While some of this has been stated in the publications cited above, here we wish to illustrate another reason for making sure tissue collections are vouchered. As we have recently learned from one of our own projects, using phylogenies to study evolution may be impossible without vouchers. Recently, during a global phylogenetic and biogeographic study involving taxa from around the world, the authors encountered some difficulty in locating voucher samples. Although most of the samples were collected with appropriate vouchers (here called Direct Vouchers), a number of Hawaiian taxa from the endemic genus Lipochaeta s.l. (Asteraceae/Compositae) were provided as DNA only. They were originally collected with local experts in the early 1990s and included multiple collections of species across all the islands. The DNA was extracted using a standard CsCl procedure (Kim & al., 1998). In the early 1990s laboratory methods were more laborious and the number of markers was necessarily small and as a result there was no resolution within Lipochaeta and the project was set aside. However, the CsCl method of DNA extraction produces stable DNA and indeed it proved an excellent source of DNA for our current study, over 20 years later. But, what could be done about the lack of vouchers? For these samples an unorthodox approach was taken to reestablish a link between the genetic sample and the population of plants it was collected from by nominating an Indirect Voucher. Unlike a Direct Voucher, which is made from the same source as the genetic material, an Indirect Voucher identifies a surrogate collection that is considered equivalent or representative of the original. An Indirect Voucher should fulfill four criteria: (1) a vouchered specimen identified as the same species, (2) collected from the same locality, (3) identified by a trusted worker, and for consistency, and if at all possible (4) gathered by the same collector as the non-vouchered sample. For the study in question many of the Hawaiian samples were also from rare or endangered species (at least today) and some were taken from living collections in Botanical Gardens that later turned out to lack a voucher. Given this, nominating an appropriate Indirect Voucher can be non-trivial. For example, the field notes state: “K-140: Lipochaeta integrifolia; cutting; 74c2094A; 27 April 1992; Accompanied by R. Hobdy” The leaf samples were collected at Waimea Falls Valley (Oahu) a public park (now Waimea Valley, WAI) from cultivated plants propagated from cuttings. No vouchers or photos were taken. However, the collector was accompanied by Robert Hobdy, a state botanist, who is intimately familiar with the flora, lending a high degree of confidence to the field 2

identification. In assigning an Indirect Voucher for this sample it is plausible that because these plants were from cultivated stock, vouchers may previously have been made. Hobdy collections are usually deposited at BISH, however we found no collections from Waimea Falls Valley at BISH. WAI herbarium (Waimea) records for the associated cutting number show that the parent plants came from “Bishop Museum, Atherton Halau” with Bishop Museum records showing that the plant in the Atherton Halau originated from the National Tropical Botanical Garden (PTBG) on Kauai. A search of BISH, HAW, PTBG, US, and WAI records and herbarium specimens did not reveal any vouchers of any of these cultivated plans and all three garden plants have since died. It was therefore necessary to designate an Indirect Voucher from the locality where these original collections were made. A subsequent database search and herbarium visit found only one collection in BISH for the same species from the same locality (Lau 1087, BISH 496231, 1987). We designated this collection as the Indirect Voucher. Designating an Indirect Voucher does not repair the chain of evidence broken by not having a Direct Voucher. It does however nominate a surrogate record that the authors believe may be equivalent. If the identity of a sample is later questioned, the Indirect Voucher offers a ready starting point for reexamination, including extraction and sequencing, for comparison with the original sample. Ideally, Indirect Vouchers should seldom be required as Direct Vouchers are part of proper sampling protocol (with exceptions for rare species, etc.), and we appeal to all workers collecting samples to be conscientious. We acknowledge that our solution is clearly not ideal, and relies on confidence in the identification of both the original sample and the Indirect Voucher, however it provides a reference point for the best concept that the authors of a study had for the identity of a sample at the time. Tying a species representative back to a common locality is also important for any biogeographical analyses, and in theory, sequencing of the Indirect Voucher could be used to test for a match if necessary in the future. At the same time the process of designating an Indirect Voucher can be time consuming and divert valuable resources (e.g. visiting distant herbaria) best used elsewhere, and even when done rigorously any degree of uncertainty in the identity of the original collection weakens the strength of any subsequent research. In our recent study of Melanthera s.l. (Compositae/ Asteraceae; Edwards & al., submitted) three of the species that are widespread were sampled from several islands in the Hawaiian chain. The phylogeny shows that these samples did not group by species. Where some of these collections misidentified? These species are fairly easy to identify, especially by a botanist from Hawaii. Is this pattern a true reflection of evolution? Are these multiple island species undergoing rapid speciation on each island? Or could it be hybridization? Without a Direct Voucher we cannot check the identification and we cannot take additional leaf material from the voucher to re-sequence. Even carefully selected Indirect Vouchers do not adequately replace proper collection of a Direct Voucher in the first instance, however they can be an important link in rebuilding a broken data chain.



TAXON — 12 Feb 2018: 3 pp.

In our example the lack of Direct Vouchers stymied further investigation into whether this is the result of misidentification, cryptic lineages, or recent radiation with incomplete lineage sorting. So these potentially interesting evolutionary questions cannot be addressed. Yes, vouchers are important for documenting the location of the collection and the correct name but they are also necessary for making sure the phylogeny can be used to study evolution. Finally, almost as bad as failing to collect a voucher, is not publishing voucher meta-data in studies that use them. While the case for hard-copy publication of voucher lists is strong (Funk & al., 2005) there are multiple repositories for lodging electronic files of voucher lists, with non-perishable urls and doi if they cannot be accommodated in the body of a manuscript or the journal-hosted supplementary material. There is really no excuse for not making voucher information available.

AUTHOR CONTRIBUTIONS VAF suggested the manuscript; VAF and RE drafted and read several versions of the manuscript; SK provided original data and VAF finalized the article. — VAF, http://orcid.org/0000-0002-7975-1450

ACKNOWLEDGEMENTS We thank the staff of several herbaria and botanical gardens for their assistance and use of the data and collections: Bernice Pauahi Bishop Museum (BISH), University of Hawaii, Manoa (HAW), National Tropical Botanical Garden (PTBG), U.S. National Herbarium, Smithsonian Institution (US), and Waimea Valley Park (WAI).

LITERATURE CITED Astrin, J.J., Zhou, X. & Misof, B. 2013. The importance of biobanking in molecular taxonomy, with proposed definitions for vouchers in a molecular context. ZooKeys 365: 67–70. https://doi.org/10.3897/zookeys.365.5875 Funk, V.A., Hoch, P.C., Prather, L.A. & Wagner, W.L. 2005. The importance of vouchers. Taxon 54: 127–129. https://doi.org/10.2307/25065309 Funk, V.A., Gostel, M., Devine, A., Kelloff, C., Wurdack, K., Tuccinardi, C., Radosavljevic, A., Peters, M. & Coddington, J. 2017. Guidelines for collecting vouchers and tissues intended for genomic work (Smithsonian Institution): Botany Best Practices. Biodivers. Data J. 5: ee11625 [38 pages including Appendices]. https://doi.org/10.3897/BDJ.5.e11625 Garbino, G.T. & Nogueira, M.R. 2017. On the mammals collected by Friedrich Sellow in Brazil and Uruguay (1814–1831), with special reference to the types and their provenance. Zootaxa 4221: 172–190. https://doi.org/10.11646/zootaxa.4221.2.2 Garg, A. 2010. Relevance of voucher specimens in palynological publications. Curr. Sci. 99: 996. Kim, H.-G., Keeley, S.C., Vroom, P.S. & Jansen, R.K.. 1998. Molecular evidence for an African origin of the Hawaiian endemic Hesperomannia (Asteraceae). Proc. Natl. Acad. Sci. U.S.A. 95: 15440–15445. https://doi.org/10.1073/pnas.95.26.15440 Krell, F.T. & Wheeler, Q.D. 2014. Specimen collection: Plan for the future. Science 344: 815–816. https://doi.org/10.1126/science.344.6186.815

Metsger, D.A. 1999. Destructive sampling and information. Pp. 301–314 in: Metsger, D.A. & Byers, S.C. (eds.), Managing the modern herbarium: An interdisciplinary approach. Washington, D.C.: Society for the Preservation of Natural History Collections; Ontario: Royal Ontario Museum. Minteer, B.A., Collins, J.P., Love, K.E. & Puschendorf, R. 2014a. Avoiding (re)extinction. Science 344: 260–261. https://doi.org/10.1126/science.1250953 Minteer, B.A., Collins, J.P. & Puschendorf, R. 2014b. Specimen collection: An essential tool—Response. Science 344: 816. https://doi.org/10.1126/science.344.6186.816-a Pleijel, F., Jondelius, U., Norlinder, E., Nygren, A., Oxelman, B., Schander, C., Sundberg, P. & Thollesson, M. 2008. Phylogenies without roots? A plea for the use of vouchers in molecular phylogenetic studies. Molec. Phylogen. Evol. 48: 369–371. https://doi.org/10.1016/j.ympev.2008.03.024 Rabeler, R.K. 2017. Making molecular specimen vouchers more accessible. Taxon 66: 537–538. https://doi.org/10.12705/663.1 Rattenbury, J.A. 1948. Chromosome number publication. Madroño 9: 257–258. Rocha, L.A., Aleixo, A., Allen, G. Almeda, F., Baldwin, C.C., Barclay, M.V.L., Bates, J.M., Bauer, A.M., Benzoni, F., Berns, C.M., Berumen, M.L., Blackburn, D.C., Blum, S., Bolaños, F., Bowie, R.C.K., Britz, R., Brown, R.M., Cadena, C.D., Carpenter, K., Ceríaco, L.M., Chakrabarty, P., Chaves, G., Choat, J.H., Clements, K.D., Collette, B.B., Collins, A., Coyne, J., Cracraft, J., Daniel, T., de Carvalho, M.R., de Queiroz, K., Di Dario, F., Drewes, R., Dumbacher, J.P., Engilis, A., Jr., Erdmann, M.V., Eschmeyer, W., Feldman, C.R., Fisher, B.L., Fjeldså, J., Fritsch, P.W., Fuchs, J., Getahun, A., Gill, A., Gomon, M., Gosliner, T., Graves, G.R., Griswold, C.E., Guralnick, R., Hartel, K., Helgen, K.M., Ho, H., Iskandar, D.T., Iwamoto, T., Jaafar, Z., James, H.F., Johnson, D., Kavanaugh, D., Knowlton, N., Lacey, E., Larson, H.K., Last, P., Leis, J.M., Lessios, H., Liebherr, J., Lowman, M., Mahler, D.L., Mamonekene, V., Matsuura, K., Mayer, G.C., Mays, H., Jr., McCosker, J., McDiarmid, R.W., McGuire, J., Miller, M.J., Mooi, R., Mooi, R.D., Moritz, C., Myers, P., Nachman, M.W., Nussbaum, R.A., Ó Foighil, D., Parenti, L.R., Parham, L.F., Paul, E., Paulay, G., Pérez-Emán, J., Pérez-Matus, A., Poe, S., Pogonoski, P., Rabosky, D.L., Randall, J.E., Reimer, J.D., Robertson, D.R., Rödel, M.-O., Rodrigues, M.T., Roopnarine, P., Rüber, L., Ryan, M.J., Sheldon, F., Shinohara, G., Short, A., Simison, W.B., Smith-Vaniz, W.F., Springer, V.G., Stiassny, M., Tello, J.G., Thompson, C.W., Trnski, T., Tucker, P., Valqui, T., Vecchione, M., Verheyen, E., Wainwright, P.C., Wheeler, T.A., White, W.T., Will, K., Williams, J.T., Williams, G., Wilson, E.O., Winker, K., Winterbottom, R. & Witt, C.C. 2014. Specimen collection: An essential tool. Science 344: 814–815. https://doi.org/10.1126/science.344.6186.814 Stafleu, F.A. 1967. Palynology, nomenclature and terminology. Rev. Palaeobot. Palynol. 3: 15–26. https://doi.org/10.1016/0034-6667(67)90035-8 Stern, W.L. & Chambers, K.L. 1960. The citation of wood specimens and herbarium vouchers in anatomical research. Taxon 9: 7–13. https://doi.org/10.2307/1217349 O’Toole, G. 2014. History does not repeat itself, but it rhymes. Quote Investigator: Exploring the origins of quotations. https://quoteinvestigator.com/2014/01/12/history-rhymes/ (accessed 12 Oct 2017). Turney, S., Cameron, E.R., Cloutier, C.A. & Buddle, C.M. 2015. Nonrepeatable science: Assessing the frequency of voucher specimen deposition reveals that most arthropod research cannot be verified. PeerJ 3: e1168. https://doi.org/10.7717/peerj.1168 Weedin, J.F. & Powell, A.M. 1978. Chromosome numbers in Chihuahuan Desert Cactaceae. Trans-Pecos Texas. Amer. J. Bot. 65: 531–537. http://www.jstor.org/stable/2442586


3