Open access

DNA barcoding the Lepidoptera inventory of a large complex tropical conserved wildland, Area de Conservacion Guanacaste, northwestern Costa Rica

Publication: Genome
28 April 2016


The 37-year ongoing inventory of the estimated 15 000 species of Lepidoptera living in the 125 000 terrestrial hectares of Area de Conservacion Guanacaste, northwestern Costa Rica, has DNA barcode documented 11 000+ species, and the simultaneous inventory of at least 6000+ species of wild-caught caterpillars, plus 2700+ species of parasitoids. The inventory began with Victorian methodologies and species-level perceptions, but it was transformed in 2004 by the full application of DNA barcoding for specimen identification and species discovery. This tropical inventory of an extraordinarily species-rich and complex multidimensional trophic web has relied upon the sequencing services provided by the Canadian Centre for DNA Barcoding, and the informatics support from BOLD, the Barcode of Life Data Systems, major tools developed by the Centre for Biodiversity Genomics at the Biodiversity Institute of Ontario, and available to all through couriers and the internet. As biodiversity information flows from these many thousands of undescribed and often look-alike species through their transformations to usable product, we see that DNA barcoding, firmly married to our centuries-old morphology-, ecology-, microgeography-, and behavior-based ways of taxonomizing the wild world, has made possible what was impossible before 2004. We can now work with all the species that we find, as recognizable species-level units of biology. In this essay, we touch on some of the details of the mechanics of actually using DNA barcoding in an inventory.


L’inventaire continu, en cours depuis 37 ans, des 15 000 espèces de lépidoptères qu’on estime se trouver au sein des 125 000 hectares terrestres de l’Area de Conservacion Guanacaste, dans le nord-ouest du Costa Rica, a permis de documenter plus de 11 000+ espèces au moyen de codes à barres de l’ADN, ainsi qu’au moins 6000+ espèces de chenilles capturées et plus de 2700+ espèces de parasitoïdes. L’inventaire a commencé au moyen de méthodes victoriennes et l’identification au niveau de l’espèce, avant d’adopter pleinement le codage à barres de l’ADN en 2004 pour l’identification des spécimens et la découverte d’espèces. Cet inventaire tropical au sein d’un réseau trophique complexe, multidimensionnel et extraordinairement riche en espèces a été réalisé en s’appuyant sur les services de séquençage fournis par le Centre canadien de codage à barres de l’ADN, les services informatiques de BOLD (« Barcode of Life Data Systems »), les outils majeurs développés par le Centre de génomique de la biodiversité à l’Institut de la biodiversité de l’Ontario. L’accès à tous ces outils a été assuré par service de messagerie et l’internet. Au fur et à mesure que l’information sur la biodiversité émerge de ces milliers d’échantillons inconnus et souvent très semblables, suite à leur transformation en produits utiles, il devient évident que le codage à barres de l’ADN, combiné aux méthodes taxonomiques traditionnelles fondées sur la morphologie, l’écologie, la micro-géographie et le comportement, permet de classifier la nature d’une manière qui n’était pas possible avant 2004. Il est maintenant possible de travailler avec toutes les espèces qui sont trouvées et d’en définir des unités reconnues à l’échelle des espèces. Dans cet essai, les auteurs abordent certains des détails de la mécanique qui rend possible l’utilisation du codage à barres de l’ADN pour réaliser un inventaire. [Traduit par la Rédaction]


Why conduct a biodiversity inventory of a large complex tropical conserved wildland? This inventory began in 1978 as a purely academic question about who are the caterpillars behind the hundreds of species of moths that came to a 25 W light bulb in 1978 in the dry forest of the then Parque Nacional Santa Rosa, today’s Sector Santa Rosa of Area de Conservacion Guanacaste (ACG) in northwestern Costa Rica (Janzen et al. 2009; Janzen and Hallwachs 2016) (Fig. 1). While our focus (by personal interest and ability) is on caterpillars and their adults, such an inventory necessarily includes the caterpillars’ parasitoids and food plants, the fellow travelers in the ACG trophic web of at least 375 000+ eukaryotic species that extends from Pacific coast dry forest to high elevation cloud forest and down into Caribbean rain forest. This ACG inventory is only a small piece of tropical biodiversity, but nonetheless includes tens of thousands of species.
Fig. 1.
Fig. 1. Contour (up to 2000 m) of Area de Conservacion Guanacaste (inside white boundary), ranging from Pacific dry-forested coastal plane to volcano-top cloud forest and down to Caribbean rain forest: Volcan Orosi (1400 m) on the left, Volcan Cacao (1500 m) more central, and Volcan Rincon de la Vieja – Volcan Santa Maria (2000–1800 m) on the right; 125 000 ha terrestrial, 43 000 ha marine. Image credit: Waldy Medina, Area de Conservacion Guanacaste.
This overview is offered by just its two authors, rather than co-authored by the literally hundreds of members of the taxasphere, DNA barcoding guild, biodiversity conservation and management guild, and many others, who have been major and minor contributors to the ACG inventory. This is quite simply because we do not have the months of time required to circulate this general description and incorporate the multitude of comments that such circulation will create. All of these persons have been and will be coauthors and acknowledged collaborators of portions of the inventory over time.
During the inventory’s 37-year ongoing travelogue, its technical purpose has remained the same: to know what they are, where they are, what they do, how to find them when you want them, and simultaneously, to get all of that information into the public internet domain for all sectors of society to reference and use. There are tens of thousands of ACG Lepidoptera, parasitoid, and plant species to find and make some sense of. But it has also necessarily evolved into addressing the broader social need of conservation of wild areas as well. This inventory is meant to be a living public library with open stacks. And we do this because such understanding and social integration is a necessary component of a conserved tropical wildland that is to have some chance of surviving indefinitely, albeit tattered and bruised, at relative peace with its society—local, national, and international (Janzen 1988, 1991, 1998, 1999, 2000a, 2000b; Janzen and Hallwachs 1994; Geddes 2015).
The ACG inventory of caterpillars and their adults, parasitoids, and food plants is carried out by a large team of on-site Costa Rican parataxonomists (Janzen et al. 1993, 2009; Janzen and Hallwachs 2011), who currently constitute a large portion of the “Programa de Parataxónomos” of ACG (ACG 2016), which is jointly funded and in-kind supported by ACG and the Guanacaste Dry Forest Conservation Fund (GDFCF 2016). These two conservation entities are in turn supported by a huge and institutionally diverse array of financial and sweat-equity supporters (see Acknowledgments). And this entire network is directly supported, guided, encouraged, and stimulated by several hundred volunteers, administrators, taxonomists, the taxasphere, institutions and others—neighboring, national, and international. To detail this network and its activities will require a 500-page book that attempts description of a constantly shifting and diverse sociology, economics, science, and politics—you can never photograph the same river twice.
Here, in reaction to the 6th International Barcode of Life Conference, global biennial symposium of the International Barcode of Life initiative held at the University of Guelph in August 2015, represented by this special issue and others, we sketch a few of the salient features and results associated with the insertion of DNA barcoding (Hebert et al. 2003; Janzen et al. 2005, 2009; Hajibabaei et al. 2006; Janzen and Hallwachs 2011; Janzen et al. 2012) into the long-running ACG Lepidoptera inventory. We make no attempt to review and compare the results with other on-going tropical biodiversity inventories (e.g., Bassett et al. 2004, 2015; Novotny and Miller 2014; Telfer et al. 2015; Miller 2015; Glover et al. 2016), as that would require an entire book-length treatment. The most salient overarching result of inserting DNA barcoding into the ACG inventory is to realize that it would be a seriously debilitating oversight to conduct a species-level arthropod biodiversity inventory of a large complex tropical biological system without having DNA barcoding and its anticipated offspring in the toolbox. Since Next Generation Sequencing (NGS) is still being explored for this ACG inventory (e.g., Shokralla et al. 2014; Gibson et al. 2014; Shokralla et al. 2015), this entire commentary relates only to DNA barcodes obtained by Sanger sequencing.


In March and September 2003, at Cold Spring Harbor Laboratory, Long Island, New York, we listened to Paul Hebert (Biodiversity Institute of Ontario, University of Guelph, Canada) present and defend the concept of adding DNA barcoding for specimen identification and species discovery to the taxasphere’s tool box (Hebert et al. 2003; Devan 2015), explicitly not for phylogenies or pushing more specimens into the museum-based taxasphere (defined in Janzen 1993). Through facilitation by the Smithsonian Institution and John Burns, the ACG’s core skipper butterfly (Hesperiidae) taxonomist, the inventory offered Hebert a leg from each of the 25-year accumulation of 484 oven-dried and museum-spread reared vouchers of “Astraptes fulgerator”, a common, showy, and well-known roadside Central American butterfly (Fig. 2), to test-drive the barcoding concept with a horribly complex fauna, but yet one of the few areas of ACG butterfly taxonomy assumed to be well understood. This ACG species was chosen because of the both morphology- and ecology-based suspicion that it was well more than one species, yet neither set of traits could consistently delineate these species among the ACG specimens.
This first example revealed 10 sympatric species (now 11 through the addition of one newly discovered by the inventory) of nearly identical but subtly variable adult butterflies (Fig. 2) hiding inside one scientific name that had been happily used for 225 years; this complex then sorted cleanly into barcode clusters that matched caterpillar color patterns, food plant clusters, and subtle adult colors on both the upper side and underside of their wings, although the adults cannot be distinguished by the usual taxonomist’s key character, their genitalia (Hebert et al. 2004). This result was followed by finding 32 (now 39) ACG species of highly host-specific tiny parasitic wasps (Braconidae, Microgastrinae), previously thought to be a “generalist” parasitoid of Hesperiidae caterpillars, that had been hiding under one morphology-based name, “Apanteles leucostigmus” (Smith et al. 2008), for more than a century; these wasps have subsequently been found to be distinguishable by overlooked morphological differences discovered by a taxonomist minutely scrutinizing the wasps’ shape and relative sizes of body and leg sculpture, color, and any other distinctive trait under the microscope, assisted by the barcodes and host records (Smith et al. 2012b; Rodriguez et al. 2013; Fernandez-Triana et al. 2014a). Each morphologically distinguished wasp species has its own species of hesperiid caterpillar hosts. Equally, 16 species of “generalist” tachinid parasitoid flies were found by barcoding to be actually 72 species (mostly specialists), only nine of which were truly generalists (Smith et al. 2006, 2007).
These results underlined the imperative of barcoding practically all of the inventory’s voucher specimens that had been saved since 1978. A major accumulated result is that the initial estimate of 9500 species of ACG Lepidoptera has been revised to 15 000, and the Canadian government, Biodiversity Institute of Ontario (BIO), Wege Foundation, and other donors have now funded the DNA barcoding (Sanger sequencing) of tens of thousands of 1–20-year-old oven-dried or ethanol-preserved insect legs. The results have driven much greater amounts of collecting, vouchering, and data-gathering than originally anticipated by the inventory; simultaneously, the inventory is now far more accurate and useful than it would have been.
In practical terms, an ACG insect leg is couriered to BIO, just as we used to ship our rolls of K25 film to Kodak. And we get back a highly usable result—the barcode and its analysis/use in BOLD (Ratnasingham and Hebert 2007; BOLD 2016)—without really understanding the guts of Kodak’s magic, nor having to sustain the administrative and dollar costs and headaches of maintaining such a biodiversity processing center. It is noteworthy that BIO and BOLD’s home, University of Guelph, is noted for its agricultural research and practical crop breeding. DNA barcoding is, indeed, a kind of Crop Science. The inventory has long found that with numerous very useful exceptions, some of which are outlined here, the DNA barcodes match extremely well the morphology-based species-level identifications that have been obtained through tens of thousands of hours of taxasphere time already invested in the ACG inventory since 1978, and for centuries before. But simultaneously the barcodes have revealed many hundreds of overlooked species and undescribed species, and allowed the inventory itself to operate a “home-made” level of taxonomic operational2 accuracy that is generally ready to use decades earlier than if the taxasphere were to be able to examine the specimens more formally, describe them, and otherwise make use of them, if indeed it ever would or could be given the immense taxonomic challenge of tropical diversity.
Fig. 2.
Fig. 2. The 10-species Astraptes fulgerator species complex reared in Area de Conservacion Guanacaste from wild-caught caterpillars. Adult uppersides, from left to right, top down, Astraptes fruticibus*, 99-SRNP-399; Astraptes procrastinator*, 97-SRNP-5143; Astraptes obstupefactus*, 97-SRNP-1804; Astraptes boreas*, 95-SRNP-8046; Astraptes favilla, 06-SRNP-55033; Astraptes augeas, 06-SRNP-36607; Astraptes audax, 02-SRNP-29798; Astraptes inflatio*, 02-SRNP-20353; Astraptes viracocha, 00-SRNP-10977; Astraptes synecdoche*, 00-SRNP-10415. Underside, same butterflies, same order. Caterpillars, same species, same order, bottom center is last in the series. All ten species are distinguishable with certainty by their barcodes, as well as by a food plant+caterpillar color combination, and in a few cases, by subtle details in adult facies. * indicates that the six species occur within the same BIN: BOLD:AAA1584 (Fig. S1)2. Image credit: Daniel Janzen, University of Pennsylvania.
An enormous array of inventory and ecological questions can be addressed with only minimal and mutually iterative reinforcement of the initial input from the taxasphere. Is this wasp Braconidae or is it Ichneumonidae? Is this caterpillar Apateledidae or Bombycidae, Erebidae, or Noctuidae? Is this biology tagged with the family name “Pyralidae” in 1900 the same entity as is tagged with “Pyralidae” today? With such taxasphere guideposts in place, the inventory can, for example, majorly grasp the species-level community traits and social uses of 500+ species of formally unidentified crambid moths, facilitate any crambid taxonomist, and simultaneously offer a massive amount of collateral information to join with those same species that are today represented largely by decades-to-centuries old museum specimens blessed with only its morphology, a date of collection, place of collection, and who caught it. The potential for inventory-taxasphere mutualisms is enormous.
Within 1–6 months of the leg’s submission to BIO there was (and still is) a DNA barcode available to the inventory in BOLD, with its within-BOLD massive out-bound and in-house comparability, for us to join with what the bioinventory is telling us about food plant, parasitoids, appearance, and micro-within-ACG location, and simultaneously to gift this information outwards to other biodiversity users. As a tiny, but illustrative example of the latter, the species behind the newly designated 20 patronyms for the Costa Rican school children who are winners of an Area de Conservacion Tempisque conservation drawing contest (Kazmier 2015a) were all located by DNA barcoding of ACG inventory Malaise-trapped wasps and their simultaneous morphology-based generic placement and corroboration; the wasps are minute and hardly noticeable braconid microgastrine ACG parasitic Promicrogaster (Fernandez-Triana et al. 2016); all 150+ members of the ACG staff have an Apanteles patronym that is most easily recognized by its barcode and corroborated by its morphology, placement in by classical key, host caterpillar, and microgeographic habitat (Fernandez-Triana et al. 2014a); even the President of Costa Rica has his very own unique Pseudapanteles luisguillermosolisi (Fernandez-Triana et al. 2014b; Kazmier 2015b).
While the ACG barcoding began with barcoding adults reared from wild-caught caterpillars, and their reared parasitoids (Janzen et al. 2009), BIO’s insatiable craving for larger and more complex (largely tropical and non-Canadian) barcoding projects led to Hebert’s 2006 offer to also barcode the ACG wild-caught adult Lepidoptera as a way to speed a “total” inventory and demonstrate that DNA barcoding “works” for a large taxonomically complex fauna. (This speeds the inventory because only about 30% of ACG Lepidoptera has been found as caterpillars, despite 30 years of work by parataxonomists.) At that time, the US National Science Foundation was unwilling to support barcoding costs and Hebert’s funds were critically needed. Two very experienced caterpillar-focused parataxonomists were then repurposed to mentor apprentice parataxonomists to become a 4-member team to blanket light-trap ACG, expressly for adults to be barcoded and museum-deposited as vouchers. This team set out to re-capture the ACG moth fauna that had been light-trapped from 1978 to 1990. In 2006, those previous voucher specimens, residing in INBio’s national insect survey collection, were deemed to be too old for easy barcoding. This INBio collection is now donated to, and melded into, the Museo Nacional de Costa Rica, and reinforces the current ACG inventory as well as that of the entire country.
Today, the barcoding of rearing and light-trapping combined has already tallied 11 000+ species of ACG Lepidoptera; the species-area curve rises abruptly each time a new locality and (or) habitat or ecosystem subset is added to the inventory, and gradually creeps upwards as more seemingly “known” specimens from well-inventoried portions of ACG are barcoded. More than 215 000 adult ACG Lepidoptera legs (and 55 000+ parasitoid legs) have been sent to BIO for attempted Sanger sequencing by the end of 2015.

Routine ACG barcoding, routine results

There are major DNA barcoding-facilitated inventory processes in the tortured and multi-path travelogue for the biodiversity in the forest to become biodiversity understanding and information in someone’s brain. When a parataxonomist finds and rears a caterpillar on its food plant (Janzen 2004; Janzen et al. 2009, 2011), or selects an adult from a light trap, the specimen’s on-site database record (in a FMP (FileMaker Pro) flat file) receives all the collateral available at that ACG biological station, including best-guess interim (or actual) species and higher taxon names. The same applies to food plants and parasitoids when they are first perceived.
Along its early journey to BIO/BOLD (or elsewhere), and to the final institutional adult voucher specimen residence, that specimen’s name may be upgraded one to many times by a subsequent handler. After several months the BOLD-generated and inventory-screened name is first applied to the record (by downloading from BOLD). It is often upgraded in later months, years, and decades as it is worked over by the taxasphere (the aggregate of taxonomists, museums, and their knowledge, Janzen 1993) and as it receives input from the inventory itself. No taxonomist has necessarily had to directly bother with this process or specimen until the time that the first BOLD results are obtained, though the processing parataxonomists and taxonomists along this trip all occasionally logic-check the interim ID and may upgrade it, as based on morphology or food plants. In this essay “morphology” includes color patterns as well as more strictly morphological traits.
At any time, the addition of a “better” name creates a pulse of confirmations and explorations by the Principle Investigator (PI) within the bioinventory and (or) the subset of relevant records in BOLD, and often, new queries back to the taxasphere. This is especially the case if there is discordance between the inventory database collateral and the collateral carried by other specimens with that name. For example, when the inventory says that Enyo ocypete (Sphingidae) caterpillars are invariably found feeding on plants in the Dilleniaceae, Onagraceae, Vitaceae, Actinidiaceae (n = 4577), and a 2015 record in BOLD lists Rubiaceae as the food plant for this moth, the inventory PI knows that there is a biological or bookkeeping problem to be resolved. The problem is found to be in bookkeeping 98%+, but not 100%, of the time. However, it is also the case that mother moths can be sloppy in oviposition or simply have oviposition “preferences” a bit broader than suggested by traditional expectations, and caterpillars may even “wander”, especially if they have defoliated their food plant and have “generalist” abilities.
For a simple illustrative example of the many possible barcode-mediated interactions between the inventory and the taxasphere, tachinid flies DHJPAR0058284 and DHJPAR0058285, reared in 1986 and 1989, respectively, were both initially named tachJanzen01 Janzen01 (general genus- and species-level place-holders for totally unknown Tachinidae). Then in 1998 they were morphologically identified as Vibrissina Wood01, meaning, at that time, morphologically placed in the genus Vibrissina (A. Fleming, personal communication) by Monty Wood, a world-level tachinid taxonomist, but not distinguishable to species (especially difficult because both are females, and tachinid fly names are often based on the more morphologically distinct males). In a major barcode sweep through the reared tachinid vouchers in 2007 they were deemed “to old” to be worth the effort to attempt DNA barcoding, but their host caterpillars were the expected sawflies (Argidae) for Vibrissina. Then in 2015, the five ACG barcode- and morphology-determined species of reared ACG Vibrissina moved into the queue for alpha-taxonomy description by Alan Fleming and Monty Wood, the two current CNC (Canadian National Collection, Ottawa) museum-based taxonomists for ACG Tachinidae.
In searching the database for specimens to be morphologically coordinated with the barcode-backed specimens in BOLD, 86-SRNP-527 and 89-SRNP-343 (caterpillar codes), later to become DHJPAR0058284 and DHJPAR0058285 (parasite codes), were re-encountered in the CNC collection in their “too old” status. Emphasized to Fleming and Wood by the inventory, the current curators of the CNC collection of specimens behind 20 000+ rearing records of ACG tachinids, they were then morphologically identified to be a sixth species of Vibrissina and therefore queued for description as Vibrissina pedroleoni, with that name duly entered in BOLD. A few days later, a leg from each was couriered to BIO to be given special treatment by Suresh Naik and Sean Prosser, and a few days later the two barcodes appeared in the already-named barcode cluster of Vibrissina robertwellsi, in a standard neighbor joining (NJ) tree. Hmm. Aha! They are both females, and the females of V. robertwellsi were unknown (and “look” quite different from the males). Mystery solved and an email was sent to BIO to upgrade the name in BOLD to be V. robertwellsi.
While it has taken 30 years for the species-level status of these two female tachinid parasites of Sericoceros vumirus and Sericoceros mexicanus (Argidae) (thank you Dave Smith of the SEL/USDA laboratory at the Smithsonian Institution NMNH for that name) to be taxonomically resolved, the genus-level morphology-based ID only took 12 years, and it would have been just 2 years more to have species-level morphology-based resolution by Wood if they had been males. Today such a travelogue is only a matter of how long is required to barcode a leg. Furthermore, within the year the inventory knows if a newly reared fly from Argidae sawfly caterpillars is one of the five now known ACG species of Vibrissina, or yet another species new to ACG, or some other genus of tachinid. Obviously it will be a major effort before the five new species of ACG Vibrissina (no others are known from ACG) will be morphologically and barcode-integrated with the other known 25 species of Neotropical and extra-tropical Vibrissina, all of which are morphologically distinguishable from the five new species from ACG (A. Fleming, personal communication) and all of which look like houseflies to the untrained eye. Now repeat the travelogue described above for many tens of thousands of ACG specimens, and 13 000+ species of Lepidoptera and their parasitoids.
The routine caterpillar search and rearing creates a steadily increasing deposit of frozen voucher specimens at each of the 11 ACG rearing stations. Each specimen of adult Lepidoptera is stored frozen in a small plastic envelope (sold in the local market for candy) with its small standard-size high-quality paper label bearing its yy-SRNP-xxxxx voucher code to its database record, the University of Pennsylvania web site URL (, and “Area de Conservacion Guanacaste”. If the specimen is later to be barcoded, it is spread and oven dried, and within about a year its pedigree can be found on Google by searching for the voucher code. The same label is still on the pinned specimen (or in its alcohol tube) where it resides in its final institutional home. At irregular monthly intervals, hundreds of these envelopes are delivered from each station to the clearing center in the Area Administrativa of Sector Santa Rosa, where the PI compares each specimen with a copy of its database record (which arrives via email at weekly intervals, sent by the parataxonomists from their rearing stations). This is the last check before getting barcodes from BOLD to ascertain that it matches what is expected for its collateral data.
Since these incoming specimens are only very rarely removed from their see-through plastic envelopes, there is almost no chance for DNA contamination during this process; each specimen was handled individually at the rearing station (specimen killing is done by freezing the rearing container). To date, the very rare cases of DNA contamination throughout this process have not created barcode confusion, though on very rare occasions there are transcription errors at the time of writing the voucher code on the specimen label or in typing the fate of the specimen into the station-specific databases.
The envelopes of adult Lepidoptera that are selected for specimen barcoding are passed to a parataxonomist who museum-spreads and oven-dries them. They are shingled tightly into light-weight insect boxes for hand-carry to Philadelphia (D.H.J and W.H. carry 32–36 of these boxes (as in Fig. 3) on each flight from ACG to Philadelphia). There appears to be opportunity for DNA contamination during the spreading on wooden spreading boards that are used for tens of thousands of moths, shingling, and later sorting, as well as later de-legging and photographing. However, this has proven to be essentially not the case; less than 1 per 10 000 of these 1–10-year-old specimens have been found to be contaminated with foreign barcodes, even though it is certain that some wing and body scales must move from one to another (wasps and flies display essentially zero contamination). When barcoding 10–20-year-old inventory specimens, there has been slightly greater frequency of contamination (perhaps 1 per 1000 to 5000 specimens), but any apparent case of contamination is usually solved by checking the photographs of each specimen against the names that are eventually displayed through a newly downloaded NJ tree. Vastly more time is spent correcting typing errors (transposition of numbers, misreading numbers off of rearing bags, accidental label-switching, etc.) than in dealing with contaminants.
Fig. 3.
Fig. 3. Standard insect box (23.5 cm × 32.5 cm) of light-trapped, oven-dried pinned, and shingled pyraloid moths readied for hand-carry to Philadelphia from Area de Conservacion Guanacaste (ACG), and thence to the National Museum of Natural History, Smithsonian Institution, as prepared and individually data based by the ACG parataxonomists. Each has only five or fewer legs because one leg has been couriered in its lysis plate to BIO at the University of Guelph, and each will get its first pass at taxasphere-level identification when its barcode appears in an NJ tree of pyraloids downloaded from BOLD by the PI. Image credit: Daniel Janzen, University of Pennsylvania.
In Philadelphia, the adult Lepidoptera are unshingled, sorted to family by eye, and de-legged and photographed in pulses of hundreds. At the time of putting a portion of one leg into its lysis plate well (to be couriered to BIO), the specimen is also quick-and-dirty digitally photographed above and below (see images of adults in The photographs are cropped to a reasonable size at their maximum resolution, and a .jpg thumbnail is loaded into the specimen record on the University of Pennsylvania web site. The same image database is used by the parataxonomists in their own laptops, as a “field guide” and audit trail, separate from the main project database. The high-resolution photograph is stored at the University of Pennsylvania for public distribution, and it is also sent to BIO as the root image for the specimen record in BOLD. At the time of de-legging, a record is made of that event in a separate database, one that is often searched to correct voucher code errors in other databases, BOLD, and the audit trails of specimens. This database is also used to create the multi-field BIO databases (“specimen record” and “plate record” in BIO terminology) that accompany the lysis plates to BIO. At this point the inventory waits until there is need for a new update on the taxonomy within some higher taxon before querying BOLD for an NJ tree and accompanying spreadsheet (e.g., Figs. S1, S22).
Parasitoids are treated as above, with pinned wasps and flies being treated as though they were pinned Lepidoptera, and ethanol-preserved smaller specimens remaining in refrigerated ethanol except when losing a leg or being sorted by eye. However, they are not photographed at the time of de-legging, but rather, this is delayed until there will be a representative specimen of each (then) known species, a specimen to be photographed in such a way as to facilitate taxonomic labors as well as auditing of the barcoding process (see, e.g., figures in Fleming et al. 2014a, 2014b, 2015; Fernandez-Triana et al. 2014a, 2014b; Hansson et al. 2015).
The above sketched protocol is modified by the collectors of light-trapped adults, the BioLep project, in a home-made air-conditioned field laboratory in Sector Santa Rosa of ACG, such that the entire array of actions are done on-site in Santa Rosa by the parataxonomists. The PIs take the specimens (Fig. 3) to Philadelphia for distribution to their institutional home and BIO directly receives the lysis plates and their associated databases and images by courier from Philadelphia. All specimens retained and processed by BioLep are DNA barcoded, in contrast to the adults now generated by the caterpillar rearing; many of the latter are discarded (but still recorded) except for (i) suspected new species, (ii) puzzling suspected species complexes, and (iii) fulfilling requests by participating taxonomists for more specimens of this or that. Rearing continues of well-known species as part of the inventory of their parasitoids.
While BOLD stores the inventory records in 1000-member dollops, called “projects”, the ACG inventory does not use this portion of the BIO system. Instead, our NJ trees from BOLD are derived across projects, question-specifically aggregated from records of various taxon levels, localities, dates, or other groupings of inventory variables. The values of these collaterals are scattered throughout tens of thousands of records entered into BOLD since 2004, and therefore the search is by strings of voucher codes, not projects; these strings are generated by searches within the inventory databases. Queries to BOLD are submitted directly by the inventory personnel and PI (or other users) by submitting a list of voucher codes if the query is likely to contain less than about 5000 sequences, or sent as an excel file list for BOLD to do by the BOLD facilitators for the inventory if in the processor-demanding 5000 – 30 000 sequence range, or if there is some special aspect to the analysis request (such as inclusion of BIN codes in the NJ tree and its accompanying Excel file, e.g., Figs. S1, S22). The inventory has been fortunate in having several members of the BIO staff be direct facilitators for these requests (and uploads to BOLD, be they single records or bulk simultaneous updates to thousands of records), currently being handled by Suresh Naik, Ramya Manjunath, Megan Milton, Sean Prosser, and the footprints of many more before them. This facilitation goes far beyond straightforward sequencing and needs to be included in evaluations of direct costs (fees) to the inventory for barcoding.

Standard NJ tree and accompanying Excel file from BOLD

The first time an NJ tree is requested from BOLD for a higher taxon, the tree usually contains many clusters of similar or apparently identical barcodes, most clusters of which are found to match a morphologically-based species name (whether interim or formally scientifically described). This is the inventory’s first real look at its species richness if the specimens are small or nondescript (the case for most parasitoids and Microlepidoptera, and many Macrolepidoptera moths). These clusters were initially called “lumps” because that is what they look like when viewing an NJ tree as a whole, and today they are generally viewed as the group of specimens backstopping a BIN code (Ratnasingham and Hebert 2013; e.g., Figs. S1, S22). This matching has been an ongoing joy, and it is the reason that the inventory was fused with the BIO DNA barcoding initiative. At its core, it meant that barcoding could potentially be able to identify anything anywhere by anyone, as an obvious pathway towards universal bioliteracy (Janzen 2005).
However, among the “known and expected” were also clusters of barcodes that were quite unexpected and very early on signaled the presence of species complexes hiding under one morphological appearance, perception, and (or) name. These clusters, distant from each other by anywhere from well less than 1% of their base pairs to as many as 10%, are initially given an interim species epithet such as is shown by Scotura leucophleps (Fig. 4). The scientific epithet “leucophleps” is retained within the interim species-level epithet for obvious inventory reasons and to assist the participating taxonomist, Isidro Chacon of the Museo Nacional de Costa Rica. When Chacon and the inventory barcoded S. leucophleps, it was found to be five species within Costa Rica (only two of the newly discovered species occur in ACG, the other three occur at least in other parts of Costa Rica). The five barcode clusters were then found to be backed up by obvious differences in genitalia (Fig. 4). This discovery does, however, create the labeling and identification problem that all the inventory specimens of S. leucophleps that were not barcoded (discarded, donated elsewhere, died of disease, etc.) will not be taxonomically resolved within the now identified species complex, and will continue to be labeled S. leucophleps, just as they currently are in museums and publications (Miller 2009) throughout the world. This problem is an extensive result of barcoding any inventory.
Fig. 4.
Fig. 4. Male genitalia paired with their respective adults of species flushed out unexpectedly by DNA barcoding the inventory: Scotura leucophlepsDHJ01 INB0003971032 CR, Alajuela, Los Chiles, Laguna Medio Queso, Lat/Long: 11032, –84.689; Scotura leucophlepsDHJ02 INBIOCRI001295625 CR, Limon, Talamanca, Gandoca Manzanillo Lat/Long: 9.633, –82.659; Scotura leucophlepsDHJ03 INB0004202460 CR, Puntarenas, Buenos Aires, Paraiso Lat/Long: 9.084, –83.303; Scotura leucophlepsDHJ04 INBIOCRI001210027 CR, Limon, Pococi, Barra de Colorado, Rio Sardinas Lat/Long: 10.64405, –83.742005; Scotura leucophlepsDHJ05 INB0004250623 CR, Heredia, Sarapiqui, La Virgen, Tirimbina, La Isla Lat/Long: 10.416, –84.13. Species DHJ01 and DHJ05 occur in Area de Conservacion Guanacaste, the other three in other parts of Costa Rica. All specimens now in the Museo Nacional de Costa Rica. Image credit: Isidro Chacon, Museo Nacional de Costa Rica.
The feedback to the parataxonomists is, “oops, from now on, we save and barcode all S. leucophleps until this situation is resolved”. The feedback to the taxasphere is “oops, which, if any, of our five matches the holotype of S. leucophleps; and please be aware that your drawer of S. leucophleps is a complex that may be resolved only by barcoding or dissection or via some other biological clue”. When this is resolved, one of the five species, perhaps, can continue to march forward as S. leucophleps and the other four have to get new names, a process that can take many years and much energy by the taxasphere (e.g., Burns et al. 2010; Chacon et al. 2012, 2013; Grishin et al. 2013a, 2013b; Phillips-Rodriguez et al. 2014; Fleming et al. 2014a, 2014b, 2015a, 2015b; Fernandez-Triana et al. 2014a, 2014b, 2015, 2016; Brown et al. 2014; Hansson et al. 2015, etc.). To reinforce this problem, Fig. 5 displays three more “well-known” and long-known examples, among many, of brightly colored (presumed aposematic) species of Arctiinae (Arctiidae of old, today a subfamily in the Erebidae) detected in the first pass barcoding of light-trapped and reared ACG moths (B. Espinoza, personal communication).
Fig. 5.
Fig. 5. How three well-known species of Costa Rican Arctiinae break up into six species when barcoded and then their genitalia examined: (a, g) Amaxia apyga, ♂, 10-SRNP-112569; (b, h) Amaxia apygaDHJ01, ♂, 08-SRNP-102359; (c, i) Amaxia osmophora, ♂, 08-SRNP-101635; (d, j) Amaxia osmophoraDHJ01, ♂, 09-SRNP–107029; (e, k) Amaxia pardalisDHJ02-BE04, ♂, 08-SRNP-105195; (f, l) Amaxia pardalisBE02, ♂, adult INB0004166127, genitalia INB0004301863. Image credit: Bernardo Espinoza, Museo Nacional de Costa Rica.
But biology is rarely so simple as knowns and unknowns, each backed by a neatly defined cluster of DNA barcodes in an NJ tree from BOLD. These are thorny but inevitable complications of trying to untangle very species-rich areas. As more ACG species, and more samples per “species” are added to BOLD, there are three recurring, case-after-case forms of speed bumps in barcode interpretation that confront the inventory and the downstream taxasphere — (i) a pair of adjacent clusters in an NJ tree, made up of morphologically indistinguishable specimens but conspicuously different barcodes (note that the S. leucophleps and Arctiinae cases cited above do not fall in this category because the clusters were found to have quite different morphology in their genitalia); (ii) clusters that diverge by only about 0.1%–1.5% of their base pairs, yet they are clearly different (no overlap); and (iii) single specimen outliers. Each of these three results petitions the inventory and the taxasphere for different actions. In general, the inventory is asked for more samples and more samples from additional biological circumstances, searching for different distinguishing collateral.

(i) A pair of adjacent clusters in an NJ tree, made up of morphologically indistinguishable specimens but conspicuously different barcodes

Amphonyx lucifer (also known as Cocytius lucifer) is a huge and well-known sphingid moth (see fig. 3 in Janzen et al. 2009) that occurs from Mexico to Argentina. In ACG its large caterpillars eat eight species of Annonaceae in five genera, and have been wild-caught 253 times by the inventory, in dry forest, rain forest, and their intergrades. As of today, 64 adults, both from light traps and reared, have been DNA barcoded, and they break cleanly into two clusters that are minimum 2.89% different in their barcodes. Intense morphological inspection has found no difference between them. Were it not for barcoding the inventory, the existence of these two species would be unknown. Amphonyx luciferDHJ02 (BOLD:AAB3608) is exclusively in ACG rain forest, as adults and caterpillars (n = 34). Amphonyx luciferDHJ01 (BOLD:AAB3605) occurs throughout dry forest during the rainy season, absent during the dry season (because it has migrated to the rain forested side of ACG), and both adults and caterpillars occur in the rain forest throughout the year (n = 32). It is unambiguous that there are two species, one “stay-at-home” in the rain forest and one that uses both ecosystems when green leaf food is available. This east–west (dry forest – rain forest) seasonal migration pattern has long been recognized for ACG (Janzen 1987a, 1987b), but it took DNA barcoding of the inventory to flush out a case of a sibling species pair, one confined to one ecosystem. There may well be hundreds of such cases among ACG Lepidoptera, since there are tens of thousands of cases where adults of what appears to be a single species (without barcoding) of insect move seasonally among tropical ecosystems and habitats.
If there was no collateral ecological data to accompany the barcode results, it would be easy to assume that the two “A. lucifer” are merely a polymorphic barcode or a true barcode paired with a pseudogene. A similar case has been recently unearthed with four ACG species of essentially identical Perichares skipper butterflies (Hesperiidae) that cannot be distinguished by their genitalia (the “normal” way that look-alike species of Hesperiidae are distinguished) but are easily distinguished by their barcodes, caterpillar colors, and in two cases, by exclusively feeding on one or the other of two genera of side-by-side understory rain forest palms (Burns et al. 2008). The most extreme ACG case unearthed to date is three sympatric rain forest and dry forest, morphologically identical (Fig. 6), new species of skippers hiding under the name Urbanus belli and distinguishable only by their barcodes and nuclear genes, some subtle ecosystem differences, and to some degree their infection by Wolbachia (Smith et al. 2012a; Bertrand et al. 2014).
Fig. 6.
Fig. 6. Reared males (columns one and two, left to right) and females (columns three and four, left to right) of three new cryptic species of Urbanus from Area de Conservacion Guanacaste, northwestern Costa Rica, in dorsal (left, greenish) and ventral (right, brown) view. Males are holotypes; females are paratypes (pins and pinholes artificially removed). (ad) Urbanus segnestami (06-SRNP-46181, 02-SRNP-1028); (eh) Urbanus bernikerni (93-SRNP-3294, 93-SRNP-3098); (il) Urbanus ehakernae (03-SRNP-10927, 01-SRNP-22184) (voucher code of each specimen in parentheses) (Bertrand et al. 2014). Images credits: John. M. Burns and Karie Darrow, Smithsonian Institution.

(ii) Clusters that diverge by only about 0.1%–1.5% of their base pairs yet are cleanly different

It is commonplace for two or more clusters to differ by only 0.1%–1.5% of their base pairs (and therefore fall in the same BIN in BOLD) and be expressed in an NJ tree as a very shallow “split”. Such shallow splits can of course be caused by barcode+pseudogene pairs (as discovered from careful further study of Telemiades chrysorrhoeaDHJ01 and T. chrysorrhoeaDHJ02 in the Hesperiidae and Thisbe ireneaDHJ01 and T. ireneaDHJ02 in the Riodinidae, Mehrdad Hajibabaei, personal communication). However, when there are more than two clusters in the complex, this possibility is discarded because two different pseudogenes have not been encountered to date. Sympatric barcode polymorphisms are also possible (e.g., Prepona philatelicaDHJ01 and P. philatelicaDHJ02 in the Nymphalidae), though experience to date suggests that they are very rare. When the sample sizes are large in each of two or more clusters only shallowly separate, there is a strong suggestion of a species complex. Whatever their cause, their emergence as the inventory progresses lead to seriously large sample sizes being barcoded as well as increased effort to obtain more collateral information.
A straightforward example is offered by the nymphalids Adelpha pseudaethalia and Adelpha melanthe (Fig. 7), two ACG species separated long ago by their very different appearance (Willmott 2003), yet falling in the same BIN (BOLD:AAA5411) and having an average barcode distance of 0.29%. They have very similar caterpillars and all the food plants of A. pseudaethalia are included within the food plant list of A. melanthe. However, the adults belong to different mimicry complexes. Their two clusters even slightly overlap within an NJ tree of just the pair, or one of all Nymphalidae, but the overlaps are likely due to reading “errors” of a single nucleotide.
Fig. 7.
Fig. 7. Adelpha melanthe and Adelpha pseudaethelia, a pair of Area de Conservacion Guanacaste fully sympatric (dry forest and rain forest) nymphalid butterflies with nearly identical, but not so, barcodes. The latter species was initially identified as Adelpha phylaca (as indicated in this historic portion of the NJ tree) but is now known to be Adelpha pseudaethelia (K. Willmott, pers. com., University of Florida, Gainesville). Image credit: Daniel Janzen, University of Pennsylvania.
A more complex example is the case of the hesperiids Phocides belus, Phocides Warren01, Phocides pigmalionDHJ01, and Phocides pigmalionDHJ02 (Fig. 8) each separated by only very shallow splits in an NJ tree (and an average distance of 0.59%) and all receiving the same BIN code (Figs. S1, S2)2 (BOLD:AAA6571). Phocides belus and P. Warren01 are nearly undistinguishable by their appearance (Fig. 8), but they fall into distinct clusters with very different ecology (P. belus feeds on Thounidium decandrum (Sapindaceae, n = 189) while P. Warren01 feeds on three species of mangroves (Conocarpus erectus, Laguncularia racemosa, Rhizophora mangle)). Equally, P. pigmalionDHJ01 and P. pigmalionDHJ02 are also nearly indistinguishable by their appearance (Fig. 8) but fall into distinct very close clusters; the former feeds on two species of native plants (and three introduced) and the latter on five species (in three plant families) of quite different native plants.
Fig. 8.
Fig. 8. Four sympatric or parapatric species of Phocides (Hesperiidae) whose barcodes differ by less than 1% and all occupy the same BIN (BOLD:AAA6571) (Fig. S1)2. Phocides belus and Phocides Warren01 fall in the same cluster but have very different facies and ecology; Phocides pigmalionDHJ02 and Phocides pigmalionDHJ01 are very shallow splits off the former cluster, but they are easily identified by their facies and non-overlapping food plants. Image credit: Daniel Janzen, University of Pennsylvania.
A particularly dramatic example is the previously mentioned case of the A. fulgerator 10-species complex (Fig. 2), all members of which being parapatric to sympatric within ACG, that triggered full-scale application of barcoding to the inventory: 6 of the 10 species, although distinguishable by food plants and caterpillar color patterns, fall in one BIN (BOLD:AAA1584), even though each species falls in non-overlapping and only shallowly separated clusters (Figs. S1, S2)2.
A particularly extreme case was first described in the early days of applying barcoding to the inventory (Burns et al. 2007). What had been long and comfortably thought to be one distinctive species of large hesperiine hesperiid butterfly—Neoxeniades luda—differed in barcode from its ACG look-alike (newly described as Neoxeniades pluviasilva) by only 0.1% (= one base pair), but the shallow split attracted attention by its ecological traits (rain forest for one side of the split and dry forest for the other side) and “variable” coloration. An NJ tree constructed of all inventory-reared specimens showed them to be mixed and overlapping in their location in the tree, and in today’s terminology, in one BIN (BOLD:AAB4945). But when the specimens with short (300–400 bp) barcodes are excluded from consideration, the two species make two barely separated clusters in the tree—in hindsight easily recognizable in the adults by slight differences in color, slight differences in genitalic morphology, and occupying parapatric rain forest and dry forest. If the inventory had only several specimens of each species, they would be lumped somewhere in the morass of terms “variable”, “slightly polymorphic barcode”, and simply called Neoxeniades luda as they have been since 1877.
The inventory has to remain skeptical of such shallow splits displayed in an NJ tree; many are probably nothing more than extant polymorphisms, or a population in the midst of a selective sweep and therefore transitioning from one super-abundant barcode to another, but of course many may also be cases like that of Neoxeniades mentioned in the previous paragraph—species in the process of splitting into two or just having done so. From a practical standpoint, however, they are a major driver for increasing inventory sample size and capturing yet more detailed collateral about the specimens. On the other hand, they do not confound the identification-by-barcode process since the two sides of the shallow split fall side-by-side in an NJ tree. What does confound is when a short barcode (often from an old specimen or one that was stored for a time in high humidity) causes the specimen to locate within another adjacent cluster of similar but unambiguously different barcodes (e.g., the Adelpha shallow split mentioned above, see Fig. 7). Such cases are often resolvable and noticed by other biological correlates (especially food plant, host caterpillar, microgeography, and (or) split reared sib clusters) before the specimen needs to be scrutinized by a morphology-based taxonomist to clarify its status.
These five examples, and many tens of others like them, have caused us to become hypercautious in treating shallow splits among sample-rich clusters as if they are “intraspecific variation”, polymorphisms, laboratory errors, and barcode-pseudogene pairs (and therefore can be ignored). Amongst the inventoried butterflies, Hesperiidae (skipper butterflies) have been the most intensively examined in this respect. If we use Hesperiidae BINs to count the number of reared ACG species to date, there are 514 species; if we use clusters explored as in the previous five examples, there are 590 species. Each of these numbers is useful, depending on the question being asked of the inventory. BINs are a vastly more rapid way of approaching species counts in a large sample of barcoded specimens of unknown taxonomy than is attempting to read the cluster morphology of NJ trees containing thousands of sequences. However, in this ACG inventory with abundant collateral data, we are interested in real-world species-level identification of independent evolutionary lineages, and this will sometimes split up a BIN. But BINs are exceedingly useful information when ecological collateral is largely absent, such as for specimens in museum drawers or Malaise trap samples, or there have been no time, taxonomist, and (or) other resources for detailed morphological examination and comparison of museum specimens.

(iii) Single specimen outliers

It is a commonplace event that a well-defined cluster in an NJ tree has one to few specimens on an immediately adjacent short branch (indicating a nearly identical barcode sequence). If these are short sequences, they often represent nothing more than the “variation” generated by a shorter (300–600 bp) barcode. However, a significantly large number of them represent cases with 600+ base pair barcodes and could represent a rare sibling species co-occurring with a common one; a member of a common sibling species that is ecologically “out of place” and therefore only rarely picked up by the inventory; two common highly sympatric species, only one of which is easily captured by the trapping techniques; true variability in barcode content within a species (several biologically real haplotypes, not laboratory errors); and simple laboratory “errors” in reading the barcode. The last case requires valuable laboratory time to resolve, while the former four cases call for more intense or different specimen capture by the inventory, something that the inventory does on a regular basis as these cases are discovered.
A conspicuous example of undercollecting a common species, collected by the inventory as adults but never by caterpillar, is illustrated by three species of ACG Adhemarius (Sphingidae). These three species cannot be reliably distinguished by eye by anyone, but two have been described by sphingid taxonomists working with light-trapped adults. Adhemarius daphne is the ACG “dry forest species” but also occupies ACG rain forest, as evidenced by capture of adults and caterpillars in both ecosystems. Adhemarius fulvescens is the ACG cloud forest species, as evidenced by capture of adults and caterpillars only in ACG cloud forest. Adhemarius gannascus, the name casually slapped onto all three of these species throughout their combined ranges from Mexico to Brazil, occurs only in ACG rain forest. The three barcode at about 4% difference. Adhemarius gannascus is evidenced by hundreds of adults in ACG light traps but after 30+ years of search on its presumed lauraceous food, the caterpillar has never been found. We hypothesize that in contrast to the caterpillars of the other two, it does not occur within 2–3 m of the ground, the stratum where the parataxonomists do most of their searching.
Were the inventory only surveying caterpillars, it would report two similar and ecologically distinct species of this Adhemarius complex, A. daphne throughout the lowlands and A. fulvescens restricted to cloud forest above. In a purely caterpillar-based inventory, the first A. gannascus caterpillar barcoded would be a definite NJ tree outlier to some cluster. Since the caterpillar inventory has many examples of such outliers that have not been (yet?) light-trapped, how are we to interpret their presence? And to complicate the Adhemarius case yet further, barcoding of the large sample of A. daphne reveals that there are A. daphneDHJ01 and A. daphneDHJ02, totally sympatric and with identical caterpillars and adults, eating the same food plants, and receiving the same BIN (BOLD:AAA2433), but cleanly 1.31% average distance between them.
An outlier example just discovered when writing this paper is Creonpyge creon, a large, showy, well-known cloud forest hesperiid butterfly (see Janzen and Hallwachs 2016). Over the past 17 years, the inventory has been able to successfully barcode 11 of them from the top of Volcan Cacao (1100–1400 m), one of the three ACG cloud forest-covered volcanoes (Fig. 1), with no barcode surprises. But in late January 2015, an exploring parataxonomist, Calixto Moraga, found three 2–3rd instar caterpillars of C. creon at 1800 m in the poorly collected cloud forest on Volcan Santa Maria; one (15-SRNP-30170) survived to eclose 6.5 months later (caterpillars from cold high elevations can be excruciating slow in development). Its 658 base pair barcode is 0.75% different (Fig. S1)2 from those on Volcan Cacao, and these two cloud forests are about 10 km apart, separated by a 400–700 m elevation (comparatively hot) valley that has been a lower-elevation rain forest divider since the last glacial retreat 10 000 – 20 000 years ago.
The next step on the inventory side of the equation is to get a larger sample of C. creon from both volcanoes, and a quick check by someone to ensure that the shallow split is not from a laboratory error. If the general split persists with increased sample size, we hope that the taxasphere will explore them for morphological differences that might be “enough” to warrant the highly subjective decision of two micro-allopatric sibling species. On the one hand, the inventory sample of C. creon can be increased most readily by simply barcoding the caterpillars (much easier to obtain than rearing through to adults), but the caterpillars do not have the traits of genitalia and wings that are classically used in species designation by hesperiid taxonomists. Furthermore, general barcoding of valuable to-be-reared caterpillars (instead of eclosed adults) brings its own problems. To further complicate the matter, could C. creon, described in 1874 and believed to range from Costa Rica to Colombia, be found to have a different “species” on each of the tens of cloud-forested volcano top sky islands over its range? If so, which one matches the holotype?

Inventory and taxasphere time economies

We cannot overemphasize the savings in time, dollars, and inestimable human resources for the inventory that is represented by the addition of on-going iterative barcoding. Barcoding has the impact of both stimulating and guiding further sampling in more ways as cryptic or “rare” species are encountered, and knowing with greater certainty which species are “done” (which varies according to the metrics of the hypotheses underlying the inventory).
But once the specimens and their data move into name-dependent aggregation and identification, the major positive effect of the barcodes for many kinds of analyses really kicks in.
For example, there were 19 254 barcodes of 599 species of Hesperiidae in the ACG NJ tree downloaded from BOLD on 19 November 2015 (very similar to Figs. S1, S22). Among them were the barcodes of 328 newly reared specimens identified to family but not species, included to be identified (largely because owing to time constraints, taxonomist unavailability, and (or) morphological crypticity, there was no other way to identify them). They had never been examined by a taxonomist, though the parataxonomists had offered their opinions at the time of caterpillar collection, and the inventory PI had sometimes done so as well at the time of making the decision to mount, oven-dry, and barcode them at the Santa Rosa clearing center in ACG (i.e., the several thousand reared specimens of hesperiids with “certain” identifications in this harvest period had been culled months before). With the NJ tree and its associated excel file available in a laptop, it took 187 min for the PI to assign 325 of these puzzling specimens to a species already recorded by the inventory, corroborate that their collateral matched what was expected for that species and morphology (in the photograph taken at the moment of de-legging), and discover three specimens of two species new to the inventory (both of which are undescribed cryptic sibling species flushed out by this barcoding session). The associated excel file was then sent to BIO for updating into BOLD.
If the collateral (including the barcode) of one of those 328 specimens had not matched what was expected, that would have initiated inspection of the specimen image. Then there would have been a search backwards to locate the source of discord, or allow the assignment of a new interim name, as occurred for the three specimens of species new to the inventory. If the clusters in the NJ tree had been tagged by their BIN codes (Ratnasingham and Hebert 2013), the assignment to a BIN would be automatic by BOLD and almost instantaneous. However, in this particular case, a BIN-based analysis (approximate 2% cut-off for species-level differences) finds 509 species in the sample and misses both of the potential new species; the slower inventory method of comparing clusters of barcodes with food plants and microgeography suggests that there are 592 species in the sample. With further biology-based refinement, some of the currently distinguished species will be found to be the result of over-splitting. However, others will be found to be the result of under splitting. Either way, the inventory has rapidly obtained detailed guidance as to where to put more (or less) collecting and analysis emphasis.
Without barcoding, this process of specimen identification of a large sample of “unidentifiable” hesperiids would have required 2–10 days of laborious examination of specimens by an extremely proficient neotropical hesperiid taxonomist (of which there are perhaps five in the world). This would include many tens of dissections and time-consuming puzzled comparisons with other specimens, all of which reside in a museum several thousand kilometers distant. There are no keys or field guides to any of these species, and if there were, we now know that the barcodes are more reliable for identification within large arrays of unknowns than is any morphology-based tool. At this point, whether the species is known by a scientific name or by an interim name is irrelevant. By this time in the identification and feedback process, no expert taxonomist has had to examine any of the newly reared or wild-caught specimens, yet the field portion of the inventory has full tracking identification of its product for feedback as it conducts further sampling.
Turning to parasitoids of the ACG inventory caterpillars as a second example, the process is the same yet more complex. The latest NJ tree of Microgastrinae wasps (Braconidae, reared and malaise trapped; 13 December 2015) contains 14 532 barcodes of 1169 confirmed or highly suspected species. Among these were 200 new barcodes from new rearings, identified by non-microscope eye to be likely “Microgastrinae” through their general appearance and the appearance of their cocoons. The NJ tree immediately demonstrated (by long branches parked at each end of the tree) that eight of them were not Microgastrinae, which were then moved to other subfamilies or families by checking backwards to their initial rearing records (to be later identified to species cluster by inclusion in NJ trees for those higher taxa). By sorting the accompanying Microgastrinae excel file for the “new” placeholder name (mgJanzen01 Janzen01), labeling these sequences yellow, and returning the file to its original order matching the NJ tree, it then cost 55 min to identify all 192 sequences to previously identified species except for the nine species new to the inventory, by simply scrolling through the Excel file and performing “fill down” or “fill up” for each colored record, after a glance at the respective part of the NJ tree. This updated Excel file was then sent to BOLD as an attachment for uploading, and therefore available a few hours to a day later.
To have conducted this routine identification by classical microscopic examination and keys (as in those generated by the taxonomic processing of the inventory through adhering to traditional huge keys, Fernandez-Triana et al. 2014a) would have been logistically and human-resources impossible. This is especially because the specimens themselves, in addition to being at best 2–3 mm long, black, and yellow, reside in 2 mL refrigerated tubes of ETOH in transit to Ottawa. There, they will be further sorted, point-mounted as needed for morphology-based species-descriptions, used to generate genus-level treatments (Fernandez-Triana et al. 2014a, 2014b, 2015), and finally deposited in the Canadian National Collection and the Smithsonian Institution, and in other institutional collections willing to receive them (e.g., the Museo Nacional de Costa Rica and The Natural History Museum in London). Of course, were the parataxonomists to have their own field-based personal DNA barcorders, most of these specimens and sample lots would not even have to start up the chain of specimen processing, thereby reserving precious taxasphere resources for the new, puzzling, and more intriguing records.
Now envision the above process taking place simultaneously for 1100+ species of parasitic flies (Tachinidae), and 1000+ species of non-microgastrine Hymenoptera, feeding on the same set of 11 000+ species of caterpillars (to date) at the same time in the same place (ACG). The specimens disappear into the taxasphere and barcoding processing system. These then reappear iteratively as, for example, the inventory and the parataxonomists now know that Astraptes augeas (Hesperiidae) caterpillars have been reared 1393 times feeding on Hampea appendiculata (Malvaceae) in different circumstances and that 10 of these were parasitized by Apanteles osvaldoespinozai (Braconidae, Microgastrinae). The inventory also then knows that this species of wasp also parasitized five other closely related sympatric species of Astraptes feeding on Fabaceae and Sapindaceae at equally low frequencies, etc. A trophic web is being constructed from the bottom up through iterative passages of barcode-facilitated information moving back and forth from the field to the taxasphere.


DNA barcoding of inventory product is enormously helpful, and the earlier in the inventory, the more helpful. The process itself is hugely question-, place-, society-, history-, taxon-, and budget-based. And its analysis differs dramatically, depending on the amount of species-level resolution desired by the user of the process and the product. DNA barcoding offers a form of bioliteracy. Just as different people use literacy in different ways in different places for different circumstances, DNA barcoding has the same cross-society potential. For many millennia, the use, science, and engineering of biodiversity at a species level has been practiced with various society-dependent levels of resolution, tailor-made to the situation. For some, the species is a bug. For others it is a butterfly. For others it is a pretty butterfly. For others it is a Blue Flasher. For others it is Astraptes. For others it is Astraptes fulgerator. For others it is Astraptes LOWHAMP. And finally, today, for others it is Astraptes augeas, its “real” name. DNA barcoding a fragment of a specimen can take anyone anywhere any time, quickly and accurately, to the end of that chain, today for a few dollars, tomorrow for a few pennies, and the day after, for no more than what it cost to read this word.


We gratefully acknowledge the unflagging support of the team of ACG parataxonomists (Janzen et al. 2009; Janzen and Hallwachs 2011) who found and reared the specimens used in this study, and the team of biodiversity managers who protect and manage the ACG forests that host these parasitoids and their caterpillar hosts. The study has been supported by U.S. National Science Foundation grants BSR 9024770 and DEB 9306296, 9400829, 9705072, 0072730, 0515699, and grants from the Wege Foundation, International Conservation Fund of Canada, Jessie B. Cox Charitable Trust, Blue Moon Fund, Guanacaste Dry Forest Conservation Fund, Area de Conservación Guanacaste, Permian Global, individual donors, and University of Pennsylvania (D.H.J. and W.H.). This study has been supported by the Government of Canada through its ongoing support of the Canadian National Collection, Genome Canada, the Biodiversity Institute of Ontario, the Ontario Genomics Institute, and the Natural Sciences and Engineering Research Council of Canada.


Supplementary data are available with the article through the journal Web site at Supplementary Material.


Area de Conservacion Guanacaste (ACG). 2016. Official web site of Area de Conservacion Guanacaste, northwestern Costa Rica. Programa de Parataxonomos. Available from
Basset Y., Mavoungou J.F., Mikissa J.B., Missa O, Miller S.E., Kitching R.L., and Alonso A. 2004. Discriminatory power of different arthropod data sets for the biological monitoring of anthropogenic disturbance in tropical forests. Biodivers. Conserv. 13: 709–732.
Basset Y., Cizek L., Cuénoud P., Didham R.K., Novotny V., Ødegaard F., et al. 2015. Arthropod distribution in a tropical rainforest: tackling a four dimensional puzzle. PLoS ONE, 10(12): e0144110.
Bertrand C., Janzen D.H., Hallwachs W., Burns J.M., Gibson J.F., Shokralla S., and Hajibabaei M. 2014. Mitochondrial and nuclear phylogenetic analysis with Sanger and next-generation sequencing shows that, in Area de Conservación Guanacaste, northwestern Costa Rica, the skipper butterfly named Urbanus belli (family Hesperiidae) comprises three morphologically cryptic species. BMC Evolutionary Biology, 14: 153.
BOLD. 2016. Biodiversity of Life Database. Available from
Brown J.W., Janzen D.H., Hallwachs W., Zahiri R., Hajibabaei M., and Hebert P.D.N. 2014. Cracking complex taxonomy of Costa Rican moths: Anacrusis Zeller (Lepidoptera: Tortricidae: Tortricinae). J. Lepid. Soc. 68: 248–263.
Burns J.M., Janzen D.H., Hajibabaei M., Hallwachs W., and Hebert P.D.N. 2007. DNA barcodes of closely related (but morphologically and ecologically distinct) species of skipper butterflies (Hesperiidae) can differ by only one to three nucleotides. J. Lepid. Soc. 61(3): 138–153.
Burns J.M., Janzen D.H., Hajibabaei M., Hallwachs W., and Hebert P.D.N. 2008. DNA barcodes and cryptic species of skipper butterflies in the genus Perichares in Area de Conservación Guanacaste, Costa Rica. Proc. Natl. Acad. Sci. 105: 6350–6355.
Burns J.M., Janzen D.H., and Hallwachs W. 2010. Of many similar species in the neotropical genus Porphyrogenes (Lepidoptera: Hesperiidae), a new one, repeatedly reared in Costa Rica, is relatively distinct. Proc. Entomol. Soc. Wash. 112: 32–42.
Chacon I.A., Montero-Ramirez J., Janzen D.H., Hallwachs W., Blandin P., Bristow C.R., and Hajibabaei M. 2012. A new species of Opsiphanes Doubleday, [1849] from Costa Rica (Nymphalidae: Morphinae: Brassolini), as revealed by its DNA barcodes and habitus. Bull. Allyn Mus. 166: 1–15.
Chacon I.A., Janzen D.H., Hallwachs W., Sullivan J.B., and Hajibabaei M. 2013. Cryptic species within cryptic moths: new species of Dunama Schaus (Notodontidae, Nystaleinae) in Costa Rica. ZooKeys, 264: 11–45.
Devan, C. 2015. What do scientists do? Part II – A tribute to taxonomists. Available from
Fernandez-Triana J.L., Whitfield J.B., Rodriguez J.J., Smith M.A., Janzen D.H., Hallwachs W., et al. 2014a. Review of Apanteles sensu stricto (Hymenoptera, Braconidae, Microgastrinae) from Area de Conservación Guanacaste, northwestern Costa Rica, with keys to all described species from Mesoamerica. ZooKeys, 383: 1–565.
Fernandez-Triana J.L., Janzen D.H., Hallwachs W., Whitfield J.B., Smith M.A., and Kula R. 2014b. Revision of the genus Pseudapanteles (Hymenoptera, Braconidae, Microgastrinae), with emphasis on the species in Area de Conservación Guanacaste, northwestern Costa Rica. ZooKeys, 446: 1–82.
Fernandez-Triana J.L., Whitfield J.B., Smith M.A., Kula R.R., Hallwachs W., and Janzen D.H. 2015. Revision of the genera Microplitis and Snellenius (Hymenoptera, Braconidae, Microgastrinae) from Area de Conservacion Guanacaste, Costa Rica, with a key to all species previously described from Mesoamerica. Dtsch. Entomol. Z. 62(2): 137–201.
Fernandez-Triana J., Boudreault C., Dapkey T., Smith M.A., Rodriguez J.J., Hallwachs W., and Janzen D.H. 2016. Revision of the genus Promicrogaster (Hymenoptera, Braconidae, Microgastrinae) from Area de Conservacion Guanacaste, Costa Rica, with a key to all species previously described from Mesoamerica. J. Hymenoptera Res. 50: 25–79.
Fleming A.J., Wood D.M., Smith M.A., Hallwachs W., and Janzen D.H. 2014a. Revision of the New World species of Houghia Coquillett (Diptera, Tachinidae) reared from caterpillars in Area de Conservación Guanacaste, Costa Rica. Zootaxa, 3858(1): 001–090.
Fleming A.J., Wood D.M., Smith M.A., Janzen D., and Hallwachs W. 2014b. A new species of Cordyligaster Macquart, reared from caterpillars in Area de Conservación Guanacaste, northwestern Costa Rica. Biodivers. Data J. 2: e4174.
Fleming A.J., Wood D.M., Janzen D.H., Hallwachs W., and Smith M.A. 2015a. Seven new species of Spathidexia Townsend (Diptera: Tachinidae) reared from caterpillars in Area de Conservación Guanacaste, Costa Rica. Biodivers. Data J. 3: e4597.
Fleming A.J., Wood D.M., Smith M.A., Janzen D.H., and Hallwachs W. 2015b. Nine new species of Itaplectops (Diptera: Tachinidae) reared from caterpillars in Area de Conservacion Guanacaste, northwestern Costa Rica, with a key to Itaplectops species. Biodivers. Data J. 3: e4596.
Geddes, P. 2015. The Yellowstone of the future. Available from
Gibson J., Shokralla S., Porter T.M., King I, van Konynenburg S., Janzen D.H., et al. 2014. Simultaneous assessment of the macrobiome and microbiome in a bulk sample of tropical arthropods through DNA metasystematics. Proc. Natl. Acad. Sci. 111: 8007–8012.
Glover A.G., Dahlgren T.G., Wiklund H., Mohrbeck I., and Smith C.R. 2016. An end-to-end benthic biodiversity survey in the Clarion-Clipperton Zone, Central Pacific Abyss. J. Mar. Sci. Eng. 4(2): 1–34.
Grishin N.V., Janzen D.H., and Hallwachs W. 2013a. Hiding behind gaudy looks, a new Central American species of Phareas (Hesperiidae: Eudaminae). J. Lepid. Soc. 67: 161–174.
Grishin N.V., Burns J.M., Janzen D.H., Hallwachs W., and Hajibabaei M. 2013b. Oxynetra: facies and DNA barcodes point to a new species from Costa Rica (Hesperiidae: Pyrginae: Pyrrhopygini). Journal of the Lepidopterists’ Society, 67: 1–14.
Guanacaste Dry Forest Conservation Fund (GDFCF). 2016. Official website of the Guanacaste Dry Forest Conservation Fund. Available from
Hajibabaei M., Janzen D.H., Burns J.M., Hallwachs W., and Hebert P.D.N. 2006. DNA barcodes distinguish species of tropical Lepidoptera. Proc. Natl. Acad. Sci. 103: 968–971.
Hansson C., Smith M.A., Janzen D.H., and Hallwachs W. 2015. Integrative taxonomy of New World Euplectrus Westwood (Hymenoptera, Eulophidae), with focus on 55 new species from Area de Conservacion Guanacaste, northwestern Costa Rica. ZooKeys, 485: 1–236.
Hebert P.D.N., Cywinska A., Ball S.L., and deWaard J.R. 2003. Biological identifications through DNA barcodes. Proc. R. Soc. London, Ser. B, 270: 313–321.
Hebert P.D.N., Penton E.H., Burns J.M., Janzen D.H., and Hallwachs W. 2004. Ten species in one: DNA barcoding reveals cryptic species in the neotropical skipper butterfly Astraptes fulgerator. Proc. Natl. Acad. Sci. 101: 14812–14817.
Janzen D.H. 1987a. When, and when not to leave. Oikos, 49: 241–243.
Janzen D.H. 1987b. How moths pass the dry season in a Costa Rican dry forest. Int. J. Trop. Insect Sci. 8: 489–500.
Janzen, D.H. 1988. Guanacaste National Park: Tropical ecological and biocultural restoration. In Rehabilitating damaged ecosystems, Vol. II. Edited by J.J. Cairns. CRC Press, Boca Raton, Fla. pp. 143–192.
Janzen D.H. 1991. How to save tropical biodiversity. Am. Entomol. 37: 159–171.
Janzen, D.H. 1993. Taxonomy: universal and essential infrastructure for development and management of tropical wildland biodiversity. In Proceedings of the Norway/UNEP Expert Conference on Biodiversity, Trondheim, Norway. Edited by O.T. Sandlund and P.J. Schei. NINA, Trondheim, Norway. pp. 100–113.
Janzen D.H. 1998. Gardenification of wildland nature and the human footprint. Science, 279: 1312–1313.
Janzen D.H. 1999. La sobrevivencia de las areas silvestres de Costa Rica por medio de su jardinificación. Ciencias Ambientales, 16: 8–18.
Janzen D.H. 2000a. Costa Rica’s Area de Conservación Guanacaste: a long march to survival through non-damaging biodevelopment. Biodiversity, 1(2): 7–20.
Janzen D.H. 2000b. Wildlands as gardens. National Parks Magazine, 74(11–12): 50–51.
Janzen D.H. 2004. Setting up tropical biodiversity for conservation through non-damaging use: participation by parataxonomists. J. Appl. Ecol. 41: 181–187.
Janzen, D.H. 2005. How to conserve wild plants? Give the world the power to read them. Forward. In Plant conservation: a natural history approach. Edited by G. Krupnick and J. Kress. University of Chicago Press, Chicago, Ill. 346 pp.
Janzen D.H. and Hallwachs W. 2011. Joining inventory by parataxonomists with DNA barcoding of a large complex tropical conserved wildland in northwestern Costa Rica. PLoS ONE, 6(8): e18123.
Janzen, D.H., and Hallwachs, W. 2016. Caterpillars, pupae, butterflies, and moths of Area de Conservacion Guanacaste, northwestern Costa Rica. Available from
Janzen D.H. and Hallwachs W. 1994. Ethical aspects of the impact of humans on biodiversity. In Man and his environment. Tropical forests and the conservation of species. Edited by G.B. Marini-Bettolo. Pontificiae Academiae Scientiarum Scripta Varia, 84: 227–255.
Janzen, D.H., Hallwachs, W., Jimenez, J., and Gámez, R. 1993. The role of the parataxonomists, inventory managers and taxonomists in Costa Rica’s national biodiversity inventory. In Biodiversity prospecting. Edited by W.V. Reid, S.A. Laird, C.A. Meyer, R. Gamez, A. Sittenfeld, D.H. Janzen, M.A. Gollin, and C. Juma. World Resources Institute, Washington, D.C. pp. 223–254.
Janzen D.H., Hajibabaei M., Burns J.M., Hallwachs W., Remigio E., and Hebert P.D.N. 2005. Wedding biodiversity inventory of a large and complex Lepidoptera fauna with DNA barcoding. Philos. Trans. R. Soc. B Biol. Sci. 360(1462): 1835–1845.
Janzen D.H., Hallwachs W., Blandin P., Burns J.M., Cadiou J., Chacon I., et al. 2009. Integration of DNA barcoding into an ongoing inventory of complex tropical biodiversity. Mol. Ecol. Resour. 9(Suppl. 1): 1–26.
Janzen D.H., Hallwachs W., Burns J.M., Hajibabaei M., Bertrand C., and Hebert P.D.N. 2011. Reading the complex skipper butterfly fauna of one tropical place. PLoS ONE, 6(8): e19874.
Janzen D.H., Hallwachs W., Harvey D.J., Darrow K., Rougerie R., Hajibabaei M., et al. 2012. What happens to the traditional taxonomy when a well-known tropical saturniid moth fauna is DNA barcoded? Invertebr. Syst. 26: 478–505.
Kazmier, R. 2015a. 20 new wasp species named after Costa Rican school children. The Tico Times, 14 December 2015. Available from
Kazmier, R. 2015b. World, meet Pseudapanteles luisguillermosolisi, a new wasp species named after Costa Rica’s president. The Tico Times, 2 August 2015. Available from
Miller J.S. 2009. Generic revision of the Dioptinae (Lepidoptera, Noctuoidea, Notodontidae). Bull. Am. Mus. Nat. Hist. 321: 1–674.
Miller, S.E. 2015. DNA barcoding in floral and faunal research. In Descriptive taxonomy: the foundation of biodiversity research. Edited by M.F. Watson, C. Lyal, and C. Pendry. Cambridge University Press, Cambridge. pp. 296–311.
Novotny V. and Miller S.E. 2014. Mapping and understanding the diversity of insects in the tropics: past achievements and future directions. Austral Entomol. 53(3): 259–267.
Phillips-Rodriguez E., Powell J.A., Hallwachs W., and Janzen D.H. 2014. A synopsis of the genus Ethmia Hubner in Costa Rica: biology, distribution, and description of 22 new species (Lepidoptera, Gelechioidea, Depressariidae, Ethmiinae), with emphasis on the 42 species known from Area de Conservacion Guanacaste. ZooKeys, 461: 1–86.
Ratnasingham, S., and Hebert, P.D.N. 2007. BOLD: The barcode of life system ( Mol. Ecol. Notes, 7: 355–364.
Ratnasingham S. and Hebert P.D.N. 2013. A DNA-based registry for all animal species: the Barcode Index Number (BIN) system. PLoS ONE, 8: e66213.
Rodriguez J.J., Fernandez-Triana J.L., Smith M.A., Janzen D.H., Hallwachs W., Erwin T.L., and Whitfield J.B. 2013. Extrapolations from field studies and known faunas converge on dramatically increased estimates of global microgastrine parasitoid wasp species richness (Hymenoptera: Braconidae). Insect Conserv. Divers. 6: 530–536.
Shokralla S., Gibson J.F., Nikbakht H., Janzen D.H., Hallwachs W., and Hajibabaei M. 2014. Next-generation DNA barcoding: using next-generation sequencing to enhance and accelerate DNA barcode capture from single specimens. Mol. Ecol. Resour. 14: 892–901.
Shokralla S., Porter T.M., Gibson J.F., Dobosz R., Janzen D.H., Hallwachs W., et al. 2015. Massively parallel multiplex DNA sequencing for specimen identification using an Illumina MiSeq platform. Scientific Rep. 5: 9687.
Smith M.A., Woodley N.E., Janzen D.H., Hallwachs W., and Hebert P.D.N. 2006. DNA barcodes reveal cryptic host-specificity within the presumed polyphagous members of a genus of parasitoid flies (Diptera: Tachinidae). Proc. Natl. Acad. Sci. 103: 3657–3662.
Smith M.A., Wood D.M., Janzen D.H., Hallwachs W., and Hebert P.D.N. 2007. DNA barcodes affirm that 16 species of apparently generalist tropical parasitoid flies (Diptera, Tachinidae) are not all generalists. Proc. Natl. Acad. Sci. 104: 4967–4972.
Smith M.A., Rodriguez J.J., Whitfield J.B., Deans A.R., Janzen D.H., Hallwachs W., and Hebert P.D.N. 2008. Extreme diversity of tropical parasitoid wasps exposed by iterative integration of natural history, DNA barcoding, morphology, and collections. Proc. Natl. Acad. Sci. 105: 12359–12364.
Smith M.A., Bertrand C., Crosby K., Eveleigh E.S., Fernandez-Triana J., Fisher B.L., et al. 2012a. Wolbachia and DNA barcoding insects: patterns, potential, and problems. PLoS ONE, 7(5): e36514.
Smith M.A., Fernández-Triana J.L., Eveleigh E., Gómez J., Guclu C., Hallwachs W., et al. 2012b. DNA barcoding and the taxonomy of Microgastrinae wasps (Hymenoptera, Braconidae): impacts after 8 years and nearly 20,000 sequences. Mol. Ecol. Resour. 13(2): 168–176.
Telfer A.C., Young M., Quinn J., Perez K., Sobel C., Sones J., et al. 2015. Biodiversity inventories in high gear: DNA barcoding facilitates a rapid biotic survey of a temperate nature reserve. Biodivers. Data J. 3: e6313.
Willmott, K.R. 2003. The genus Adelpha: its systematics, biology and biogeography (Lepidoptera: Nymphalidae: Limenitidini). Scientific Publishers, Gainesville, Fla. 322 pp.

Supplementary Material

Supplementary data (gen-2016-0005suppla.pdf)
Supplementary data (gen-2016-0005supplb.xls)

Information & Authors


Published In

cover image Genome
Volume 59Number 9September 2016
Pages: 641 - 660
Editor: Vlad Dincă


Received: 16 January 2016
Accepted: 2 April 2016
Accepted manuscript online: 28 April 2016
Version of record online: 28 April 2016


This paper is part of a special issue entitled Barcodes to Biomes.

Key Words

  1. conservation
  2. taxasphere
  3. barcode
  4. parataxonomists
  5. tropics


  1. conservation
  2. taxasphère
  3. codage à barres
  4. parataxonomiste
  5. tropiques



Daniel H. Janzen [email protected]
Department of Biology, University of Pennsylvania, Philadelphia, PA 19104, USA.
Winnie Hallwachs
Department of Biology, University of Pennsylvania, Philadelphia, PA 19104, USA.


Copyright remains with the author(s) or their institution(s). This work is licensed under a Creative Commons Attribution 4.0 International License (CC BY 4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original author(s) and source are credited.

Metrics & Citations


Other Metrics


Cite As

Export Citations

If you have the appropriate software installed, you can download article citation data to the citation manager of your choice. Simply select your manager software from the list below and click Download.

Cited by

1. To and fro in the archipelago: Repeated inter-island dispersal and New Guinea’s orogeny affect diversification of Delias, the world’s largest butterfly genus
2. Forest leaf litter beetles of Taiwan: first DNA barcodes and first insight into the fauna
3. Cyphoedma n. gen. now described after a century of use, with the addition of a new species from Central America (Geometridae, Ennominae)
4. The wing interference patterns (WIPs) of Parapanteles (Braconidae, Microgastrinae): demonstrating a powerful and accessible tool for species-level identification of small and clear winged insects
5. A GLP1-receptor-agonistáktól a glükagonreceptor-agonizmussal kiegészített hármashormonreceptor-aktiválásig.
6. A revision of the parasitoid wasp genus Alphomelon Mason with the description of 30 new species (Hymenoptera, Braconidae)
7. Integrative taxonomy of a new Belciana species (Lepidoptera, Noctuidae, Dyopsinae) from northern Thailand
8. Evolutionary history of Euteliidae (Lepidoptera, Noctuoidea)
9. Revision of Belvosia Robineau-Desvoidy (Diptera, Tachinidae) and 33 new species from Area de Conservación Guanacaste in northwestern Costa Rica with a key to known North and Mesoamerican species
10. Taxonomic identification accuracy from BOLD and GenBank databases using over a thousand insect DNA barcodes from Colombia
11. Deadly and venomous Lonomia caterpillars are more than the two usual suspects
12. DNA Barcode Library of Megadiverse Lepidoptera in an Alpine Nature Park (Italy) Reveals Unexpected Species Diversity
13. Cryptic biodiversity of tropical hesperiid caterpillar-attacking parasitoid wasps: three new species of Creagrura Townes (Hymenoptera, Ichneumonidae, Cremastinae) from Costa Rica and Perú
14. Multiple Stresses of Wheat in the Detection of Traits and Genotypes of High-Performance and Stability for a Complex Interplay of Environment and Genotypes
15. Going for a Ridens Evans (Lepidoptera: Hesperiidae: Eudaminae): A New Species Reared in Area De Conservacion Guanacaste, Northwestern Costa Rica
16. Revision of the “celia clade” of Pseudodebis Forster, 1964, with Two New Species and Notes on Papilio phorcys Fabricius, 1793 (Lepidoptera: Nymphalidae: Satyrinae)
17. More discussion of minimalist species descriptions and clarifying some misconceptions contained in Meier et al. 2021
18. Many evolutionary roads led to virus domestication in ichneumonoid parasitoid wasps
19. Introduction: The Reality of Insect Declines
20. The Problems of Interpreting Changes
21. Improving Understanding of Insect Diversity
22. Defining and Countering Threats
23. Addendum to a minimalist revision of Costa Rican Braconidae: 28 new species and 23 host records
24. High resolution DNA barcode library for European butterflies reveals continental patterns of mitochondrial genetic diversity
25. A molecular phylogeny of the parasitoid wasp subfamily Rogadinae (Ichneumonoidea: Braconidae) with descriptions of three new genera
26. Revison of Metaplagia Coquillett (Diptera: Tachinidae) with description of five new species from Area de Conservación Guanacaste in northwestern Costa Rica
27. Editorial: DNA Barcodes: Controversies, Mechanisms, and Future Applications
28. Phylogenomics of Ichneumoninae (Hymenoptera, Ichneumonidae) reveals pervasive morphological convergence and the shortcomings of previous classifications
29. Response to Zamani et al. (2020): The omission of critical data in the pursuit of “revolutionary” methods to accelerate the description of species
30. A Novel Origin of Pteridivory among the New World Noctuoidea: Fern-Feeding “Litter Moths” (Erebidae: Herminiinae)
31. Minimalist revision and description of 403 new species in 11 subfamilies of Costa Rican braconid parasitoid wasps, including host records for 219 species
32. To us insectometers, it is clear that insect decline in our Costa Rican tropics is real, so let’s be kind to the survivors
33. A global food plant dataset for wild silkmoths and hawkmoths and its use in documenting polyphagy of their caterpillars (Lepidoptera: Bombycoidea: Saturniidae, Sphingidae)
34. DNA barcodes for Aotearoa New Zealand Pyraloidea (Lepidoptera)
35. A taxonomic and molecular survey of the pteridophytes of the Nectandra Cloud Forest Reserve, Costa Rica
36. Resolving Food-Web Structure
37. Área de Conservación Guanacaste, northwestern Costa Rica: Converting a tropical national park to conservation via biodevelopment
38. Using DNA-barcoded Malaise trap samples to measure impact of a geothermal energy project on the biodiversity of a Costa Rican old-growth rain forest1
39. A five-gene molecular phylogeny reveals Parapanteles Ashmead (Hymenoptera: Braconidae) to be polyphyletic as currently composed
40. Revision of Telothyria van der Wulp (Diptera: Tachinidae) and twenty-five new species from Area de Conservación Guanacaste in northwestern Costa Rica with a key to Mesoamerican species
41. Genomic evidence suggests Mesapamea remmi is an imaginary species (Lepidoptera: Noctuidae)
42. DNA barcode library for European Gelechiidae (Lepidoptera) suggests greatly underestimated species diversity
43. Annotated and illustrated world checklist of Microgastrinae parasitoid wasps (Hymenoptera, Braconidae)
44. Assessing insect biodiversity with automatic light traps in Brazil: Pearls and pitfalls of metabarcoding samples in preservative ethanol
45. Asturodes Amsel (Lepidoptera: Crambidae: Spilomelinae): Three New Species from the Western Hemisphere and Food Plant Records from Area de ConservaciÓn Guanacaste, Costa Rica
46. Contrasting Patterns of Temporal Diversification in Neotropical Butterflies: An Overview
48. A species-level taxonomic review and host associations of Glyptapanteles (Hymenoptera, Braconidae, Microgastrinae) with an emphasis on 136 new reared species from Costa Rica and Ecuador
49. Host Records for Tortricidae (Lepidoptera) Reared from Seeds and Fruits in a Thailand Rainforest
50. Monotypy Justified: Genitalia and DNA Move Ephyriades eugramma (Mabille) to New Genus Neomorphuncus Burns (Lepidoptera: Hesperiidae: Pyrginae)
51. Cryptic diversity and ranavirus infection of a critically endangered Neotropical frog before and after population collapse
52. A revolutionary protocol to describe understudied hyperdiverse taxa and overcome the taxonomic impediment
53. Aprica: a new genus and life history for the pteridivore Xanthia patula Druce, 1898 (Lepidoptera, Noctuidae)
54. Twenty-two new species in the genus Hyphantrophaga Townsend (Diptera: Tachinidae) from Area de Conservación Guanacaste, with a key to the species of Mesoamerica
55. Perspective: Where might be many tropical insects?
56. Is molecular evolution faster in the tropics?
57. A new species of Trismegistomya Reinhard (Diptera: Tachinidae) from Area de Conservación Guanacaste in northwestern Costa Rica
58. A revision of Dolichogenidea (Hymenoptera, Braconidae, Microgastrinae) with the second mediotergite broadly rectangular from Area de Conservación Guanacaste, Costa Rica
59. Genomes of skipper butterflies reveal extensive convergence of wing patterns
60. Characterization and comparison of poorly known moth communities through DNA barcoding in two Afrotropical environments in Gabon1
62. A review of Leucosigma Druce, 1908: a newly discovered case of fern-feeding and descriptions of three new species (Lepidoptera, Noctuidae)
63. Lonomia caterpillar envenoming in French Guiana reversed by the Brazilian antivenom: A successful case of international cooperation for a rare but deadly tropical hazard
64. Insect diversity in the Saharo-Arabian region: Revealing a little-studied fauna by DNA barcoding
65. Fifty Years of Mountain Passes: A Perspective on Dan Janzen’s Classic Article
66. A new species of Cotesia Cameron (Hymenoptera, Braconidae, Microgastrinae) reared from the hickory horned devil, Citheronia regalis, and luna moth, Actias luna, in east Texas
67. Revision of the Mesoamerican species of Calolydella Townsend (Diptera: Tachinidae) and description of twenty-three new species reared from caterpillars in Area de Conservación Guanacaste, northwestern Costa Rica
68. Systematics, Biology, and Evolution of Microgastrine Parasitoid Wasps
69. Aquatic Larvae of Two Acentropines, Usingeriessa onyxalis (Hampson) and Oxyelophila callista (Forbes) (Lepidoptera: Crambidae)
70. A new species of Voria Robineau-Desvoidy (Diptera: Tachinidae) from Area de Conservación Guanacaste in northwestern Costa Rica
71. Molecular methods to detect Spodoptera frugiperda in Ghana, and implications for monitoring the spread of invasive species in developing countries
72. Profile of Daniel H. Janzen
73. Nuclear genomes distinguish cryptic species suggested by their DNA barcodes and ecology
74. Descriptions of Four New Species of Struthoscelis Meyrick (Lepidoptera: Oecophoridae: Oecophorinae), One from Area De Conservación Guanacaste, Northwestern Costa Rica, Providing the First Known Biology for the Genus, and Discovery of a Novel Wing Morphology in Males
75. Upgrading protected areas to conserve wild biodiversity
76. Mapping global biodiversity connections with DNA barcodes: Lepidoptera of Pakistan
77. A biodiversity hotspot for Microgastrinae (Hymenoptera, Braconidae) in North America: annotated species checklist for Ottawa, Canada
78. From Barcodes to Biomes: Special Issues from the 6th International Barcode of Life Conference
79. From Barcodes to Biomes: Special Issues from the 6th International Barcode of Life Conference

View Options

View options


View PDF

Get Access

Login options

Check if you access through your login credentials or your institution to get full access on this article.


Click on the button below to subscribe to Genome

Purchase options

Purchase this article to get full access to it.

Restore your content access

Enter your email address to restore your content access:

Note: This functionality works only for purchases done as a guest. If you already have an account, log in to access the content to which you are entitled.





Share Options


Share the article link

Share on social media

Cookies Notification

We use cookies to improve your website experience. To learn about our use of cookies and how you can manage your cookie settings, please see our Cookie Policy.