Recipe for Reusability: Biodiversity Data Journal integrated with Profeza’s CREDIT Suite

Through their new collaboration, the partners encourage publication of dynamic additional research outcomes to support reusability and reproducibility in science

In a new partnership between open-access Biodiversity Data Journal (BDJ) and workflow software development platform Profeza, authors submitting their research to the scholarly journal will be invited to prepare a Reuse Recipe Document via CREDIT Suite to encourage reusability and reproducibility in science. Once published, their articles will feature a special widget linking to additional research output, such as raw, experimental repetitions, null or negative results, protocols and datasets.

A Reuse Recipe Document is a collection of additional research outputs, which could serve as a guidelines to another researcher trying to reproduce or build on the previously published work. In contrast to a research article, it is a dynamic ‘evolving’ research item, which can be later updated and also tracked back in time, thanks to a revision history feature.

Both the Recipe Document and the Reproducible Links, which connect subsequent outputs to the original publication, are assigned with their own DOIs, so that reuse instances can be easily captured, recognised, tracked and rewarded with increased citability.

With these events appearing on both the original author’s and any reuser’s ORCID, the former can easily gain further credibility for his/her work because of his/her work’s enhanced reproducibility, while the latter increases his/her own by showcasing how he/she has put what he/she has cited into use.

Furthermore, the transparency and interconnectivity between the separate works allow for promoting intra- and inter-disciplinary collaboration between researchers.

“At BDJ, we strongly encourage our authors to use CREDIT Suite to submit any additional research outputs that could help fellow scientists speed up progress in biodiversity knowledge through reproducibility and reusability,” says Prof. Lyubomir Penev, founder of the journal and its scholarly publisher – Pensoft. “Our new partnership with Profeza is in itself a sign that collaboration and integrity in academia is the way to good open science practices.”

“Our partnership with Pensoft is a great step towards gathering crucial feedback and insight concerning reproducibility and continuity in research. This is now possible with Reuse Recipe Documents, which allow for authors and reusers to engage and team up with each other,” says Sheevendra, Co-Founder of Profeza.

First-ever fern checklist for Togo to help decision makers in the face of threats to biodiversity

Maidenhair fern (Adiantum schweinfurthii) occurring in dense forests.

Ferns and their allied species, which together comprise the pteridophytes, are vascular non-flowering plants that reproduce via spores. Many of their species are admired for their aesthetics.

However, despite being excellent bioindicators that allow for scientists and decision-makers to monitor the state of ecosystems in the face of climate change and global biodiversity crisis, these species are too often overlooked due to their relatively small size and lack of vivid colours.

Spike moss (Selaginella versicolor) with a preference for very humid and shaded forests.

To bridge the existing gaps in the knowledge about the diversity of ferns and their allied species, while also seeking to identify the ways these plants select their habitats and react to the changes occurring there later on, a research team from Togo and France launched an ambitious biodiversity project in 2013. As for the setting of their long-term study, they chose Togo – an amazingly species-rich country in Western Africa, whose flora expectedly turned out to be hugely understudied.

Having concluded their fern project in 2017, scientists Komla Elikplim Abotsi and Kouami Kokou from the Laboratory of Forestry Research, University of Lomé, Togo, who teamed up with Jean-Yves Dubuisson and Germinal Rouhan, both affiliated with the Institute of Systematics Evolution and Biodiversity (UMR 7205), France, have their first findings published in a taxonomic paper in the open access Biodiversity Data Journal.

In this first-of-a-kind checklist of Togolese ferns, the researchers record as many as 73 species previously not known to inhabit the country, including 12 species introduced for horticultural purposes. As a result of their 4-year study, the pteridophyte diversity of Togo – a country barely taking up 56,600 km² – now counts a total of 134 species.

Still, the authors believe that there are even more species waiting to be discovered on both national and global level.

“Additional investigations in the difficult to access areas of the far north of the country, and Togo Mountains are still needed to fill possible biodiversity data gaps and enable decision-makers to make the right decisions,” say the researchers.

The triangular staghorn species Platycerium stemaria living on a coffee tree branch.

In addition to their taxonomic paper, the authors are also set to publish an illustrated guide to the pteridophytes of Togo, in order to familiarise amateur botanists with this fascinating biodiversity.

 

Original source:
Abotsi KE, Kokou K, Dubuisson J-Y, Rouhan G (2018) A first checklist of the Pteridophytes of Togo (West Africa). Biodiversity Data Journal 6: e24137. https://doi.org/10.3897/BDJ.6.e24137

A new hope: One of North America’s rarest bees has its known range greatly expanded

The Macropis Cuckoo Bee is one of the rarest bees in North America, partly because of its specialized ecological associations. It is a nest parasite of oil-collecting bees of the genus Macropis which, in turn, are dependent on oil-producing flowers of the genus Lysimachia.

In fact, the cuckoo bee – which much like its feather-bearing counterpart does not build a nest of its own, but lays its eggs in those of other species instead – is so rare that it was thought to have gone extinct until it was collected in Nova Scotia, Canada, in the early 2000s. As a result, the Macropis Cuckoo Bee was brought to the attention of the Committee on the Status of Endangered Wildlife in Canada (COSEWIC).

Recently, an individual reported from Alberta, Canada, brought new hope for the survival of the species. In addition to previously collected specimens from Ontario, this record greatly expands the known range of the cuckoo.

Scientists Dr Cory S Sheffield, Royal Saskatchewan Museum, Canada, who was the one to rediscover the “extinct” species in Nova Scotia, and Jennifer Heron, British Columbia Ministry of Environment & Climate Change Strategy, present their new data, and discuss the conservation status of this species in their paper, published in the open access journal Biodiversity Data Journal.

“This species has a very interesting biology,” they say, “being a nest parasite – or cuckoo – of another group of bees that in turn have very specialized dietary needs.”

Image 2 Macropis on flower

The hosts, bees of the genus Macropis (which themselves are quite rare) are entirely dependent on plants of the primrose genus Lysimachia. Moreover, they only go after those Lysimachia species whose flowers produce oil droplets, which the insects collect and feed to their larvae. Thus, Macropis bees require these oil-producing flowers to exist just like Macropis cuckoo bees need their hosts and their nests. Curiously, this reliance, as suggested by previous studies on related European species, has made the female cuckoos develop the ability to find their host’s nests by the smell of the floral oils.

“This level of co-dependence between flower, bee, and cuckoo bee, makes for a very tenuous existence, especially for the cuckoo,” the authors comment. “The recent specimen from Alberta lets us know that the species is still out there, and is more widespread than we thought.”

In conclusion, the authors suggest that continuing to monitor for populations of rare bees, and documenting historic records, are crucial for conservation status assessments of at-risk species.

Biodiversity Data Journal provides a great venue to share this type of information with our colleagues for regional, national, and international efforts for species conservation,” they note.

###

Original source:

Sheffield C, Heron J (2018) A new western Canadian record of Epeoloides pilosulus (Cresson), with discussion of ecological associations, distribution and conservation status in Canada. Biodiversity Data Journal 6: e22837. https://doi.org/10.3897/BDJ.6.e22837

Dispatch from the field II: Students describe an elusive spider while stationed in Borneo

A mystery has long shrouded the orb-weaving spider genus Opadometa, where males and females belonging to one and the same species look nothing alike. Furthermore, the males appear to be so elusive that scientists still doubt whether both sexes are correctly linked to each other even in the best-known species.

Such is the case for Opadometa sarawakensis – a species known only from female specimens. While remarkable with their striking red and blue colors and large size, the females could not give the slightest hint about the likely appearance of the male Opadometa sarawakensis.

The red and blue female Opadometa sarawakensis
The red and blue female Opadometa sarawakensis

Nevertheless, students taking part in a recent two-week tropical ecology field course organized by the Naturalis Biodiversity Center and Leiden University, and hosted by the Danau Girang Field Centre (DGFC) on the island of Borneo, Malaysia, found a mature male spider hanging on the web of a red and blue female, later identified as Opadometa sarawakensis. Still quite striking, the male was colored in a blend of orange, gray, black, and silver.

At the brink of a long-awaited discovery and eager to describe the male, the students along with their lecturers and the field station scientific staff encountered a peril – with problematic species like the studied orb weaver they were in need for strong evidence to prove that it matched the female from the web. Furthermore, molecular DNA-based analysis was not an option at the time, since the necessary equipment was not available at DGFC.

On the other hand, being at the center of the action turned out to have advantages no less persuasive than DNA evidence. Having conducted thorough field surveys in the area, the team has concluded that the male’s observation on that particular female’s web in addition to the fact that no other Opadometa species were found in the area, was enough to prove they were indeed representatives of the same spider.

Adapting to the quite basic conditions at the DGFC laboratory, the students and their mentors put in use various items they had on hand, including smartphones paired up with headlights mounted on gooseneck clips in place of sophisticated cameras.

In the end, they gathered all the necessary data to prepare the formal description of the newly identified male.

Once they had the observations and the data, there was only one question left to answer. How could they proceed with the submission of a manuscript to a scholarly journal, so that their finding is formally announced and recognised?

submitting

Thanks to the elaborated and highly automated workflow available at the peer-reviewed open access Biodiversity Data Journal and its underlying ARPHA Writing Tool, the researchers managed to successfully compile their manuscript, including all underlying data, such as geolocations, and submit it from the field station. All in all, the authoring, peer review and publication – each step taking place within the ARPHA Platform‘s singular environment – took less than a month to complete. In fact, the paper was published within few days after being submitted.

This is the second publication in the series “Dispatch from the field”, resulting from an initiative led by spider taxonomist Dr Jeremy Miller. In 2014, another team of students and their mentors described a new species of curious one-millimetre-long spider from the Danau Girang Field Center. Both papers serve to showcase the feasibility of publication and sharing of easy to find, access and re-use biodiversity data.

“This has been a unique educational experience for the students,” says Jeremy. “They got to experience how tropical field biologists work, which is often from remote locations and without sophisticated equipment. This means that creativity and persistence are necessary to solve problems and complete a research objective. The fact that the students got to participate in advancing knowledge about this remarkable spider species by contributing to a manuscript was really exciting.”

###

Original source:

Miller J, Freund C, Rambonnet L, Koets L, Barth N, van der Linden C, Geml J, Schilthuizen M, Burger R, Goossens B (2018) Dispatch from the field II: the mystery of the red and blue Opadometa male (Araneae, Tetragnathidae, Opadometa sarawakensis). Biodiversity Data Journal6: e24777. https://doi.org/10.3897/BDJ.6.e24777

New parasitoid wasp likely uses unique saw-like spines to break out of its host body

About the size of a sesame seed, a new species of wasp from Costa Rica, named Dendrocerus scutellaris, has elaborate branched antennae that could be used for finding mates. Or hosts.

The new insect is described by PhD candidate Carolyn Trietsch, Dr. István Mikó and Dr. Andrew Deans of the Frost Entomological Museum at Penn State, USA, together with Dr. David Notton of the Natural History Museum in London, UK. Their study is published in the open access Biodiversity Data Journal.

The wasp is a parasitoid, meaning that its larvae feed on a live host insect. There are two types of parasitoids: ectoparasitoids, which lay their eggs on or near the host, so that the hatchling larvae can attach to and feed on the insect from the outside; and endoparasitoids, which lay their eggs directly inside the host, so that the larvae can eat them from the inside out.

Unfortunately, to puzzle out the new wasp’s lifestyle, the researchers could only rely on specimens collected back in 1985, which had spent the past few decades stored in the collections of the Natural History Museum of London before being loaned to the Frost Museum at Penn State for research.

What can you learn about a wasp’s lifestyle from specimens that are over 30 years old? Even though the new species has never been observed in the wild, researchers managed to learn a lot by looking at the wasps’ morphology, concluding that the species is likely an endoparasitoid.

The larva of an endoparasitoid wasp needs a safe place to develop and mature, so when it is done feeding on its host, it may stay inside the host’s body where it can develop undisturbed. Once it is fully grown, the adult wasp either chews or pushes its way out, killing the host if it isn’t already dead.

Unlike its close relatives, the new species does not have pointed mandibles for chewing. Instead, it has a series of spines along its back. While the wasp is emerging, it may rub these spines against the host and use them like a saw to cut open the body. Once emerged, it flies off to mate and continue the cycle.

“While their lives may sound gruesome, parasitoid wasps are harmless to humans and can even be helpful,” explain the scientists. “Depending on the host they parasitize, parasitoids can benefit agriculture by controlling pest insects like aphids that damage crops.”

It is currently unknown what the new species feeds upon, but naming the species and bringing it to attention is the first step in learning more about it.

###

Original source:

Trietsch C, Mikó I, Notton D, Deans A (2018) Unique extrication structure in a new megaspilid, Dendrocerus scutellaris Trietsch & Mikó (Hymenoptera: Megaspilidae). Biodiversity Data Journal 6: e22676. https://doi.org/10.3897/BDJ.6.e22676

New ‘big-armed fly’ species named after former California governor Arnold Schwarzenegger

New species can be named for all types of attributes, but Natural History Museum of Los Angeles County entomologist Brian Brown knew exactly what name to give a bizarre new fly species he discovered in the Brazilian Amazon.

“As soon as I saw those bulging legs, I knew I had to name this one after Arnold,” says Brown. “Not only is he a major cultural icon and an important person in the political realm, his autobiography gave me some hope that I could improve my body as a skinny teenager.” For these reasons, Brown says, the former governor deserves to have the new fly named in his honor.

His research article is published in the open access Biodiversity Data Journal.

The fly is impressive in other ways, Brown explains. “It is known only from one female specimen that we almost overlooked because it is so incredibly small.”

In fact, it is the world’s smallest known fly, according to Brown, who should know, since he had previously described what was formerly the world’s smallest fly, at 0.400 mm in body length. The new fly, named Megapropodiphora arnoldi, is just a fraction smaller, coming in at 0.395 mm.

MegapropodiphoraHowever, unlike the enlarged forelegs that prompted the naming, the mid- and hind legs appear to be highly reduced, and the wings reduced to tiny stubs.

Even though the fly has not been observed in the wild, Brown concludes that it is clearly a parasitoid, probably of ants or termites, based on its pointed, sharp ovipositor. He further speculates that these flies probably grab onto the hosts and “hold on for dear life” until they reach a nest or colony where they can parasitize their victims more effectively.

Brown has had considerable success finding new species of tiny flies, which he says are “the continuing frontier for insect discovery.”

Some of the more obvious, larger insects might have already been described, but by looking at smaller specimens, especially from remote, tropical sites, the entomologist finds that almost everything is new.

Even in his home city of Los Angeles, Brown and collaborators found that almost half of the phorid flies were previously unknown.

For example, it was last year that they finally figured why a secretive fly had been observed around mushrooms with no clear explanation for nearly 30 years. The revelation occurred when L.A. Bed & Breakfast owners Patsy Carter and Lisa Carter-Davis decided to alert entomologists about a phenomenon happening in their yard.

Back in 2016, Brian and his team described at once a total of twelve scuttle fly species new to science after ‘field’ trips in the backyards of houses around the city of angels.

###

Original source:

Citation: Brown B (2018) A second contender for “world’s smallest fly” (Diptera: Phoridae). Biodiversity Data Journal 6: e22396. https://doi.org/10.3897/BDJ.6.e22396

Citizen scientists discover 6 new species of beetles in Borneo

Scientists estimate that 80% of the world’s animal and plant species are still unknown. Although the work of taxonomists (whose job is to describe and name those) is appreciated by the general public, funding for taxonomy is dwindling. Moreover, while the areas hosting most of the unknown biodiversity are under threat, time is running out.

To help solve this problem, Taxon Expeditions has become the first organisation in the world to initiate field courses for citizen scientists in biodiversity hotspots, with the aim of discovering, describing, naming, and publishing new species under the slogan “You can be Darwin too”.

“Relying on extra hands means that unknown species can be discovered faster and,” says Taxon Expeditions director and biologist Dr. Iva Njunjic, “for some of that work, you don’t even need to be a trained taxonomist.”

Taxon Expedition’s first field course to the remote 30-kilometre-wide Maliau Basin in Malaysian Borneo, yielded six new species. Three of those, all tiny beetles living in rainforest leaf litter, are published today in the Biodiversity Data Journal. The other three, belonging to the family Elmidae (riffle beetles) will be published next year.

Citizen scientists discovered these species during a field exercise employing the method of ‘Winkler extraction’. Using this technique, dead leaves are collected from the rainforest floor before being sieved, so that hundreds of tiny soil-dwelling insects can be revealed.

Professor Menno Schilthuizen recognised three of those as new species. Under his guidance, the participants studied, photographed and drew the specimens in the expedition’s field lab, extracted their DNA and finalised a draft ready for publication.

The participants also came up with the original names for the three new species. English teacher Sean Otani from Japan decided to name Colenisia chungi after Malaysian entomologist Arthur Chung. The names for Clavicornaltica sabahensis and Dermatohomoeus maliauensis referring to the studied sites were suggested by staff and rangers of Maliau Conservation Area during the farewell party for the course.

All collected samples are deposited in the insect collection of Universiti Malaysia Sabah and the rest of the results – in online databases. This way, these discoveries will help other biologists working on Borneo’s biodiversity.

In March 2018, Taxon Expeditions will again head for Maliau Basin with a new group of participants, hopefully discovering more new species for science. Meanwhile, this year’s team look back on having contributed to real scientific discoveries.

“I had no idea how different, how exciting, how interesting it was going to be. It has been an amazing experience,” says retired corporate account manager Mary Erickson from Canada.

###

Original source:

Schilthuizen M, Seip LA, Otani S, Suhaimi J, Njunjic I (2017) Three new minute leaf litter beetles discovered by citizen scientists in Maliau Basin, Malaysian Borneo (Coleoptera: Leiodidae, Chrysomelidae). Biodiversity Data Journal 5: e21947. https://doi.org/10.3897/BDJ.5.e21947

Chance record of an annual mass emergence of enigmatic mantis-flies

Being neither mantids nor flies, the peculiar mantis-flies are in fact predatory lacewings which use their mantis-like forelegs to catch prey. While most mantis-flies are known to feed on spider eggs in their immature stages, the larval lifestyle of one subfamily–the Drepanicinae–has remained a mystery.

That is until James Dorey, an accomplished insect photographer and student of entomology at the University of Queensland, encountered a mass emergence of rarely-seen adults in his father’s macadamia orchard in eastern Australia, as reported in the open access Biodiversity Data Journal.

Right before James’ eyes and camera, the pupae were coming out from under the ground to grasp hold of tree trunks and undergo the final moult to emerge as adults.

Not only did he take some beautiful photos of the insects, but he also managed to capture a stunning time-lapse video of the pupae moulting into the adult stage.

The young researcher then contacted his university lecturer, David Merritt, to find out more about the phenomenon he had encountered. Together, they approached experts on mantis-flies, including Trevor Lambkin who was able to identify the insects as belonging to the mystery group, whose immature stages have so long remained unknown.

James managed to obtain some eggs from a fertile female and observed them hatch to produce tiny larvae that were immediately photographed and described in detail for the first time in a scientific journal.

“It is likely that the immature stage of these mantis-flies’ life cycle takes place underground in moist or forested habitats, perhaps explaining why it has never been recorded before,” explain the authors.

Yet to answer is the question what the immature mantis-flies feed on: is it spider eggs just like their relatives or some other underground-dwelling insect or arachnid?

“Perhaps some digging around in the macadamia orchard at the right time of year will reveal the answer,” say the researchers.

James is currently at Flinders University in Adelaide, carrying out a genetic analysis of the diversification of Fiji’s native bees.

###

Original source:

Dorey JB, Merritt DJ (2017) First observations on the life cycle and mass eclosion events in a mantis fly (Family Mantispidae) in the subfamily Drepanicinae. Biodiversity Data Journal 5: e21206. https://doi.org/10.3897/BDJ.5.e21206

Poison ivy an unlikely hero in warding off exotic invaders?

Dozens of studies have looked at the effects of Japanese knotweed on natural communities in Europe and North America. Yet Bucknell University professor Chris Martine still felt there was something important to learn about what the plant was doing along the river in his own backyard.

“The more time I spent in the forests along the Susquehanna River, the more it seemed like something was really going wrong there,” said Martine. “In addition to the prevalence of this single invasive species, it looked like the very existence of these forests was under threat.”

What Martine noticed was similar to what local nature lovers and biologists with the Pennsylvania Natural Heritage Program were also starting to see: these forests, specifically those classified as Silver Maple Floodplain Forests, were not regenerating themselves where knotweed had taken a foothold.

In a new study published in the open access Biodiversity Data Journal, Martine and two recent Bucknell alumni conclude that Japanese knotweed has not only excluded nearly all of the native understory plant species in these forests, but it has prevented the trees already established in the canopy from leaving behind more of themselves.

“If you were to fly over these forests, or even look at a Google Earth image, you’d see a nice green canopy along the river consisting of mature silver maples, river birches, and sycamores,” explained Martine. “But below that canopy there is almost nothing for tens of feet before you reach an eight-to-twelve-foot-tall thicket of knotweed. Few new trees have been able to grow through that in the last 50-60 years and our surveys found that seedlings of these species are quite rare.”

The authors suggest that as mature trees die of natural causes over the next several decades and are not replaced, these systems will shift from tree-dominated riverbank habitats to “knotweed-dominated herbaceous shrublands” incapable of supporting a rich diversity of insects, birds, and other wildlife. Loss of trees in these habitats could likely also lead to riverbank erosion and increase the severity of flood events.

The few places where knotweed has not taken over offer a bit of hope, however, from an unlikely hero: poison-ivy, which Martine calls “perhaps the least popular plant in America.”

“What we see in the data is that poison-ivy often trades understory dominance with knotweed. That is, when knotweed isn’t the big boss, poison-ivy usually is. The difference is that whereas knotweed knocks everyone else out of the system, poison-ivy is more of a team player. Many other native plants can co-occur with it and it even seems to create microhabitats that help tree seedlings get established.”

The prevalence of poison-ivy in these sites didn’t go unnoticed by undergraduate Anna Freundlich, who collected most of the plant community data — more than 1,000 data points — in a single summer as a research fellow.

“Anna developed a pretty serious methodology for avoiding a poison-ivy rash that included long sleeves, long pants, gloves, duct tape, and an intense wash-down protocol,” said her research advisor, “and even after crawling through the plant for weeks she managed to never once get a rash.”

Martine cautions against too much optimism regarding the chances of one itch-inducing native plant saving the day, however.

“Righting this ship is going to require eradicating knotweed from some of these sites, and that won’t be easy work. It will take some hard manual labor. But it’s worth doing if we want to avoid the imminent ecological catastrophe. These forests really can’t afford another half-century of us letting knotweed run wild.”

Freundlich is a now pursuing a Master’s degree in plant ecology at the University of Northern Colorado. Lead author Matt Wilson, a Bucknell Master’s student at the time of the study who analyzed the dataset, now works for the Friends of the Verde River in Cottonwood, AZ.

###

Original source:

Wilson M, Freundlich A, Martine C (2017) Understory dominance and the new climax: Impacts of Japanese knotweed (Fallopia japonica) invasion on native plant diversity and recruitment in a riparian woodland. Biodiversity Data Journal 5: e20577. https://doi.org/10.3897/BDJ.5.e20577

###

About Japanese knotweed:

Japanese knotweed is considered to be one of the toughest, most damaging and insidious plants in the world. Native to East Asia, the species has already established successfully in many parts throughout North America and Europe, where it can easily grow and invade private properties and homes. It is hardy enough to penetrate patios, house foundations and concrete. Given it spreads easily and can grow underground to a depth of 3 metres with a horizontal range of up to 7 metres, it is extremely difficult to eradicate and its treatment requires special attention. To find advice on recognition, hazards and treatment, you can check out The Ultimate Japanese Knotweed Guide.

Artificial neural networks could power up curation of natural history collections

Deep learning techniques manage to differentiate between similar plant families with up to 99 percent accuracy, Smithsonian researchers reveal

Millions, if not billions, of specimens reside in the world’s natural history collections, but most of these have not been carefully studied, or even looked at, in decades. While containing critical data for many scientific endeavors, most objects are quietly sitting in their own little cabinets of curiosity.

Thus, mass digitization of natural history collections has become a major goal at museums around the world. Having brought together numerous biologists, curators, volunteers and citizens scientists, such initiatives have already generated large datasets from these collections and provided unprecedented insight.

Now, a study, recently published in the open access Biodiversity Data Journal, suggests that the latest advances in both digitization and machine learning might together be able to assist museum curators in their efforts to care for and learn from this incredible global resource.

A team of researchers from the Smithsonian Department of BotanyData Science Lab, and Digitization Program Office recently collaborated with NVIDIA to carry out a pilot project using deep learning approaches to dig into digitized herbarium specimens.

Smithsonian researchers classifying digitized herbarium sheets.
Smithsonian researchers classifying digitized herbarium sheets.

Their study is among the first to describe the use of deep learning methods to enhance our understanding of digitized collection samples. It is also the first to demonstrate that a deep convolutional neural network–a computing system modelled after the neuron activity in animal brains that can basically learn on its own–can effectively differentiate between similar plants with an amazing accuracy of nearly 100%.

In the paper, the scientists describe two different neural networks that they trained to perform tasks on the digitized portion (currently 1.2 million specimens) of the United States National Herbarium.

The team first trained a net to automatically recognize herbarium sheets that had been stained with mercury crystals, since mercury was commonly used by some early collectors to protect the plant collections from insect damage. The second net was trained to discriminate between two families of plants that share a strikingly similar superficial appearance.

Sample herbarium specimen image of stained clubmoss
Sample herbarium specimen image of stained clubmoss.

The trained neural nets performed with 90% and 96% accuracy respectively (or 94% and 99% if the most challenging specimens were discarded), confirming that deep learning is a useful and important technology for the future analysis of digitized museum collections.

“The results can be leveraged both to improve curation and unlock new avenues of research,” conclude the scientists.

“This research paper is a wonderful proof of concept. We now know that we can apply machine learning to digitized natural history specimens to solve curatorial and identification problems. The future will be using these tools combined with large shared data sets to test fundamental hypotheses about the evolution and distribution of plants and animals,” says Dr. Laurence J. Dorr, Chair of the Smithsonian Department of Botany.

 

###

Original source:

Schuettpelz E, Frandsen P, Dikow R, Brown A, Orli S, Peters M, Metallo A, Funk V, Dorr L (2017) Applications of deep convolutional neural networks to digitized natural history collections. Biodiversity Data Journal 5: e21139. https://doi.org/10.3897/BDJ.5.e21139