The Barcode Blog

In July 2008 Wildlife Management researchers from Smithsonian Institution report on identifying otherwise unidentifiable remnants from bird-aircraft collisions (hereafter birdstrikes). Authors Dove et al point out ”birdstrikes are a serious safety hazard and a major expense for the industry”. The US Federal Aviation Agency Wildlife Mitigation site shows about 600 incidents a month over the past year, peaking in late summer and early fall, presumably coincident with fall migration. The Smithsonian Institution has been identifying birdstrike species for military and civil aviation industries since the 1960s, analyzing specimens which range from whole carcasses to bits of feathers, tissue, or blood. Prior to availability DNA testing, identifications have relied on expert examination of detailed feather morphology with comparisons to Smithsonian’s vast bird specimen collection.

Of 1,715 birdstrike samples sent to Smithsonian Insitution during 4 months in fall 2006, 821(47.9%) contained only blood or tissue. Of these, 554 (67.5%) had amplifiable mtCOI DNA, and 535 (96.6%) with DNA led to species-level identifications based on reference sequences in Barcode of Life Database (BOLD). DNA barcoding identified 128 species representing 14 orders of birds, plus 2 bat species. 19 cases were deemed inconclusive as DNA barcode matched to a set of 2 or more closely-related species with overlapping barcodes, or the recovered sequence did not meet their 98% match criteria when compared to BOLD.

There was much better success recovering DNA from dry samples (70%) than from samples collected with a wet paper towel (about 23%), which had been the standard method, pointing the way toward improving yield of DNA-based ID. The authors conclude with a call for applying “a combination of morphological and molecular methods such as DNA barcoding for efficient, cost-effective birdstrike identifications”.

Just as in CSI television series, DNA-based identification can make possible what would otherwise be impossible; in this case, identifying birds from bits of tissue and blood and making birdstrike identifications available to those without access to Smithsonian’s experts or vast collections. In addition to helping airlines, birdstrike ID will inform our knowledge of bird migration routes. There are many exciting discoveries ahead.

Like the creatures Sigourney Weaver battles in Alien, parasitoids are organisms whose larva develop in other species, usually leading to the death of the host. Insect parasitoids are widely used as biological control agents; sometimes these efforts go awry, threatening non-pest species in local ecosystems. Widespread introduction of tachnid fly parasitoid Compsilura concinnata has failed to control Gyspy moth Lymantra dispar outbreaks in eastern US, but has led to dramatic declines in large, showy Silk moths including the beautiful Luna moth Actias luna (Elkinton and Boettler. 2004). 

About 10% of named insect species are parasitoids, mostly wasps, but recognizing these often minute insects can be tricky. In November 2007 Conservation Genetics researchers from Czech Academy of Sciences and University of South Bohemia, Czech Republic; Natural History Museum, London; and Imperial College London apply COI DNA sequencing to identify wasps parasitizing Canary Islands Large White butterfly Pieris cheiranthi, which is restricted to local endemic ecosystem of relict laurel forests. Lozan et al reared 55 P. cheiranthi caterpillars from 2 Canary Island sites, and found half of the larva from forest margin and none from central forest were parasitized with what appeared to be Cotesia glomerata, native to Europe and introduced elsewhere as biocontrol agent although not in Canary Islands. 

3 of 600 C. glomerata-like adult wasps reared from Canary Island White larvae and 2 of 700 C. glomerata reared in Czech Republic from European Large White P. brassicae larvae were analyzed and found to have identical 5′ COI DNA sequences (this is the same region selected as a DNA barcode for animals). The authors conclude that European C. glomerata has been accidentally introduced to Canary Islands and is threatening a local endemic butterfly already under pressure from habitat loss. Without mentioning DNA barcoding by name, the authors conclude with a call for “increased effort to sequence morphological Costesia spp. from a broad geographical range…enabling the regular testing of species hypotheses…and the incorporation of all life stages using a single character set”. I hope that the authors can join forces and enable their sequences and associated metadata (eg collection location, specimen photographs, voucher information) from this and future Cotesia spp work to be usefully combined with growing COI barcode database (>415,000 COI barcode records from >41,000 species in BOLD so far, including 514 records from 89 named and provisional Cotesia spp). Looking ahead, routine application of DNA-based identification to parasitoids will help establish host ranges of potential biocontrol agents and detect inadvertent introduction of broad-range parasitoids that damage local ecosystems.

In June 2008 PLoS ONE, thirteen researchers from nine Canadian universities, museums, and federal agencies report on mtDNA sequences from 1360 individuals representing 195 (95%) of Canada’s 205 freshwater fish species. Hubert et al follow “best practices” established for DNA barcode records (similar criteria would enhance the value of other genetic reference data as well), namely each sequence is derived from a vouchered specimen and the barcode record includes:

• “Bi-directional sequences of at least 500 base-pairs from the approved barcode region of COI, containing no ambiguous sites
• Links to electropherogram trace files available in the NCBI Trace Archive
• Sequences for the forward and reverse PCR amplification
• Species names that refer to documented names in a taxonomic publication or other documentation of the species concept used
• Links to voucher specimens using the approved format of institutional acronym:collection code:catalog ID number”

The researchers analyzed an average of 7.6 specimens/species, with an effort to sample across species ranges. A first pass look at genetic distances among and within Canadian freshwater fish shows results similar to those of other animal groups: average variation within species, 0.3%; average minimum distance between congeneric species (nearest neighbor), 8.3%; species with overlapping mtDNA sequences, 7% (4 species pairs and 1 flock of 5 species; one of the overlapping species pairs represents probable introgression. ) Five species showed divergent clusters differing by 1-2% in different parts of their geographic ranges, and 2 species showed larger divergences (3%, 7%); some or all of these might represent distinct species. 

A challenge for science publishing is disseminating the large data sets that are increasingly generated. Restricting publication to only those studies with novel findings can lead to a kind of distortion, sometimes with serious consequences. The bias against negative studies, for example, is one factor contributing to the misculation of risks of medicines. As biodiversity genetics moves forward, we need ways to ensure high-quality work, receive appropriate academic credit, and disseminate results in a timely manner.  PLoS ONE describes itself as “an international, peer-reviewed, open-access, online publication…that welcomes reports on primary research from any scientific discipline.” It seems to me that this sort of forum with a focus on quality rather than novelty is needed as a home for publication of large genetic data sets including DNA barcode records. Making this information available in a timely manner will in turn help drive development of analytic and display tools and enable scientific applications, such as identification of fish eggs and larva shown above. 

How many giraffes were onboard the Ark?  Giraffes are classified as a single species, Giraffa camelopardalis, with five to nine subspecies proposed based on regional variation in pelage (coat pattern). In 21 dec 2007 BMC Biology (open access) researchers from University of California, Los Angeles; Center for Conservation Research, Omaha Zoo; and Mpala Research Centre, Kenya, investigate genetic variation in giraffes across African continent. 

Using biopsy darts, the authors collected skin specimens from 266 giraffes at 19 localities in West, East, and South Africa. A 654 nucleotide region of mtDNA spanning cytb and control sequences was analyzed, revealing 35 haplotypes, and the remainder of the cytb gene (1709 bp total) was sequenced from one individual from each of the 35 haplotypes. The mtDNA sequences clustered into six reciprocally monophyletic lineages, which corresponded to groupings according to pelage pattern and regional location, and were largely concordant with subspecies designations.  Genetic distances suggested these groups have been reproductively isolated for 0.3 to 1.6 MY, similar to calculated divergence times among other closely-related mammals.

Analysis of 14 nuclear microsatellites from 381 individuals at 18 locations (it is not clear whether these are the same individuals as above) recovered the same six groups and suggested additional genetic subdivisions within some groups. Although at least some of the genetically and pelage-defined clusters have overlapping or adjacent ranges without geographic barriers, only three (0.8%) of individuals were identified as hybrids. These findings raise interesting questions about giraffe biology; for example, is there behavioral isolation perhaps based on visual recognition of pelage patterns? 

It is impressive that species can be overlooked in such large, boldly patterned, iconic animals.  Might there be similar divisions within the numerous species of small, brown, rarely seen mammals? Routine DNA analysis of a standardized mtDNA region (aka DNA barcoding) will help discover how finely divided animal biodiversity is.  Wilson and Reeder’s Mammal Species of the World, Third Edition lists 5,419 species, so this appears to be an achievable goal for our mammalian kin (list available online http://nmnhgoph.si.edu/msw/).  I hope the authors include barcode region COI in their next analyses, so their data can be easily combined with other data sets, including the 28,560 mammalian barcode records in BOLD to date. 

High school students in San Diego are using DNA barcoding to survey life in San Diego Bay, ranging from invasive mussels, to gastropod egg masses on eel grass, zooplankton and endangered species. Under leadership of Dr. Jay Vavra, students developed a simplified protocol for DNA extraction and amplification that can be performed in the high school’s biotechnology laboratory, and successfully identified dried jerky meat from ostrich, turkey, and beef. They have established a collaboration with East African graduate students to apply this approach to identifying bushmeat from endangered species in local African markets.

Just two years ago, in Syst Biol 55: 844, 2006 some taxonomists worried whether DNA barcoding would ever be useful: ”The truth is that DNA barcoding will not have any meaningful use for the general public and even when a portable barcoder becomes available it will not lead to any increase in the biological literacy of the man in the street.” Authors Cameron et al might want to visit their local high school!

For high school students DNA barcoding seems as natural as texting. You can analyze DNA to identify species? Sure. You only need a trace sample, like a hair or a bit of dried skin? Sure, just like CSI shows. On the other side, many identification keys are not practical for most persons who would like to identify what is in their backyard.

“It is impossible to describe biological diversity with traditional approaches. Molecular methods are the way forward–especially, perhaps in the form of DNA barcodes” observed Mark Blaxter in a 2003 Nature commentary, “Counting angels with DNA“, on the first paper proposing “DNA barcodes” as a standardized method for identifying species.

Five years on, how do things look? I believe the scientific and practical value of molecular (ie DNA-based) identification of species is established. Of course visual methods will often be the method of choice to identify specimens in the lab and in the field, but the standardized genetic libraries (aka DNA barcode databases) linked to specimens stored in museums are an increasingly valuable reference for assigning specimens to known species and as a means of species discovery (for more, see www.barcoding.si.edu; www.barcodinglife.org).

In addition to helping identify what is already known, DNA analysis can reveal what would otherwise remain hidden. In 16 May 2008 Science, researchers from Cornell College, Smithsonian Institution, US Department of Agriculture, University of Maryland, and Ithaca College use DNA to reveal hidden diversity in Blepharoneura, a neotropical genus of tephritid fruit flies that feeds within the flowers or fruits of plants in the cucumber family (Cucurbitaceae). To skip to the conclusion, mtCOI sequencing of 2857 flies reared from 24 cucurbit host species collected in six locations in Central and South America revealed 52 morphologically similar species (most were entirely indistinguishable) with “highly conserved patterns of specificity to host taxa and host parts.” Nuclear genes showed the same pattern of genetic clustering as mitochondrial COI.   

This report highlights an exciting scientific challenge raised by genetic surveys of biodiversity, including DNA barcoding: there are far more species, each with biologically specialized traits, than anyone has recognized. Condon et al report ”diversity exceeding the original morphologic estimates by an order of magnitude” but conclude this must be an underestimate because of limited sampling (usually along single transects in one season at 6 sites in 5 countries), considering the vast expanse of neotropical forests in Central and South America. Also they used a conservative 4% mtCOI divergence as a cutoff (if 1% cutoff were used, an additional 10 species would be recognized, and several generalist species would be split into narrowly specialized ones).

In closing, I wish the authors had sequenced the barcode region of COI (they analyzed a 693 bp fragment from the 3′ end of the gene which does not overlap with the 5′ DNA barcode region). It would be interesting for example to compile these results with data from the tephritid fly initiative, which aims to collect DNA barcodes from the 4500 known tephritid species. Perhaps these valuable Blepharoneura DNA samples can be reanalyzed for barcode region COI.

On 15 may 2008 an international assembly of bee experts gathered at York University and announced a new initiative to DNA barcode world’s bees. Some snippets from news reports:

“According to York biology professor Laurence Packer, who’s leading the group’s efforts, precisely 19,231 different kinds of bees have been identified. But he thinks there might be another 5,000 or more species out there waiting.” (Toronto Star) 

“There are two reasons why bee species would benefit from a barcode name tag. “Most of the specimens in museums are not identified, and the ones that are identified are only 60 to 70 percent correctly identified,” says Dr. Packer.” (University Affairs)

Early results, as in figure from poster at right, suggest that DNA barcoding will help sort out bee taxonomy and speed discovery of new species, with benefits to society and science.  Bees are essential pollinators for many of the world’s plants, including many endangered species, and approximately 1/3 of world’s food is derived directly or indirectly from bee-pollinated plants.

In common with other Hymenoptera (bees, wasps, ants), bees exhibit haplodiploidy (males haploid, females diploid), many species are eusocial (live in large colonies with single queens), and have greatly accelerated rates of mitochondrial evolution. Are these factors causally linked? Looking at differences among and within world’s bees may help provide fundamental insight into mitochondrial biology and evolution.  

As Bruegel the Elder recognized in 1557, ”big fish eat little fish”. Determining exactly what eats what remains a fundamental question in modern ecology and this task is particularly challenging for biologists studying small organisms, which make up the bulk of the biomass in ecosystems. To add interest, a number of these tiny creatures with unknown diets are medically and/or economically important disease vectors.

In Mol Ecol Resources May 2008 researchers from University of California, Irvine, Kenya Medical Research Insitute, and Kanazawa University, Japan, apply DNA analysis identify gut contents of larval Anopheles gambiae complex mosquitoes, the major malaria vector in Africa. 

Authors Garros, Ngugi, Githeko, Tuno, and Yan collected anopheline larva near Kisumu in western Kenya, dissected stomach contents of third and fourth instar forms, extracted DNA, and amplified an 800 bp fragment of nuclear 18s rRNA. A separate PCR assay was used to confirm species identity (five were A. gambiae s.s. and 68 were sister species A. arabiensis). According to authors, 18s rRNA was analyzed rather than COI because “more sequences are available [for 18s than for COI] in databases for plants, fungi, and protists”. I note there are now many research groups working on “plants, fungi, and protists” so it should be possible to achieve greater resolution in this sort of study as the DNA barcode libraries are built up.

The PCR products from gut contents were first screened with a restriction endonuclease known to digest mosquito but not algal 18s.  355 PCR products from eight randomly selected larvae were screened, yielding 14 unique non-mosquito sequences. Best matches in GenBank blast results clustered into 3 main clades: green algae (7), fungi (5), and unknown eukaryotes (2). The authors conclude “the method presented in this study may be a promising tool to investigate natural diets of [anopheline] larvae”. Looking ahead, “such studies will not only improve our understanding of Anopheles larval ecology, but also provide fundamental information to facilitate the develpment of novel larval control tools.”

This study is one demonstration of how routine DNA analysis combined with expanding DNA barcode libraries will help reveal and monitor changes in a multiplicity of tiny food webs. More generally, routine DNA analysis combined with reliable DNA reference libraries opens wide avenues for rapid progress in understanding how the diversity of organisms interact, with benefits to society and science. Continued development of robust, inexpensive methods for analyzing DNA from the various types of biological samples and methods of matching the results to well-curated DNA reference libraries will speed this along.   

Knowledge of how species are distributed is essential for understanding evolution and ecology, and monitoring enables detecting invasive species and recognizing effects of biological and physical environmental change. That’s easy to say, but many species are small, secretive, or difficult to distinguish from one another, so mapping species distributions requires enormous human effort and ongoing monitoring requires even more. I venture a guess that we have good monitoring for 10,000 or so plant and animal species, mostly large animals and those plants and animals of economic importance, and static distribution maps for about 100,000 species, out of a total of 1.7 million named species and not counting the projected total of 10 million species that might eventually be recognized when surveying biodiversity approaches completion. 

Just as high-resolution satellite mapping has surpassed most ground surveys in accuracy, speed, and cost, we need efficient technologies that can help detect and monitor species from environmental samples. In 9 April 2008 Early Online Biol Lett researchers from Universite Joseph Fourier and Universite de Savoie, France, and Universita Milano Bicocca, Italy, apply high-sensitivity DNA analysis to detect presence or absence of American Bullfrog Rana catesbeiana, a globally invasive species. PCR amplification of a diagnostic 79 bp fragment of mitochondrial gene cyt b using species-specific primers (no amplification of samples from the 5 locally native Rana sp). Three 15 mL water samples were collected from each of 9 ponds (surface area 1000-10,000 m^2) in France, including ”three ponds where bullfrogs were present at low density (one to two adults seen, no reproduction), three ponds where bullfrogs were present at high density (dozens of adults and thousands of tadpoles), and three ponds where bullfrogs have never been detected.” Each sample was analyzed 3-5 times, giving 9-15 repeats per pond. R. catesbiana DNA was never detected in the ponds with no bullfrogs and was detected in water samples from all three high-density ponds, with most (79%) of replications positive. Bullfrog DNA was also detected in all low-density ponds, although fewer of the replications were postive ( 37%).

Ficetola et al observe “our approach allows the reliable detection of secretive organisms in wetlands without direct observation.”  The authors conclude “The ongoing effort to develop DNA barcodes for identifying species  from degraded DNA (Hajibabaei et al 2006; Taberlet et al 2007) will make our approach applicable to more and more plant and animals species…These factors will soon make possible the assessment of the current biodiversity of macro-organisms from environmental samples.” 

Like satellite mapping 20 years ago, DNA-based environmental monitoring of biodiversity, aided by growing DNA barcode libraries, is set to expand rapidly.

Alien species sometimes damage native landscapes. In Voyage of the Beagle, in entry dated September 19, 1832, Darwin describes the spread of an introduced European thistle Cyanara cardunculus in Banda Oriental, now Uruguay: ”very many (probably several hundred) square miles are covered by one mass of these prickly plants, and are impenetrable by man or beast. Over the undulating plains, where these great beds occur, nothing else can live…I doubt whether any case is on record, of an invasion on so grand a scale of one plant over the aborigines.”

The challenge is to recognize invasive species before they become established. In 11 January 2008 Polar Biology researchers from Stellenbosch University and University of Western Ontario apply DNA barcoding to otherwise unrecognizable moth larvae on sub-Antarctic Marion Island. The indigenous Lepidoptera on Marion Island comprises 2 or 3 flightless moths, and the occassional adult winged moths or butterflies have been assumed to be transients arrived on fresh produce.

In April 2004, 3 noctuid moth larvae were found in an abandoned Wandering Albatross nest, a common habitat for one of the indigenous moth species. The larvae could be tentatively identified only to genus level and so rearing was attempted, with one larva dying after several months of pupating (as an aside, this is one example of how morphologic identifications can be laborious and/or incomplete, even for experts). The final larva was killed and preserved for DNA study; COI DNA barcode region was amplified using standard Folmer primers. The Marion Island moth larva barcode clustered with the 40 or so Black Cutworm Agrotis ipsilon sequences in BOLD, and was distinct from COI sequences of the other 18 Agrotis species in BOLD. Agrotis ipsilon is a common pest that feeds on a wide variety of plants. The authors conclude that Agrotis ipsilon is an established alien species with the potential to disrupt local ecosystems and that ”steps be taken to eradicate the species from Marion Island.”

It is easy to predict that rapid identification of potential invasive alien species will be a major application of DNA barcoding, with direct economic and ecosystem benefits.