Syst. Biol.

56(6):875–878, 2007
Copyright
c Society of Systematic Biologists
ISSN: 1063-5157 print / 1076-836X online
DOI: 10.1080/10635150701748506

Species Delimitation: New Approaches for Discovering Diversity

J OHN J. WIENS
Department of Ecology and Evolution, Stony Brook University, Stony Brook, NY 11794-5245, USA;
E-mail: [email protected]

Systematics can be considered to have two major goals: Sullivan. Of course, the field of species delimitation is too

Downloaded from https://academic.oup.com/sysbio/article/56/6/875/1653079 by guest on 15 October 2021

(1) to discover and describe species and (2) to determine large and diverse to be adequately covered by 6 or even
the phylogenetic relationships of these species. But even 11 papers. Nevertheless, the present papers provide an
a quick perusal of titles in Systematic Biology through- overview of some of the important ideas and a diversity
out most of its 15-year history will confirm an obvious of cutting-edge research in this area.
fact: most articles focus on reconstructing phylogenies We begin with a paper by Kevin de Queiroz on species
and on using these phylogenies to address evolution- concepts. Species delimitation obviously depends on
ary or biogeographic questions, with few papers on the having some idea of what species are, and it is easy to
methodology underlying this first major goal of system- imagine a symposium on species delimitation degener-
atics research (Fig. 1). ating into endless disagreements about species concepts.
Nevertheless, species delimitation, the process by Not so. In fact, there has been real progress made in think-
which species boundaries are determined and new ing about species concepts, which now makes some gen-
species are discovered, may finally be emerging as a ma- eral agreement seem possible. Part of this progress comes
jor topic in modern systematics (e.g., Sites and Marshall, from recognizing a distinction between what species are
2003). New methods for species delimitation are being and the evidence used to recognize them (e.g., Frost and
developed (e.g., Puorto et al., 2001; Templeton, 2001; Kluge 1995; Mayden 1997; de Queiroz 1998). Here, de
Wiens and Penkrot, 2002; Morando et al., 2003; Pons Queiroz (2007) discusses the underlying unity of species
et al., 2006) and compared (e.g., Marshall et al., 2006; concepts and the implications this has for species de-
Pons et al., 2006). Ambitious proposals have been made, limitation. As evidence of some of this progress, there
with the goal of accelerating the rate at which new species seems to be general agreement among almost all par-
are discovered and described, including DNA barcoding ticipants that species are lineages (e.g., de Queiroz, 2007;
(e.g., Hebert et al., 2003, 2004), DNA taxonomy (Tautz Knowles and Carstens, 2007; Shaffer and Thomson, 2007;
et al., 2003), and Web-based taxonomy (e.g., Godfray, Raxworthy et al., 2007; Rissler and Apodaca, 2007, this
2002; Scoble, 2004; Knapp et al., 2007). Controversies paper). This lets us focus on the challenge of how we go
are raging in the literature over these proposals (espe- about delimiting those lineages.
cially over DNA barcoding: Will et al., 2005; Brower, 2006; The paper by Knowles and Carstens (2007) addresses
Hickerson et al., 2006; Meier et al., 2006) and other aspects how molecular data (i.e., gene trees from DNA sequence
of species delimitation and over the consequences of new data) can be used in species delimitation. I suspect that
approaches in species delimitation for other fields, such this will be the single topic of greatest interest to most
as ecology and conservation (Agapow et al., 2004; Isaac readers of Systematic Biology. They propose a new method
et al., 2004; Padial and de la Riva, 2006). This new em- that uses coalescent simulations to test hypotheses about
phasis on species delimitation has arisen (at least in part) species limits. Their method is particularly valuable in
because of growing concern over threats to biodiversity that it can incorporate data from multiple loci and does
and the desire to describe as many species as possible as not require that species have diverged to the point of
quickly and accurately as possible before they disappear. being reciprocally monophyletic. This paper also show-
However, an increased interest in species delimitation is cases the increasing importance of population genetics
barely reflected in the pages of Systematic Biology, if at to species delimitation.
all (Fig. 1). This disconnect was my major motivation for The paper by Shaffer and Thomson (2007) focuses on
organizing a symposium for the Society of Systematic another topic of considerable interest and practical im-
Biologists on this topic. portance: how can we obtain a large number of unlinked
nuclear markers for use in delimiting species? These au-
T HE S YMPOSIUM thors describe an approach for generating SNPs (sin-
At the evolution meetings in Stony Brook last year gle nucleotide polymorphisms) using genomic resources
(2006), we held the first symposium in the history of the and demonstrate this approach with a case study involv-
Society of Systematic Biologists dedicated to the topic ing Australian turtles. They also describe some of the
of species delimitation. After a brief introduction, there methods that might be used to apply SNP data in species
were 11 speakers. Six of these led to the papers you see delimitation. Although Shaffer and Thomson (2007) em-
now; presenting but not publishing were Keith Crandall, phasize nonphylogenetic methods for species delimita-
Paul Hebert, Marshal Hedin, Loren Rieseberg, and Jack tion (unlike Knowles and Carstens, 2007), both papers

875
876 SYSTEMATIC BIOLOGY VOL. 56

even more fundable, by moving information on the tax-

onomy of every group entirely to the Web. They provide
a model for how this might work and discuss the pos-
sible pros and cons. Although some more phylogeny-
focused readers may be tempted to skip this paper, they
should note that these authors present reasonable argu-
ments for dramatic and sweeping changes in the field of
systematics.

CONTINUING CHALLENGES IN S PECIES D ELIMITATION

Again, I must reiterate that these six chapters cannot

Downloaded from https://academic.oup.com/sysbio/article/56/6/875/1653079 by guest on 15 October 2021

possibly cover the entire field of species delimitation.
They are merely a sampler. Below, I list some major areas
that were not covered and some topics for future research
FIGURE 1. The number of papers published in Systematic Biology that is needed.
that deal (at least in part) with the topic of species delimitation (closed
circles), relative to the total number of papers published in the jour-
nal (open circles), from 1992 to 2006. Both major articles and Points of Morphological Data
View were considered, but not book reviews, editorial comments, or A major omission in this symposium is the lack of pa-
brief symposium introduction essays. Given that there were so few pa-
pers that could potentially be considered as dealing with the theory or
pers on methods for delimiting species using morpho-
practice of species delimitation, I was generous about assigning papers logical data. Superficially, this may seem trivial. After
to this category. all, the majority of recognized species presumably have
been delimited and described based on morphological
differences. Why worry about methodology? In general,
clearly agree on the importance of incorporating popu- species are delimited based on one or more qualitative
lation genetics into species delimitation. or quantitative morphological characters that show no
Two papers (Raxworthy et al., 2007; Rissler and overlap with other species. This criterion is very tradi-
Apodaca, 2007) pioneer a novel use of ecological data tional but also makes sense biologically. If two species
in species delimitation. Specifically, they describe how are consistently distinguished by one or more diagnos-
GIS-based analyses of climatic data can be used, in com- tic morphological differences, then presumably there is
bination with molecular and morphological data, to help no gene flow between them (given some assumptions,
delimit species. I find this to be a particularly exciting such as the idea that each morphological difference has
development in species delimitation. Every species has a genetic basis). But what if one has only a single spec-
a geographic range, and the ecological and evolutionary imen? How can one be sure that all the individuals of
processes that limit the geographic range may be crucial that species share that diagnostic character(s)? Many re-
for creating new species, particularly allopatric species searchers might hesitate to describe a species based only
(e.g., Wiens, 2004a, 2004b; Wiens and Graham, 2005). Just on a single specimen for this reason. But is having two
think about this simple example. One could easily invest specimens really that much better than having only one?
hundreds of hours and thousands of dollars into field- What about three? With this problem in mind, Wiens
work and lab work to show that there is no or only limited and Servedio (2000) examined the sample sizes needed
gene flow between two sets of populations that repre- to have statistical confidence that a given diagnostic char-
sent two putative species by using molecular markers. acter is truly fixed (frequency = 100%) within a species.
But if the habitat separating these sets of populations In fact, being reasonably certain (i.e., allowing for a 5%
is clearly unsuitable and uncrossable for both putative error rate) that a trait is truly fixed within a species is
species (e.g., an arid valley separating mesic montane basically impossible, even if hundreds or thousands of
species), then there cannot be ongoing gene flow between individuals are sampled. Even allowing for some level of
them. In theory, this hypothesis could be tested using polymorphism in a diagnostic character (e.g., frequency
GIS-based analyses with a few hours work and almost of 95% rather than 100%), very large sample sizes may
no cost. Although previous authors might have infor- still be required to be reasonably certain that the trait
mally considered the geographic distribution of suitable is diagnostic at the desired level (Wiens and Servedio,
habitats and differences between habitats when mak- 2000).
ing species decisions, the new methods described by What about other approaches to delimiting species
Raxworthy et al. (2007) and Rissler and Apodaca (2007) with morphology? Using phylogenetic analyses of mor-
provide a way to test and incorporate this sort of “bi- phology to make species decisions remains largely un-
ological intuition” with rigorous data and statistical explored (e.g., Wiens and Penkrot, 2002). Furthermore,
analyses. the few studies that have compared the results of this ap-
The final contribution presents a discussion of the idea proach to those using diagnostic morphological charac-
of Web-based taxonomy (Godfray et al., 2007). These au- ters found these methods can give quite different species
thors argue that taxonomy could be made easier to do, limits in some cases (e.g., Wiens and Penkrot, 2002; Doan
more widely accessible to nonspecialists, and perhaps and Castoe, 2003).
2007 WIENS—SPECIES DELIMITATION 877

My point here is not that species limits based on mor- tists who actually work on describing new species (e.g.,
phological characters are often wrong or that any par- Rodman and Cody, 2003; Agnarsson and Kuntner, 2007)?
ticular approach is better than any other. My point is In summary, this symposium highlights some of the
simply that our methods for delimiting species with mor- exciting research that is going on in the area of species
phological data remain woefully understudied. This is delimitation. But I think that each paper will show that
particularly ironic in that many recent papers that have we are only beginning to address the important questions
protested the problems of “DNA taxonomy” seem to in this field.
implicitly assume that species delimitation with mor-
phology is straightforward and uncomplicated. ACKNOWLEDGMENTS
I thank the Society of Systematic Biologists for financially sup-
Molecular Data porting the symposium, all of the participants who spoke in the sym-
posium, and the six sets of authors who carried their papers through

Downloaded from https://academic.oup.com/sysbio/article/56/6/875/1653079 by guest on 15 October 2021

The development of methods for delimiting species to the final stage you see today. I also thank the reviewers for their
with DNA data is an active area of research (e.g., Sites and invaluable input on these contributions.
Marshall, 2003; Pons et al., 2006; Knowles and Carstens,
2007). A wish list for such a method might include the R EFERENCES
following that it considers both incomplete lineage sort- Agapow, P. M., O. R. P. Bininda-Emonds, K. A. Crandall, J. L. Gittleman,
ing and gene flow among populations, can integrate data G. M. Mace, J. C. Marshall, and A. Purvis. 2004. The impact of species
from multiple loci, can determine species limits without concept on biodiversity studies. Q. Rev. Biol. 79:161–179.
Agnarsson, I., and M. Kuntner. 2007. Taxonomy in a changing world:
having those limits defined a priori (i.e., it will allow Seeking solutions for a science in crisis. Syst. Biol. 56:531–539.
one to discover unanticipated species from the molec- Brower, A.V.Z. 2006. Problems with DNA barcodes for species delimi-
ular data), and can allow one to estimate the statistical tation: “Ten species” of Astraptes fulgerator reassessed (Lepidoptera:
support for species-level decisions. Clearly, we are not Hesperiidae). Syst. Biodiv. 4:127–132.
there yet. Most existing methods contain one or more of de Queiroz, K. 1998. The general lineage concept of species, species cri-
teria, and the process of speciation: A conceptual unification and
these elements, but a single method that combines all of terminological recommendations. Pages 57–75 in Endless forms:
them may still be some ways off. Species and speciation (D. J. Howard and S. H. Berlocher, eds.).
Two critical issues that are particularly relevant to Oxford University Press, Oxford, UK.
species delimitation with molecular data were brought de Queiroz, K. 2007. Species concepts and species delimitation. Syst.
Biol. 56: This issue.
up during the symposium but did not make it to the Doan, T. M., and T. A. Castoe. 2003. Using morphological and molecu-
publication stage. One is that of dealing with naturally lar evidence to infer species boundaries within Proctoporus bolivianus
fragmented populations (by Marshall Hedin). The other Werner (Squamata: Gymnophthalmidae). Herpetologica 59:433–
is making decisions about lineages that are currently 450.
introgressing (by Jack Sullivan). Both problems under- Frost, D. R., and A. G. Kluge. 1995. A consideration of epistemology
in systematic biology, with special reference to species. Cladistics
score the idea that species delimitation sometimes re- 10:259–294.
quires making cut-offs involving continuous processes Godfray, H. C. J. 2002. Challenges for taxonomy. Nature 417:17–19.
that are generating (or eliminating) species. Both topics Godfray, H. C. J., B. R. Clark, I. J. Kitching, S. J. Mayo, and M. J. Scoble.
clearly would benefit from further study in the context 2007. The Web and the structure of taxonomy. Syst. Biol. 56: This
issue.
of species delimitation. Hebert, P. D. N., A. Cywinska, S. L. Ball, and J. R. DeWaard. 2003.
Biological identifications through DNA barcodes. Proc. R. Soc. Lond.
Other Issues 270:313–321.
Hebert, P. D. N., E. H. Penton, J. Burns, D. J. Janzen, and W. Hallwachs.
There are dozens of other important and unresolved 2004. Ten species in one: DNA barcoding reveals cryptic species in the
issues that go beyond what will be covered in this sympo- neotropical skipper butterfly, Astraptes fulgerator. Proc. Natl. Acad.
sium or what I can summarize here. For example, how Sci. USA 101:14812–14817.
should asexual species be delimited? Should we even Hickerson, M. J., C. Meyer, and C. Moritz. 2006. DNA barcoding will
fail to discover new animal species over broad parameter space. Syst.
call them species? What insights might we gain from Biol. 55:729–739.
studies of speciation in the field of evolutionary biology Isaac, N. J., J. Mallet, and G. M. Mace. 2004. Taxonomic inflation:
(i.e., in terms of the evolution of intrinsic reproductive Its influence on macroecology and conservation. Trends Ecol. Evol.
isolating mechanisms) that can help inform our studies 19:464–469.
Knapp, S., A. Polaszek, and M. Watson. 2007. Spreading the word.
of species delimitation? What is the best way to inte- Nature 446:261–262.
grate data from DNA sequences and other types of data Knowles, L. L., and B. Carstens. 2007. Delimiting species without mono-
(e.g., morphology, behavior, ecology, allozymes, chromo- phyletic gene trees. Syst. Biol. 56: This issue.
somes) for delimiting species? Given that many, if not Marshall, J. C., E. Arèvalo, E. Benavides, J. Sites, and J. W. Sites, Jr. 2006.
most, new species continue to be discovered primarily Delimiting species: Comparing methods for Mendelian characters
using lizards of the Sceloporus grammicus (Squamata: Phrynosomati-
through new fieldwork, how might we accelerate the dae) complex. Evolution 60:1050–1065.
rate at which new species are actually encountered in Mayden, R. L. 1997. A hierarchy of species concepts: The denouement
the field? Raxworthy et al. (2003) describe how GIS-based in the saga of the species problem. Pages 381–424 in Species. The units
methods facilitated their discovery of new lizard species of biodiversity (M. F. Claridge, H. A. Dawah, and M. R. Wilson, eds.).
Chapman and Hall, London.
in Madagascar, but so far these methods have only been Meier, R., R. K. Shiyang, G. Vaidya, and P. K. L. Ng. 2006. DNA barcod-
used on a relatively small geographic and taxonomic ing and taxonomy in Diptera: A tale of high intraspecific variability
scale. How might we increase the number of systema- and low identification success. Syst. Biol. 55:715–728.
878 SYSTEMATIC BIOLOGY VOL. 56

Morando, M., L. J. Avila, and J. W. Sites, Jr. 2003. Sampling strate- Rodman, J. E., and J. H. Cody. 2003. The taxonomic impediment over-
gies for delimiting species: Genes, individuals, and populations come: NSF’s Partnerships for Enhancing Expertise in Taxonomy
in the Liolaemus elongatus-kriegi complex (Squamata: Liolaemi- (PEET) as a model. Syst. Biol. 52:428–435.
dae) in Andean-Patagonian South America. Syst. Biol. 52:159– Scoble, M. J. 2004. Unitary or unified taxonomy? Phil. Trans. R. Soc.
185. Lond. B 359:699–710.
Padial, J. M., and I. de la Riva. 2006. Taxonomic inflation and the stabil- Shaffer, H. B., and R. C. Thomson. 2007. Species, SNPs, and systematics:
ity of species lists: The perils of ostrich’s behavior. Syst. Biol. 55:859– Defining species in a post-genomic age. Syst. Biol. 56: This issue.
867. Sites, J. W., Jr., and J. C. Marshall. 2003. Delimiting species: A Renais-
Pons, J., T. G. Barraclough, J. Gomez-Zurita, A. Cardoso, D. P. Duran, S. sance issue in systematic biology. Trends Ecol. Evol. 18:462–470.
Hazell, S. Kamoun, W. D. Sumlin, and A. P. Vogler. 2006. Sequence- Tautz, D., P. Arctander, A. Minelli, R. H. Thomas, and A. P. Vogler. 2003.
based species delimitation for the DNA taxonomy of undescribed A plea for DNA taxonomy. Trends Ecol. Evol. 18:70–74.
insects. Syst. Biol. 55:595–609. Templeton, A. R. 2001. Using phylogeographic analyses of gene trees
Puorto, G., M. G. Salomao, R. D. G. Theakston, R. S. Thorpe, D. A. War- to test species status and processes. Mol. Ecol. 10:779–791.
rell, and W. Wuster. 2001. Combining mitochondrial DNA sequences Wiens, J. J. 2004a. Speciation and ecology revisited: Phylogenetic niche

Downloaded from https://academic.oup.com/sysbio/article/56/6/875/1653079 by guest on 15 October 2021

and morphological data to infer species boundaries: Phylogeogra- conservatism and the origin of species. Evolution 58:193–197.
phy of lanceheaded pitvipers in the Brazilian Atlantic forest, and the Wiens, J. J. 2004b. What is speciation and how should we study it? Am.
status of Bothrops pradoi (Squamata: Serpentes: Viperidae). J. Evol. Nat. 163:914–923.
Biol.14:527–538. Wiens, J. J. and C. H. Graham. 2005. Niche conservatism: Integrating
Raxworthy, C. J., C. M. Ingram, and R. G. Pearson. 2007. Species de- evolution, ecology, and conservation biology. Ann. Rev. Ecol. Evol.
limitation applications for ecological niche modeling: A review and Syst. 36:519–539.
empirical evaluation using Phelsuma day gecko groups from Mada- Wiens, J. J., and T. L. Penkrot. 2002. Delimiting species based on DNA
gascar. Syst. Biol. 56: This issue. and morphological variation and discordant species limits in spiny
Raxworthy, C. J., E. Martinez-Meyer, N. Horning, R. A. Nussbaum, G. E. lizards (Sceloporus). Syst. Biol. 51:69–91.
Schneider, M. A. Ortega-Huerta, and A. T. Peterson. 2003. Predicting Wiens, J. J., and M. R. Servedio. 2000. Species delimitation in system-
distributions of known and unknown reptile species in Madagascar. atics: Inferring diagnostic differences between species. Proc. R. Soc.
Nature 426:837–841. Lond. 267:631–636.
Rissler, L. J., and J. J. Apodaca. 2007. Ecological niche models and phy- Will, K. W., B. D. Mishler, and Q. D. Wheeler. 2005. The perils of DNA
logeography help uncover cryptic biodiversity. Syst. Biol. 56: This barcoding and the need for integrative taxonomy. Syst. Biol. 54:844–
issue. 851.