C

J

a

b

c

d

6e

RA

A

2asccltgma©

K

I

goauau

0d

Available online at www.sciencedirect.com

European Journal of Protistology 48 (2012) 96–102

oncepts in protistology: Species definitions and boundaries

ens Boenigka,∗, Marc Ereshefskyb, Kerstin Hoef-Emdenc, James Malletd, David Basse

General Botany, University Duisburg-Essen, Universitätsstr, 5, 45117 Essen, GermanyDepartment of Philosophy, University of Calgary, 618 Campus Place N.W., Calgary, CanadaBotanical Institute, Cologne Biocenter, University of Cologne, Zülpicher Str. 47b, 50674 Cologne, GermanyGalton Laboratory, Department of Genetics, Evolution and Environment (GEE), University College London, Gower Street, London WC1EBT, United KingdomDepartment of Zoology, The Natural History Museum, Cromwell Road, London SW7 5BD, United Kingdom

eceived 24 August 2011; received in revised form 28 November 2011; accepted 28 November 2011vailable online 29 December 2011

bstract

This paper summarises the Symposium ‘Concepts in Protistology’, during the VI European Congress of Protistology, Berlin,5–29 July 2011. There is an increasing focus on cataloguing the number of species on earth, species barcoding initiatives,nd the increasing need to reconcile molecular with morphological data in protists within a taxonomic framework. We identifyeveral obstructions to defining species in protists, including the high incidence of asexuality, high levels of both morphologicalonservation and evolutionary convergence, high levels of genetic diversity that cannot so far be correlated with phenotypicharacters, conflicting signals between both genetic and phenotypic taxonomic markers, and different requirements and chal-enges of species definition in different protist groups. We assert that there is no species ‘category’ for protists, and recommendhat a working definition of species is clarified on a case-by-case basis. Thus, a consensus approach may emerge within protistroups, but any one approach is unlikely to encompass a wide phylogenetic range. However, as long as clarity of intent and

ethod is maintained, the utility of the term ‘species’ in protists will also be maintained as a reproducible and convenient (if

rtificial) way of referring to particular lineages within a tightly defined context.2011 Elsevier GmbH. All rights reserved.

OTU

lulnae

eywords: Species concept; Protist; SSU rDNA; ITS rDNA; COI;

ntroduction

Although the high-level taxonomy of protists is a well-rounded (although not resolved) area of research, thepposite end of the taxonomic spectrum, alpha-taxonomy,nd what constitutes protist species are issues fraught with

ncertainty and disagreement. The fact that there is no gener-lly accepted basis for delimiting species in protists has manynfortunate consequences, prime among them being (1) a

∗Corresponding author.E-mail address: jens.boenigk@uni-due.de (J. Boenigk).

((ort

‘

932-4739/$ – see front matter © 2011 Elsevier GmbH. All rights reserved.oi:10.1016/j.ejop.2011.11.004

ack of basic communicability about fundamental biologicalnits (with obvious negative implications for barcoding), (2)ack of clarity regarding their evolutionary and ecological sig-ificance, and (3) a drastic underestimation of protist diversitynd importance in more general biodiversity papers. A goodxample of the latter is shown by Mora et al. (2011), in whichby no fault of theirs) estimates of species numbers of protistsin particular) are unrealistically depressed, in part becausef the problem with defining ‘species’ and also because of the

apidly changing and relatively unstructured nature of protistaxonomy overall.

The ECOP workshop did not set out to ‘decide’ on thebest’ species concept to apply to protists, but rather to assess

rnal of

tasaoa

Tm

awbliM“a(t1fdloi

asnkookao‘cmtpmps

fiu2Otcwdi

fpao

A

cofowihavblc

zssvaD–tttrcsdfdrpe

atbiahmaono

J. Boenigk et al. / European Jou

o what extent such a concept is possible, or even desirable,nd what the best working bases for protist alpha-taxonomyhould be. The main body of this paper summarises the issuesnd outcomes of the interactive process culminating in theral presentations. The individual contributions of the authorsre reproduced in the three supplementary files.

he historical context: shifts in the conception ofegasystematics and species

The era of protistology has seen major methodologicaldvances starting with the invention of simple microscopes,hich allowed the first visualization of individual microbesy Leeuwenhoek in 1674 (cf. Dobell 1932). Early observersumped the small microorganisms into broad categories, fornstance as “Vermis punctiformis” (cf. the genus Monas

üller 1773), as “Punktthierchen”, “Kugelthierchen” andOvalthierchen” (von Gleichen and Freiherr 1778; comparelso Monas punctum Ehrenberg 1838), or in the genus ChaosLinnaeus 1758). The general system achieved by the end ofhe 19th century (cf. Bütschli 1880–89; Doflein 1916; Kent880–81; Pascher and Lemmermann 1914) remained in placeor most of the 20th century. The general quality of the taxoniagnoses of protists improved little during this time and wasargely based on light microscopy, even though the inventionf electron microscopy in the 1930s (cf. Agar 1996) furtherncreased taxonomic resolution.

The advent of molecular methods in the 1990s providedvery different perspective on microbial eukaryotic diver-

ity. Many new genetic lineages were detected, a substantialumber of which were highly distinct from those previouslynown, suggesting novel elements of biodiversity at high tax-nomic ranks. At the other end of the scale, very high levelsf genetic diversity were found within and around alreadynown lineages, suggesting an abundance of cryptic speciesnd sister taxa. Another revelation was that protist morphol-gy is highly and often surprisingly convergent – for exampleamoebae’ are found over most of the tree of life, classi-al Heliozoa and Radiolaria are several fold polphyletic, andany lineages that were thought to be fungal were shown

o branch elsewhere on the tree when placed by molecularhylogenetics (fungal analogues such as oomycetes, plas-odiophorids, labyrinthulids, etc.). Thus another layer of

oorly resolved complexity was added to the already muddledtory emerging from earlier morphological studies.

A recent approach for overcoming the drawbacks of insuf-cient taxonomic coverage of the studied diversity is these of operational taxonomic units (OTUs) (e.g. Green et al.004). Especially in sequence-based diversity studies, suchTUs are defined by using sequence similarity or distance

hresholds (e.g., Schloss and Handelsman 2005). In many

ases these OTUs were treated synonymous to species, andere used, for instance, for the estimation of species richnessata. The taxonomic power of lumping together organismsn OTUs based on sequence similarities was never evaluated

ssW

Protistology 48 (2012) 96–102 97

or eukaryotic microorganisms. However, in the case ofrokaryotes the underestimation of real species numbers bypproaches defining OTUs based on 16S rRNA sequences isbvious (Stackebrandt and Ebers 2006).

consistent species concept for protists?

Biodiversity research and ecology rely on safe identifi-ations of species and on reproducible species counts, butften this requirement is not met. Various problems resultrom partially inconsistent species concepts, insufficient tax-nomic coverage (undersampling), and uncertainty abouthich characters should best be used as bases for decid-

ng species boundaries. Recent methodological progress hasighlighted severe inconsistencies between the conceptualnd the practical historic approaches to species and biodi-ersity. The dispute is currently stirred up by inconsistenciesetween molecular phylogenies on the one hand and morpho-ogical species denominations and traditional classificationoncepts on the other.

The conceptual conflict embraces the differences betweenoological, botanical, and microbiological concepts ofpecies. Due to ambiguous, contradictory and/or inconsistentpecies descriptions, species numbers obtained from biodi-ersity surveys using the traditional morphospecies conceptre not comparable to those obtained from environmentalNA. As different methodological approaches are – in partlinked to different concepts of species and of diversity

he dispute on protist species, protist diversity, and pro-ist systematics often fails to differentiate differences inhe conceptual basis from methodological limitations andeal variation. For instance, the existence of newly revealedryptic species obviously results in increasing biodiver-ity estimates. By contrast, taxon-independent (OTU-based)iversity studies as often applied for microorganisms tend toail to resolve species and thus tend to underestimate bio-iversity. This basic problem has many consequences, andepresents a serious obstacle to understanding key aspects ofrotist biology and ecology, for example the ‘everything isverywhere’ debate and the perception of protist biodiversity.

In this paper we review the difficulties and challenges oflpha-taxonomy and species delineation. We do not attempto review each protist group in an attempt to decide how theasic taxonomic units in each group should be defined; thats up to the experts working on them. However, examplesre given from some of the groups in which the authors doave expertise, and the inclusion of some very well-studiedetazoa provides a phylogenetically distant but conceptu-

lly relevant and informative perspective on issues which areften thought to be particularly problematic in protists. A sig-ificant element of our discussion is not associated with anyrganismal group, but considers the ‘problem’ from a philo-

ophical/logical standpoint, which we feel offers perhaps thetrongest direction and encouragement for alpha-taxonomy.e see the key issues in protist alpha-taxonomy to be:

9 rnal of

(

(

(

(

(

(

M

ssdptw

(

(

(

lTmatd

R

bobMbtdggvIwicsemdchitdoei

seirtcmrstosttabhm‘gol

8 J. Boenigk et al. / European Jou

1) confusion arising from high levels of evolutionary con-vergence and morphological conservation in protists,

2) morphological plasticity in many protist groups can leadto unreliable morphological diagnoses,

3) different relative rates of phenotypic and genotypicdivergence among protist groups, meaning that a singlegenetic marker will provide very inconsistent taxonomicsignals across protists as a whole,

4) uncertainty about which characters are best used to dis-tinguish species,

5) high levels of diversity revealed by molecular techniquesthat are not associated with known cells, and genotypicdiversity being much higher than (easily measured) mor-phological diversity,

6) limited genetic information about many protist lineages.

aterials and Methods

Boenigk and Bass invited three contributors to their ses-ion, whose work is concerned with different aspects ofpecies delineation, definitions, and concepts. Overall, weid not attempt to be taxonomically comprehensive; rather torovide an intellectual framework for the future considera-ion of ‘the species problem’ in protists. The invited speakersere

1) James Mallet, Professor of Biological Diversity at Uni-versity College London (Galton Laboratory, Departmentof Genetics, Evolution and Environment (GEE)). Oneof Mallet’s core research questions is how new speciesarise; his group focuses on the evolution and ecology ofbutterflies and moths (Lepidoptera), particularly near thespecies boundary.

2) Marc Ereshefsky, Professor of Philosophy at the Univer-sity of Calgary, Canada, who has written many papersand books on species and taxonomy, and lists his researchinterests as: Species; Natural Kinds; Kinds and Individ-uals in Biology; Homology; Historicity.

3) Kerstin Hoef-Emden from the Botanical Institute, Uni-versity of Cologne, Germany; an expert in the taxonomyand biology of cryptophytes, and recipient of the TygeChristensen Prize from the International PhycologicalSociety in recognition of the best algal paper publishedin Phycologia during 2007.

Each contributor provided a written document for this pub-ication as well as an oral presentation at the Symposium.hese (edited) documents are provided as supplementaryaterials (Suppl. texts 1–3), and are referred to in the Results

nd Discussion, which attempts to synthesise and summariseheir content with that of the oral presentations and the groupiscussion that ended the Symposium.

isii

Protistology 48 (2012) 96–102

esults and Discussion

Many of the problems in defining species in protists havee attributed to differences between protists and multicelullarrganisms (multicells), in which species boundaries haveeen thought to be in many cases clearer and confirmable.ulticells are often morphologically distinctive, and it was

ased on discrete character sets between sets of populationshat the concept of ‘species’ was generated. Protists are small,ifficult to observe, and morphologically relatively homo-eneous (certainly within groups, but also between manyroups that we now know to be only distantly related), pro-iding relatively few characters with which to separate them.n addition, they can be phenotypically very variable evenithin a clonal line. Combined with the difficulty of observ-

ng a single specimen over a period long enough to properlyharacterise it, this means that many light microscopy-basedpecies descriptions are at best incomplete and frequentlyffectively useless or misleading. The advent of molecularethods and the use of marker genes to count and measure

iversity has in some respects made the situation even moreomplicated. This showed that most protist morphospeciesarbour high levels of genetic diversity that often cannot eas-ly be associated with phenotypic differences, either becausehe latter are too small or difficult to detect, or because theirection and scale of one kind of distance has a complicatedr contradictory relationship with the other (i.e. very high lev-ls of morphological conservation and convergent evolutionn protists).

However, molecular findings have also complicatedpecies delineation in multicells. Mallet (Suppl. text 1)xplains that analyses of sequence data have tended toncrease the number of species in animal groups, by ‘re-ecognising’ varieties and races that had been lumpedogether as polytypic species by the biological species con-ept (BSC). One of the intentions of the BSC was partly toinimise confusion caused by the multitude of geographical

aces/varieties within biological species by not recognisinguch varieties as separate species unless they were reproduc-ively isolated in sympatry. Mallet explains that at least 10%f all (metazoan) species hybridise with at least one otherpecies (and not just sister species), and that it is a fallacyhat species delimination in macro-organisms is more robusthan in microbes. There is often a continuum of intermedi-te types, which means that a consistent species concept wille elusive. The genomic consequences of introgression andorizontal gene transfer mean that the use of single geneticarkers (or barcodes) is fundamentally flawed. Different

good’ species can share the same barcode, and hybrid intro-ression can spread barcodes among species. There is a lackf correspondence between phenotypic characters, genome-evel divergence, and individual marker genes. Therefore, its necessary to integrate multiple lines of evidence to resolve

pecies boundaries. The broader implications of these find-ngs lead Mallet to argue that the idea of a ‘true species reality’s unsound, partly because different scientists believe that

rnal of

ds2lpt

sihbicoltrtpictTsDwTpilttpa

etoptcsptaasip

(p(Ttf

edgawagttiIaailG1hiv

calcsseooftgmlMscsdtdla

uhulmsc

J. Boenigk et al. / European Jou

ifferent processes are important as bases for delimitingpecies. Therefore, attempting to select from one of the c.4 existing species concepts is futile and ignores the bio-ogical realities. Species delimitation should be based on theracticalities of each case; i.e. the meaning of ‘species’ is inheir use, not as absolute entities.

These conclusions are strongly concordant with Ereshef-ky’s (Suppl. text 2) arguments. Both authors agree that theres no species category. Scientists have been trying in vain forundreds of years to find the correct definition of ‘species’,ut the simplest conclusion is that one does not exist. Themportant thing to notice about the variety of species con-epts of which Ereshefsky gives a partial but illuminatingverview (Suppl. text 2) is that they pick out different types ofineages, by defining their collection in different ways. Whathis should tell us is not that we need to look harder for the cor-ect definition of ‘species’, but that there are many differentypes of fundamental lineages in the world: ‘species’ com-rise a multitude of different types of taxa. Therefore, if theres no species category, then we should stop trying to find theorrect definition of ‘species’. However, this does not meanhat species do not exist – they are real lineages in nature.he key requirement is that scientists are explicit about whichpecies delineation approach is being taken. Ereshefsky citesarwin: “I mean by species, those collections of individuals,hich have commonly been so designated by naturalists”.his applies to all life, but at the beginning of the 21st Centuryerhaps with particular and valid force to protists, which aren the dawn of their true elucidation at the alpha-taxonomicevel, as so many multicellular taxa were when Darwin wrotehose words. Furthermore, the advent of molecular data resetshe taxonomic clock in many important respects, and com-els us to reconsider how species may be most effectivelynd informatively delineated.

These considerations have already been adopted andmbraced to an extent by many protistologists. The contribu-ion of Hoef-Emden and Bass (Suppl. text 3) gives examplesf how alpha-taxonomic decisions have been made in crypto-hytes, and cercomonad and glissomonad Cercozoa, wherehe phenotypic characters discernable by light microscopyan be difficult to distinguish and/or highly plastic. A con-ensus approach is emerging that combines and reconcileshenotypic with genotypic data to delineate species, in whichhe importance and influence of sequence-based charactersre particularly emphasised in cases where phenotypic sep-ration is difficult. Hoef-Emden and Bass advocate highampling within and between closely related lineages to max-mise consistency and robustness of inference, and a goodhylogenetic model on which to map phenotypic characters.

Unlike in many multicellular groups, where mitochondrialmtDNA) genes are commonly used as barcodes/markers, inrotists the genes and spacer regions on the ribosomal RNA

rRNA) gene cistron are currently most commonly used.here are several reasons for this: this region (principally 18S,

he small subunit rRNA gene) has been routinely sequencedrom culture collection and new strains, as it is relatively

pl(t

Protistology 48 (2012) 96–102 99

asily PCR-amplified and sequenced, primer sites are welleveloped, and SSU rDNA can be used to generate relativelyood single-gene phylogenies. However, the rDNA regionslso have disadvantages, principally that they are multicopyith variable levels of intragenomic variation among groups,

nd the SSU is generally considered too conserved to distin-uish among species. Therefore, in the better-known groupshe faster evolving ITS rDNA is used for species delimita-ion, taking the cue from Annette Coleman’s investigationsnto the utility of compensatory base changes in particularTS2 regions as good markers of biological species bound-ries. The mitochondrial COI gene is espoused by some asgood ‘species’ marker in some protist groups. However,

ts widespread use is currently not feasible as many groupsack information about it. For example, a rough assessment ofenBank entries for Cercozoa show ∼3700 entries for SSU,10 for LSU, 500 for the ITS regions, and 10–20 for COI. Theigh incidence of derived mitochondria and anaerobic groupsn protists relative to other eukaryotes also preclude the uni-ersal use of mitochondrial markers as protist barcodes.

It is important to realise that any ‘species level’ markerurrently under consideration can only ever be exactly that –marker for taxonomic boundaries, rather than a directly

inked indicator. None of these is associated with sexualompatibility or genetic isolation. However, we do not con-ider this to be a fatal problem as the benefits of usinguch highly sampled and widely accessible (i.e. relativelyasily sequenced regions from a wide phylogenetic rangef protists) markers outweighs the theoretical desirabilityf a ‘functionally relevant’ marker. If (when) the latter areound they are almost certainly going to (and should) beaxon group-specific. Research on the vast majority of protistroups is not yet so advanced. Importantly, the sexuality ofost protist groups is unknown – many are assumed to be

ong-term asexuals. These are discussed in more detail byallet (Suppl. text 1); however we note here that their pre-

umed asexuality does not preclude the use of any of theommonly used marker regions (including the ‘biologicalpecies marker’ ITS2 rDNA) being used as one of the specieselimitation criteria. As Mallet states, “It seems most sensibleo call them [asexual protist lineages] separate species if theyisplay major disease-causing differences or have other eco-ogical or biological differences that seem worth recognisings species”.

Mallet also points out that the very large effective pop-lation sizes of protists should theoretically harbour veryigh levels of neutral or nearly neutral polymorphism in sex-al species. In these cases, coalescence times will be veryarge, which could lead to much sharing of ancestral poly-

orphism between recent species. Consequently, the use ofingle markers to delimit species becomes very risky, as thisould create many apparent species that bear no relation with

atterns elsewhere in the genome. Asexual species with veryarge populations can become highly genetically structuredhigh levels of disequilibrium), also risking overestimatinghe number of units that can reasonably be treated as species.

1 rnal of

Tscm‘Egtctsppasomemponpbasd

pwiTdtmevfmttdfepkgaHCuehudc

aFassmamtpubfiplefaits

ibcSmaeobwpmgtaiTgtwa

C

aibn

00 J. Boenigk et al. / European Jou

hese features of protist populations may result in an evenmoother continuum of differences among what might beonsidered species, and between what previously (with evenore incomplete knowledge) may have been recognised as

good’ species (Fenchel and Finlay 2006). However, as Hoef-mden and Bass show, it is the case in at least some protistroups that very closely related lineages as measured byhe most commonly used markers (SSU and ITS rDNA)an show significant phenotypic and ecological differenceshat reasonably justify them being referred to as differentpecies, as it is likely that they have different functionalrofiles. These and other studies refute the idea that ‘mor-hospecies’ are insufficiently resolving as a basis for protistlpha-taxonomy. However, we believe that a ‘good’ specieshould be genotypically and phenotypically distinct from allthers; ‘phenotype’ is a much more inclusive category thanorphology alone. Blindly creating species on the basis of

quivalent genetic (marker) differences along can lead to aassive proliferation of ‘species’ for which there is no known

henotypic basis. As sequence data are often much easier tobtain quickly than the more labour-intensive job of phe-otypic/ecological characterisation this is a real temptation,ending further investigation. We feel that in these cases theest approach is to designate strain names/codes, linking anctual organism to its genetic data, and only to create newpecies when some other line of evidence reinforces thatecision.

The advent of next generation sequencing (NGS; princi-ally 454 Sequencing and Illumina platforms at the time ofriting) arguably constitutes a second ‘molecular revolution’

n the way that biological diversity is measured and modelled.hese technologies have allowed us to sequence much moreeeply into genomes and transcriptomes of organisms acrosshe tree of life and into the lineage diversity of (particularly)

icrobial communities using culture/specimen-independentnvironmental molecular probing. Genome projects for indi-idual protist lineages are increasingly growing in number,acilitated by NGS technologies and associated develop-ent of analytical methods. These are revealing fundamental

ruths about many aspects of the biology and evolution ofhe sequenced organisms. However, multilocus genome-levelata can be generated in less intensive than reconstructingull genomes, and these offer a powerful way of delin-ating fundamental taxonomic units, avoiding many of theitfalls outlined above in this paper. Statistical packagesnown as ‘assignment tests’ identify genetically distin-uishable populations using multilocus genotypic data, andssign individuals to those populations (Pritchard et al. 2000;uelsenbeck and Andolfatto 2007; Falush et al. 2003, 2007;orander et al. 2003, 2004). All of these implementationsse inhomogeneities in the data, particularly linkage dis-quilibrium, to detect groups; one does not expect to see

eterozygote deficits or linkage disequilibrium within pop-lations (see Pritchard et al. 2000). The results are usuallyisplayed as a bar graph, with the bars for each individualoloured to represent Bayesian posterior probabilities, the

rabm

Protistology 48 (2012) 96–102

ssignment probabilities, of belonging to each group (seeigure in Suppl. text 1). If the samples are all taken fromsingle area, species delimitation via genotypic clusters in

ympatry consists of detecting inhomogeneities of populationtructure in multilocus genetic data; the resulting assign-ent probabilities represent probabilities of belonging toparticular species. Mallet regards assignment tests as theethod of choice when implementing the genotypic clus-

er idea of species as a delimitation procedure. Their use inrotists, despite the caveats described above about large pop-lations sizes and large numbers of asexual taxa, is untestedut could be very informative. The small amount of genomengerprinting work done on protist populations suggests thatopulations are much more highly structured than the preva-ent marker-based assessments show, and that this diversity iscologically informative (e.g. Logares et al. 2009), and there-ore taxonomically relevant. Therefore, NGS can open upgenomic perspective in taxonomic studies, allowing more

nclusive and adaptively/functionally informative differenceso be measured among strains and inform taxon delineationtudies.

Another application of NGS technology to protistologys environmental marker tag sequencing to measure micro-ial assemblage structure and response to environmentalonditions (e.g. Creer et al. 2010; Medinger et al. 2010;toeck et al. 2010), when environmental variables are co-easured with nucleic acid sample collection. Although this

pproach does not give much information about each lin-age detected, it does provide a very detailed perspectiven community structure and diversity. Methods are currentlyeing developed to integrate such sequence diversity dataith environmental parameters relating to the same sam-les, so that of ‘ecotype’ genotypes can be defined by theirulti-dimensional ecological cluster profiles. Ecotypes are

enetically consistent functionally informative units belowhe level of morphospecies (Finlay 2004; Finlay et al. 2006),nd can form the basis of future genomic/transcriptomicnvestigation once they are isolated from the environment.his approach, an ecological analogue to the genomic fin-erprinting/assignment tests described above, could providehe basis for delineating ecologically defined ‘species’, evenhen no other information about the organisms involved is

vailable (Suppl. text 3).

onclusions

The overriding outcome of the Symposium was thatlthough a species category does not exist in nature (and theres no reason why it should), and that species are not fixediological points in time or space, the concept of species isot an outmoded or unhelpful one. However, what scientists

efer to as species will differ between biological situations,nd therefore the methods used to delineate species shoulde made clear in each study. Sometimes this will requireore clarification than in others. An increasing battery of

rnal of

sotkstp

rcSntuaclvabaasb

to

(

(

(

(

(

A

iaE

A

cj

R

A

B

C

C

C

D

DE

F

F

J. Boenigk et al. / European Jou

ampling and analytical tools and approaches offer manypportunities for achieving this. As the application of theerm ‘species’ to a group of organisms always implies someind of hypothesis (as we are imposing these artificial or tran-ient boundaries on nature), an informed and flexible use ofhe term should actually enhance and inform the scientificrocess.

The central proposal therefore is to skip the search for a cor-ect general concept but in each case firstly to clearly state theoncept (basis for taxonomic subdivision) that is being used.econdly, for biodiversity studies based on molecular data weeed to accept that such data target different levels of resolu-ion in different groups of organisms. Thirdly, it is generallynlikely that current genetic marker-based diversity studiesre serendipitously ‘counting’ species, although the markerhosen may be a close approximation to a sensible species-evel distinction. The extent to which marker vs. phenotypics. genomic differences vary within apparently well-definednd homogeneous groups is also largely unknown, and muste accounted for as much as possible. The idea of the ‘species’s a basic unit of diversity is in many cases problematic andlternative, complementary views and measures of biodiver-ity, functional diversity, and evolutionary complexity shoulde brought into play.

To address the issues outlined at the end of the Introduc-ion, we recommend the following guidelines for the futuref protist alpha-taxonomy:

A) Species may be delineated in a variety of ways. Thedelineation method can be determined by the case inpoint and may be based entirely on phenotype or geno-type, if unavoidable. In such cases, it might be prudentto wait until complementary data are available beforeformally describing species to avoid later changes.

B) Ideally, species will be delineated based on more thanone line of evidence, and their distinguishing featuresmade explicit in a diagnosis. Such species need not cor-respond to any existing species ‘definition’, but the basisfor delineation should be made clear by the authors.

C) Genome-level profiling (genomic fingerprinting,sequencing, multi-gene datasets) has great potentialfor describing and quantifying differences betweenlineages, particularly as single-cell genomics is nowpossible in many cases. This is not currently an approachaccessible to many protistologists, but should be activelydeveloped as a taxonomic tool. In addition, generatinggenomic/transcriptomic studies of multiple protistlineages from all branches of the tree of life should bean ongoing objective of high priority for the researchcommunity.

D) A robust phylogenetic framework is essential for correctinterpretation of phenotypic characters and divergence.

This should be a base standard for the establishment ofnew species.

E) Environmental sequence data offers great potential forassessing species diversity in nature. However, this

F

F

Protistology 48 (2012) 96–102 101

approach requires knowledge of how genotypic and phe-notypic divergence is related in the groups detected inorder to be reliable and informative. Methods shouldbe developed to integrate ecological and genotypic datafrom environmental studies to enable robust delineationof ecotypes and (surrogate) species.

cknowledgments

We would like to thank Prof. Dr. Klaus Hausmann, Pres-dent of VI ECOP, and his organizing team, for funding andssisting us to organize and present this symposium at VICOP, Berlin, July 2011.

ppendix A. Supplementary data

Supplementary data associated with this articlean be found, in the online version, at doi:10.1016/.ejop.2011.11.004.

eferences

gar, A.W., 1996. The story of European commercial electronmicroscopes. In: Mulvey, T. (Ed.), The growth of electronmicroscopy. Advances in Imaging and Electron Physics 96. Aca-demic Press, London, pp. 415–584.

ütschli, O., 1880–89. Protozoa, Bronns Klassen und Ordnungenim Tierreich. CF Winter, Leipzig.

orander, J., Waldmann, P., Sillanpaa, M.J., 2003. Bayesian analysisof genetic differentiation between populations. Genetics 163,367–374.

orander, J., Waldmann, P., Marttinen, P., Sillanpää, M.J., 2004.BAPS 2: enhanced possibilities for the analysis of genetic pop-ulation structure. Bioinformatics 20, 2363–2369.

reer, S., Fonseca, V.G., Porazinska, D.L., Giblin-Davis, R.M.,Sung, W., Power, D.M., Packer, M., Carvalho, G.R., Blaxter,M.L., Lambshead, P.J.F., Thomas, W.K., 2010. Ultrasequenc-ing of the meiofaunal biosphere: practice, pitfalls and promises.Mol. Ecol. 19 (s1), 4–20.

obell, C., 1932. Antony van Leeuwenhoek and his ‘little animals’.Bale and Danielsson, London.

oflein, F., 1916. Lehrbuch der Protozoenkunde. Fischer, Jena.hrenberg, C.G., 1838. Die Infusionsthierchen als Vollkommene

Organismen. Voss, Leipzig.alush, D., Stephens, M., Pritchard, J.K., 2003. Inference of popu-

lation structure using multilocus genotype data: linked loci andcorrelated allele frequencies. Genetics 164, 1567–1587.

alush, D., Stephens, M., Pritchard, J.K., 2007. Inference of popula-tion structure using multilocus genotype data: dominant markersand null alleles. Mol. Ecol. Notes 7, 574–578.

enchel, T., Finlay, B.J., 2006. The diversity of microbes: resurgenceof the phenotype. Philos. Trans. R. Soc. B 361, 1965–1973.

inlay, B.J., 2004. Protist taxonomy: an ecological perspective. Phi-los. Trans. R. Soc. B 359, 599–610.

1 rnal of

F

G

H

K

L

L

M

M

M

P

P

S

S

S

02 J. Boenigk et al. / European Jou

inlay, B.J., Esteban, G.F., Brown, S., Fenchel, T., Hoef-Emden,K., 2006. Multiple cosmopolitan ecotypes within a microbialeukaryote morphospecies. Protist 157, 377–390.

reen, J.L., Holmes, A.J., Westoby, M., et al., 2004. Spatial scalingof microbial eukaryote diversity. Nature 432, 747–750.

uelsenbeck, J.P., Andolfatto, P., 2007. Inference of popula-tion structure under a Dirichlet process model. Genetics 175,1787–1802.

ent, W.S., 1880–81. A manual of the Infusoria, Vol. 1–3. DavidBogue, London.

innaeus, C., 1758. Systema Naturae, 10th ed. Laurentii Salvii,Stockholm.

ogares, R., Boltovskoy, A., Bensch, S., Laybourn-Parry, J., Renge-fors, K., 2009. Genetic diversity patterns in five protist speciesoccurring in lakes. Protist 160, 301–317.

edinger, R., Nolte, V., Vinay Pandey, R., Jost, S., Ottenwälder, B.,Schlötterer, C., Boenigk, J., 2010. Diversity in a hidden world:potential and limitation of next-generation sequencing for sur-veys of molecular diversity of eukaryotic microorganisms. Mol.

Ecol. 19 (s1), 32–40.

ora, C., Tittensor, D.P., Adl, S., Simpson, A.G.B., Worm, B., 2011.How many species are there on earth and in the ocean? PLoSBiol. 9, e1001127.

v

Protistology 48 (2012) 96–102

üller, O.F., 1773. Vermium Terrestrium et Fluviatilium. Hauniaeet Lipsiae.

ascher, A., Lemmermann, E., 1914. Die Süsswasserflora vonDeutschland, Österreich und der Schweiz. Band 1: Flagellatae.Gustav Fischer, Jena.

ritchard, J.K., Stephens, M., Donnelly, P., 2000. Inference of pop-ulation structure using multilocus genotype data. Genetics 155,945–959.

chloss, P.D., Handelsman, J., 2005. Introducing DOTUR, a com-puter program for defining operational taxonomic units andestimating species richness. Appl. Environ. Microbiol. 71,1501–1506.

tackebrandt, E., Ebers, J., 2006. Taxonomic parametersrevisited: tarnished gold standards. Microbiol. Today 33,152–155.

toeck, T., Bass, D., Nebel, M., Christen, R., Jones, M.D.M.,Breiner, H.-W., Richards, T.A., 2010. Multiple marker paralleltag environmental DNA sequencing reveals a highly complexeukaryotic community in marine anoxic water. Mol. Ecol. 19

(s1), 21–31.

on Gleichen, Freiherr, W.F., 1778. Abhandlung über die Saamen-und Infusionsthierchen- und über die Erzeugung. Winter-schmidt, Nürnberg.