old museum samples and recent taxonomy: a taxonomic ... · formaldehyde. on the other hand, museum...
TRANSCRIPT
ORIGINAL ARTICLE
Old museum samples and recent taxonomy: A taxonomic,biogeographic and conservation perspective of the Niphargustatrensis species complex (Crustacea: Amphipoda)
Cene Fišer & Charles Oliver Coleman &
Maja Zagmajster & Benjamin Zwittnig &
Reinhard Gerecke & Boris Sket
Received: 13 December 2008 /Accepted: 21 July 2009 /Published online: 9 March 2010# Gesellschaft für Biologische Systematik 2010
Abstract Natural history museum collections harbourvaluable information on species. The usefulness of suchdata critically depends on the accurate identification ofspecies, which has been challenged by introduction ofmolecular techniques into taxonomy. However, mostcollections may suffer from DNA degradation, due to ageand/or improper preservation; hence the identification ofspecimens depends solely on morphological features. Thisstudy explores how and to what extent morphological datacan help to solve ambiguous taxonomic cases based onselected species concepts and with the use of operationalcriteria in a species-hypothesis testing procedure. Thestudied taxon, the Niphargus tatrensis species complex,comprises freshwater subterranean amphipods, distributedacross Central Europe, the taxonomic status of which wasdebated extensively between 1930 and 1960. Using thegeneral species concept, character- and tree-based opera-
tional criteria reveal northern and southern diagnosable andexclusive lineages identified here as N. tatrensisWrześniowski, 1888 and N. scopicauda sp. n., respectively.The remaining populations represent the non-exclusive N.aggtelekiensis Dudich, 1932, which occurs from the easternAlps to Hungary. In the entire complex, altitudinaldistribution is largely limited to areas above 400 m, wherethe mean annual temperature never exceeds 9°C. Seeminglywell-defined distributional ranges of N. tatrensis and N.aggtelekiensis are fragmented in an ecological sense, whichraises the question whether two of the three speciesrecognised here actually consist of several unidentifiedtaxa. Morphological similarity between the species, numer-ous polymorphic features, and the association with cooltemperatures lead to a hypothesis in which fragmentation ofthe ancestral range occurred during post-Pleistocene climatewarming, reducing gene flow across lowland populationsdue to niche conservatism of the ancestral species and/or toinvasion of competitive species along the Danube andDrava rivers. The results are discussed regarding how oldmuseum samples are conducive to more detailed molecular-taxonomic and conservation studies.
Keywords General species concept . Population aggregationanalysis . Tree based species delimitation .Mantel test .
Museum collections . Polymorphic characters
Introduction
Taxonomy has progressed remarkably in the past50 years, both conceptually and technically. Moleculartechniques have been found to be a powerful tool for thedetection of morphologically similar (‘cryptic’ or ‘sib-
C. Fišer (*) :M. Zagmajster : B. SketOddelek za Biologijo, Biotehniška Fakulteta,Univerza v Ljubljani,PO box 2995, 1001 Ljubljana, Sloveniae-mail: [email protected]
C. O. ColemanMuseum für Naturkunde der Humboldt-Universität zu Berlin,10115 Berlin, Germany
B. ZwittnigAkademska in raziskovalna mreža Slovenije,PO box 7, 1001 Ljubljana, Slovenia
R. GereckeBiesingerstr. 11,72070 Tübingen, Germany
Org Divers Evol (2010) 10:5–22DOI 10.1007/s13127-010-0006-2
ling’) species, and we are about to realize how much thediversity of some groups has been underestimated (e.g.Bickford et al. 2007). As a result, many previousidentifications of old samples appear questionable—atleast in some groups—and their value for science becomeslimited (Graham et al. 2004). Advances in moleculartechniques allow DNA extraction from rather old samples(Stuart et al. 2006; Wandeler et al. 2007), but only if thelatter were preserved properly. This is usually not the casewith old crustacean samples; to our knowledge DNAextraction has been unsuccessful in many such cases(authors’ experience; pers. comm. from other researchers).The same is true for all samples preserved at hightemperatures or with inappropriate preservatives such asformaldehyde. On the other hand, museum collectionsharbour specimens that are rare and hard to be matched byfresh samples; the many reasons for this include taxonextinction as the result of “a changing world” (Wheeler2004; Winston 2007). Consequently, the key to obtainingbiologically relevant information from such specimens isthe comparative analysis of their morphology (Wiens2004a). Progress in debates of species concepts (e.g. deQueiroz 2007) and the development of operational criteriafor species delimitation (DeSalle et al. 2005; Sites andMarshall 2003, 2004; Wiens 1999, 2007) set a frameworkthat should help interpret ambiguous patterns of morpho-logical data. It is expected that a systematic study of themorphology of specimens from old collections should leadto more convincing taxonomic conclusions, or at least to abetter understanding of the geographic distribution ofmorphological variation. In this paper we explore howmuch and what kind of information can be grasped fromold museum samples that were taxonomically re-evaluatedusing the general species concept (GSC) and operationalcriteria developed since the early 1990s.
The material studied belongs to the amphipod genusNiphargus Schiödte, 1849, the members of which mainlyinhabit groundwater in Europe and the Middle East (Fišer etal. 2009a; Karaman and Ruffo 1986). Apart from evolu-tionary and ecological issues, the high number of species inthis genus (Väinölä et al. 2008) and the high degree ofendemism in most of them (Fišer et al. 2006b; Trontelj etal. 2009) suggest that Niphargus could play an importantrole in conservation planning for European groundwaters(see, e.g., Bininda-Emonds et al. 2000; Funk et al. 2002;Myers et al. 2000). Moreover, the capacity of niphargids toaccumulate heavy metals implies that they could be used asbiological indicators (Coppellotti Krupa and Guidolin2003).
These topics have remained virtually unexplored mainlydue to the problematic taxonomy of the genus (Ruffo1972). The genetic diversity in Niphargus implies that thetaxonomy, as inferred from morphology so far, is largely
unsatisfactory (Lefébure et al. 2006, 2007; Trontelj et al.2009). Morphological characters either are highly homo-plastic (Fišer et al. 2006a, 2007a), were overlooked or theirtaxonomic importance was underestimated (Fišer andZagmajster 2009); many cryptic species may be expected(Trontelj et al. 2009). Furthermore, taxonomist-subjectiveassignment of species rank (Fišer et al. 2009b) can be asource of bias in subsequent ecological, biodiversity andconservation studies (e.g. Agapow et al. 2004).
The primary targets in niphargid taxonomy, therefore, are‘species’with large ranges that possibly represent complexesof morphologically similar species (Trontelj et al. 2009). Thefocal ‘species’ of this study is N. tatrensis Wrześniowski,1888, originally described from southern Poland, which isimportant for several reasons. Its name was establishedearly, thus has a long and stable standing in nomenclature.Second, the wide range of the taxon encompasses theterritories of Poland, Slovakia, the Czech Republic, Austriaand Hungary. Finally, the variability of the ‘species’ hasattracted the attention of taxonomists, who have disagreedabout its taxonomic status. After the original description,seven ‘forms’ from Austria, Poland and Hungary weredistinguished (Schellenberg 1935, 1937, 1938). [The namesnewly proposed for these ‘forms’ by Schellenberg (1935,1937) unequivocally denoted infrasubspecific entities. Sincethey were never used as valid for a species or subspeciesprior to 1985, they are nomenclaturally unavailable underICZN (1999) Code Article 45.6.4.] Schellenberg’s divisionswere criticized by Jersche (1963), who explicitly rejectedthree forms from the eastern Alps, and by Micherdziński(1956), who recognised only the “North-Alpine”, “Silesian(Sudety)” and “Tatry” forms. Both authors based theirconclusions on “considerable plasticity and wide range ofvariability” (Micherdziński 1956) that is larger within thanbetween populations. More recently, Sket (1994) reportedon an undescribed niphargid from northern Slovenia,informally named N. “scopicauda n.n.”, that proved to bemorphologically and genetically similar to N. tatrensis(Fišer et al. 2008b; Trontelj et al. 2009).
The first aim of the present study is a taxonomic (re-)evaluation of variation in the ‘species’ considered as N.tatrensis and populations informally named N. scopicauda.We relied on the general species concept (GSC; de Queiroz1998, 2007) that was tested with character- and tree-basedoperational criteria. In a second step, the taxonomicconclusions are critically reviewed with the aid of correla-tion distance matrix analysis, and with an analysis of thedistribution of some recent ecological factors. The secondaim is to address the role of collections. Based on thetaxonomic conclusions drawn here, we illustrate howmorphology-based taxonomic conclusions affect furtherconservation studies, and what the current impediments tomorphology-based taxonomy are.
6 C. Fišer et al.
Material and methods
Samples
The specimens examined derived from the Schellenbergcollection held in the Museum für Naturkunde Berlin(totalling 45 samples from Poland, the Czech Republic,Austria and Hungary), and from the collection of Odd-elek za biologijo, Biotehniška fakulteta Univerza vLjubljani (10 samples from Slovenia, Austria and thetype locality of N. tatrensis). Distribution data wereobtained from the collections and supplemented withinformation from the literature (Micherdziński 1956;Skalski 1972). Several samples consisted of juvenilespecimens, which together with literature data remainedidentified only to the level of a species group. Altogether17 populations distributed across the entire range of thecomplex were re-examined morphologically.
In the first phase of research, the type population of N.tatrensis, voucher material for the seven ‘forms’ describedby Schellenberg (1935, 1937, 1938), and one sample of N.scopicauda (Table 1) were reviewed for a large set ofmorphological characters traditionally considered as infor-mative in niphargid taxonomy (Fišer et al. 2009b, c; seealso http://niphargus.info). In order to minimize variationdue to differences in age or sex (Fišer et al. 2008a), onlyadult specimens were examined (males with differentiateduropods III; females with oostegites). Only characters thatshowed between-population variation (Table 1) werereviewed for a larger number of adult individuals. Wepresent and discuss these characters in the “Results” sectionbelow, together with remarks on how observations weretreated in the character-coding process.
Species concepts
Sites and Crandall (1997) emphasized the importance ofviewing the diagnosis of species boundaries as ahypothesis-testing procedure, with clear statements of thecriteria used. Despite the long history of dispute overspecies concepts, Mayden (1997) and de Queiroz (1998,2005a, b, 2007) found that all contemporary speciesconcepts and definitions share common a element relatedto the speciation process that (according to de Queiroz1998, 2005a, b, 2007) is a key to reconciling the alternativespecies concepts. As pointed out, differences among speciesconcepts arise on account of different defining (necessary)properties that function as cut-off criteria in speciesdelimitation. By contrast, all of the concepts explicitly orimplicitly equate species with separately evolving (seg-ments of) metapopulation lineages, where a metapopulationis an inclusive population made up of a set of subpopula-tions, and a lineage (at the population level) is a population
extended through time or ancestral-descendant series oftime-limited (instantaneous) populations. In this view, deQueiroz identified a separate evolution of metapopulationlineages as a property necessary for the postulation of aspecies hypothesis, whereas any of the properties requiredby different concepts become a sufficient pointer forpostulation of a species hypothesis. The focus of theanalyses in the present work is the identification of two ormore metapopulations, among which gene flow has beenrestricted. The operational criteria for species delimitationused together with underlying assumptions are outlined inthe next section.
Assumptions and operational criteria
All approaches outlined below require an assumption thateach sample (population) from each locality consists of asingle species. The assumption is based on the premise thattwo morphologically similar, closely related species com-pete for the same food and space, which will result in acompetitive exclusion of the weaker competitor in arelatively short period of time (Fišer et al. 2007b; Ginet1965; but see also Fišer 2006b). Next, as pointed out, theterm monophyly may not be applicable at the species level;for this reason we operate with the related term ‘exclusive’(e.g. Graybeal 1995).
The species hypotheses were built using the followingprotocol. In the first step we separately postulated specieshypotheses using character-based and tree-based opera-tional criteria described in more detail below. The finalspecies hypotheses are the conservative consensus ofresults in order to minimize the flaws of both methods.In the second step, we searched for additional support forspecies hypotheses in correlation between morphologicalvariation and geographic distribution. Finally, the avail-able ecological information was used to explain currentmorphological variation and geographic distribution ofthe complex within a historical perspective and theniche-conservatism concept (Wiens 2004b; Wiens andGraham 2005).
Character-based species delimitation
Character-based species delimitation has been formalised in aprocedure called population aggregation analysis (PAA; Davisand Nixon 1992). PAA involves systematically comparingcharacter state distributions among populations, aggregatingsets of populations which differ only in polymorphic traits,and considering sets of populations which differ by at leastone seemingly fixed difference (or which shares no states forthat character) to be different species. The usage of themethod has been extended to the nonoverlapping ranges ofquantitative characters (Wiens and Penkrot 2002). The
Old museum samples and recent taxonomy: A taxonomic, biogeographic and conservation perspective 7
Tab
le1
Distributionof
characterstates
intheNipha
rgus
tatrensisaggregate
SpeciesPop
ulation
Locality
(num
berof
specim
ens
exam
ined)a
Maxilliped
innerlobe
PpIIIb
PpIV
bPpVc
PpVI
PpVII
Telson
apical
Telson
lateral
Telson
mesial
Telson
dorsal
Telson
lobes
setae
(A,R)
extrasp.
(F)
extrasp.
(F)
extrasp.
(F)
extrasp.
(F)
extrasp.
(F)
spines
(A,
R)
spines
(A,
R)
spines
(A,
R)
spines
(A,R)
apical
shape
N.tatrensis
type
Zakop
ane(3)
3.33
(3–4
)0.00
0.00
0.00
0.67
0.67
3.50
(3–4)
1.75
(0–3
)0.33
(0–1
)1.00
(1)
narrow
ByčiSkala
(15)
3.50
(3–4
)0.72
0.76
0.00
00.16
3.00
(2–4)
1.47
(1–2
)0.83
(0–2
)1.07
(1–2
)narrow
Dem
anow
skaHöh
le(15)
3.00
(3–4
)0.08
0.14
0.00
0.04
0.45
3.04
(2–4)
1.64
(1–2
)0.07
(0–1
)1.57
(1–3
)narrow
“bei
Rajec”(15)
3.50
(3–5
)0.07
0.03
0.00
00.29
3.03
(3–4)
1.73
(1–3
)0.07
(0–1
)1.17
(1–3
)narrow
“f.
reyersdo
rfensis”
ReyersdorferHöh
le(16)
3.00
(1–3
)0.30
0.30
0.00
00
3.10
(3–4)
1.30
(1–2
)0.40
(0–1
)1.03
(1–2
)narrow
“f.
schn
eebergensis”
PatzeltHöh
le(10)
3.00
(3– 4
)0.00
0.00
0.00
0.18
0.8
3.00
(2–4)
1.85
(1–3
)0.00
(0)
1.00
(1)
narrow
N.ag
gtelekiensis
type*
Agg
telekerHöh
le(18)
4.50
(4–5
)0.84
0.95
0.06
0.04
0.69
3.00
(3)
0.51
(0–2
)0.00
(0)
1.03
(1–2
)wide
Ostmark(6)
2.80
(2–3
)0.08
0.00
0.00
00.25
4.00
(4)
0.92
(0–2
)0.08
(0–1
)1.08
(1–2
)wide
Kop
penb
ruellerSinon
ykapelle
(10)
3.10
(3–4
)0.00
0.00
0.00
00
4.40
(4–5)
1.05
(0–2
)0.20
(0–1
)1.00
(1)
wide
Gesäuse
NationalPark(8)
3.00
(3)
0.73
0.71
0.10
00.5
3.88
(3–5)
0.81
(0–2
)0.31
(0–1
)1.00
(1)
wide
“f.lunzensis”
Lun
z,4samples
(17)
2.75
(2–3
)0.97
0.97
0.31
0.65
0.97
3.79
(3–5)
0.83
(0–3
)0.37
(0–1
)1.94
(1–3
)wide
“f.lurensis”
Lurhö
hle,
Sem
riach(16)
2.00
(1–3
)0.88
0.87
0.00
0.97
0.97
4.25
(3–5)
1.25
(1–3
)0.07
(0–1
)1.00
(1)
wide
“f.oetscherensis”
OetscherHöh
le(3),Nixhö
hle
(6)
2.50
(2–3
)0.67
0.77
0.23
0.25
0.82
3.86
(3–4)
0.57
(0–2
)0.00
(0)
1.14
(1–2
)wide
“f.salzbu
rgensis”
Schenko
fenh
öhle,Sulzak(4)
3.00
(3–4
)0.72
0.72
0.07
00.40
4.00
(4)
1.25
(1–2
)0.80
(0–1
)1.08
(1–2
)wide
N.scop
icau
da
Belojača(4)
3.00
(3)
0.13
0.00
0.00
0.40
0.43
3.88
(3–4)
0.38
(0–1
)0.00
(0)
0.00
(0)
wide
Habidov
agrapa(3)
2.50
(2–3
)0.00
0.00
0.00
0.00
0.00
4.25
(4–5)
0.38
(0–1
)0.00
(0)
0.00
(0)
wide
Hud
aLuk
nja(10)
2.50
(2–3
)0.04
0.04
0.00
0.18
0.48
4.30
(4–5)
0.26
(0–1
)0.04
(0–1
)0.00
(0)
wide
(A,R)averageor
rang
e,extrasp.
extraspine,(F)frequency,
Pppereop
odaWereview
edtheleftandrigh
tside
ofeach
specim
en;individu
alsin
somepo
pulatio
nsareextrem
elyasym
metrical
bHerewepresentraw
frequencies.In
theanalysiswetreatedextraspineson
pereop
odsIII–IV
asasing
lecharacter,calculated
asaveragefrequency(see
text)
cThese
frequenciesof
extraspine
presence
onpereop
odV
seem
tobe
toolow
tobe
inform
ative,
thus
wereexclud
edfrom
furtheranalyses
8 C. Fišer et al.
approach is advantageous in that there is a clear relationshipbetween fixed differences and gene flow—if the diagnostictraits are genetically based and are truly fixed in eachspecies, there is unlikely to be any gene flow between thespecies. However, given finite sample sizes, determiningwith certainty whether traits are truly fixed is almostimpossible. We followed Wiens and Servedio (2000) inorder to evaluate statistical confidence in distinctness of setsof populations that were delimited using the character-basedapproach. Hence, we tested whether sample sizes (for agiven set of populations) were large enough to argue that atleast one of the diagnostic characters was not polymorphicabove a selected frequency cut-off. The maximum frequencyof the alternative character state allowed in the diagnosticcharacter used as a cut-off value was 0.1. Failing this testmeant that there was greater than 5 % probability that all ofthe diagnostic characters for a given taxon were actuallypolymorphic, with the unobserved character state occurringat a frequency of 0.1 or greater.
The characters used in our study are presented in the“Results” section and in Fig. 1. The presence of additional
strong and spiniform setae on the pereopod dactyls and theshape of the telson were treated as qualitative characters,which were expressed in frequencies, whereas the numbersof strong and spiniform setae on maxilliped and telson weretreated as quantitative traits and expressed as value ranges.Overlap in trait values between populations was treated asequivalent to polymorphism in qualitative traits.
Tree-based analyses
Davis and Nixon (1992) stated that phylogenetic analysisshould not be attempted below the species level, becausegene flow among populations and recombination amongcharacters should break up hierarchical patterns withinspecies inferred from morphological data. By contrast,Wiens and Penkrot (2002) advocated a tree-based approach,using the assumption that the absence of hierarchicrelationships below the species level leads to weaklysupported population-level trees that are discordant withgeography, and that this lack of signal can be detected andused to infer species limits in cases where these limits are
Fig. 1 Morphological charac-ters included in the analysis.(a–c) Inner lobe of maxilliped,showing flattened spines; aNiphargus aggtelekiensis “f.lurensis”, b N. tatrensis “f.schneebergensis”, c N.. aggtele-kiensis, topotype. (d, e) Dactyluswith (d) and without (e) addi-tional spiniform seta. (f–h)Niphargus tatrensis, telson nar-rowed and concave in apicalpart (f = “forma schneebergen-sis”, g = topotype, h = “formareyersdorfensis”). (i–l) Niphar-gus aggtelekiensis, telson broadconvex (i = topotype, j = “formaoethscerensis”, k = “forma lun-zensis”, l = “formasalzburgensis”)
Old museum samples and recent taxonomy: A taxonomic, biogeographic and conservation perspective 9
unknown. Sets of populations that were strongly supportedas ‘exclusive’, and were geographically coherent, wererecognized as potentially distinct species. This approachuses properties of populations rather than properties ofindividuals as terminal units, because morphological char-acters are generally thought to be genetically unlinked, andusing individuals will inappropriately treat all polymor-phisms shared between populations as homoplasies ratherthan potential synapomorphies.
We followed the above assumptions. Qualitative poly-morphic characters were coded using the frequency-basedstep-matrix approach (for performance see Wiens 1995;Wiens and Servedio 1998); quantitative characters werecoded using step-matrix gap-weighting (Wiens 2001). Themethod uses absolute differences between the standardizedraw data (summarized as an average value) as weights. Themaximum number of steps for any step-matrix character was100. In order to maintain balanced character weights, i.e. thesame cost for a transition between fixed character states forall morphological and molecular characters, all other charac-ters were weighted by 100. Step matrices were constructedusing a Pearl script (available at http://www.niphargus.info/).
As bootstrapping of nine characters makes virtually nosense, we tested clade support by modifying weightsbetween character states in a step matrix by modificationof difference between raw data by square and rootfunctions. Furthermore, in accordance with existing molec-ular studies (Fišer et al. 2008b; Trontelj et al. 2009) weincluded Niphargus schellenbergi S. Karaman, 1932 as thesole outgroup taxon.
Correlated distance matrices
We tested whether morphological similarity and geographicdistance matrices correlate with each other. Our two-stepprocedure was based on two assumptions. (1) Pheneticdiversity, when significantly correlated with geography,may present evidence either for a morpho-cline or forseparate metapopulations consistent with geography. (2)The absence of significant correlation may indicatepanmictic populations. In the first step, calculations wereperformed on the whole set of data that were morpholog-ically examined, whereas in the second step correlationswere repeated for the populations of the three putativespecies, namely N. tatrensis, N. aggtelekiensis and N.scopicauda.
Localities were georeferenced as precisely as possiblewith the use of various cartographic sources. The studiedregion was projected in Lambert Conformal ConicalProjection for Europe (central meridian 10°, standardparallels 43° and 62°, latitude of origin 30°), which wasused for measures of geographic distances. In total, 27localities were georeferenced (Appendix A). Correlation
analysis with morphological distances includes only those17 populations, from which sufficiently large samplesallowed an estimation of morphological variation.
Dissimilarity coefficients for pairs of populations weresummed from standardized absolute differences in frequen-cies (quantitative characters) and average values (qualita-tive characters) for all characters using the programpackage DELTA (Dallwitz et al. 2007). Geographicdistances among localities were calculated with Hawth’sAnalysis Tool, ver. 3.27 (Beyer 2008) in ArcMap (ArcGIS,ver. 9.1; ESRI).
To test whether specimens closer to each other in spaceare more alike we calculated correlation of the two matricesusing a Mantel test. We used the ADE4 package (Chessel etal. 2008) in the R environment (www.r-project.org).Statistical significance was tested with 1000 randomizations(Chessel et al. 2008).
Use of ecological data
In this part of the analyses we included and mappedadditional localities from the literature. Relying on pub-lished data and preliminary analyses we eliminated param-eters that do not affect the distribution of the speciescomplex—presence of karst (own data); pH and carbonatehardness (Micherdziński 1956)—and constrained this partof the analyses to the temperatures. The temperature of thesubterranean environment roughly corresponds to the meanannual temperature of the area (Gams 1974), a parameterrelated to altitude and latitude. We overlaid the distribu-tional data over a map of mean annual temperature layersfor Europe (Hijmans et al. 2005; layers available via http://www.worldclim.org/) in order to see whether distributioncorresponds with either of the two parameters. The spatialresolution of the original layer is a square kilometre. Thetemperatures at localities that could not be georeferencedprecisely were obtained by averaging the mean annualtemperature values within a diameter of 10 kilometres inorder to minimize possible error.
Results
Characters
The morphological characters reviewed in this sectionshowed outstanding between-population variation and formthe basis of all analyses. The maxilliped consists of an innerlobe, outer lobe and palpus. In the apical part of the innerlobe there are two types of setae: cylindrical ones with finehair, and stronger flattened setae (character 1; Fig. 1a). Thenumber of the latter varies between 2 and 5 per lobe; it islower in juvenile than in adult specimens and can differ
10 C. Fišer et al.
between the left and right body sides of a specimen.However, Schellenberg (1938) reported an increased num-ber of these flattened setae in N. tatrensis “forma”aggtelekiensis.
The dactyls of pereopods III–VII are regularly armedwith a single dorsal (convex side) plumose seta and withone weak and one spiniform seta close to the insertion ofthe nail. In addition, one, rarely two additional spiniformsetae can be developed in the proximal part of the dactylusin the N. tatrensis aggregate (Fig. 1b). Schellenberg’s(1935, 1937) definitions of his ‘forms’ were largely basedon the presence of these additional spiniform setae.However, close examination has revealed that their pres-ence varies even in fully developed (i.e. the least variable)specimens. Additional spiniform setae develop on differentpereopods at different frequencies; the armature of the leftand right appendages can differ. A comparison with other,unrelated niphargid species shows that such extra setaedevelop either on all pereopods (pp), exclusively on pp IIIand IV, on pp V–VII, on pp VI and VII, or exclusively onpereopod VII (e.g. Fišer et al. 2006a). Assuming indepen-dent evolution of extra setae on different pereopods, wetreated pp III–IV, pereopod VI and pereopod VII as threeseparate, independent characters (characters 2–4) ratherthan as a single putative serial homologue. Pereopod V wasomitted from the analyses, since it rarely carries extra setae(Table 1). We examined the left and right sides of eachanimal; each extra seta was counted as ‘present’. Usingthese data we estimated the frequency of presence of extraspiniform setae (or extraspines) for each character in allpopulations. Frequencies calculated for pereopods III andIV differed minimally; hence, character 2 is the average ofthe frequencies calculated for pereopods III and IV.
The telson is bilobed. The shape of the lobes differsamong populations (Schellenberg 1937). The apical part ofa lobe is either broad with clearly convex outer margins ormore or less narrowed with concave or ‘broken’ outermargins (character 5; Fig. 1c, d). Furthermore, four groupsof spiniform setae that seem to have evolved independently(Fišer et al. 2006a) can be identified on the telson (Fig. 1d):an apical (character 6), lateral (character 7), mesial (locatedin the telson’s cleft; character 8) and a dorsal group(character 9). The number of spiniform setae depends onthe specimen’s age and can be left-right asymmetric.Schellenberg (1935, 1937) noted varying numbers of dorsalspiniform setae, whereas the remaining spiniform setaeremained unstudied. In contrast with pereopod dactyls, forwhich we express the presence of extra spiniform setae infrequencies, the numbers of spiniform setae on (the left andright lobe of) the telson in a population were pooled andsubsequently treated as average values (tree-based opera-tional criteria) or ranges (character-based operationalcriteria).
Species hypotheses
The studied characters are highly variable (Table 1). ThePAA merged populations into three groups on the basis oftwo characters. Populations aggregated into N. tatrensissensu stricto share narrowed and apically concave telsonlobes (Fig. 1c, d), whereas in the remaining populations theapical lobes are wide and convex. The three southernmostpopulations treated as N. scopicauda, in contrast to the restof the populations, lack strongly spiniform setae on thedorsal surface of the telson. A weak point of these results isthe number of specimens analysed: According to theprobabilistic approach by Wiens and Servedio (2000), aminimum of 30 specimens should be analysed to achieve95 % confidence that seemingly rare frequencies of the twoout of nine studied characters are fixed below the value of0.1. Even when the number of specimens was increasedartificially by treating left and right sides of the animalsseparately, the number of analysed individuals still wasinsufficient in nine out of seventeen populations. However,assuming that aggregated populations constitute the samemetapopulation, the total number of the analysed specimensis high enough to assume fixations of the two diagnosticcharacters.
All five parsimony analyses resulted in a similar hierarchicstructure (Fig. 2; linear weighting: length = 1925.3, CI =0.467, RI = 0.661; quadrate weighting: length = 1723.3, CI =0.522, RI = 0.665; square-root weighting: length = 2016.8,CI = 0.446, RI = 0.655). Regardless of the weightingscheme, the aggregate consists of a southern (N. scopicauda)and a northern lineage. The latter is further split in twopopulations and two large groups; one of them being N.tatrensis s. str. Considering the size of the dataset, thequestion arises whether exclusivity assumed for these fivelineages is true or not.
PAA and parsimony analysis agree that the northernmostand the southernmost populations are diagnosable andexclusive. We find it plausible to consider the two groupsof populations as separate species, to which we thereforeapply the names N. tatrensis and N. scopicauda, respec-tively. The remaining populations constitute three para-phyletic lineages. In contrast to the results from theparsimony analysis, the PAA implies gene flow betweenthese presumable lineages. Therefore, they could beconsidered as a non-exclusive species (“ferespecies” sensuGraybeal 1995), for which the name N. aggtelekiensisDudich, 1932 would be available.
Some additional support for the consensus between thegenerated character- and tree-based species hypotheses isprovided by the geographic distribution of the putativespecies (Fig. 3). Niphargus tatrensis is distributed north-ward from the Danube (Donau) river, across mountainridges along the southern Polish and northern Czech and
Old museum samples and recent taxonomy: A taxonomic, biogeographic and conservation perspective 11
Slovakian borders; N. aggtelekiensis is widespread in theeastern Alps (south of the Danube), with an isolatedpopulation in northwestern Hungary (north of the Danube).Niphargus scopicauda is limited to a narrow prealpine areain northeastern Slovenia, south of the Drava (Drau) river.Interestingly, the distance matrix based on morphologicalvariation showed a highly significant correlation (coeffi-cient = 0.418, p=0.00099) with the distance matrix basedon geographic distribution, when all 17 populations wereincluded. By contrast, the correlations largely decreasedand became non-significant when populations of N.tatrensis (coefficient = 0.009, p=0.411) and N. aggtele-kiensis (coefficient = 0.227, p=0.240) were analysed alone.
The subset of N. scopicauda was too small, thus was notanalysed further.
The present-day distribution corresponds to altitudesabove 400 m (with minor exceptions), at which the meanannual temperature never exceeds 9°C (Fig. 3, Appendix A;note that two localities at which temperatures exceed 8°Crefer to Salzburg, not to any caves in the region).Considering low temperatures as a factor affecting distri-bution (see “Discussion” for different views of this pattern),the ranges of N. tatrensis and N. aggtelekiensis appearfragmented. Distributional data thus unambiguously sup-port only the species hypothesis for N. scopicauda, and to alesser extent the species hypothesis for N. tatrensis,
Fig. 2 Most parsimonious treebased on morphological charac-ters; except for branch-lengthdifferences, identical topologywas obtained regardless of theweighting scheme
12 C. Fišer et al.
whereas the status of N. aggtelekiensis is questionable.Compared to the Alpine populations the Hungarian oneexhibits a higher number of flattened spines on the innerlobe of the maxilliped (4–5 and 2–3, only rarely 4,respectively), but independent evolution of the two (groups)of populations is not supported by a clear and unambiguousgap in morphology. Unfortunately, from that part of theaggregate range we had a single sample of adult specimens,which does not allow us to test the exclusivity.
To summarize, we hypothesize the existence of threespecies, keeping in mind that the diversity of the aggregatemight be underestimated. No data reject the hypothesis ofNiphargus scopicauda as a separate species which isformally established in Appendix B.
Discussion
Evidence for recent speciation
Our view on the history of the aggregate is built on (1)morphological similarity among lineages coupled with a highnumber of polymorphic characters, and (2) geographic
distribution of the aggregate that corresponds to low temper-atures. Morphological similarity and lack of exclusivity in N.aggtelekiensis favour the hypothesis of relatively recentemergence two exclusive lineages (N. tatrensis, N. scopi-cauda) over the alternative view, i.e. that speciation tookplace in ancient karstic massifs. Low-temperature dependentgeographic distribution implies that Quaternary glaciationsplayed an important role in diversification of the N. tatrensiscomplex. The present data are insufficient to resolve whetherindividual populations survived ice ages under glaciers(Kristjánsson and Svavarsson 2004) or recolonized theempty niches after retreat of the glaciers (Folquier et al.2008). Nevertheless, some hypotheses on causes for theseparation of lineages can be discussed. The most straight-forward explanation assumes that temperatures act as adistribution-limiting ecological factor, so that post-Pleistocene climate warming negatively affected fitness oflowland populations and reduced gene flow across theseareas. “Niche conservatism” (Wiens 2004b; Wiens andGraham 2005) of N. tatrensis—inability to adapt to hightemperatures that limits dispersal—explains its distribution.
This hypothesis, however, is in strong contrast to thegeneral view that niphargids, including N. tatrensis (Jersche
Fig. 3 Distribution of the Niphargus tatrensis aggregate in relation to mean annual temperature (below 8°C in the shaded areas)
Old museum samples and recent taxonomy: A taxonomic, biogeographic and conservation perspective 13
1963; Micherdziński 1956), are generalists. Althoughtemperature preferences have never been studied in detail,researchers have argued that various niphargid species areeuryhaline (Jersche 1963; Sket 1977) or able to cope withscarce food (Hervant et al. 1997, 1999a), low oxygenconcentrations (Hervant and Mathieu 1995; Hervant et al.1999b; Sket 1977) and low pH (Simčič and Brancelj 2006).By contrast, niphargids seem to be poor competitors (Fišeret al. 2007b; Sket 1981), which supports an alternativehypothesis to explain fragmentation of the ancestralpopulation. This view assumes a widely distributed ances-tral population living in both surface and subterraneanwaters (Fišer et al. 2006b; b; Sket 1958; 1981). Post-Pleistocene warming could have resulted in extensive changesto the crustacean fauna along the Danube and Drava rivers,including the arrival of ‘paleo-invasive’ species. The repro-ductive potential of gammarid species seems to be higher thanthat of niphargids (Fišer et al. 2007b), which could haveformed the basis for competitive exclusion of the studiedniphargids (Fišer et al. 2007b; Grabowski et al. 2007;Savage 1981) and consequently led to fragmentation of therange of the ancestral population. Other Niphargus specieswithin the area (Nesemann et al. 1995) also could haveplayed an important role in competition by occupyingpossible refugia or distributional pathways such as interstitialspaces in large rivers. The cessation of gene flow betweensubpopulations of the ancestral population could have been aresult of interactions resembling those observed todaybetween invasive and native species (Jażdżewski et al.2004; Sax et al. 2007; Vellend et al. 2007).
The collections and modern taxonomy: implications
Whenever dealing with poorly differentiated species,taxonomic conclusions largely depend on the selectionand number of observed characters, abundance and numberof samples, as well as on species concepts and methodsused. The latter two factors play rather important roles—after all, we reanalysed Schellenberg’s samples usingvirtually the same characters as Schellenberg (1935, 1937,1938) had. Cases in which the evidence for separation oflineages is weak will always be debatable and will largelydepend on the particular species concept applied. Thegeneral species concept advocated by de Queiroz (1998,2005a, b, 2007) minimizes the number of properties neededfor the definition of species, and therefore inherently forcesa splitter’s taxonomy. It can be expected that sometaxonomists will be critical of taxonomic conclusionsrelying on the GSC. On the other hand, rather thandiscussing the appropriateness of various species concepts,paying more attention to indications of lineage separationmight result in higher ranking of taxonomy amonghypothesis-driven fields in biology. For instance, the
species hypotheses presented here would also fulfil thecriteria of related evolutionary species concepts (e.g.Mayden 1997) or of (some) phylogenetic species concepts(Davis and Nixon 1992), whereas the evaluation whetherthe members of the three groups of populations couldinterbreed successfully (biological species concept) remainsa challenge to future research.
The most important taxonomic issue that needs to beresolved is the status of N. aggtelekiensis. Ecological data,especially when interpreted within the historical context,disagree with the conclusions based on morphology alone.This discrepancy can be the result either of incompletesampling in the lowland area and an inappropriatetemperature-dependence hypothesis, or of incomplete tax-onomic conclusions which underestimated both, the ob-served morphological differences between Hungarian andAustrian populations (number of flattened spiniform setaeon maxilliped inner lobes) and the slightly higher temper-ature regime at the Hungarian locality (see Raxworthy et al.2007 for species identification using the accuracy ofecological models). While the first alternative seems lessprobable, the second could be tested with molecular and/orecological data. Similarly, the postulated range of N.tatrensis is fragmented in an ecological sense. The questionwhether this fragmentation affects the taxonomic conclu-sions presented here remains unsolved.
In short, even in taxonomic puzzles such as Niphargus,morphology alone renders some taxonomic conclusions. Aformalized species concept / operational criteria procedureseems to be a key to unlocking the doors of collections withmaterial of even the most problematic taxa the DNA ofwhich has been damaged. From that point of view wewelcome both, additional development of methodology andprojects promoting the exploration of collections, such asSynthesys (http://www.synthesys.info/).
The role of collections: guidelines for conservation
In conservation biology, the present study yields the mostinteresting effects if the subject of conservation is variationon a subspecific level (evolutionary significant units; seeBininda-Emonds et al. 2000; Crandall et al. 2000; Moritz1994). The major concern with this approach is themaintenance of species fitness and its capacity for evolu-tionary response to environmental change. As pointed out,subspecific variation depends on long-term historicalisolation (vicariance) and adaptive evolution (phenotypicvariation) (Moritz 1994; 2002). (However, it remainsunclear how long-isolated populations defined as recipro-cally monophyletic groups differ from some, e.g. phyloge-netic, species concepts. The GSC—if accepted—maysimplify a starting position in this branch of conservationbiology.) We share the opinion of Crandall et al. (2000)
14 C. Fišer et al.
who state that “the potential for species evolutionary successis maximized through the maintenance of adaptive diversity,by preserving the maximum diversity of functionally diver-gent gene copies across the geographic range of a species.”However, if the entire aggregate is treated as a single species,the orientation of phenotypic diversity follows a north-southdirection. By contrast, considering three groups of populationsas three species, the longest axes of their ranges are reorientedroughly perpendicularly to follow a west-east direction.Although the studied characters show no geography-dependent subspecific variability, it can be assumed thatgenetic diversity increases with geographic distance, thusalong the longest axis in the range. This view strengthenshabitat patches arranged in a west-east direction withfavourable temperatures occupied by N. tatrensis. From thispoint of view, reconsideration of old museum sampleschanges both, sampling strategy in studies aiming to evaluategenetic diversity of species and eventual conservational acts.
Acknowledgements Our thanks go to Elzbieta Dumnicka (Krakow,Poland) for providing additional samples from Poland. Peter Trontelj
and two anonymous reviewers critically commented on an earlierversion of the manuscript. The first author received support from theSYNTHESYS Project (http://www.synthesys.info/) financed by Euro-pean Community Research Infrastructure Action under the FP6“Structuring the European Research Area” Programme (DE-TAF-3960; visit to the Museum für Naturkunde der Humboldt-Universitätzu Berlin).
Appendix A
Geographic origin, temperature and voucher informationfor the known material
Asterisks denote samples used for analyses in the presentstudy. Locality names are given exactly as by the respectiveoriginal source (reference or label); any additions areenclosed in square brackets. Georeferences often refer tolarger geographic entities rather than to the detailed locality.Abbreviations for country names: A = Austria, CZ = CzechRepublic, H = Hungary, POL = Poland, SK = Slovakia,SLO = Slovenia.
Species Lat, Long Temperature Source / referenceCountry: locality (WGS84) mean (10 km range)
Niphargus aggtelekiensis Dudich
A: Gesaeuse National Park [several samples]* 14.726342, 47.599389 5.9 (5.6–6.0) BF
A: Koppenbrueller Höhle, Dachstein Gebirge* 13.719097, 47.568692 7 (6.6–7.0) ZMB 25100
A: Koppenbrueller Sinonykapelle 13.719097, 47.568692 7 (6.6–7.0) ZMB 25546, ZMB 25481,ZMB 25482
A: Herdengelhöhle, Lunz am See[“Ostmark” in publications]*
15.050000, 47.850000 5.3 (5.1–5.5) ZMB 25479
A: Lunz, aus einem Seitenarm des Seebaches* 15.050000, 47.850000 5.3 (5.1–5.5) ZMB 24487, ZMB 24488
A: Lunz 15.050000, 47.850000 5.3 (5.1–5.5) ZMB25240
A: Wilhelminenhöhle bei Lunz 15.050000, 47.850000 5.3 (5.1–5.5) ZMB 24489, ZMB 25244,ZMB 25483, ZMB 24865
A: Semiriacher Eingang, Lurhöhle* 15.402756, 47.217814 6.3 (6.1–6.5) ZMB 24484
A: Lurhöhle bei Peggau 15.350000, 47.200000 8 (7.6–8.0) ZMB 26239
A: Lurhöhle b. Semiriach 15.402756, 47.217814 6.3 (6.1–6.5) ZMB 25245, ZMB 25272,ZMB 25181
A: “Nischöhle” :(=Nixhöhle?) 15.316667, 47.966667 6.9 (6.6–7.0) ZMB 25480
A: Nixhöhle b. :Frankenfels 15.316667, 47.966667 6.9 (6.6–7.0) ZMB 24839
A: Ötscher Höhle, :Nieder Oesterreich* 15.100000, 47.933333 7.4 (7.1–7.5) ZMB 24490
A: Schenkofenhöhle :bei Sulzau 13.166667, 47.500000 7.6 (7.6–8.0) ZMB 24491
A: Schenkofenhöhle, :Hage-Gebirge bei Salzburg* 13.033333, 47.800000 8.9 (8.6–9.0) ZMB 25730
A: Steinkeller bei :Lunz; Gang beim Wasser 14.745000, 46.984722 4.7 (4.6–5.0) ZMB 24492
A: Ybbs Gebiet (Weis :Ois, Aquelle Langau, Muhlbach) 14.916667, 47.800000 6.6 (6.6–7.0) ZMB 25817
A: Baernhoehle, :Hagen-Gebirge bei Salzburg 13.033333, 47.800000 8.9 (8.6–9.0) ZMB 25730
H: Aggteleker Höhle* 20.483333, 48.483389 8.3 (8.1–8.5) ZMB 24864
H: Höhle von Aggtelek 20.483333, 48.483389 8.3 (8.1–8.5) ZMB 25793
SK: Domica Höhle [Jaskyna Domica] 20.472500, 48.476667 8.4 (8.1–8.5) ZMB 24865
Old museum samples and recent taxonomy: A taxonomic, biogeographic and conservation perspective 15
Appendix B Taxonomic section
Here we present (1) characteristics needed for diagnosingthe entire aggregate and (2) a full description of N.scopicauda sp. n. Terminology and detailed descriptionsof landmarks used in measurements follow Fišer et al.(2009c). All morphological data presented here are freelyavailable at http://www.niphargus.info/, as are photographsof specimens from the populations of Schellenberg’s‘forms’ of N. tatrensis. Anyone using data referring to thisstudy is kindly asked to cite the present paper.
Niphargus tatrensis species complex
Diagnosis
Body length up to 20 mm. Pereonite VII with 2–4 slenderpostero-ventral setae, pleonites I–III with 6–8 setae along
dorso-posterior margin. Epimeral plates II–III withdistinct postero-distal corner, but never produced. Uro-somite I postero-dorso-laterally with 1 weak seta;urosomite II postero-dorso-laterally with 2–4 spiniformsetae; urosomite III without setae. Telson roughlyquadrangular (length/width = 0.80–0.95), with deep cleft(65–75% of length); lobes apically of variable shape.Telson armed with long spines (40–60% of telsonlength); apical and lateral strong, spiniform setae alwayspresent, mesial and dorsal spiniform setae only in certainspecies/populations (Table 1). Antenna I 40–50% of bodylength, maxilla I with 7 uni-, bi- or pluri-toothed stoutsetae; inner lobe with 2–3 (rarely 4) setae. Propods ofgnathopods I–II small to middle-sized (gnathopod IIpropodus length + palm length + diagonal length = 15–20% of body length), subquadrate, palm convex andslightly inclined; propodus II larger than propodus I.Gnathopod dactyls with a row of single setae along
N. scopicauda sp. n.
SLO: Belojača* 15.654633, 46.299612 8.8 (8.6–9.0) BF
SLO: Habidova Grapa* 15.529364, 46.567097 8.1 (8.1–8.5) BF
SLO: Huda luknja* 15.174250, 46.414464 7.2 (7.1–7.5) BF
N. tatrensis WrześniowskiCZ: Höhle Byči skala [Morawsko]* 16.694722, 49.307500 7.9 (7.6–8.0) ZMB 24790, ZMB 24857,
ZMB 24889
POL: Hausbrunnen [house well] inReyersdorf [Radechów]*
16.879717, 50.350000 7 (6.6–7.0) ZMB 25181
POL: Patzelt Höhle, Glatzer Schneeberg [Patzeltovajeskina, Śnieżnik Kłodzki (POL) or Králický Sněžník (CZ)]*
16.799619, 50.122667 6.1 (6.1–6.5) ZMB 23906
POL: Quarzloecher Glatzer Schneeberg 16.799619, 50.122667 6.1 (6.1–6.5) ZMB 23989
POL: Reyersdorfer Höhle [also reported as ReyersdorferTropfsteinhöhle, Bad Landeck, Schlesien; now JaskiniaRadochowska, Lądek-Zdrój]
16.879717, 50.350000 7 (6.6–7.0) ZMB 24486
POL: Salzlöcher b. Habelschwerdt [= Bystrzyca Kłodzka] 16.650000, 50.300000 7.3 (7.1–7.5) ZMB 19368, ZMB 24908
POL: Zakopane* 19.948694, 49.300239 4.8 (4.6–5.0) BF
SK: Demänovska Höhle [Demänovská jaskyňa, Tatry]* 19.584233, 48.997700 4.1 (4.1–4.5) ZMB 24866
SK: südl. Tatra u. bei Rajec* 18.633333, 49.083333 7.5 (7.1–7.5) ZMB 25168/71
N. tatrensis group
POL: Jaskinia Mrozna 19.871111, 49.269444 4.5 (4.1–4.5) Micherdziński (1956)POL: Lodowe źrodło [“Icy Spring”], Kościeliska dolina 19.869514, 49.261667 4.5 (4.1–4.5) Micherdziński (1956)POL: Jaskinia Malinowska 18.986111, 49.655556 4.3 (4.1–4.5) Micherdziński (1956)POL: Jaskinia Miŕtusia 19.906944, 49.263333 3.5 (3.0–3.5) Micherdziński (1956)POL: Babia gora, Beskidy Mountains 19.566292, 49.584375 3.3 (3.0–3.5) Skalski (1972)
POL: Poronin 20.006667, 49.335833 5.9 (5.6–6:.0) Skalski (1972)
POL: Jurkow 20.234047, 49.683506 7 (6.6–7.0) Skalski (1972)
POL: Wielka Puszcza 19.259167, 49.821389 6.1 (6.1–6.5) Skalski (1972)
POL: Pewel Mala 19.277778, 49.662500 7.8 (7.6–8.0) Skalski (1972)
POL: Bialka Tatrzanska 20.105556, 49.387778 6.2 (6.1–6.5) Skalski (1972)
POL: Romanka 19.277778, 49.662500 7.8 (7.6–8.0) Skalski (1972)
POL: Menczot 19.277778, 49.662500 7.8 (7.6–8.0) Skalski (1972)
POL: Barania Gora 18.981497, 49.601364 4.4 (4.1–4.5) Skalski (1972)
POL: Leskowiec estimated from figure 6.8 (6.6–7.0) Skalski (1972)
POL: Gorce estimated from figure 3.1 (3.0–3.5) Skalski (1972)
16 C. Fišer et al.
anterior margins. Pereopod dactyli III–VII ventrally with1, occasionally 2 spiniform setae. Coxa VII with 2–4posterior setae. Bases V–VII narrow with straight poste-rior margins, length/width = 0.50–0.65; pereopod VII 40–50% of body length. Pleopods I–III with 2-hookedretinacles. Uropod I endopodite/exopodite length ratio =0.90–1.05; endopodite distally with bunches of longsetae. Male uropod III up to 45% of body length.Endopodite 40–50% of protopodite length; exopodite ofuropod III rod-shaped. Endopodite distal article 40–55%of proximal article length in females, 40–100% in males.
Remarks
The combination of characters listed above is unique, buteach of the individual character states can be found inother niphargid species as well. Some features are lessfrequent among the remaining niphargid species thanothers. Noteworthy is the elongated distal article of theuropod III exopodite in females. The elongation of thedistal article in uropod III is common among niphargids,but occurs almost exclusively in males. Although the N.tatrensis species complex has retained sexual dimorphismin this character, the degree of elongation of this articleappears to be exceptional in females. According to ourobservations (75 species) and published information(unfortunately often incomplete), the length of the distalarticle in females does not exceed 30% of the proximalarticle, thus is much shorter than in the species studiedhere.
The monophyly of the aggregate is well supportedmorphologically as well as by sequences of 28 S DNAfragments (N. tatrensis: from type locality; N. scopicauda:from type locality and Habidova grapa). The sister-taxon ofthis species complex remains unknown. Schellenberg(1937) clarified similarities and differences between N.tatrensis and N. stygius Schiödte, 1847; Jersche (1963)noted similarities with N. foreli Humbert, 1876. The latterspecies shows not only morphological but also someecological similarities with the N. tatrensis aggregate. Itshares with N. tatrensis the long spiniform setae on thetelson, and narrowed pereopod V–VII bases. It is distrib-uted across Italy, Switzerland and Germany within theAlpine range, thus within the area with low mean annualtemperatures (Karaman and Ruffo 1993). In addition, themorphologically related N. setiferus Schellenberg, 1937(originally described as N. foreli setiferus) from the westernAlps has similar bunches of setae on the apical part of therami of uropod I. However, neither N. foreli nor N. stygiusappeared to be more closely related to the N. tatrensisaggregate than any other niphargid species in morphology-based and DNA-based phylogenetic analyses, respectively(Fišer et al. 2008b).
Niphargus tatrensis Wrześniowski, 1888
Diagnosis
Member of the N. tatrensis aggregate with apically concaveand narrowed telson lobes; telson always with dorsalspiniform setae. Distributed in Poland, the Czech Republicand Slovakia. Comprises the population at the type loaclity,and Schellenberg’s (1935, 1937) ‘forms’ “f. reyersdorfen-sis” and “f. schneebergensis”. The taxonomy of the speciesmay be incomplete (see Discussion).
Niphargus aggtelekiensis Dudich, 1932
Diagnosis
Member of the N. tatrensis aggregate with apically wide,convex telson lobes; telson always with dorsal spiniformsetae. Distributed in Austria, Slovakia and Hungary. Com-prises the population at the type locality, and Schellenberg’s(1935, 1937) ‘forms’ N. tatrensis “f. lurensis”, N. t. “f.lunzensis”, N. t. “f. ötscherensis” and N. t. “f. salzburgensis”.The taxonomy of the species may be incomplete (seeDiscussion).
Niphargus scopicauda sp. n
Etymology
The species epithet refers to the bunches of long setae onthe rami of uropod I (Latin ‘scopa’ meaning ‘broom’ or‘brush’). It is to be treated as a noun in apposition for thepurposes of nomenclature.
Material examined
Holotype. Male, 14.5 mm; SLOVENIA, puddles in Hudaluknja cave near Gornji Dolič, 16.11.2002, leg. C. Fišer;deposited in the collection of Oddelek za biologijo, Biotehniškafakulteta, Univerza v Ljubljani (Department of Biology, Bio-technical Faculty, University of Ljubljana). Paratypes. 5 malesand 5 females from type locality (undissected); male and femalefrom Habidova grapa (dissected); all other data as for holotype.
Diagnosis
Member of the N. tatrensis aggregate with apically wide,convex telson lobes; telson never with dorsal spiniformsetae. Distributed in Slovenia.
Description of holotype and dissected paratypes
Head and trunk (Figs. 4, 9). Body length up to 14.5 mm.Head length up to 10% of body length; rostrum absent.
Old museum samples and recent taxonomy: A taxonomic, biogeographic and conservation perspective 17
Pereonites I–VI without setae; pereonite VII with 2–4slender postero-ventral setae.
Pleonites I–III with 6–8 setae along dorso-posteriormargin. Epimeral plate II ventral and posterior marginsroughly perpendicular, posterior and ventral margins con-vex, ventro-postero-distal corner distinct but blunt and notproduced; along ventral margin 1–2 spiniform setae; alongposterior margin 5–8 thin setae. Epimeral plate III ventraland posterior margins at sharp to perpendicular angle,posterior and ventral margin slightly convex, ventro-postero-distal corner distinct but not produced; alongventral margin 2 spiniform setae; along posterior margin7–8 thin setae.
Urosomite I postero-dorso-laterally with 1 weak seta;urosomite II postero-dorso-laterally with 2 spiniform setae;urosomite III without setae. Near insertion of uropod I 1spiniform seta.
Telson length : width as 1.0 : 0.80–0.95 ; cleft 65–75%of length; lobes apically widely rounded. Telson spines (perlobe): 4–5 apical spines of 40–55% telson length; lateraland mesial margins exceptionally with 1 spine; dorsalsurface without spines (see also Table 1). Pairs of plumosesetae inserted mid-laterally.
Antennae (Fig. 5). Antenna I 35–40% of body length.Flagellum with up to 25 articles; each article with 1 longaesthetasc. Peduncle article proportions 1.0 : 0.90 : 0.35–0.40. Proximal article of peduncle dorso-distally slightlyproduced. Accessory flagellum biarticulated; distal articleshorter than one-half of proximal article.
Lengths of antennae I : II as 1.0 : 0.50. Flagellum ofantenna II with 10–11 articles; each article with setae andelongate sensilla of unknown function. Lengths of pedunclearticles 4 : 5 as 1.0 : 0.90–0.95; flagellum 70–80% ofpeduncle length (articles 4 + 5).
Mouthparts (Fig. 5). Inner lobes of labium longer thanhalf of outer lobes.
Left mandible: incisor with 5 teeth, lacinia mobilis with4 teeth; between lacinia and molar a row of thick, serratedsetae, long seta at base of molar. Right mandible: incisorprocessus with 4 teeth, lacinia mobilis with several smalldenticles, between lacinia and molar a row of thick, serrated
setae. Proportions of mandibular palp articles 2 : 3 (distal) =1.0 : 1.15–1.20. Proximal palp article without setae; secondarticle with up to 5–7 setae; distal article with 1 group of 5–7 A setae; 3–4 groups of B setae; 20–28 D setae; 4–5 Esetae.
Maxilla I distal palp article with 5–7 apical andsubapical setae. Outer lobe of maxilla I with 7 uni-, bi-orpluri-toothed spines; inner lobe with 2 setae.
Maxilla II inner lobe slightly smaller than outer lobe;both of them setose apically and subapically.
Maxilliped palp article 2 with 6–9 rows of setae alonginner margin; distal article with dorsal seta and group of
Fig. 5 Niphargus scopicauda sp. n.; holotype, antennae (aI, aII) andmouthparts (lb = labium; mdL, mdR = left and right mandibles; mxI,mxII = maxillae I and II; mxpe = maxilliped)
Fig. 4 Niphargus scopicaudasp. n.; holotype, pleopodsomitted
18 C. Fišer et al.
Fig. 8 Niphargus scopicauda sp. n.; holotype, pereopods (pp) VI andVII
Fig. 7 Niphargus scopicauda sp. n.; holotype, pereopods (pp) III and IV
Fig. 6 Niphargus scopicauda sp. n.; holotype, gnathopods (gp) I and II
Fig. 9 Niphargus scopicauda sp. n.; holotype, pleon details (f-upIII =uropod III of female paratype; plp = pleopod II; T = telson; up =uropod)
Old museum samples and recent taxonomy: A taxonomic, biogeographic and conservation perspective 19
small setae at base of nail. Maxilliped outer lobe with 9–10flattened, thick setae and 5–7 serrated setae; inner lobe with3–4 flattened, thick setae apically and 6–8 serrated setae.
Coxal plates, gills (Figs. 4, 6, 7 and 8). Coxal plate I offlattened rhomboid shape, antero-ventral corner sub-rounded; anterior and ventral margin of coxa I with 6–9setae. Coxal plate II width : depth as 1.00 : 1.00–1.20;anterior and ventral margin with 7–10 setae. Coxal plate IIIwidth : depth as 1.00 : 1.10–1.25; along antero-ventralmargin 6–7 setae. Coxal plate IV width : depth as 1.00 :1.10–1.30; posteriorly slightly concave (5–10% of coxawidth); along antero-ventral margin 7–6 setae. Coxal platesV–VI: anteriorly developed lobe; along posterior margin 3–4setae. Coxal plate VII half-egg shaped, along posteriormargin 2–3 setae. Gills II–VI ovoid, reaching to mid-basis.
Gnathopod I (Fig. 6). Ischium with up to 11 postero-distal setae. Carpus length 60–70% of basis and 75–80% ofpropodus. Anterior margin of carpus with distal group ofsetae; carpus posteriorly with transverse rows of setaeproximally and a row of lateral setae; postero-proximalbulge large (1/3 of carpus length), positioned proximally.Propodus subquadrate, palm convex and slightly inclined.Along posterior margin 7–8 rows of denticulated setae.Anterior margin with 18–23 setae in 4 groups, antero-distalgroup with 12–13 setae. Group of 4–5 facial setae below(proximal of) palmar spine; several groups of surface setaepresent. Palmar corner with strong palmar spine, singlesupporting spine on inner surface, and 3 denticulated, thickspiniform setae on outer side. Nail length 30–35% of totaldactylus length; along anterior margin 5–6 single setae;along inner margin a row of short setae.
Gnathopod II (Fig. 6). Basis width : length as 1.0 : 0.30–0.33. Ischiumwith 4–7 postero-distal setae. Carpus length 55–60% of basis and 80–85% of propodus. Anterior margin ofcarpus with distal row of setae; carpus posteriorly withtransverse rows of setae proximally, a row of lateral setae;postero-proximal bulge large (1/3 of carpus length), posi-tioned proximally. Propodus medium-sized (sum of length,diagonal and palm length measures up to 20% of body length)and larger than propodus of gnathopod I (1.0 : 0.85). Propodusrectangular, palm convex and more inclined than palm ofgnathopod I. Posterior margin with 10–12 rows of denticulat-ed setae. Anterior margin with 10–24 setae in 3–5 groups;antero-distal group with 11–16 setae. Group of 4–6 facialsetae below (proximal of) palmar spine; individual surfacesetae present. Palmar corner with strong palmar spine, singlesupporting spine on inner surface, and 2 denticulated, thickspiniform setae on outer side. Nail length 30–35% of totaldactylus length. Along anterior margin 5–7 single setae; alonginner margin few short setae.
Pereopods III–IV (Fig. 7). Proportions of pereopods III :IV as 1.0 : 0.95. Dactylus IV 40–45% of propodus IV; naillength 50–60% of total dactylus length. Dactyli III–IV with
dorsal plumose seta; at base of nail a tiny seta and aspiniform seta; additional spiniform seta present in less than5% of individuals (see Table 1).
Pereopods V–VII (Fig. 8). Proportions of pereopods V :VI : VII as 1.00 : 1.40–1.45 : 1.55. Pereopod VII length40–45% of body length.
Bases V–VII narrow, with straight or slightly concaveposterior margins, length : width as 1.00 : 0.55–0.65;posterior margin straight, without distal lobes; posteriorly9–10, 10–14 and 13–15 setae, respectively; anteriorly 5, 5–6 and 5–6 slender spiniform setae, respectively. DactylusVII length 25–30% of propodus VII length; nail length 30–35% of total dactylus length. Dactyli V–VII with dorsalplumose seta; at base of nail a tiny seta and a spinform seta;up to 50% of specimens have an additional spinifom setaon dactyli VI–VII (Table 1).
Pleopods and uropods (Fig. 9). Pleopods I–III with 2-hooked retinacles. Pleopod II rami of 9–13 articles each.
Uropod I protopodite with 7 dorso-lateral and 3–4 dorso-medial spinifom setae. Length ratio endopodite : exopodite as1.00 : 0.90–1.00; rami straight. Endopodite with 9–26 longsetae in 5–8 groups; apically 5 spinifom setae. Exopodite with9–38 setae in 5–15 groups; apically 5 spinifom setae.
Uropod II endopodite : exopodite length as 1.00 : 1.05–1.15.
Uropod III up to 45% of body length. Protopodite with6–12 lateral setae and 8–12 apical spiniform and thin setae.Endopodite 40–50% of protopodite length, apically with 2–4 thin-flexible and spiniform setae; laterally 1–3 thin andspiniform setae. Exopodite of uropod III rod-shaped, distalarticle 55–100% of proximal article length. Proximal articlewith 5–6 groups of plumose, thin-flexible and spiniformsetae along inner margin; 4–6 groups of thin-flexible andspiniform setae along outer margin. Distal article with 2–5lateral setae groups along each side; apically 3–5 setae.
References
Agapow, P. M., Bininda-Emonds, O. P. R., Crandall, K. A., Gittleman,J. L., Mace, G. M., Marshall, J. C., et al. (2004). The impact ofspecies concept on biodiversity studies. The Quarterly Review ofBiology, 79, 161–179.
Beyer, H. L. (2008). Hawth’s Analysis Tools for ArcGIS. SpatialEcol-ogy.com. http://www.spatialecology.com/htools. Accessed 15June 2008.
Bickford, D., Lohman, D. J., Sodhi, N. S., Ng, P. K. L., Meier, R.,Winker, K., et al. (2007). Cryptic species as a window ondiversity and conservation. Trends in Ecology and Evolution, 22,148–155.
Bininda-Emonds, O. R. P., Vázquez, D. P., & Manne, L. L. (2000).The calculus of biodiversity: integrating phylogeny and conser-vation. Trends in Ecology and Evolution, 15, 92–94.
Chessel, D., Dufour, A.-B., & Dray, S. (2008). The ade4 Package.Biometry and Evolutionary Biology Lab, University Lyon 1.http://pbil.univ-lyon1.fr/ADE-4. Accessed 20 June 2008.
20 C. Fišer et al.
Coppellotti Krupa, O., & Guidolin, L. (2003). Responses ofNiphargus montellianus and Gammarus balcanicus (Crustacea,Amphipoda) from karst waters to heavy metal exposure. Journalde Physique IV, 107, 323–326.
Crandall, K. A., Bininda-Emonds, O. R. P., Mace, G. M., & Wayne,R. K. (2000). Considering evolutionary processes in conservationbiology. Trends in Ecology and Evolution, 15, 290–295.
Dallwitz, M. J., Paine, T. A., & Zurcher, E. J. (2007). DELTA—DEscription Language for TAxonomy [computer software andmanuals]. DELTAWebsite. http://delta-intkey.com/. Accessed 15May 2007.
Davis, J. I., & Nixon, K. C. (1992). Populations, genetic variation, andthe delimitation of phylogenetic species. Systematic Biology, 42,421–445.
DeSalle, R., Egan, M. A., & Siddall, M. (2005). The unholly trinity:taxonomy, species delimitation and DNA barcoding. Philosoph-ical Transactions of the Royal Society, Series B, BiologicalSciences, 360, 1905–1916.
de Queiroz, K. (1998). General lineage concept of species, speciescriteria, and the process of speciation. In D. J. Howard & S. H.Berlocher (Eds.), Endless forms: Species and speciation (pp. 57–75). Oxford: Oxford University Press.
Queiroz, K. de (2005a). Ernst Mayr and the modern concept ofspecies. Proceedings of the National Academy of Sciences of theUnited States of America, 102, 6600–6607.
de Queiroz, K. (2005b). Different species problems and theirresolution. BioEssays, 27, 1263–1269.
de Queiroz, K. (2007). Toward an integrated system of clade names.Systematic Biology, 56, 956–974.
Fišer, C., & Zagmajster, M. (2009). Cryptic species from crypticspace: the case of Niphargus fongi sp. n. (Crustacea, Amphipoda,Niphargidae). Crustaceana, 82, 593–614.
Fišer, C., Trontelj, P., & Sket, B. (2006a). Phylogenetic analysis of theNiphargus orcinus species-aggregate (Crustacea: Amphipoda:Niphargidae) with description of new taxa. Journal of NaturalHistory, 40, 2265–2315.
Fišer, C., Sket, B., & Stoch, F. (2006b). Distribution of four narrowlyendemic Niphargus species (Crustacea: Amphipoda) in thewestern Dinaric region with description of a new species.Zoologischer Anzeiger, 245, 77–94.
Fišer, C., Zakšek, V., Zagmajster, M., & Sket, B. (2007a). Taxonomyand biogeography of Niphargus steueri (Crustacea: Amphipoda).Limnology, 8, 297–309.
Fišer, C., Keber, R., Kereži, V., Moškrič, A., Palandačić, A., Potočnik,H., et al. (2007b). Coexistence of two species of two amphipodgenera: Niphargus timavi (Niphargidae) and Gammarus fossarum(Gammaridae). Journal of Natural History, 41, 2641–2651.
Fišer, C., Bininda-Emonds, O. P. R., Blejec, A., & Sket, B. (2008a).Can heterochrony help explain the high morphological diversitywithin the genus Niphargus (Crustacea: Amphipoda)? Organ-isms, Diversity and Evolution, 8, 146–162.
Fišer, C., Sket, B., & Trontelj, P. (2008b). A phylogenetic perspectiveon 160 years of troubled taxonomy of Niphargus (Crustacea:Amphipoda). Zoologica Scripta, 37, 665–680.
Fišer, C., Çamur-Elipek, B., & Özbek, M. (2009a). The subterraneangenus Niphargus (Crustacea, Amphipoda) in the Middle East: afaunistic overview with descriptions of two new species.Zoologischer Anzeiger, 248, 137–150.
Fišer, C., Sket, B., Turjak, M., & Trontelj, P. (2009b). Public onlinedatabases as a tool of collaborative taxonomy: a case study onsubterranean amphipods. Zootaxa, 2095, 47–56.
Fišer, C., Trontelj, P., Luštrik, R., & Sket, B. (2009c). Toward aunified taxonomy of Niphargus (Crustacea: Amphipoda): areview of morphological variability. Zootaxa, 2061, 1–22.
Folquier, A., Malard, F., Lefébure, T., Douady, C. J., & Gibert, J.(2008). The imprint of Quaternary glaciers on the present-day
distribution of the obligate groundwater amphipod Niphargusvirei (Niphargidae). Journal of Biogeography, 35, 552–564.
Funk, V. A., Sakai, A. K., & Richardson, K. (2002). Biodiversity: theinterface between systematics and conservation. SystematicBiology, 51, 235–237.
Gams, I. (1974). Kras. Zgodovinski, naravoslovni in geografski oris.Ljubljana: Slovenska Matica.
Ginet, R. (1965). Expérience de colonisation souterraine aquatique parNiphargus (Crustacé amphipode); premiers résultats biologiques.Bulletin de la Société Zoologique de France, 90, 581–588.
Grabowski, M., Bacela, K., & Konopacka, A. (2007). How to be aninvasive gammarid (Amphipoda: Gammaroidea)—comparison oflife history traits. Hydrobiologia, 590, 75–84.
Graham, C. H., Ferrier, S., Huettman, F., Moritz, C., & Townsend,P. A. (2004). New developments in museum-based informaticsand applications in biodiversity analysis. Trends in Ecology andEvolution, 19, 497–503.
Graybeal, A. (1995). Naming species. Systematic Biology, 44, 237–250.
Hervant, F., & Mathieu, J. (1995). Ventilatory and locomotory activitiesin anoxia and subsequent recovery of epigean and hypogeancrustaceans. Comptes Rendus de l’Académie des Sciences, Paris,Série 3, Sciences de la vie / Life Sciences, 318, 585–592.
Hervant, F., Mathieu, J., Barré, H., Simon, K., & Pinon, C. (1997).Comparative study on the behavioral, ventilatory, and respiratoryresponses of hypogean and epigean crustaceans to long-termstarvation and subsequent feeding. Comparative Biochemistry andPhysiology A, Molecular & Integrative Physiology, 118, 1277–1283.
Hervant, F., Mathieu, J., & Barré, H. (1999a). Comparative study onthe metabolic responses of subterranean and surface-dwellingamphipods to long-term starvation and subsequent refeeding.Journal of Experimental Biology, 202, 3587–3595.
Hervant, F., Garin, D., Mathieu, J., & Freminet, A. (1999b). Lactatemetabolism and glucose turnover in the subterranean crustaceanNiphargus virei during post-hypoxic recovery. Journal ofExperimental Biology, 202, 579–592.
Hijmans, R. J., Cameron, S. E., Parra, J. L., Jones, P. G., & Jarvis, A.(2005). Very high resolution interpolated climate surfaces for globalland areas. International Journal of Climatology, 25, 1965–1978.
ICZN=International Commission on Zoological Nomenclature(1999). International Code of Zoological Nomenclature, fourthedition. International Trust for Zoological Nomenclature. http://www.iczn.org/iczn/index.jsp. Accessed 15 April 2009.
Jażdżewski, K., Konopacka, A., & Grabowski, M. (2004). Recentdrastic changes in the gammarid fauna (Crustacea, Amphipoda)of the Vistula River deltaic system in Poland caused by alieninvaders. Diversity and Distributions, 10, 81–87.
Jersche, G. (1963). Zur Artfrage und Variabilität von Niphargustatrensis Wrzesniowski (Ein Beitrag zur postembryonalen Verän-derung taxonomischer Merkmale). Zeitschrift für ZoologischeSystematik und Evolutionsforschung, 1, 240–276.
Karaman, G. S., & Ruffo, S. (1986). Amphipoda: Niphargus-group(Niphargidae sensu Bousfield, 1982). In L. Botosaneau (Ed.),Stygofauna mMundi (pp. 514–534). Leiden: E. J. Brill/Dr. W.Backhuys.
Karaman, G. S., & Ruffo, S. (1993). Niphargus foreli Humbert, 1876 andits taxonomic position (Crustacea Amphipoda, Niphargidae). Bollet-tino del Museo Civico di Storia naturale di Verona, 17, 57–68.
Kristjánsson, B. K., & Svavarsson, J. (2004). Crymostygidae, a newfamily of subterranean freshwater gammaridean amphipods(Crustacea) recorded from subarctic waters. Journal of NaturalHistory, 38, 1881–1894.
Lefébure, T., Douady, C. J., Gouy, M., Trontelj, P., Briolay, J., &Gibert, J. (2006). Phylogeography of a subterranean amphipodreveals cryptic diversity and dynamic evolution in extremeenvironments. Molecular Ecology, 15, 1797–1806.
Old museum samples and recent taxonomy: A taxonomic, biogeographic and conservation perspective 21
Lefébure, T., Douady, C. J., Malard, F., & Gibert, J. (2007). Testingdispersal and cryptic diversity in a widely distributed groundwa-ter amphipod (Niphargus rhenorhodanesis). Molecular Phyloge-netics and Evolution, 42, 676–686.
Mayden, R. L. (1997). A hierarchy of species concepts: thedenouncement in the saga of the species problem. In M. F.Claridge, H. A. Dawah, & M. R. Wilson (Eds.), Species: theunits of biodiversity (pp. 381–424). London: Chapman and Hall.
Micherdziński, W. (1956). Taksonomia i ekologia Niphargus tatrensisWrześniowski 1888 (Amphipoda). Annales Zoologici, 16, 81–134.
Moritz, C. (1994). Defining ‘Evolutionary significant units’ forconservation. Trends in Ecology and Evolution, 9, 373–375.
Moritz, C. (2002). Strategies to protect biological diversity and theevolutionary processes that sustain it. Systematic Biology, 51,238–254.
Myers, N., Mittermeier, R. A., Mittermeier, C. G., da Fonseca, G. A.B., & Kent, J. (2000). Biodiversity hotspots for conservationpriorities. Nature, 403, 853–858.
Nesemann, H., Pöckl, M., & Wittmann, K. J. (1995). Distribution ofepigean Malacostraca in the middle and upper Danube (Hungary,Austria, Germany). Miscellanea Zoologica Hungarica, 10, 49–68.
Raxworthy, Ch J, Ingram, C. M., Rabibisoa, N., & Pearson, R. G.(2007). Applications of ecological niche modeling for speciesdelimitation: a review and empirical evaluation using day Geckos(Phelsuma) from Madagascar. Systematic Biology, 56, 907–923.
Ruffo, S. (1972). Actes du Ier Colloque Internationale sur le genreNiphargus, Verona 15–16 aprile 1969. Verona: Museo Civico diStoria Naturale di Verona.
Savage, A. A. (1981). The Gammaridae and Corixidae of an inlandsaline lake from 1975–1978. Hydrobiologia, 76, 33–44.
Sax, D. F., Stachowicz, J. J., Brown, J. H., Bruno, J. F., Dawson,M. N., Gaines, S. D., et al. (2007). Ecological and evolutionaryinsights from species invasions. Trends in Ecology and Evolu-tion, 22, 465–471.
Schellenberg, A. (1935). Schlüssel der Amphipodengattung Niphar-gus mit Fundortangaben und mehreren neuen Formen. Zoolog-ischer Anzeiger, 111, 204–211.
Schellenberg, A. (1937). Bemerkungen zu meinem Niphargus-Schlüssel und zur Verbreitung und Variabilität der Arten, nebstBeschreibung neuer Niphargus-Formen. Mitteilungen aus demZoologischen Museum in Berlin, 22, 1–30.
Schellenberg, A. (1938). Alters-, Geschlechts- und Individualunter-schiede des Amphipoden Niphargus tatrensis f. aggtelekiensisDudich. Zoologische Jahrbücher, Abteilung für Systematik,Ökologie und Geographie der Tiere, 71, 191–202.
Simčič, T., & Brancelj, A. (2006). Effects of pH on electron transportsystem (ETS) activity and oxygen consumption in Gammarusfossarum, Asellus aquaticus and Niphargus sphagnicolus. Fresh-water Biology, 51, 686–694.
Sites, J. W., & Crandall, K. A. (1997). Testing species boundaries inbiodiversity studies. Conservation Biology, 11, 1289–1297.
Sites, J. W., & Marshall, J. C. (2003). Delimiting species: arenaissance issue in systematic biology. Trends in Ecology andEvolution, 18, 462–470.
Sites, J. W., & Marshall, J. C. (2004). Operational criteria fordelimiting species. Annual Review of Ecology Evolution andSystematics, 35, 199–227.
Skalski, A. W. (1972). Distribution des amphipodes souterrains enPologne, avec notes sur la variabilité du Niphargus tatrensisWrześniowski. In S. Ruffo (Ed.), Actes du Ier Colloque Internatio-
nale sur le genre Niphargus, Verona 15–16 aprile 1969 (pp. 47–53). Verona: Museo Civico di Storia Naturale di Verona.
Sket, B. (1958). Prispevek k poznavanju naših amfipodov. BiološkiVestnik, 6, 67–75.
Sket, B. (1977). Niphargus im Brackwasser. Crustaceana Supplement,4, 188–191.
Sket, B. (1981). Distribution, ecological character, and phylogeneticimportance of Niphargus valachicus. Biološki Vestnik, 29, 87–103.
Sket, B. (1994). Distribution patterns of some subterranean Crustaceain the territory of the former Yugoslavia. Hydrobiologia, 287,65–75.
Stuart, B. L., Dugan, K. A., Allard, M. W., & Kearney, M. (2006).Extraction of nuclear DNA from bone of skeletonized and fluid-preserved museum specimens. Systematics and Biodiversity, 4,133–136.
Trontelj, P., Douady, C., Fišer, C., Gibert, J., Gorički, Š., Lefébure, T.,et al. (2009). A molecular test for hidden biodiversity ingroundwater: how large are the ranges of macro-stygobionts?Freshwater Biology, 54, 727–744.
Väinölä, R., Witt, J. D. S., Grabowski, M., Bradbury, J. H.,Jażdżewski, K., & Sket, B. (2008). Global diversity ofamphipods (Amphipoda; Crustacea) in freshwater. Hydrobiolo-gia, 595, 241–255.
Vellend, M., Harmon, L. J., Lockwood, J. L., Mayfield, M. M.,Hughes, A. R., Wares, J. P., et al. (2007). Effects of exoticspecies on evolutionary diversification. Trends in Ecology andEvolution, 22, 481–488.
Wandeler, P., Hoeck, P. E., & Keller, L. F. (2007). Back to the future:museum specimens in population genetics. Trends in Ecologyand Evolution, 22, 634–642.
Wheeler, Q. D. (2004). Taxonomic triage and the poverty ofphylogeny. Philosophical Transactions of the Royal Society ofLondon, Series B, Biological Sciences, 359, 571–583.
Wiens, J. J. (1995). Polymorphic characters in phylogenetic system-atics. Systematic Biology, 44, 482–500.
Wiens, J. J. (1999). Polymorphism in systematic and comparativebiology. Annual Review of Ecology, Evolution, and Systematics,30, 327–362.
Wiens, J. J. (2001). Character analysis in morphological phyloge-netics: problems and solutions. Systematic Biology, 50, 689–699.
Wiens, J. J. (2004a). The role of morphological data in phylogenyreconstruction. Systematic Biology, 53, 653–661.
Wiens, J. J. (2004b). Speciation and ecology revisited: phylogenetic nicheconservatism and the origin of species. Evolution, 58, 193–197.
Wiens, J. J. (2007). Species delimitation: new approaches fordiscovering diversity. Systematic Biology, 56, 875–878.
Wiens, J. J., & Graham, C. H. (2005). Niche conservatism: integratingevolution, ecology, and conservation biology. Annual Review ofEcology, Evolution, and Systematics, 36, 519–539.
Wiens, J. J., & Penkrot, T. A. (2002). Delimiting species using DNAand morphological variation and discordant species limits inspiny lizards (Sceloporus). Systematic Biology, 51, 69–91.
Wiens, J. J., & Servedio, M. R. (1998). Phylogenetic analysis andintraspecific variation: performance of parsimony, likelihood, anddistance methods. Systematic Biology, 47, 228–253.
Wiens, J. J., & Servedio, M. R. (2000). Species delimitation insystematics: inferring diagnostic differences between species.Proceedings of the Royal Society of London / B, 267, 631–636.
Winston, J. E. (2007). Archives of a small planet: The significance ofmuseum collections and museum based research in invertebratetaxonomy. Zootaxa, 1668, 47–54.
22 C. Fišer et al.