Cryptic species in Lepidocyrtus lanuginosus (Collembola: Lepidocyrtidae) are sorted by habitat types

Abstract

High intraspecific genetic variance in Collembola indicates that cryptic species are widespread and this chal- lenges the delimitation of morphologically defined species. Lepidocyrtus lanuginosus (Gmelin, 1788) is a widely distributed habitat generalist with high genetic variance between populations from different locations in Europe. In this study we investigated the genetic variance of L. lanuginosus from three dominant habitat types in Central Europe, i.e. forests, grasslands and arable fields, using four molecular markers (ribosomal subunit 28S rDNA D1–2 domain, elongation factor 1-α, cytochrome c oxidase subunit I and subunit II). The results suggest that L. lanuginosus separates into three major genetic lineages with one of these lineages being close to Lepidocyrtus cyaneus Tullberg, 1871. The phylogenetic tree based on the concatenated data set of four genes suggests that all lineages of L. lanuginosus are monophyletic. Selective colonization of the three habitats by these lineages in- dicates that they are sorted by habitats: one lineage was common and occurred in each of the three habitat types but preferentially in arable land, the second was restricted to forest, and the third, although rare, preferentially occurred in grassland. Our results indicate that genetic markers allow delineating cryptic species in Collembola, which are widespread, morphologically coherent and differ in the habitats they colonize. The existence of cryptic species/lineages in widely distributed Collembola species that sort by habitat type calls for studies integrating genetic structure and ecological traits.

Full text

Contents lists available at ScienceDirect
Pedobiologia - Journal of Soil Ecology
jou rna l h ome pa ge: www.e lse vier. com /loca te/pedob i
Cryptic species in (Collembola: Entomobryidae) areLepidocyrtus lanuginosus
sorted by habitat type
Bing Zhang
a,⁎,1
,Ting-Wen Chen
a 1,
,Eduardo Mateos
b
,Stefan Scheu
a c,
,Ina Schaefer
a
a
University of Göttingen, J.F. Blumenbach Institute of Zoology and Anthropology, Animal Ecology, Untere Karspüle 2, 37073, Göttingen, Germany
b
Departament de Biologia Evolutiva, Ecologia i Ciències Ambientals, Facultat de Biologia, Universitat de Barcelona, Avinguda Diagonal 643, 08028, Barcelona, Spain
c
University of Göttingen, Centre of Biodiversity and Sustainable Land Use, Von-Siebold-Str. 8, 37075, Göttingen, Germany
A R T I C L E I N F O
Keywords:
DNA barcoding
Genetic variation
Lepidocyrtus cyaneus
Mitochondrial genes
Pigmentation
Springtail
A B S T R A C T
High intraspeci c genetic variance in Collembola indicates that cryptic species are widespread and this chal-fi
lenges the delimitation of morphologically de ned species. (Gmelin, 1788) is a widelyfiLepidocyrtus lanuginosus
distributed habitat generalist with high genetic variance between populations from di erent locations in Europe.ff
In this study we investigated the genetic variance of from three dominant habitat types in CentralL. lanuginosus
Europe, i.e. forests, grasslands and arable elds, using four molecular markers (ribosomal subunit 28S rDNAfi
D1 2 domain, elongation factor 1- , cytochrome c oxidase subunit I and subunit II). The results suggest that–αL.
lanuginosus Lepidocyrtusseparates into three major genetic lineages with one of these lineages being close to
cyaneus Tullberg, 1871. The phylogenetic tree based on the concatenated data set of four genes suggests that all
lineages of are monophyletic. Selective colonization of the three habitats by these lineages in-L. lanuginosus
dicates that they are sorted by habitats: one lineage was common and occurred in each of the three habitat types
but preferentially in arable land, the second was restricted to forest, and the third, although rare, preferentially
occurred in grassland. Our results indicate that genetic markers allow delineating cryptic species in Collembola,
which are widespread, morphologically coherent and di er in the habitats they colonize. The existence of crypticff
species/lineages in widely distributed Collembola species that sort by habitat type calls for studies integrating
genetic structure and ecological traits.
1. Introduction
The development of molecular methods in the past decade triggered
the discovery of cryptic species ( ;Bickford et al., 2007 Emerson et al.,
2011), in particular among soil organisms. Soil animal species show
exceptionally high genetic variability as documented for earthworms
( ; ), oribatid mites (James et al., 2010 King et al., 2008 Rosenberger
et al., 2013 Schä er et al., 2010 Cicconardi et al.,;ff) and springtails (
2013, 2010). High genetic variability of morphospecies suggests the
existence of cryptic species, which may substantially contribute to the
diversity of belowground invertebrates ( ). CrypticPorco et al., 2012a
species are morphologically indistinguishable but show genetic di er-ff
entiation, suggesting that they di er in their ecological nichesff
( ; ) as a result of niche-Bidochka et al., 2001 Davidson-Watts et al., 2006
di erentiation within a habitat or adaptation to di erent habitatsff ff
( ; ). Environmental di erencesEisenring et al., 2016 Tarjuelo et al., 2001 ff
can impede gene flow across habitat boarders, promoting population
divergence (isolation by environment; ).Wang and Summers, 2010
Springtails (Hexapoda: Collembola) are among the most abundant and
diverse soil invertebrates and occur in virtually any terrestrial habitat
( ). Currently, about 8,600 species have been described.Hopkin, 1997
Traditionally, delimitation of Collembola species relies on morpholo-
gical characters, mainly chaetotaxy, a method that examines arrange-
ment of chaetae on di erent body parts ( ). MolecularffKatz et al., 2015
studies, however, indicated that the number of existing Collembola
species likely is much higher as many species comprise a number of
cryptic species ( ; ;Emerson et al., 2011 Porco et al., 2012a Soto-Adames,
2002). Notably, cryptic species appear to be common in widely dis-
tributed and locally abundant Collembola species, such as Orchesella
cincta Parisotoma(Linnaeus, 1758) ( ) andTimmermans et al., 2005
notabilis (Schä er, 1896) ( ;ffPorco et al., 2012b von Saltzwedel et al.,
2017).
Lepidocyrtus lanuginosus (Gmelin, 1788) is one of the most widely
distributed Collembola species in Central and Western Europe (Salmon
et al., 2014). It rapidly colonizes new habitats and is characterized by
long legs and antenna, developed furcula and complete visual apparatus
https://doi.org/10.1016/j.pedobi.2018.03.001
Received 26 September 2017; Received in revised form 25 March 2018; Accepted 26 March 2018
⁎
Corresponding author.
1
These authors contributed equally to this paper.
E-mail addresses: bzhang3@gwdg.de tchen2@gwdg.de emateos@ub.edu sscheu@gwdg.de ischaef@gwdg.de(B. Zhang), (T.-W. Chen), (E. Mateos), (S. Scheu), (I. Schaefer).
Pedobiologia - Journal of Soil Ecology 68 (2018) 12–19
0031-4056/ © 2018 Elsevier GmbH. All rights reserved.
T

with eight ocella per eye spot ( ).Ponge et al., 2006 Lepidocyrtus lanu-
ginosus has been considered as habitat generalist ( )Auclerc et al., 2009
associated with arable elds ( ) and grasslandsfiQuerner et al., 2013
( ; ), i.e. anthropogenic habitatsAuclerc et al., 2009 Heiniger et al., 2015
characterized by disturbances of varying intensity. However, the spe-
cies also colonizes forests ( ), which are ratherCicconardi et al., 2010
stable habitats. In forests of the Mediterranean region, L. lanuginosus
displays high genetic variation, suggesting the existence of cryptic
species or impeded gene ow between forests ( ).flCicconardi et al., 2010
It is therefore likely that of di erent habitats also com-L. lanuginosus ff
prises cryptic species or lineages with distinct genetic structure.
In this study we investigated phylogenetic relationships and genetic
distances between populations of from three di erentL. lanuginosus ff
habitats in Central Europe, i.e. arable elds, grasslands and forests thatfi
were replicated at six locations. Geographic distances between sam-
pling locations were considerably larger than between habitat types
allowing to identify habitat speci c genotypes. Intraspeci c geneticfi fi
variation was measured using two mitochondrial (COI and COII) and
two nuclear genes (28S ribosomal DNA D1 2 domain and elongation–
factor 1- ). Mitochondrial COI is commonly used for animal DNAα
barcoding and reliably distinguishes between Collembola species as
well as cryptic species ( ; ;Hebert et al., 2003 Hogg and Hebert, 2004
Porco et al., 2012a), while COII has been used to reconstruct phylo-
genetic relationships of Collembola including the genus Lepidocyrtus
( ; ; ).Cicconardi et al., 2010, 2013 Frati et al., 2000 Stevens et al., 2007
The two nuclear genes represent markers independent from mi-
tochondrial genes strengthening the delineation of phylogenetic re-
lationships among individuals of each lineage of L. lanuginosus. We
checked if the phylogenetic structure of is best explainedL. lanuginosus
by habitat types or by sampling locations, i.e. geographic distance. If L.
lanuginosus comprises a single generalist species individuals should sort
by geographic distances due to limited dispersal, or generate a random
pattern of relatedness in the phylogenetic tree if locations and habitats
were colonized multiple times. In contrast, if comprisesL. lanuginosus
cryptic species that colonize certain habitat types, genetic lineages
should sort by habitat types rather than locations. Further, we com-
pared intraspeci c genetic variability of with geneticfiL. lanuginosus
distances between L. lanuginosus and ve other species of the samefi
genus. These species included Tullberg, 1871Lepidocyrtus cyaneus
which is morphologically very similar to and belongs toL. lanuginosus
the group (L. lanuginosus Mateos, 2012), but it di ers fromffL. lanugi-
nosus by its purple body color ( ; ;Fjellberg, 2007 Hopkin, 2007 Mateos,
2008).
2. Material and methods
2.1. Sampling of animals and determination
The Collembola species and were collectedL. lanuginosus L. cyaneus
from six locations near the city of Göttingen, Germany; distances be-
tween sampling locations were between 4 and 28 km (Supplementary
Fig. S1). Each sampling location encompassed three types of habitats:
arable eld, grassland and forest that were close to each other withfi
distances below 1 km. Specimens in grasslands and arable elds werefi
collected using an aspirator (diameter of opening 14 cm). In forests
Collembola were collected from litter and extracted by heat (Kempson
et al., 1963). The collected litter from forest soil covered an area that
was similar to the aspirator opening used in grasslands and arable
fields. Animals were preserved in 96% EtOH and stored at 20 °C until−
further analyses. Collembola were sorted using a dissecting microscope
and identi ed using . In total 58 individuals offiHopkin (2007) L. lanu-
ginosus L. cyaneus, six individuals of and ve to six individuals of fourfi
other species ( cf Traser, 2000,Lepidocyrtus L. . arrabonicus L. paradoxus
Uzel, 1890, cf Lubbock, 1873 and cf Hüther,L. . violaceus L. . weidneri
1971) were analyzed. The sites at which the analyzed specimens of L.
lanuginosus were sampled are given in . Prior to DNA extraction,Table 1
we cut o the head of each specimen and only placed thorax and ab-ff
domen into the extraction solution. After the DNA extraction, the cu-
ticle of the thorax and abdomen was rescued, rinsed with 96% EtOH
and mounted on slides together with the head of the respective in-
dividual for inspection of the cephalic and dorsal chaetotaxy under a
phase contrast microscope (Zeiss Axio Scope A1, Jena, Germany) using
Mateos (2008). Microscopic slides are kept in the collection of the J.F.
Blumenbach Institute of Zoology and Anthropology, University of
Göttingen, Germany.
2.2. DNA extraction and PCR
DNA of the thorax and abdomen of single individuals was extracted
using the DNeasy
®
Blood and Tissue Kit (Qiagen, Hilden, Germany)
following the manufacturer s protocol. Puri ed DNA was eluted in 30 l’fiμ
HPLC water for PCR. PCRs of the nuclear markers elongation factor 1-α
(EF1- ) and 28S rDNA (D1 2 region), and the mitochondrial markersα–
COI and COII were performed separately in 25 l volumes containingμ
12.5 l SuperHot Taq Mastermix (Genaxxon Bioscience GmbH, Ulm,μ
Germany) with 1.5 l of each primer (10 pM), 3μ μl H
2
O, 1.5 l MgClμ
2
(25 mM) and 5 l template DNA. A 208 bp fragment of the nuclear EF1-μ
αexon was ampli ed using the primers EFLcuJ: 5 -ATG GGG GCA AGAfi′
TAG CGT CAA-3 and EFLcuN: 5 -TGA AGG CTG AAC GTG AAC GTG G-′ ′
3 ( ). The 760 bp fragment domain of D1 and′Cicconardi et al., 2010 ∼
D2 loop of the nuclear 28S rDNA was ampli ed using the primers C1 :fi′
5 -ACC CGC TGA ATT TAA GCA T-3 and D2coll: 5 - ACC ACG CAT GCW′ ′ ′
TTA GAT TG 3 ( ). A 709 bp fragment of the mi-−′D Haese, 2002’
tochondrial COI gene was ampli ed using four primers LCO1490: 5 -fi′
GGT CAA CAA ATC ATA AAG ATA TTG G-3 and HCO2198: 5 -TAA′ ′
ACT TCA GGG TGA CCA AAA AAT CA-3 ( ); ColFol-′Folmer et al., 1994
for: 5 -TTT CAA CAA ATC ATA ARG AYA TYG G-3 and ColFol-rev: 5 -′ ′ ′
TAA ACT TCN GGR TGN CCA AAA AAT CA-3 (′Ramirez-Gonzalez et al.,
2013). A 681 bp fragment of the mitochondrial COII gene was ampli edfi
using primers tRNA-K-LcuJ: 5 -GAG CGT ATT ATA AAG CGG TTT AAG-′
3 and tRNA-′ L-LcuN: 5 -CAG ACT AGT GCC ATG AAT TTA AGC-3′ ′
( ). PCR conditions included one initial activationCicconardi et al., 2010
step at 95° C for 15 min, followed by 35 ampli cation cycles of dena-fi
turation at 94° C (COI and 28S D1 2) or 95° C (COII and EF1- ) for 30 s,–α
annealing at 45° C (COI) or 56° C (COII and EF1- ) for 30 s or 50° C (28Sα
D1 2) for 45 s, and elongation at 72° C for 30 s, and a nal elongation–fi
step at 72° C for 6 min. PCR products were puri ed with Genaxxon PCRfi
Puri cation Kit (Genaxxon Bioscience GmbH, Ulm, Germany) followingfi
the manufacturer s protocol, eluted in 30 l HPLC water and sent for’μ
sequencing in the Göttingen Genome Laboratory (Institute for Micro-
biology and Genetics, University of Göttingen). All sequences are
available at NCBI GenBank (MH143920-MH144050, MH160100-
MH160185, and MH177993-MH178033, Supplementary Table S1).
2.3. Data analyses
Sequences were checked by eye aided by the chromatograms and
ambiguous positions were corrected by hand. Nucleotide sequences
Table 1
Numbers of individuals of analyzed from three habitatsLepidocyrtus lanuginosus
at the six sampling sites. The lineage that the sampled specimens were ascribed
to is given in brackets. Names of sampling sites follow nearby villages; for exact
geographic locations see Supplementary Fig. S1.
Site No. Sampling site Arable land Grassland Forest
1 Herberhausen 4 (L1) 5 (L3) 5 (L2)
2 Deppoldshausen 5 (L1) 0 5 (L2)
3 Ellershausen 3 (L1) 0 5 (L2)
4 Ossenfeld 1 (L1) 0 3 (L2)
5 Waake 2 (L1)/3 (L3) 0 5 (L2)
6 Billingshausen 3 (L1) 4 (L1) 5 (L1)
B. Zhang et al. Pedobiologia - Journal of Soil Ecology 68 (2018) 12–19
13

were aligned separately for each marker using ClustalW (Thompson
et al., 1994 Hall, 1999) implemented in BioEdit v7.2.5 ( ). To analyze
relationships and identify lineages among all 58 individuals of L. la-
nuginosus, phylogenetic trees were constructed based on 28S D1 2 and–
COII with (Geo roy, 1764) as outgroup. The best tOrchesella villosa ff fi
model of sequence evolution for each marker was estimated by jMo-
delTest v2.1.4 ( ). Phylogenetic trees were inferred usingPosada, 2008
Bayesian Inference in MrBayes v3.2 ( ). For Baye-Ronquist et al., 2012
sian Inference lset parameters were nst = 2 and rates = gamma for 28S
D1 2 and nst = 2 and rates = propinv for COII, the MCMC (Markov–
Chain Monte Carlo) chains were run for six million generations that
were sampled every 6,000th generation. For the 1,000 sampled gen-
erations a burnin of 250 was used, eliminating the rst 25% of thefi
remaining generations. Intra- and inter-lineage and intra- and inter-
speci c sequence divergences of 28S D1 2, EF1- , COI and COII werefi–α
calculated based on Kimura two-parameter distances (K2P) and un-
corrected p-distances using Mega 5.1 ( ). IndividualsTamura et al., 2011
with identical 28S D1 2 sequences and less than 4% K2P distances of–
COII were de ned as one lineage ( ).fiPorco et al., 2012b
We used six individuals from each lineage of andL. lanuginosus L.
cyaneus and one individual of the extra four species from genus
Lepidocyrtus Orchesella villosato construct a phylogenetic tree. was
chosen as outgroup. In a 2,282 bp concatenated alignment four genes
were merged: 28S D1 2, COII, COI and EF1- of a length of 740 bp,–α
681 bp, 651 bp and 210 bp, respectively. The partitioned dataset was
analyzed using MrBayes v3.2.6 on online CIPRES services (Miller et al.,
2010). All three coding positions of the protein-coding genes COI, COII
and EF1- were included in the analyses. Best- tting substitutionαfi
models were assessed for each locus (partition) under the BIC criterion
in PartitionFinder 2.1.1 ( ).Lanfear et al., 2012
Habitat preference of each lineage of was character-L. lanuginosus
ized using the IndVal index ( ) whichDufrêne and Legendre, 1997
combines the speci city of a lineage for a habitat type (maximizedfi
when the lineage is only present in the habitat type analyzed) and its
fidelity to this habitat (maximized when the lineage is present in all
samples of the habitat type analyzed): Ind
ij
= A
ij
× B
ij
× 100, with A
ij
,
the average abundance of lineage in samples of habitat divided by thei j
sum of the average abundance of lineage in all habitats, andi B
ij
, the
number of samples of habitat where the lineage was present dividedj
by the total number of samples of habitat .j
Ind
ij
ranges from 0, when lineage is absent from habitat , to 100,i j
when lineage is present in all samples of habitat and absent in alli j
other habitat samples. We obtained three IndVal values for each lineage
of , i.e. IndA, IndG and IndF indicating habitat preferenceL. lanuginosus
for arable land, grassland and forest, respectively.
To test whether there is a signi cant habitat association of eachfi
lineage of we simulated the habitat distributions of theL. lanuginosus,
58 examined individuals by randomizing their habitats 9,999 times.
The observed number of a lineage in a habitat was compared to that
derived from the simulations and a p-value was calculated based on the
rank of the observed number as compared to the simulated ones. A p-
value smaller than 0.05 indicated signi cant association (or avoidance)fi
of the respective habitat by the lineage.
3. Results
3.1. Habitat sorting of lineages of Lepidocyrtus lanuginosus
All individuals studied under phase contrast microscope shared the
characters of and as indicated inL. lanuginosus L. cyaneus Mateos
(2008). Genetic analyses based on the nuclear marker 28S D1 2 and the–
mitochondrial marker COII both revealed three distinct lineages of L.
lanuginosus, L. lanuginosusreferred to as lineage 1, 2 and 3 (L1, L2, L3)
( ). Among the 58 sequenced specimens, 21 individuals were fromFig. 1
arable elds, 9 from grasslands and 28 from forests. Most individualsfi
(n = 27) belonged to L1, which occurred in each of theL. lanuginosus
three habitat types but was signi cantly associated with arable landfi
(IndA = 66.7, IndG = 2.5, IndF = 3.1; ).Fig. 2 Lepidocyrtus lanuginosus
L2 also was common (n = 23), but was signi cantly associated withfi
forest (IndA = 0, IndG = 0, IndF = 83.3). L3Lepidocyrtus lanuginosus
was rare (n = 8), but was signi cantly associated with grasslandfi
(IndA = 6.3, IndG = 10.4, IndF = 0).
At least two lineages were present in each habitatL. lanuginosus
( and ). Arable elds were dominated by L1,Fig. 1 Table 1 fiL. lanuginosus
while L3 only occurred in one of the six arable elds. InL. lanuginosus fi
grassland L1 and L3 were present but both were rare;L. lanuginosus
each lineage occurred in only one location. In forest, L2L. lanuginosus
dominated and L1 occurred in only one of the forest lo-L. lanuginosus
cations.
3.2. Genetic distances and variances and phylogenetic analysis
In each of the four markers, K2P and uncorrected p-distances be-
tween lineages were considerably larger than within lineages ( ).Table 2
For the mitochondrial markers, mean K2P and uncorrected p-distances
within L1, L2 and L3 were below 1.2%; both nuclearL. lanuginosus
markers were identical within lineages. By contrast, mean K2P between
the three lineages were similar and ranged betweenL. lanuginosus
22.2% and 27.5% for both mitochondrial markers (19% and 22.8% for
uncorrected p-distances), between 1.3% and 2.8% for 28S D1 2 (1.2%–
and 2.9% for uncorrected p-distances), and between 1.5% and 5.1% for
EF1- (1.2% and 4.9% for uncorrected p-distances).α
Notably, mean K2P distances of mitochondrial markers between L.
cyaneus L. lanuginosusand L1 were 16.8% 17.7%, i.e. about one third–
lower than between and the other two lineagesL. cyaneus
(22.8% 25.5% for L2 and 21.6% 25.9% for L3). Similarly, genetic– –
distances of the nuclear markers were lowest between andL. cyaneus L.
lanuginosus L1 (0.1% in 28S and 1.5% in EF1- ), but were larger be-α
tween and L2 (2.8% in 28S and 4.5% in EF1- )L. cyaneus L. lanuginosus α
and L3 (2.5% in 28S D1 2 and 4.6% in EF1- ). Patterns for uncorrected–α
p-distances were similar ( ). K2P distances of COI between theTable 2
three lineages of were larger than 20% (larger than 17.8%L. lanuginosus
for uncorrected p-distances). This resembled distances between the
three lineages of and four other well-de ned species ofL. lanuginosus fi
Lepidocyrtus L. . arrabonicus L. paradoxus L. . violaceus L. ., i.e. cf , , cf and cf
weidneri ( ). By contrast, for COII, 28S D1 2 and EF1- meanTable 2 –α
genetic distances between the three lineages of wereL. lanuginosus
lower than the distances between these three lineages and the other
four species.Lepidocyrtus
The phylogenetic tree based on the combined matrix of the four
genes (28S D1– 2, EF1- ,α COI and COII) showed that all lineages of L.
lanuginosus are monophyletic (Bayesian posterior probabilities
(PP) = 1; ). Applying the standard mutation rate of COI for ar-Fig. 3
thropods of 1.5 2.3% sequence divergence per million years (–Brower,
1994 Avise, 1994; ) to the mean K2P distances suggests that the three L.
lanuginosus lineages diverged 15.9 9.7 million years ago (Mya) in the–
Miocene (13.4 8.2 Mya based on p-distances).–
4. Discussion
4.1. Habitat sorting
Results of the present study suggest that the morphotype of L. la-
nuginosus does not represent a single monophyletic species but rather
they suggest the existence of three distinct genetic lineages in a narrow
region in Central Europe. One lineage occurred in each of the three
habitat types (L1) and might be regarded as a habitat generalist but
with a preference for arable elds. The other two lineages were re-fi
stricted either to forests (L2) or to grasslands (L3), suggesting more
specialized ecotypes of . The proximity of the di erentL. lanuginosus ff
habitats at the six sites supports di erential colonization of forest andff
open habitats (arable elds and grassland) indicating that the di erentfi ff
B. Zhang et al. Pedobiologia - Journal of Soil Ecology 68 (2018) 12–19
14

lineages were sorted by environmental factors. The high genetic dis-
tances between lineages suggest impeded gene ow either by isolationfl
by distance ( ) or isolation by environmentvan der Wur et al., 2005ff
( ). Dispersal limitation between habitats isWang and Summers, 2010
unlikely because is a mobile and surface active speciesL. lanuginosus
( ; ), exhibiting high mobility inPonge et al., 2006 Salmon et al., 2014
grasslands as well as forests ( ). Rather, sorting ofAuclerc et al., 2009
lineages into di erent habitats due to lineage-speci c preferences orff fi
physiological constraints to habitat characteristics is more likely. For-
ests are stable habitats usually with thick organic layers and high soil
carbon content providing a variety of resources ( ). InHopkin, 1997
contrast, open habitats such as arable elds and grasslands typically arefi
characterized by disturbances due to agricultural practices and stronger
fluctuations in temperature and soil moisture (Batlle-Aguilar et al.,
2011). Di erences in tolerance to disturbances among lineages areff
possible, but ecophysiological experiments are needed to test this hy-
pothesis. Our results suggest habitat preferences of genetic lineages in
L. lanuginosus but this needs to be con rmed by more extensive sam-fi
pling of di erent habitats over a larger geographic area in particular forff
the rare genotype L3.L. lanuginosus
Fig. 1. Bayesian phylogenetic tree of based on single molecular markers: COII (left) and 28S D1-2 (right). Labels of OTUs represent samplingLepidocyrtus lanuginosus
locations 1-6 and habitat types: A, arable land (red); G, grassland (green); F, forest (yellow). (For interpretation of the references to color in this gure legend, thefi
reader is referred to the web version of this article.)
B. Zhang et al. Pedobiologia - Journal of Soil Ecology 68 (2018) 12–19
15

4.2. Genetic distances
High genetic distance between the three lineagesL. lanuginosus
suggests the existence of cryptic species in the collected morphologi-
cally coherent specimens. Each of the four genetic markers allowed to
separate the three lineages of and thus can be used asL. lanuginosus
barcoding marker. Inter-lineage K2P distances in COI were larger than
21%, which is similar to the inter-lineage distances of cryptic species
described by , and almost twice as high as thePorco et al. (2012a)
suggested boundary of intraspeci c variability of Collembola (fiAnslan
and Tedersoo, 2015 Porco et al., 2014; ). The inter-lineage genetic dis-
tances of the nuclear marker 28S D1 2 also exceeded the previously–
suggested minimum interspeci c genetic distance between di erentfi ff
Collembola species ( ). In addition, K2P andAnslan and Tedersoo, 2015
uncorrected p-distances in both of the mitochondrial and nuclear
markers between the three lineages of exceeded thoseL. lanuginosus
between and L1 emphasizing the existence ofL. cyaneus L. lanuginosus
cryptic species within .L. lanuginosus
Lepidocyrtus lanuginosus is widespread and frequently reported in
soil ecological studies, as we obtained 355 hits for this species in Google
Scholar for publications before 2016. Cryptic species within L. lanugi-
nosus without clearly delimiting morphological characters suggest that
the diversity of Collembola is underestimated. Our study adds to recent
evidence that cryptic species commonly exist within morphospecies of
Collembola such as (Moniez, 1889),Heteromurus major Deutonura
monticola Parisotoma(Cassagnau, 1954) ( ), andPorco et al., 2012a
notabilis (Schä er, 1896) ( ). Generalist CollembolaffPorco et al., 2012a,b
species colonize a variety of habitats ( ;Auclerc et al., 2009 Heiniger
et al., 2015 Carapelli et al., 1995). However, except for few studies ( ;
Porco et al., 2014) only single habitat types were investigated
( ; ; ;Cicconardi et al., 2013 Porco et al., 2012b Timmermans et al., 2005
von Saltzwedel et al., 2016), thereby disregarding environmental as-
sociations of cryptic species. Taking habitat types into account when
investigating genetic diversity of ubiquitous Collembola may lead to the
Fig. 2. Frequency of randomized numbers of individuals (58) of three lineages of in the three habitats studied. Red lines indicate observedLepidocyrtus lanuginosus
numbers of individuals; bars show the frequencies of simulated habitat distribution derived from 9,999 permutations. Stars indicate p-values based on the rank of
observed number compared to the simulated numbers: ***, p < 0.001; **, p < 0.01; NS, not signi cant. (For interpretation of the references to color in this gurefi fi
legend, the reader is referred to the web version of this article.)
B. Zhang et al. Pedobiologia - Journal of Soil Ecology 68 (2018) 12–19
16

discovery of cryptic or new species and provide a more accurate picture
of local and regional diversity.
4.3. Systematics
According to identi cation keys ( ; ;fiFjellberg, 2007 Hopkin, 2007
Mateos, 2008) all three lineages in this study unequivocally belong to L.
lanuginosus L.with common main characters, di ering clearly fromff
cyaneus in pigmentation. Earlier studies based on COI proposed that
color patterns are valid to discriminate species in NorthLepidocyrtus
America ( ). However, as shown in the concatenatedSoto-Adames, 2002
phylogenetic tree of this study, the three lineages of areL. lanuginosus
well separated, at least in Central Europe. At our study sites, L. cyaneus
is more closely related to L1 than L1 to L2L. lanuginosus L. lanuginosus
and L3, suggesting that may form part ofL. cyaneus L. lanuginosus
complex. Further taxonomic investigations are needed to explore
morphological di erences of the lineages found in ourffL. lanuginosus
study and in the Mediterranean ( ), e.g. bodyCicconardi et al., 2010
chaetotaxy ( ). Unequivocal identi cation of lineages ofMateos, 2012 fiL.
lanuginosus with each of the four genetic markers used in this study
Table 2
Mean inter- and intraspeci c K2P and uncorrected p-distances (%) between the three lineages of (L1 L3) and between these lineages and vefiLepidocyrtus lanuginosus –fi
other species of the genus . Upper rows represent distances in COI (left panel) and 28S D1 D2 (right panel), lower rows highlighted in grey representLepidocyrtus –
distances in COII (left panel) and elongation factor 1- (right panel). For each lineage/species the number of individuals analyzed is given (N).α
B. Zhang et al. Pedobiologia - Journal of Soil Ecology 68 (2018) 12–19
17

demonstrates that molecular markers are a reliable and fast tool to
detect cryptic species, which may facilitate taxonomic research of
Collembola species and the identi cation of possible species-speci cfi fi
morphological characters.
5. Conclusions
Our ndings show that even at a narrow geographical scale di erentfi ff
genetic lineages or cryptic species of occur. LineageL. lanuginosus
sorting by habitat suggests that specimens from di erent habitats inff
close vicinity may belong to di erent cryptic lineages. Genetically theff
three lineages are as distant to each other as good species.L. lanuginosus
Future studies need to explore the physiological characters responsible
for habitat sorting of cryptic species/lineages of the L. lanuginosus
species complex.
Acknowledgements
We thank Jo-Fan Chao for her help in laboratory work. The study
was supported by the German Ministry of Education, Science and
Technology (BMBF). Bing Zhang was supported by the China
Scholarship Council.
Appendix A. Supplementary data
Supplementary data associated with this article can be found, in the
online version, at .https://doi.org/10.1016/j.pedobi.2018.03.001
References
Anslan, S., Tedersoo, L., 2015. Performance of cytochrome c oxidase subunit I (COI),
ribosomal DNA large subunit (LSU) and internal transcribed spacer 2 (ITS2) in DNA
barcoding of Collembola. Eur. J. Soil Biol. 69, 1 7.–http://dx.doi.org/10.1016/j.
ejsobi.2015.04.001.
Auclerc, A., Ponge, J.F., Barot, S., Dubs, F., 2009. Experimental assessment of habitat
preference and dispersal ability of soil springtails. Soil Biol. Biochem. 41, 1596 1604.–
http://dx.doi.org/10.1016/j.soilbio.2009.04.017.
Avise, J.C., 1994. Molecular Markers, Natural History and Evolution. Springer Science &
Business Media, New York.
Batlle-Aguilar, J., Brovelli, A., Porporato, A., Barry, D.A., 2011. Modelling soil carbon and
nitrogen cycles during land use change. A review. Agron. Sustain. Dev. 31, 251 274.–
http://dx.doi.org/10.1051/agro/2010007.
Bickford, D., Lohman, D.J., Sodhi, N.S., Ng, P.K.L., Meier, R., Winker, K., Ingram, K.K.,
Das, I., 2007. Cryptic species as a window on diversity and conservation. Trends Ecol.
Evol. 22, 148 155. .–http://dx.doi.org/10.1016/j.tree.2006.11.004
Bidochka, M.J., Kamp, A.M., Lavendar, M.T., Dekoning, J., De Croos, J.N.A., 2001.
Habitat Association in two genetic groups of the insect-pathogenic fungus
Metarhizium anisopliae: uncovering cryptic species? Appl. Environ. Microbiol. 67,
1335 1342. .–http://dx.doi.org/10.1128/AEM.67.3.1335-1342.2001
Brower, A.V., 1994. Rapid morphological radiation and convergence among races of the
butter y inferred from patterns of mitochondrial DNA evolution.flHeliconius erato
Proc. Natl. Acad. Sci. 91, 6491 6495. .–http://dx.doi.org/10.1073/pnas.91.14.6491
Carapelli, A., Frati, F., Fanciulli, P.P., Dallai, R., 1995. Genetic di erentiation of sixff
sympatric species of (Collembola, Isotomidae) is there any di erence inIsotomurus –ff
their microhabitat preference? Eur. J. Soil Biol. 31, 87 99 –.
Cicconardi, F., Nardi, F., Emerson, B.C., Frati, F., Fanciulli, P.P., 2010. Deep phylogeo-
graphic divisions and long-term persistence of forest invertebrates (Hexapoda:
Collembola) in the North-Western Mediterranean basin. Mol. Ecol. 19, 386 400.–
http://dx.doi.org/10.1111/j.1365-294X.2009.04457.x.
Cicconardi, F., Fanciulli, P.P., Emerson, B.C., 2013. Collembola, the biological species
concept and the underestimation of global species richness. Mol. Ecol. 22,
5382 5396. .–http://dx.doi.org/10.1111/mec.12472
D Haese, C.A., 2002. Were the rst springtails semi-aquatic? A phylogenetic approach by’fi
means of 28S rDNA and optimization alignment. Proc. R. Soc. B 269, 1143 1151.–
http://dx.doi.org/10.1098/rspb.2002.1981.
Davidson-Watts, I., Walls, S., Jones, G., 2006. Di erential habitat selection byffPipistrellus
pipistrellus Pipistrellus pygmaeusand identi es distinct conservation needs for crypticfi
species of echolocating bats. Biol. Conserv. 133, 118 127.–http://dx.doi.org/10.
1016/j.biocon.2006.05.027.
Dufrêne, M., Legendre, P., 1997. Species assemblages and indicator species: the need for a
flexible asymmetrical approach. Ecol. Mono. 67, 345 366–.
Eisenring, M., Altermatt, F., Westram, A.M., Jokela, J., 2016. Habitat requirements and
ecological niche of two cryptic amphipod species at landscape and local scales.
Ecosphere 7, 1 13. .–http://dx.doi.org/10.1002/ecs2.1319
Emerson, B.C., Cicconardi, F., Fanciulli, P.P., Shaw, P.J.A., 2011. Phylogeny, phylogeo-
graphy, phylobetadiversity and the molecular analysis of biological communities.
Philos. Trans. R. Soc. B: Biol. Sci. 366, 2391 2402.–http://dx.doi.org/10.1098/rstb.
2011.0057.
Fjellberg, A., 2007. The collembola of fennoscandia and Denmark. Part II: etomo-
bryomorpha and symphypleona. Fauna Entomol. Scand. 42, 1 266–.
Folmer, O., Black, M., Hoeh, W., Lutz, R., Vrijenhoek, R., 1994. DNA primers for am-
pli cation of mitochondrial cytochrome c oxidase subunit I from diverse metazoanfi
Fig. 3. Bayesian phylogenetic tree of three lineages of and ve other species of based on concatenated data of four molecularLepidocyrtus lanuginosus fiLepidocyrtus
markers: 28S D1-2, elongation factor 1- , COI and COII. Bayesian posterior probabilities are indicated on nodes. Labels of OTUs represent sampling locations (refer toα
Table 1) and habitats (A, arable land; G, grassland; F, forest).
B. Zhang et al. Pedobiologia - Journal of Soil Ecology 68 (2018) 12–19
18

invertebrates. Mol. Mar. Biol. Biotechnol. 3, 294 299 –.
Frati, F., Dell Ampio, E., Casasanta, S., Carapelli, A., Fanciulli, P.P., 2000. Large amounts’
of genetic divergence among Italian species of the genus (Insecta,Orchesella
Collembola) and the relationships of two new species. Mol. Phylogenet. Evol. 17,
456 461. .–http://dx.doi.org/10.1006/mpev.2000.0854
Hall, T.A., 1999. ioEdit: a user-friendly biological sequence alignment editor and analysis
program for Windows 95/98/NT. Nucleic Acids Symp. Ser. 41, 95 98 citeulike-ar-–
ticle-id:691774.
Hebert, P.D.N., Cywinska, A., Ball, S.L., deWaard, J.R., 2003. Biological identi cationsfi
through DNA barcodes. Proc. Biol. Sci. 270, 313 321.–http://dx.doi.org/10.1098/
rspb.2002.2218.
Heiniger, C., Barot, S., Ponge, J.F., Salmon, S., Meriguet, J., Carmignac, D., Suillerot, M.,
Dubs, F., 2015. Collembolan preferences for soil and microclimate in forest and
pasture communities. Soil Biol. Biochem. 86, 181 192.–http://dx.doi.org/10.1016/j.
soilbio.2015.04.003.
Hogg, I.D., Hebert, P.D.N., 2004. Biological identi cation of springtails (Hexapoda:fi
Collembola) from the Canadian Arctic, using mitochondrial DNA barcodes. Can. J.
Zool. 82, 749 754. .–http://dx.doi.org/10.1139/z04-041
Hopkin, S.P., 1997. Biology of the Springtails (Insecta: Collembola). Oxford University
Press, Oxford, UK.
Hopkin, S.P., 2007. A Key to the Collembola (Springtails) of Britain and Ireland. Field
Studies Council.
James, S.W., Porco, D., Decaëns, T., Richard, B., Rougerie, R., Erséus, C., 2010. DNA
barcoding reveals cryptic diversity in Lumbricus terrestris L. 1758 (Clitellata): res-
urrection of L. herculeus (Savigny. PLoS One 5, e15629. http://dx.doi.org/10.1371/
journal.pone.0015629.
Katz, A.D., Giordano, R., Soto-Adames, F.N., 2015. Operational criteria for cryptic species
delimitation when evidence is limited, as exempli ed by North AmericanfiEntomobrya
(Collembola: Entomobryidae). Zool. J. Linn. Soc. 173, 818 840.–http://dx.doi.org/
10.1111/zoj.12220.
Kempson, D., Lloyd, M., Ghellardi, R., 1963. A new extractor for woodland litter.
Pedobiologia 3, 1 21–.
King, R.A., Tibble, A.L., Symondson, W.O.C., 2008. Opening a can of worms: un-
precedented sympatric cryptic diversity within British lumbricid earthworms. Mol.
Ecol. 17, 4684 4698. .–http://dx.doi.org/10.1111/j.1365-294X.2008.03931.x
Lanfear, R., Calcott, B., Ho, S.Y.W., Guindon, S., 2012. PartitionFinder: combined selec-
tion of partitioning schemes and substitution models for phylogenetic analyses. Mol.
Biol. Evol. 29, 1695 1701. .–http://dx.doi.org/10.1093/molbev/mss020
Mateos, E., 2008. The european bourlet, 1839 (Collembola: Entomobryidae).Lepidocyrtus
Zootaxa 59, 35 59–.
Mateos, E., 2012. The European group (Collembola:Lepidocyrtus lanuginosus
Entomobryidae), defi nition and description of a new species from Spain. Zootaxa 81,
69 81–.
Miller, M.A., Pfei er, W., Schwartz, T., 2010. Creating the CIPRES Science Gateway forff
inference of large phylogenetic trees. Institute of Electrical and Electronics Engineers
(Ed.), Proceedings of the Gateway Computing Environments Workshop (GCE) 1 8–.
Ponge, J.F., Dubs, F., Gillet, S., Sousa, J.P., Lavelle, P., 2006. Decreased biodiversity in
soil springtail communities: the importance of dispersal and landuse history in het-
erogeneous landscapes. Soil Biol. Biochem. 38, 1158 1161.–http://dx.doi.org/10.
1016/j.soilbio.2005.09.004.
Porco, D., Bedos, A., Greenslade, P., Janion, C., Skarzy ski, D., Stevens, M.I., Jansen van
Vuuren, B., Deharveng, L., 2012a. Challenging species delimitation in Collembola:
cryptic diversity among common springtails unveiled by DNA barcoding. Invertebr.
Syst. 26, 470 477. .–http://dx.doi.org/10.1071/IS12026
Porco, D., Potapov, M., Bedos, A., Busmachiu, G., Weiner, W.M., Hamra-Kroua, S.,
Deharveng, L., 2012b. Cryptic diversity in the ubiquist species Parisotoma notabilis
(Collembola, Isotomidae): a long-used chimeric species? PLoS One 7, e46056. http://
dx.doi.org/10.1371/journal.pone.0046056.
Porco, D., Skarzy ski, D., Decaëns, T., Hebert, P.D.N., Deharveng, L., 2014. Barcoding the
Collembola of Churchill: a molecular taxonomic reassessment of species diversity in a
sub-Arctic area. Mol. Ecol. Resour. 14, 249 261.–http://dx.doi.org/10.1111/1755-
0998.12172.
Posada, D., 2008. jModelTest: phylogenetic model averaging. Mol. Biol. Evol. 25,
1253 1256. .–http://dx.doi.org/10.1093/molbev/msn083
Querner, P., Bruckner, A., Drapela, T., Moser, D., Zaller, J.G., Frank, T., 2013. Landscape
and site e ects on Collembola diversity and abundance in winter oilseed rape elds inff fi
eastern Austria. Agric. Ecosyst. Environ. 164, 145 154.–http://dx.doi.org/10.1016/j.
agee.2012.09.016.
Ramirez-Gonzalez, R., Yu, D.W., Bruce, C., Heavens, D., Caccamo, M., Emerson, B.C.,
2013. PyroClean: denoising pyrosequences from protein-coding amplicons for the
recovery of interspeci c and intraspeci c genetic variation. PLoS One 8, e57615.fi fi
http://dx.doi.org/10.1371/journal.pone.0057615.
Ronquist, F., Teslenko, M., van der Mark, P., Ayres, D.L., Darling, A., Höhna, S., Larget, B.,
Liu, L., Suchard, M.A., Huelsenbeck, J.P., 2012. Mrbayes 3.2: E cient Bayesianffi
phylogenetic inference and model choice across a large model space. Syst. Biol. 61,
539 542. .–http://dx.doi.org/10.1093/sysbio/sys029
Rosenberger, M., Maraun, M., Scheu, S., Schaefer, I., 2013. Pre- and post-glacial di-
versi cations shape genetic complexity of soil-living microarthropod species.fi
Pedobiologia 56, 79 87. .–http://dx.doi.org/10.1016/j.pedobi.2012.11.003
Salmon, S., Ponge, J.-F., Gachet, S., Deharveng, L., Lefebvre, N., Delabrosse, F., 2014.
Linking species, traits and habitat characteristics of Collembola at European scale.
Soil Biol. Biochem. 75, 73 85. .–http://dx.doi.org/10.1016/j.soilbio.2014.04.002
Schä er, S., Pff fi ngstl, T., Koblmüller, S., Winkler, K.A., Sturmbauer, C., Krisper, G., 2010.
Phylogenetic analysis of European mites (Acari, Oribatida,Scutovertex
Scutoverticidae) reveals paraphyly and cryptic diversity: a molecular genetic and
morphological approach. Mol. Phylogenet. Evol. 55, 677 688.–http://dx.doi.org/10.
1016/j.ympev.2009.11.025.
Soto-Adames, F.N., 2002. Molecular phylogeny of the Puerto Rican andLepidocyrtus
Pseudosinella (Hexapoda Collembola), a validation of Yoshii s color pattern species.’
Mol. Phylogenet. Evol. 25, 27 42–.
Stevens, M.I., Frati, F., McGaughran, A., Spinsanti, G., Hogg, I.D., 2007. Phylogeographic
structure suggests multiple glacial refugia in northern Victoria Land for the endemic
Antarctic springtail (Collembola, Isotomidae). Zool. Scr. 36,Desoria klovstadi
201 212. .–http://dx.doi.org/10.1111/j.1463-6409.2006.00271. x
Tamura, K., Peterson, D., Peterson, N., Stecher, G., Nei, M., Kumar, S., 2011. MEGA5:
Molecular evolutionary genetics analysis using maximum likelihood, evolutionary
distance, and maximum parsimony methods research resource. Mol. Biol. Evol. 28,
2731 2739. .–http://dx.doi.org/10.1093/molbev/msr121
Tarjuelo, I., Posada, D., Crandall, K.A., Pascual, M., Turon, X., 2001. Cryptic species of
Clavelina (Ascidiacea) in two di erent habitats: harbours and rocky littoral zones inff
the northwestern Mediterranean. Mar. Biol. 139, 455 462.–http://dx.doi.org/10.
1007/s002270100587.
Thompson, J.D., Higgins, D.G., Gibson, T.J., 1994. CLUSTAL W: improving the sensitivity
of progressive multiple sequence alignment through sequence weighting, position-
speci c gap penalties and weight matrix choice. Nucleic Acids Res. 22, 4673 4680.fi–
http://dx.doi.org/10.1093/nar/22.22.4673.
Timmermans, M.J.T.N., Ellers, J., Mariën, J., Verhoef, S.C., Ferwerda, E.B., van Straalen,
N.M., 2005. Genetic structure in (Collembola): strong subdivision ofOrchesella cincta
European populations inferred from mtDNA and AFLP markers. Mol. Ecol. 14,
2017 2024. .–http://dx.doi.org/10.1111/j.1365-294X.2005.02548.x
van der Wur , A.W.G., Gols, R., Ernsting, G., van Straalen, N.M., 2005. Population ge-ff
netic structure of (Collembola; Hexapoda) in NW Europe, as revealedOrchesella cincta
by microsatellite markers. Pedobiologia 49, 167 174.–http://dx.doi.org/10.1016/j.
pedobi.2004.10.004.
von Saltzwedel, H., Scheu, S., Schaefer, I., 2016. Founder events and pre-glacial di-
vergences shape the genetic structure of European Collembola species. BMC Evol.
Biol. 23, 81 87. .–http://dx.doi.org/10.1007/s13398-014-0173-7.2
von Saltzwedel, H., Scheu, S., Schaefer, I., 2017. Genetic structure and distribution of
Parisotoma notabilis (Collembola) in Europe: cryptic diversity, split of lineages and
colonization patterns. PLoS One 12, e0170909. http://dx.doi.org/10.1371/journal.
pone.0170909.
Wang, I.J., Summers, K., 2010. Genetic structure is correlated with phenotypic divergence
rather than geographic isolation in the highly polymorphic strawberry poison-dart
frog. Mol. Ecol. 19, 447 458.–http://dx.doi.org/10.1111/j.1365-294X.2009.04465.x.
B. Zhang et al. Pedobiologia - Journal of Soil Ecology 68 (2018) 12–19
19