http://dx.doi.org/10.5852/eit.2013.66

www. europeani ournaloftaxonomy. eu

2013 • Martens K., Halse S. & Schon I.

This work is licensed under a Creative Commons Attribution 3.0 License.

Monograph

urn:lsid:zoobank.org:pub:5270B024-84Cl-4CD5-BC18-4AF37D2E1045

On the Bennelongia barangaroo lineage (Crustacea, Ostracoda) in

Western Australia, with the description of seven new species

Koen MARTENS 1 - 2 - 5 , Stuart HALSE 3 - 6 & Isa SCHON 1 - 4 - 7

1 Royal Belgian Institute of Natural Sciences, Freshwater Biology, Vautierstraat 29,

B-1000 Brussels, Belgium. Corresponding author: [email protected]

2 University of Ghent, Department of Biology, K.L. Ledeganckstraat 35, B-9000 Gent, Belgium.

3 Bennelongia Environmental Consultants, 5 Bishop Street, Jolimont WA 6014, Australia.

4 University of Hasselt, Research Group Zoology, Agoralaan Building D,

B-3590 Diepenbeek, Belgium

5 urn:lsid:zoobank.org:author:9272757B-A9E5-4C94-B28D-F5EFF32AADC7

6 urn:lsid:zoobank.org:author:676014BF-5979-49EC-BC8F-811214170111

7 urn:lsid:zoobank.org:author:94232F10-7092-4E90-9071-64C4FDABE691

Abstract. The ostracod genus Bennelongia De Deckker & McKenzie, 1981 is endemic to Australia

and New Zealand. Extensive sampling in Western Australia (WA) revealed a high specific and largely

undescribed diversity. Here, we describe seven new species belonging to the B. barangaroo lineage:

B. timmsi sp. nov., B. gnamma sp. nov., B. hirsuta sp. nov., B. ivanae sp. nov., B. mcraeae sp. nov.,

B. scanloni sp. nov. and B. calei sp. nov., and confirm the presence of an additional species, B. dedeckkeri ,

in WA. For five of these eight species, we could construct molecular phylogenies and parsimonious

networks based on COI sequences. We also tested for cryptic diversity and specific status of clusters

with a statistical method based on the evolutionary genetic species concept, namely Birky’s 4 theta

rule. The analyses support the existence of these five species and a further three cryptic species in the

WA B. barangaroo lineage. The molecular evidence was particularly relevant because most species

described herein have very similar morphologies and can be distinguished from each other only by the

shape, size and position of the antero-ventral lapel on the right valve, and, in sexual populations, by

the small differences in shape of the hemipenes and the prehensile palps in males. Four species of the

WA B. barangaroo lineage occur in small temporary rock pools (gnammas) on rocky outcrops. The

other four species are mainly found in soft bottomed seasonal water bodies. One of the latter species,

B. scanloni sp. nov., occurs in both claypans and deeper rock pools (pit gnammas). All species, except

for B. dedeckkeri , originally described from Queensland, have quite clearly delimited distributions in

WA. With the seven new species described here, the genus Bennelongia now comprises 25 nominal

species but several more await formal description.

Keywords. Taxonomy, evolution, cryptic species, biodiversity, Western Australia.

Martens K., Halse S. & Schon I. 2013, On the Bennelongia barangaroo lineage (Crustacea, Ostracoda) in Western

Australia, with the description of seven new species. European Journal of Taxonomy 66: 1-59. http://dx.doi.

org/10.5852/eit.2013,66

1

European Journal of Taxonomy 66 : 1-59 ( 2013 )

Introduction

The ostracod genus Bennelongia is endemic to Australia and New Zealand. Extensive sampling in

Western Australia (WA) revealed a high specific and largely undescribed diversity of Bennelongia (Halse

2002), leading to the taxonomic research reported here. The present paper is the fourth in a recent series

of taxonomic contributions on Australian Bennelongia. Martens et al. (2012) described nine new species

in three different lineages within the genus, all based on collections from WA. Shearn et al. (2012) re¬

described several extant species and described three new species, all from Eastern Australia (mostly from

Queensland). In addition, De Deckker & Martens (2013) described the unusually strong morphological

changes in valve morphology during the last 3-4 ontogenetic stages in several Bennelongia-species, and

showed that these changes can be different between the various lineages within the genus. The first two

papers, together with the earlier work by De Deckker (1981a,b, 1982) and De Deckker & McKenzie

(1981), brought the number of nominal species in the genus to 18, but both recent papers also recognised

that some additional cryptic species exist, as determined by molecular methods.

Here, we describe seven new species of Bennelongia from WA and report on the occurrence of an

additional described species in WA, namely B. dedeckkeri Shearn et al., 2012. All of these eight species

belong to the B. barangaroo lineage within the genus and for five of the species, their specific status

is confirmed with molecular methods. Two of the new nominal species furthermore comprise several,

sometimes sympatric, cryptic species that could not be recognised using either valve or soft part

morphologies.

Material and methods

Collections

Ostracods were collected from pans, lakes and rock pools with a hand net with mesh size of 250 pm

during several field trips (see below). Material for morphological analyses originated from both these

‘new’ collections and from earlier collections from all over WA, mostly collected by SH and preserved

in a collection housed at the Department of Environment and Conservation, now DPaW (Woodvale,

Perth). The molecular analyses were successful only with newly collected material, using either living

specimens or specimens sorted directly in the field and preserved in 100% ethanol. Consequently,

molecular analyses were limited to five of the eight species (four of the seven new ones). Locations

of populations used for the present paper are indicated on the map in Fig. 1. Type material of the new

species is deposited in the Western Australian Museum, Perth, WA (WAM numbers) and in the Ostracod

Collection of the Royal Belgian Institute of Natural Sciences, Brussels, Belgium (OC numbers) (see

Table 1).

Morphological analyses

Ostracods were dissected with valves stored dry in micropalaeontological slides and soft parts in

glycerine in sealed slides. Drawings of soft parts were made with a camera lucida with a compound

microscope (Leica, DM 2500 at Bennelongia Environmental Consultants, Perth). Valves were illustrated

and measured using scanning electron microscopy (Philips XL30 SEM at RBINS, Brussels).

Molecular analysis

The Qiagen Blood and Tissue extraction kit was used following the manufacturer’s protocol to extract

DNA from 99 ostracods representing four nominal and 2 cryptic species of the Bennelongia barangaroo

lineage. Universal PCR primers (Folmer et al. 1994) were applied to amplify part of the mitochondrial

COI region in a T personal Thermoblock (Biometra) with the following conditions: 25 pi volumes of the

HotStar Master Mix (Qiagen; 1.5 mM MgCfy 0.1 pM primer, 200pM dNTP, Tris Cl, KC1, (NHfySCfy

1.25 U Taq) and 15 min at 95°C, 40 cycles of 1 min at 95°C, 1 min at 44° C, 1 min at 72° C, followed

2

MARTENS K., HALSE S. & SCHON L, The Bennelongia barangaroo lineage in Western Australia

by a final extension step for 10 min at 72° C. Agarose gel electrophoresis and staining of gels with

Gelred™ was carried out to check for successful PCR amplifications. PCR products were cleaned with

the GFX™ PCR DNA and gel band purification kit (GE Healthcare) according to the manufacturer’s

protocol and sequenced in both directions with the universal COI primers and the Big Dye kit (ABI) on

an ABI 3130X following the manufacturer’s protocol.

Fig. 1 . Map of Western Australia with localities of Bennelongia - species described in the present paper.

3

European Journal of Taxonomy 66 : 1-59 ( 2013 )

No fresh (living) material of three of the seven new nominal species in the barangaroo lineage could be

obtained (namely B. gnamma sp. nov., B. hirsuta sp. nov. and B. mcraeae sp. nov.); these species are not

represented in the molecular phylogenetic tree and network.

Analyses of sequence data

Sequence chromatograms were visualised with BioEdit (Hall 2007). For each individual, the forward

and reverse strand were aligned with ClustalX (Larkin et al. 2007), subsequently checked by eye for

ambiguities, corrected and finally trimmed to obtain sequences of equal lengths. BLAST searches

(Altschul et al. 1990) were used to confirm the identity of the obtained sequences in Genbank. We

applied jModeltest 2.1.1 (Darriba et al. 2012) to identify the optimal model of molecular COI evolution

using 88 or 24 models and the AICc criterion. Two different methods were used for phylogenetic

reconstructions, Bayesian Inference (BI) in Mr Bayes 3.2 (Ronquist et al. 2011; with 4 million generations,

sampling every 100 th generation, a burn-in of 25% and the parameters identified by jModeltest for 24

different models) and the Ma xim um-Likelihood method in PhyML (Guindon & Gascuel 2003; with

1000 bootstrap replicates and the parameters of jModeltest for all 88 models). We also constructed a

parsimonious network at the 95% probability limit with TCS 1.21 (Clement et al. 2000) to connect

different sequences (or haplotypes) and to illustrate genetic diversities and genetic relationships within

and between populations. Selected sequences of all species have been submitted to Genbank (accession

numbers KF724982-KF725015; see Table 1).

Testing for cryptic diversity

We applied the 4 theta rule (Birky et al. 2010; Birky 2011) based on the evolutionary genetic species

concept (Birky & Barraclough 2009) to identify species boundaries and unravel cryptic diversities.

This technique has been successfully used in bdelloid rotifers (Fontaneo et al. 2007, 2009; Birky &

Barraclough 2009; Birky etal. 2011), asexual ostracods (Schon etal. 2012) including other Bennelongia

ostracod species (Martens et al. 2012; Shearn et al. 2012), and a wide range of asexual prokaryotes

(Birky et al. 2010).

We used the COI phylogenetic tree to identify statistically supported clades, which could potentially

be different species according to the evolutionary genetic species concept. We then estimated sequence

diversities within and between these phylogenetic clades with MEGA 5.0 (Tamura et al. 2011) using

the number of differences (p) and the Tamura-3 parameter model with gamma distribution (allowing

for multiple hits, different transition and transversion rates and GC bias) and 1000 bootstrap replicates.

Following the procedure by Birky et al. (2010), sequence diversities were subsequently corrected for

sample size and sequence lengths. In order to fulfill the criteria of the 4 theta rule, the sequence diversity

between two sister clades must be at least 4 to 4.3 times larger than within the two clades, depending on

the number of samples per clade (Birky et al. 2010).

Abbreviations used in text and figures

Cp

= carapace

CpD/V

= carapace in dorsal/ventral view

CpRL

= carapace in right lateral view

F

= female in Table 1

H

= height of valves

il

= inner list

KMWA

= original working numbers given to specimens dissected and illustrated by the first

author (KM)

K25

= electrical conductivity standardised to a water temperature of 25°C

L

= length of valves

Lpp

= left prehensile palp

4

MARTENS K., HALSE S. & SCHON I., The Bennelongia barangaroo lineage in Western Australia

Is

LV/LVe/LVi

Mext/Fext

ms

M

NT

OC

OS

Rpp

RV/RVe/RVi

SA

QLD

WA

WAMC

lateral shield of hemipenis

left valve/left valve exterior/left valve exterior

external views of valves of males/females

medial shield of hemipenis

male in Table 1

Northern Territory

Ostracod Collection in the Royal Belgian Institute of Natural Sciences (Brussels,

Belgium)

ostracod slide dissected by Stuart Halse, retrieved from the voucher collection of

DEC, now DPaW (Woodvale, Perth)

right prehensile palp

right valve/right valve exterior/right valve interior

South Australia

Queensland

Western Australia

Western Australian Museum, Crustacean Collection (Perth, WA)

Specimens in bold in Table 1 are the holotypes of the species.

Chaetotaxy of the limbs follows the model proposed by Broodbakker & Danielopol (1982), revised

for A2 by Martens (1987). The higher taxonomy of the Ostracoda follows the synopsis by Horne et al.

( 2002 ). '

Results

Results of molecular screening

We obtained 680 nucleotide-long sequences for part of the mitochondrial COI region for all 99 ostracods

subjected to DNA extraction. jModeltest selected the TPMluf+I+G model with the following parameters

among 88 models: freqA= 0.33; freqC = 0.19; freqG = 0.13; freqT = 0.35; [AC] = 1.00; [AG] = 24.84,

[AT] =2.91; [CG] =2.91; [CT] =24.84; [GT] = 1.00; p-inv = 0.56; gamma shape = 1.26. For Bayesian

Inference, the HKY+I+G model was selected among 24 models.

The phylogenetic tree (Fig. 2) had a similar topology with ML and BI methods for tree construction.

The two clades FI and F2 (belonging to B. ivanae sp. nov. and B. sp. nov. F2 respectively) group

together with high statistical support. They form the most basal branch and are separated from all other

Bennelongia specimens. Within the phylogenetic cluster containing the other species, B. dedeckkeri

forms the most basal branch, followed by clade B1 morphologically forming the new species B. calei

sp. nov. The remaining specimens can be divided into three subgroups with strong statistical support for

the tips but less support for the basal nodes of the subgroups themselves. Subgroup one consists of clade

B2 as well as E1 and TST, which form sister clades and all belong morphologically to the new species

B. scanloni sp. nov. The second subgroup is composed of clade E2, which is morphospecies B. sp. nov.

E2 and sister clades A3 and Al, while the third subgroup contains clades A4, A2 and A5. According to

their morphology, all A clades belong to the new species B. timmsi sp. nov.

We then used the topology of the phylogenetic tree (Fig. 2) to test whether clades phylogenetically

closest to each other are in fact different genetic species by applying the 4 theta rule (see Table 2). Most

sister clades represent different genetic species according to the criteria of the 4 theta rule (Birky et al.

2010), with the majority of these genetic species matching the morphologically identified species. This

is the case for sister clades FI and F2 representing B. ivanae sp. nov. and B. sp. nov. F2, respectively,

and for B1 corresponding to B. calei sp. nov.. Likewise, morphology and genetics are concordant for the

described species B. dedeckkeri and clade E2 (B. spec nov.). However, within B. timmsi sp. nov. and

5

European Journal of Taxonomy 66 : 1-59 ( 2013 )

B. scanloni sp. nov., there are phylogenetic clades representing different genetic species that cannot be

distinguished morphologically. Of the five clades within B. timmsi sp.nov, only 3 are genetically distinct

species (clades A1 and A3 on the one hand and A2+A4+A5 on the other). Within B. scanloni sp. nov.,

the cryptic clades El/TST and B2 are good genetic species (see below for further remarks on species B.

timmsi sp. nov. and B. scanloni sp. nov.).

If the topology of the tree in Fig. 2 is correct, then B. timmsi sp. nov. may be a polyphyletic species, as

clades A1 and A3 on the one hand, and clades A2+A4+A5 on the other belong to different clusters of

the tree. However, the support of the bifurcation between the two groups of B. timmsi clades is doubtful

(bootstrap value of 77), so the apparent polyphyly of the species may be an artifact. As also no clear

morphological characters, distinctive of the clusters, could be found (see below), we decided not to

describe these two groups as distinct nominal species. If at some stage in the future this decision is to be

reverted, then B. timmsi sp. nov. is defined by clade A5.

B. cf nimala sp. nov.

93/0.97

100 / 1.0

69/0.80

100 / 1.0

84/1.0

F2

<F1

B. sp. nov. F2

B. ivanae

B. dedeckkeri

64 / 1 . 0 .^ 02

75/0.82

100 / 1.0

95/1

◄

BtEI

TST

100 / 1.0

E2

100 / 1.0

77/0.95'

i A3

100 / 1 .

^A1

99/1.0

96/1.C

A4

A2

4A5

100/0.89

B1

90/1.0

B. calei

B. scanloni

B. sp. nov. E2

B. timmsi

Fig. 2. Phylogenetic tree constructed with Bayesian Inference (BI) and Ma xim um Likelihood (ML)

methods from COI sequences of 99 ostracods from the Bennelongia barangaroo lineage and with B.

nimala as outgroup. Numbers above nodes illustrate statistical support for this particular node. Numbers

before the hash (/) are % bootstrap values of ML analyses with 1000 replicates, numbers after the hash

(/) are Bayesian posterior probabilities (ranging from 0 to 1). Both methods, BI and ML, resulted in the

same tree topology. Different phylogenetic clades are indicated by different colours (as in Fig. 3, page

15). The asterisks indicate two nodes (al and a2) that are wealdy supported (see Discussion, p. 55).

6

MARTENS K., HALSE S. & SCHON I., The Bennelongia barangaroo lineage in Western Australia

o

>

o

>

o

>

o

>

o

>

§

C

&

n

2

o

n

o

n

oj

n

O

UJ

n

UJ

O

n

UJ

o

n

o

n

UJ

n

u>

u,

Ol

o.

UJ

o.

UJ

Ol

UM

Ul

UJ

u,

UM

Oi

u>

UJ

Oi

z

1 —‘

to

t — 1

to

t — 1

IO

1 —t

IO

to

IO

to

t—*

H—k

to

Ol

to

40

to

vo

IO

to

IO

oo

to

IO

to

o

to

"!

UJ

IO

40

4J.

4^

-|u.

4U.

4^

40

UJ

Oi

40

UJ

4D

vo

oo

-o

ON

LTx

4U

UJ

to

o

vo

\D

VO

nO

VO

vo

■Jo

4^

UJ

VO

vo

VO

vo

40

UJ

\D

VO

V D

oo

OO

oo

UJ

o

vo

oo

UJ

ON

to

IO

O

VO

oo

-o

ON

o

ON

un

Ul

UJ

oo

Ol

4^

UJ

to

40

ON

40

>

E

£

g 1 ?

to

'O

-o

o

& o ix

to

IO

to

3^cr

o.

Ol

Ul

un

cr S

o

-o

on

Ol

4^

UJ

to

^ B to

to

!>

o*

r*

!>

r*

JS s

g 1

g'

g -

g‘

g -

s'

g'

g‘

g'

s'

S'

s'

S'

s'

S'

s'

S'

2

s

2

Co

0_

05_

Oj

tg_

Co_

Oj

Co_

Cj_

a §

s,

S’

to

<

o*

H

to

H

n

—

i—‘

i—*

H-*

h—t

H-k

h —k

1 —*

1 —‘

1 —*

1 —‘

H-k

l—o

i—‘

i—*

o'

o

p

O

p

??

o

c5

o

O

o

c5

o

o-

o.

01

Ol

un

■Jo

4-

4J.

4^

UJ

to

2

Mext

to

Mext

2

Fext

to

M/F

un

40

un

UJ

Ol

UJ

Ul

UJ

un

40

t—^

40

to

o

^1

O

oo

40

oo

IO

4^

ON

-o

o

H—^

oo

IO

VO

O

40

o

O

o

O

o

O

3

oo

OO

>0

oo

vo

oo

vo

oo

o

vo

oo

o

OO

to

VO

oo

1 —*

IO

ON

UJ

H-^

o

t—‘

UJ

4^

o

40

to

UJ

40

oo

UJ

o

ON

to

oo

ON

UJ

to

ur>

UJ

vo

^1

o

UJ

to

! .

, .

) .

, .

! .

, .

un

Ol

ON

40

ON

4^

UJ

4^

ON

4U

UJ

un

Ol

4o

to

40

Ol

hH

Ol

UJ

o

UJ

ON

Ol

4^

oo

H—‘

UJ

4^

oo

UJ

vo

4o

ON

OO

IO

o

O

o

O

o

O

o

O

o

to

<

VO

OO

vo

OO

oo

vo

oo

vo

VO

oo

o

oo

VD

to

IO

ON

Ol

oo

-o

4^

IO

UJ

oo

4^

IO

4^

UJ

oo

Ul

oo

o

un

VO

un

40

to

40

VO

UJ

to

o

to

UJ

40

'O

1530

to

n

to

50

oc

to

nO

, .

un

o

un

oo

o

to

O

VO

vo

UJ

Ol

7

Table 1 . Individual measurements of specimens used for the present descriptions. All measurements were done using SEM (see Material and methods).

If a molecular sequence was available for the same specimen, the GenBank registration number is also given. However, some specimens were used

as whole animals for DNA sequencing, and thus no measurements are available. The present table therefore does not list all 99 specimens for which

sequences are available. Specimens in bold are holotypes.

European Journal of Taxonomy 66 : 1-59 ( 2013 )

1

o

>

o

>

o

>

o

>

o

§

c

Wl

n

2

n

s

o

n

s

2

n

tO

IO

n

o

n

IO

n

to

n

to

n

to

IO

to

'J\

to

y

to

IO

to

IO

lo>

!—k

1—k

IO

H-*

to

IO

t—^

to

H-‘

to

lo>

to

to)

IO

to

o

to

IO

lo>

to

IO

to

Ot

to

o

to

oo

^1

Ot

to

40

to

o

oc

to

o

Ot

to

oo

to

o

to

oo

to

oo

'O

^1

to

o

Ot

to

40

to

o

to

oo

o

to

o

Ot

to

40

to

>

71

g 1 C

to

to'

to

~o

'O

-o

o

-o

5 O '4

to

S * n 1

cr 5

2 o ts

to

o

to

o

to

oo

~ P !C

to

rN

'to-

Cto-

'to-

"to-

'to-

'to

5g S

s'

« a

n re,

s

§

2

s

2

§

2

s

2

s

2

o

C<3

03

X*)

03

to

03

r> «

a

«*4.

**•4.

*■4.

*-4.

’-4.

*44.

*•4.

*44.

to.

**•4.

^4.

S’

o

to

Co

to

C/3

<

ST

H

n

W

\

as

i— 1

I— 1

l— 1

I— 1

i— 1

I— 1

i— 1

I— 1

f— 1

p

o

p

to

o

o'

o

O

o

o'

o

o'

o

o'

o

>

oo

Ot

to

Oi

to

2

to

2

s

to

2

to

Fext

to

Fext

to

£

S

2

M/F

-t-

to

40

to

4^

to

40

to

Oi

to

40

oo

to

IO

H-k

O

o

to

t—k

oo

IO

o

oo

o

-o

to

40

oo

to

o

to

o

O

o

to

o

oc

o

to

§

c

\D

oo

to

oo

to

o

to

-o

to

H-‘

H-k

40

■_

o

to

oo

to

o

Ot

to

oc

to

Ot

IO

to

40

to

o

oo

to

oo

to

o

oo

, ,

! ,

, ,

1 ,

, ,

! .

, .

! ,

, ,

, .

1 ,

to

40

to

Ot

40

to

o

40

H-k

o

to

Ot

to

-o

o

-o

o

oo

Ot

to

oo

to

40

to

o

O

o

-o

to

oc

to

oo

to

oo

to

oo

oc

to

Oi

o

-|o

40

IO

40

oo

o

40

t—k

!— k

to

40

to

o

to

o

to

oo

to

oo

to

o

to

o

to

, ,

to

oo

to

n

73

oo

o

to

oo

o

to

o

to

o

, .

, ,

, .

, ■

, .

to

Ot

to

oo

H-1k

to

oo

oc

o

n

■a

to

o

■—1

o

to

oo

Ot

oo

■£-

oo

40

oo

to

■-J

oo

o

40

8

MARTENS K., HALSE S. & SCHON L, The Bennelongia barangaroo lineage in Western Australia

OC.3328

WAMC52287

WAMC52286

WAMC52285

WAMC52284

WAMC52282

WAMC52281

OC.3327

OC.3326

WAMC52280

WAMC52279

OC.3324

WAMC52273

OC.3323

WAMC52269

WAMC52274

WAMC 52270

WAMC 52272

WAMC 52277

WAMC 52276

WAMC 52275

WAMC52263

WAMC52266

OC.3322

Mus Nr

1076

180

OS007

904

444

1003

1002

1001

852

851

903

1102

1103

1106

1105

1104

1101

227

226

225

224

OS178

223

222

Genbank

KMWA accession

number

KF725015

KF725014

mcraeae

spec F2

ivanae

hirsuta

gnamma

Bennelongia

. locality

species

OSTR014B

BYT/10/09

OSTR013F

DJC/02

BYT/10/09

OSTR012D

OSTR012A

2

Fext

Mext

s

Tl

ti

Tl

ti

Fext

Tl

Fext

Tl

2

£

Tl

2

Tl

Fext

M/F RV

1390

1608

1465

1480

1400

1456

1460

1452

1475

1217

1240

1260

1390

1400

1550

1482

r

847

1002

oc

-4

h-k

878

842

904

910

904

915

719

743

738

813

842

915

960

X

1450

1663

1523

1560

1490

1513

1542

1512

1533

1302

1300

1320

1450

1470

1620

1497

r

LV

859

1033

885

896

865

935

947

940

948

719

759

763

824

845

991

1018

X

1488

1410

1300

1450

1430

r

CpRL

931

803

737

837

826

X

1452

1498

1447

1500

1430

1560

r

n

o

-

1025

855

875

896

1000

9

European Journal of Taxonomy 66: 1-59 (2013)

WAMC52298

WAMC52297

o

c/3

r —K

WAMC 52304

WAMC 52310

OC.3334

WAMC 52323

WAMC 52322

OC.3338

WAMC 5232 8

WAMC52327

OC.3340

OC.3339

WAMC 52326

WAMC 52325

WAMC52325

WAMC 52324

WAMC 522 89

Mus Nr

838

837

832

831

856

855

919

918

917

916

1107

438

437

194

299

297

295

1078

Genbank

KMWA accession

number

KF724984

KF724983

KF724982

scanloni

mcraeae

Bennelonsia

. locality

species

DJC/11

DJC/09

TST

OSTR013C

BRYDE7

OSTR014B

dd

Fext

dd

Fext

dd

Fext

dd

Fext

dd

2

Mext

2

Fext

dd

Fext

dd

Fext

dd

Fext

dd

M/F RV

1352

1367

1360

1380

1421

1465

1446

1438

1294

1394

1248

1230

1242

1260

1303

1310

1322

1330

1438

1460

1448

1460

r

789

803

773

797

898

923

890

900

752

810

756

688

738

751

765

775

803

807

860

863

873

X

1438

1458

1433

1463

1477

1525

1490

1485

1396

1487

1331

1320

1303

1310

1350

1380

1412

1044

1533

1550

1560

r

LV

808

807

796

807

910

933

898

930

833

769

709

749

761

782

788

822

839

883

887

992

895

X

1489

1650

r

CpRL

878

1000

X

1448

1497

1352

1455

r

n

O

832

910

744

858

10

MARTENS K., HALSE S. & SCHON L, The Bennelongia barangaroo lineage in Western Australia

WAMC52306

OC.3332

WAMC52321

WAMC52320

WAMC52319

OC.3337

OC.3336

WAMC52315

WAMC52314

WAMC52313

WAMC52312

WAMC52311

OC.3335

WAMC52303

WAMC52302

WAMC52301

WAMC52300

OC.3331

WAMC52292

WAMC52294

WAMC52293

WAMC52291

OC.3329

WAMC52296

WAMC52295

WAMC52299

Mus Nr

908

907

1021

1020

1018

797

1031

1030

1029

1026

1025

1023

1022

1013

1012

1011

1010

1009

o

oc

1007

1006

1005

1004

842

841

839

Genbank

KMWA accession

number

KF724991

KF724990

KF724989

KF724988

KF724987

KF724986

KF724985

scanloni

Bennelongia

. locality

species

BYT/10/10

DJC/23

DJC/19

DJC/11

dd

dn

dd

Fext

dd

2

s

2

dd

M/F RV

1587

1460

1498

1560

1573

-

1446

1485

1383

1471

1504

1312

1263

1190

1244

1223

1204

r

989

929

919

979

973

-

885

883

827

883

900

111

752

696

723

694

714

X

1646

1546

-

1633

1670

1660

1519

oc

1463

1562

1583

1398

1356

1263

1327

1294

1277

r

LV

998

952

-

987

1015

broken

906

846

889

919

792

775

708

735

714

737

X

1502

1625

1515

1290

1493

r

CpRL

898

983

887

730

867

X

1608

1721

1713

1475

1222

1302

r

O

■a

O

948

1035

1038

-

745

731

3

11

European Journal of Taxonomy 66: 1-59 (2013)

WAMC52344

WAMC52343

OC.3347

OC.3346

WAMC52349

WAMC52354

WAMC52353

WAMC52356

WAMC52355

WAMC52338

WAMC52337

OC.3345

WAMC52341

WAMC52340

WAMC52339

OC.3344

WAMC52335

WAMC52336

OC.3333

WAMC52308

WAMC52307

Mus Nr

810

809

808

807

870

196

195

238

237

236

830

829

828

1017

1016

1015

1014

823

822

\D

910

909

Genbank

KMWA accession

number

KF724997

KF724996

KF724995

KF724994

KF724993

KF724992

calei

scanloni

Bennelongia

. locality

species

BYT/11/04

BVT/11/04

DJC/10

SPM017B

DJC/18

BVT/10/10

dd

Fext

dd

Fext

dd

►n

Fext

dd

Fext

dd

Fext

dd

£

dd

M/F RV

1307

1400

1422

1167

1170

1145

1150

1244

1335

1490

1446

1452

1480

broken

1340

LLZl

1602

1415

r

792

837

833

692

699

682

686

729

775

864

835

844

857

792

802

-

994

885

1382

1508

1522

1252

1260

1232

1240

1337

1408

1588

1531

1552

1555

1419

1432

1350

1656

1465

r

LV

797

848

855

713

720

703

748

864

873

848

862

865

814

818

831

1015

898

X

1397

1330

1290

1477

r

CpRL

783

756

735

815

X

1425

1427

1325

1180

1290

1518

1477

r

n

O

ILL

782

718

695

729

802

827

3

12

MARTENS K., HALSE S. & SCHON L, The Bennelongia barangaroo lineage in Western Australia

WAMC52365

WAMC52364

OC.3352

OC.3351

WAMC52357

WAMC52360

WAMC52359

OC.3350

WAMC52352

WAMC52351

WAMC52350

OC.3349

OC.3348

WAMC52331

WAMC52330

OC.3343

WAMC52329

WAMC52348

WAMC52347

WAMC52346

WAMC52345

Mus Nr

681

680

679

678

193

192

'O

190

882

881

880

879

874

1083

1082

1081

o

oo

o

816

815

814

813

Genbank

KMWA accession

number

KF725000

KF724999

KF724998

dedeckkeri

calei

Bennelongia

. locality

species

SIKE9

KIES 10

DJC/36

DJC/15

BVT/11/05

dd

^d

►n

dd

Fext

dd

M/F RV

1125

1110

1345

-

1371

1383

r

672

676

817

-

812

X

1188

1190

1433

1587

1449

1460

r

LV

715

701

833

906

832

X

1210

1110

1410

1497

1540

1443

r

CpRL

717

650

780

850

889

808

X

1145

1187

o

oo

o

1100

1430

1513

1506

1546

1417

1473

r

n

O

660

652

651

641

830

843

842

773

828

3

13

European Journal of Taxonomy 66: 1-59 (2013)

Table 2. Results of tests for genetic species boundaries using the 4 theta method for six species of the

Bennelongia barangaroo lineage.

Phylogenetic sister

Species

max. 0

D (between

Ratio D/0

n 1 , n 2

clades

(within clades)

clades)

A1-A3

B. timmsi

0.0076

0.114

15.00

8,7

0.0078

0.139

17.82

A2-A4

B. timmsi

0.0138

0.052

3.77

9,3

0.0141

0.058

4.11

A2-A5

B. timmsi

0.0138

0.048

3.48

8, 8

0.0141

0.052

3.69

A4-A5

B. timmsi

0.0138

0.049

3.55

3, 5

0.0141

0.054

3.83

Al-El

B. timmsi-

0.0133

0.131

9.85

8, 13

B. scanloni

0.0138

0.169

12.25

A3-El

B. timmsi-

0.0133

0.140

10.53

7, 13

B. scanloni

0.0138

0.182

13.19

B1 -B. dedeckkeri

B. calei -

0.004

0.126

31.5

23,

B. dedeckkeri

0.004

0.161

40.25

25

B2-E1

B. scanloni

0.020

0.113

5.65

15,

0.021

0.141

6.71

13

B2-DJC11

B. scanloni

0.050

0.040

0.80

15,2

0.053

0.044

0.83

El-TST

B. scanloni

0.0133

0.032

2.41

13,2

0.0138

0.034

2.46

B2-E2

B. scanloni -

0.020

0.103

5.15

15,2

B. sp. nov. E2

0.020

0.127

6.35

B. scanloni -

0.0015

0.109

72.67

E2-TST

B. sp. nov. E2

0.0015

0.135

90.00

2,2

F1-F2

B. ivanae -

0.0058

0.061

10.52

3,4

B. sp. nov. F2

0.0059

0.067

11.36

0 = population genetic parameter theta, indicating genetic variability within populations. D = genetic

distance between sister clades. n 1 , n 2 = number of sequences for each sister clade. 0 and D were either

calculated as p distance or with the Tamura-3 parameter model (in italics). Nearest neighbours or sister

clades were defined from the COI tree constructed with Bayesian Inference and Maximum Likelihood

methods (see Fig. 2). In order to fulfill the criteria of the 4 theta rule for species status, the ratio of

the mean sequence diversity within as compared to between the two nearest neighbours of one sister

clade needs to be 4 or more, depending on the number of specimens per clade (Birky et al. 2010).

Comparisons, for which these criteria are fulfilled, are printed in bold.

14

Fig. 3. Parsimonius network, based on COI sequences of the Bennelongia barangaroo lineage. Squares represent ancestral sequences (or haplotypes),

small circles missing haplotypes. The size of squares and large ovals is proportional to the number of individuals with the same sequence in the analysed

population. The network was constructed at the 95% probability limit, which includes up to 8 mutation steps for connecting different sequences or

haplotypes. Different phylogenetic clades are indicated by different colours, which match those used in Fig. 2 (page 6).

MARTENS K., HATSE S. & SCHON I., The Bennelongia barangaroo lineage in Western Australia

15

European Journal of Taxonomy 66: 1-59 (2013)

The structure of the most parsimonious networks in Fig. 3 also reflects the higher genetic than

morphological diversity. However, there are two species with a single haplotype each, B. dedeckkeri

and B. spec. E2 sp. nov. (clade E2). For the latter, we obtained COI sequences of two individuals only

while the same (identical) haplotype is found in 11 specimens of B. dedeckkeri. The three individuals

in clade F2 (and B. spec. F2 sp. nov.) share two haplotypes, while the five specimens of B. ivanae sp.

nov. possess 4 haplotypes, of which one remains unconnected. The remaining three species display

two ( B. calei sp. nov.) to five (B. scanloni sp. nov.) and six (B. timmsi sp. nov.) unconnected haplotype

networks, respectively. The different genetic species within B. timmsi sp. nov. and B. scanloni sp. nov.

form unconnected haplotype networks.

Taxonomic descriptions

Class Ostracoda Latreille, 1806

Subclass Podocopa G.O. Sars, 1866

Order Podocopida G.O. Sars, 1866

Suborder Cypridocopina Baird, 1845

Superfamily Cypridoidea Baird, 1845

Family Cyprididae Baird, 1845

Subfamily Bennelongiinae Martens et al. , 2012

Genus Bennelongia De Deckker & McKenzie, 1981

Diagnosis

See Martens et al. (2012)

Bennelongia barangaroo lineage

Remarks

De Deckker (1981a) described B. barangaroo from Lake Buchanan (QLD - Type locality), but also

reported the same species from other localities in QLD, NSW, SA, WA and New Zealand. However, as

in De Deckker’s (1981a) re-description of B. australis (Brady, 1886) (see Martens et al. 2012), at least

two different species within this lineage were illustrated under the same name. The (type) specimens

of B. barangaroo in De Deckker’s (1981a) figure 7 from Lake Buchanan have a short, sub-quadrate,

slightly ventrally pointed lapel on the antero-ventral side of the RV. The specimens from a pool 25 km

N of Cue (WA) (De Deckker 1981a: fig. 9), however, appear to have an elongated lapel, much as is

the case in B. calei sp. nov. (see below). We thus decided previously (Shearn et al. 2012) that the true

B. barangaroo needs to be established based on new material from the type locality. Fortunately, De

Deckker (1981a) illustrated the valves and soft parts of the holotype male, which facilitated identification,

and allowed Shearn et al. (2012) to confidently describe B. dedeckkeri as a different species within the

B. barangaroo lineage. Shearn et al. (2012) also described B. mckenziei as a second new species from

QLD, characterised by a total absence of the lapel on the RV.

Diagnosis of the B. barangaroo lineage

All species of the B. barangaroo lineage (re-)described here share a number of features: all have

relatively elongated and wide (in dorsal view) carapaces, mostly green in colour, relatively smooth

(but hirsute) in adults. The RV has an internal eyelet at the posteroventral internal side, mostly situated

directly internally of the lapel. This eyelet is best visible with transparent light; although in most species

it is also visible on SEM micrographs (see various illustrations below).

16

MARTENS K., HALSE S. & SCHON I., The Bennelongia barangaroo lineage in Western Australia

Bennelongia timmsi sp. nov.

Figs 4-11

um:lsid:zoobank.org:act:4F6A6E8F-5636-4290-85A4-B234D5DA4466

Diagnosis

Valves in inner view (Fig. 4A-B, D-E) relatively high, with greatest height situated well in front of the

middle; ventral margin anteriorly with well-pronounced mandibular curve. FV (Fig. 4A, D) with anterior

il not overlapping. RV (Fig. 4B, E) with antero-ventral lapel subtriangular, asymmetrically produced

with a ventral point (Fig. 4K-M). Carapace in dorsal and ventral views (Fig. 4G-J) with greatest width in

the middle, hirsute, anteriorly with a mild rostrum; in lateral views (Fig. 4C, F) showing a clear anterior

FV>RV overlap.

Hemipenes (holotype: Fig. 8F) mostly symmetrical, with Is protruding well beyond ventral tip of ms,

Is with broad base, ventrally bluntly beak-shaped. Right prehensile palp (holotype: Fig. 8D) with distal

segment elongated, with dorsal margin evenly rounded. Feft prehensile palp (holotype: Fig. 8E) with

distal segment elongated, reaching beyond ventro-apical margin of proximal segment with at least half

of its length.

Etymology

This species is named after Prof. Brian V. Timms (Newcastle, Australia), in recognition of his vast

contribution to the knowledge of Australian non-marine crustaceans in general, and of phyllopods from

temporary pools in particular. Prof. Timms also collected the material of the present species from a

series of pools on various rocky outcrops in WA.

Measurements (all measurements in pm - see Table 1 for measurements of all specimens illustrated

with SEM)

Holotype 3 (WAMC52228): RV: F = 1318, H = 783. FV: F = 1378, H = 817.

Allotype $ (WAMC52229): RV: F = 1510, H = 913. FV: F = 1600, H = 965.

Type locality

Rock pools on Wave Rock, WA, ca. 2 km E of Hyden. Approximate coordinates: 32° 27’S 118° 54’ E

(WGS 84). Material handpicked from pools by B.V. Timms on 23 Jul. 2010 (sample BVT/10/05).

Type material

Holotype

3 (WAMC52228), with soft parts dissected in a sealed slide and valves stored dry in a micropalaeonto-

logical slide.

Allotype

5 (WAMC52229), with soft parts dissected in a sealed slide, and valves stored dry in a micropalaeonto-

logical slide.

Paratypes

Numerous males and females from the type locality, either dissected and stored as the holotype, or as

carapaces used for SEM (WAMC52230-52237, OC.3312-3316). Ca. 60 33 and $$ in EtOH as bulk

paratypes (WAMC52238).

Other material investigated

All material from WA, collected by B.V. Timms.

17

European Journal of Taxonomy 66: 1-59 (2013)

Fig. 4. Bennelongia timmsi sp. nov., all represent paratypes from Wave Rock (BVT/10/05 - type

locality). A. $, LVi (OC.3313). B. $, RVi (idem). C. $, CpRL (WAMC52235). D. $ holotype, LVi

(WAMC52228). E. 3 holotype, RVi (idem). F. 3, CpRL (WAMC52236). G. ?, CpD (WAMC52234).

H. $, CpV (OC.3315). I 3 , CpV (OC.3316). J. 3 , CpD (WAMC52237). K. 3 holotype, RVi, detail

anterior (WAMC52228). L. 3 holotype, RVi, detail anterior, tilted (WAMC52228). M. §, LVi, detail

anterior, tilted (OC.3313). Scales: A-J = 1 mm; K-M = 200 pm.

18

MARTENS K., HALSE S. & SCHON I., The Bennelongia barangaroo lineage in Western Australia

Fig. 5. Bennelongia timmsi sp. nov., all males with hemipenes and prehensiles palps illustrated (see

Figs 8-10). A-L = paratypes from Wave Rock (BVT/10/05), M-0 = non-types from King Rocks

(BVT/10/06). —A-C. WAMC52232: A. RVi. B. RVi, detail anterior, tilted. C. RVi, detail anterior. —

D-F. WAMC52231: D. LVi. E. RVi, detail anterior, tilted. F. RVi. — G-I. OC.3312: G. LVi. H. RVi,

detail anterior, tilted. I. RVi. — J-L. WAMC52230: J. LVi. K. RVi, detail anterior, tilted. L. RVi. —

M-O. WAMC52252: M. LVi. N. RVi, detail anterior, tilted. O. RVi. Scales: A, D, F-G, I-J, L-M, O = 1

mm; B-C, E, H, K, N = 200 pm.

19

European Journal of Taxonomy 66: 1-59 (2013)

Fig. 6. Bennelongia timmsi sp. nov., all represent females with molecular data available. — A-C.

Grahams Rock (BVT/10/02 - WAMC52243 - cryptic species Al): A. LVi. B. RVi, detail anterior, tilted.

C. RVi. — D-F. Grahams Rock (BVT/10/02 - WAMC52241 - cryptic species A2): D. LVi. E. RVi, detail

anterior, tilted. F. RVi. — G-I. King Rocks (BVT/10/06 - WAMC52251 - cryptic species A2): G. LVi.

H. RVi, detail anterior, tilted. I. RVi. — J-L. Wave Rock (BVT/10/05 - allotype WAMC52229 - cryptic

species A3): J. LVi. K. RVi, detail anterior, tilted. L. RVi. — M-O. Mt Madden Rocks (BVT/10/08 -

WAMC52254 - cryptic species A5): M. LVi. N. RVi, detail anterior, tilted. O. RVi. Scales: A, C-D, F-G,

I-J, L-M, 0=1 mm; B, E, H, K, N = 200 pm.

20

MARTENS K., HALSE S. & SCHON I., The Bennelongia barangaroo lineage in Western Australia

Fig. 7. Bennelongia timmsi sp. nov. — A-C. Mt Madden Rocks (BVT/10/08 - 3, WAMC52255 -

cryptic species A5). A. LVi. B. RVi, detail anterior, tilted. C. RVi. — D-F. Anderson Rock (BVT/10/03 -

9, WAMC52245). D. LVi. E. RVi, detail anterior, tilted. F. RVi. — G-L Paynes Find Rock (BVT/10/01

- $, WAMC52239). G. LVi. H. RVi, detail anterior, tilted. I. RVi. — J-L. Grahams Rock (BVT/10/02

- 9, OC.3317). J. LVi. K. RVi, detail anterior, tilted. L. RVi. — M. Burracopin Rock (BVT/10/04 - 9,

OC.3318), RVi, detail anterior, tilted. — N. King Rocks (BVT/10/06 - 9, WAMC52250), RVi, detail

anterior, tilted. Scales = 1 mm unless otherwise indicated.

21

European Journal of Taxonomy 66: 1-59 (2013)

Fig. 8. Bennelongia timmsi sp. nov., male type specimens from type locality (Wave Rock, BVT/10/05).

— A-C. Paratype M (OC.3312): A. Hemipenis (both hemipenes symmetrical in this specimen). B. Right

prehensile palp. C. Left prehensile palp. —D-F. Holotype $ (WAMC52228): D. Right prehensile palp. E.

Left prehensile palp. F. Hemipenis (both hemipenes symmetrical in this specimen). Scale: A-F = 92 pm.

22

MARTENS K., HALSE S. & SCHON L, The Bennelongia barangaroo lineage in Western Australia

Fig. 9. Bennelongia timmsi sp. nov., paratypes from type locality (Wave Rock, BVT/10/05). — A-C.

Paratype S (WAMC52231): A. Hemipenis (both hemipenes symmetrical in this specimen). B. Right

prehensile palp. C. Left prehensile palp. — D-F. Paratype S (WAMC52230): D. Left prehensile palp. E.

Right prehensile palp. F. Hemipenis (both hemipenes symmetrical in this specimen). Scale: A-F= 92 pm.

23

European Journal of Taxonomy 66: 1-59 (2013)

Fig. 10. Bennelongia timmsi sp. nov., male paratype (WAMC52232) from type locality (Wave Rock,

BVT/10/05) Aberrant specimen. A. Hemipenis. B. Hemipenis. C. Left prehensile palp. C\ Idem, detail

of ventroapical part of first segment, showing two lobes and a sensory organ. C”. Idem, detail of distal

part of second segment, showing aberrant, bilobed morphology. D. Right prehensile palp. Scale: A-D =

92 pm; C’, C” = 37 pm.

24

MARTENS K., HALSE S. & SCHON L, The Bennelongia barangaroo lineage in Western Australia

Fig. 11. Bennelongia timmsi sp. nov., non-type males. — A-D. Grahams Rock (WAMC52240,

BVT/10/02): A. Hemipenis. B. Hemipenis. C. Right prehensile palp. D. Left prehensile palp.

D\ Idem, detail of ventro-apical part of first segment. — E-F. King Rocks (WAMC52252, BVT/10/06):

E. Hemipenis. F. Left prehensile palp. G. Right prehensile palp. Scale: A-G = 92 pm; D’ = 37 pm.

25

European Journal of Taxonomy 66: 1-59 (2013)

Paynes Find Rocks. Approximate coordinates: 29°10’ S, 117°40’ E (sample BVT/10/01), collected by

B.V. Ti mm s on 23 Jul. 2010 (one $ - WAMC52239).

Grahams Rock. 32°28’ S, 119°03’ E (sample BVT/10/02), collected by B.V. Timms on 23 Jul. 2010 (six

33 and ??, WAMC52240-52244; OC.3317).

Anderson Rock. 32°10’ S, 118°5 V E (sample BVT/10/03), collected by B.V. Timms on 23 Jul. 2010 (one

3 , WAMC52245).

Burracopin Rock. 31°24’ S, 118°27’ E (sample BVT/10/04), collected by B.V. Timms on 26 Jul. 2010

(six 33 and WAMC52246-52249; OC.3318-3319).

King Rocks. 32°19’ S, 119°09’ E (sample BVT/10/06), collected by B.V. Timms on 23 Jul. 2010 (one 3

and two ??, WAMC52250-52252).

Yorkrakine Rocks. 31°25’ S, 117°30’ E (sample BVT/10/07), collected by B.V. Timms on 27 Jul. 2010.

Mt Madden Rock. 33°14’ 22” S, 119°50’ 33” E (sample BVT/10/08), collected by B.V. Timms on 01

Aug. 2010 (11 33 and ??, WAMC52253-52262; OC.332-3321).

Differential diagnosis

Bennelongia timmsi sp. nov. can be distinguished from most species of the B. barangaroo lineage by

the triangular and ventrally pointed lapel on the RV and the strongly sinuous ventral valve margins. The

lapel of B. scanloni sp. nov. is also subtriangular and ventrally protruding, but it is rounded, thus looking

drop-shaped in internal (non-tilted) view.

Bennelongia timmsi sp. nov. can moreover be distinguished from B. gnamma sp. nov. by the less high

and less rounded valves and by the Is of the hemipenes, which protrudes well beyond the ms (subequal

in B. gnamma sp. nov.).

Additional notes on cryptic species

As was described above, five genetic clusters are recognised in this species (Al-5, Fig. 2). According to

the calculations of the 4 theta rule, three cryptic species were found in B. timmsi sp. nov. with molecular

methods, but no morphological diagnostic features could be found. Cryptic species A1 occurred in

BVT/10/02, 03 and 07. Cryptic species A3 was found in BVT/10/02, 04, 06 and 08. Cryptic species

A2+A4+A5 occurred in BVT/10/03 and 05 and is used here to characterize B. timmsi sp. nov. with

BVT/10/05 (Wave Rock) as type locality. Note that BVT/10/02 (Grahams Rock) and BVT/10/03

(Anderson Rock) hold at least two sympatric clades/cryptic species each. In order to establish beyond

reasonable doubt that the specimens belonging to these clusters and cryptic species are indeed

morphologically indistinguishable, long series of specimens are illustrated.

Sample BVT/10/05 from pools on Wave Rock appeared to contain only one genetic cluster and cryptic

species and, for this reason, Wave Rock was chosen as type locality. We then proceeded with two

different approaches: (1) to dissect a series of males from this sample to test whether male reproductive

organs (hemipenes, prehensile palps) showed uniformity within one cluster/cryptic species; (2) we

checked for potential differences in the morphology of the valves of specimens belonging to different

populations and/or shown to belong to different clusters/cryptic species.

Type specimens

Valves and carapaces of males and females of the type population (in sample BVT/10/05) were illustrated

(Fig. 4) and this morphology defines the species. We then dissected several males from the same sample

and population and illustrated the soft part and valve morphology. Shape of valves and size and shape of

the antero-ventral lapel on the RV were most similar and indeed almost indistinguishable (Fig. 5). In all

male specimens the valves have the shape described in the diagnosis above. The lapels are all elongated

subtriangular, with a more or less serrated distal margin. In tilted perspective some lapels appear to be

26

MARTENS K., HALSE S. & SCHON I., The Bennelongia barangaroo lineage in Western Australia

shorter than others (e.g., the lapel in Fig. 5B appears shorter than in 5E), but this is almost entirely a

matter of distorted perspective depending on how the valves were positioned when the photographs were

taken (the same lapels appear almost equally long in non-tilted views - Fig. 5A and 5F, respectively).

However, there are significant differences in soft part morphology. Whereas the shapes of the hemipenis-

outline and of the left prehensile palps are fairly uniform in the different specimens (Figs 8A, C, E-F;

9A, C-D, F), the second segment of the right prehensile palps ranges from elongated sub-triangular with

almost equally rounded distal margin (Fig. 8B, D), to sub-rectangular with a clear blunt comer in this

margin (Fig. 9B, E). It is not clear to what extend these differences are a biological reality, or whether

the differences are distortions of the limbs caused by different positions in the slides. The differences are

sufficiently small to be accepted as part of intra-specific variability, yet future investigations should take

this variability into account. The morphology in the holotype (WAMC52228 - right prehensile palp in

Fig. 8D) determines the specific morphology.

One male (WAMC52232 - Fig. 10A-D) had an aberrant morphology, with the terminal segment of the

right prehensile palp (Fig. 10D) being even more elongated and with especially the terminal segment of

the left prehensile palp being distally bilobed (Fig. IOC, C”), a morphology never before encountered

in Cyprididae. Nevertheless the valves of this male show no differences with other type specimens (Fig.

5A-C).

Morphology within different clades/cryptic species

Valve morphology of specimens for which molecular clades are known (Al: Fig. 6A-C; A2: Fig. 6D-

I; A3: Fig. 6J-F; A5: Fig. 6M-0, 7A-C) and for specimens from different populations for which no

molecular data were available (Fig. 7D-N) again show no constant differences that could be used as

identifying characters. There is some variability in size, shape and degree of crenulation of the lapel, but

insufficiently so to use such features to characterise different clades/cryptic species.

One male specimen from BVT/10/02 and thus belonging to either cryptic species Al or A3, had a right

prehensile palp with a terminal segment clearly showing a blunt angle on the distal margin (Fig. 11C),

while a male from B VT/10/06, and thus most likely belonging to clade A2, had a more elongated segment

there with a more rounded distal margin (Fig. 11G) as in the holotype. In both of these specimens, the

terminal segment of the left prehensile palp is slightly shorter than in the type specimens (Fig. 11D, F).

Hemipenis outlines (Fig. 11A-B, E) are indistinguishable from those in the types.

Ecology and distribution

Bennelongia timmsi sp. nov. is a typical rock pool species and occurs in fresh water in gnammas on

various rocky outcrops in the south/central part of western Australia. Although it appears to be limited

to this restricted area, it seems to be quite common there.

Bennelongia gnamma sp. nov.

Figs 12, 13A-E

urn:lsid:zoobank.org:act:6931D617-1443-4776-891B-C02752E6C0BE

Diagnosis

Valves in internal view (Fig. 12A-B, E-F) high, with greatest height situated in front of the middle;

ventral margin a lm ost evenly curved except for middle third. FV (Fig. 12A, E) with anterior il slightly

overlapping. RV (Fig. 12B, F) with antero-ventral lapel subtriangular, asymmetrically produced with a

dorsal point (Fig. 12H-K).

27

European Journal of Taxonomy 66 : 1-59 ( 2013 )

Fig. 12. Bennelongia gnamma sp. nov., type material from type locality (Cairn Rock, OSTR012A).

A. $ paratype, LVi (OC.3322). B. $ paratype, RVi (idem). C. $ paratype, LVe (idem). D. § paratype,

RVe (idem). E. § holotype, LVi (WAMC52263). F. § holotype, RVi (idem). G. $ paratype, CpD

(WAMC52266). H. $ paratype, RVi, detail anterior (OC.3322). I. $ paratype, RVi, detail anterior, tilted

(idem ). J. $ holotype, RVi, detail anterior (WAMC52263). K. $ holotype, RVi, detail anterior, tilted

(idem). Scales: A-G = 1 mm; H-K = 200 pm.

28

MARTENS K., HALSE S. & SCHON I., The Bennelongia barangaroo lineage in Western Australia

Fig. 13. Bennelongia gnarnma sp. nov. (A-E, type specimens from type locality, Cairn Rock, OSTR012A)

and Bennelongia hirsuta sp. nov. (F-I type specimens from type locality, Styles Rock, OSTR012D).

Males. — A-E. B. gnarnma sp. nov.: A. Hemipenis (both hemipenes symmetrical in this specimen,

WAMC52264). B. Apical part of hemipenis (WAMC52265). C. Right prehensile palp (WAMC52264).

D. Left prehensile palp (WAMC52264). E. Idem, detail of ventro-apical part of first segment (different

specimen, WAMC52265). —F-I. B. hirsuta sp. nov. (WAMC52271): F. Hemipenis. G. Right prehensile

palp. FI. Left prehensile palp. H’. Idem, detail of ventro-apical part of first segment. I. Hemipenis. Scale:

A-D, F-I = 92 pm; E, H’ = 37 pm.

29

European Journal of Taxonomy 66 : 1-59 ( 2013 )

Valves in external lateral view (Fig. 12C-D) high and rounded on all sides, even ventrally to some

extent; hirsute and wealdy pitted. Carapace in dorsal view (Fig. 12G) anteriorly with a mild rostrum.

Hemipenes (Fig. 13 A) mostly symmetrical, with length of Is subequal to that of ms, in one specimen tip

of ms even extending beyond that of Is (Fig. 13B); Is with broad base, ventrally bluntly beak-shaped.

Right prehensile palp (Fig. 13C) with distal segment broad, with anterior margin straight, distal margin

bilobed. Left prehensile palp (Fig. 13D-E) with distal segment short, reaching beyond ventro-apical

margin of proximal segment with less than half its length.

Etymology

The present species is named after the Australian term for small to middle-sized rock pools, namely

gnammas, in which it occurs.

Measurements (all measurements in pm - see Table 1 for measurements of all specimens illustrated

with SEM)

Holotype ? (WAMC52264): RV: L = 1550, H = 915. LV: L = 1620, H = 991.

Type locality

Rock pools (gnammas) on Cairn Rock, WA, ca. 67 km SE of Merredin. Approximate coordinates:

31°51’3r S, 118°50’39” E (WGS 84). All material (voucher sample OSTR012A; locality code SPS059)

collected by J. McRae & A. Pinder on 24 Sep. 1997 with a sweep net.

Type material

Holotype

5 (WAMC52263), with soft parts dissected in a sealed slide and valves stored dry in a micropalaeonto-

logical slide.

Allotype

S (WAMC52264), with soft parts dissected in a sealed slide, and valves stored dry in a micropalaeon-

tological slide.

Paratypes

2 $ $ (WAMC52266; OC.3322) and one $ (WAMC52265); 3 ? ? in bulk in EtOH (WAMC52267).

Other material investigated

? 1 $ (WAMC52268) from Yanneymooning Rocks (identification uncertain).

Differential diagnosis

Bennelonga gnamma sp. nov. can be distinguished from all other species in the lineage by the high and

rounded shape of the valves and especially by the subequal Is and ms in the hemipenes.

Ecology and distribution

This species is thus far known with certainty only from its type locality, a set of rock pools on Cairn

Rock.

30

MARTENS K., HALSE S. & SCHON I., The Bennelongia barangaroo lineage in Western Australia

Bennelongia hirsuta sp. nov.

Figs 13F-I, 14

urn:lsid:zoobank.org:act:CDE0DlE7-52E6-4F18-B6FF-B3AD48C6E4B4

Diagnosis

Valves elongated, with greatest height situated well in front of the middle, dorsal margin evenly sloping

towards the posterior side; ventral margin anteriorly with pronounced mandibular curve. FV (Fig. 14A,

D) with antero-ventral inner list large, well-overlapping the dorsal il. RV (Fig. 14B, E) with antero-

ventral lapel long, narrow and weakly crenulated (Fig. 14J-M).

Carapace in dorsal and ventral view (Fig. 14G-I) with greatest width in the middle, most hirsute,

anteriorly with a clear rostrum; in lateral views (Fig. 14C, F) anteriorly with a clear FV>RV overlap.

Hemipenes asymmetrical (Fig. 13F, I), Is with slender base, ventrally sharply beak-shaped and pointed,

only slightly protruding beyond ventral tip of ms. Right prehensile palp (Fig. 13G) with distal segment

stout and subquadrate, anterior margin straight, dorsal margin sinuous. Feft prehensile palp (Fig. 13H,

FT) with distal segment short and sickle-shaped, reaching beyond ventro-apical margin of proximal

segment with less than a third of its length.

Etymology

Named after the hirsute (‘hairy’) nature of this species. All species of the B. barangaroo lineage are

hirsute to some extent, but the present species is more so, with the entire carapace set with long and stiff

setae.

Measurements (all measurements in pm - see Table 1 for measurements of all specimens illustrated

with SEM)

Holotype 3 (WAMC52269): RV: F = 1260, H = 738. FV: F = 1320, H = 763.

Allotype $ (WAMC52270): RV: F = 1400, H= 842. FV: F = 1470, H = 845.

Type locality

Rock pools (gnammas) on Styles Rock, WA, ca. 80 km N of Esperance. Approximate coordinates:

33°07’35” S, 121°48’02” E (WGS 84). All specimens (voucher sample OSTR012D; locality code

SPS139), collected on 07 Sep. 1998 by J. McRae & A. Pinder with a sweep net.

Type material

Holotype

3 (WAMC52269), with soft parts dissected in a sealed slide and valves stored dry in a micropalaeon-

tological slide.

Allotype

5 (WAMC52270), with soft parts dissected in a sealed slide, and valves stored dry in a micropalaeon-

tological slide.

Paratypes

Numerous S3 and 55, as valves or carapaces (WAMC52271-52277; OC.3323-3325). Nine SS and

5 5 in bulk in EtOH (WAMC52278).

Other material investigated

? One 5 (WAMC52279) from Filian Stokes Rocks (BVT/10/09) (identification uncertain).

31

European Journal of Taxonomy 66 : 1-59 ( 2013 )

Fig. 14. Bennelongia hirsuta sp. nov., type specimens from type locality (Styles Rock, OSTR012D). A.

? allotype, LVi (WAMC52270). B. $ allotype, RVi (idem). C. ? paratype, CpRL (WAMC52277). D. $

holotype, LVi (WAMC52269). E. $ holotype, RVi (idem). F. $ paratype, CpRL (WAMC52272). G. ?

paratype, CpD (WAMC52275). H. $ paratype, CpV (WAMC52276). I. $ paratype, CpD (WAMC52272).

J. $ allotype, RVi, detail anterior (WAMC52270). K. § allotype, RVi, detail anterior, tilted (idem). L. $

holotype, RVi, detail anterior (WAMC52269). M. $ holotype, RVi, detail anterior, tilted (idem). Scales:

A-I = 1 mm; J-M = 300 pm.

32

MARTENS K., HALSE S. & SCHON I., The Bennelongia barangaroo lineage in Western Australia

Differential diagnosis

Bennelongia hirsuta sp. nov. can be distinguished from all other species in the B. barangaroo lineage

by the pointed shape of the Is in the hemipenis. Also the shapes of the distal segments of the prehensile

palps are distinctive. In valve morphology, the species is easily recognisable by the long and narrow

lapel on the RV (which is nevertheless very difficult to see with a normal binocular microscope) and the

large antero-ventral il.

Ecology and distribution

The species is thus far only known with certainty from its type locality, a set of rock pools on Styles

Rock.

Bennelongia ivanae sp. nov.

Fig. 15

urn:lsid:zoobank.org:act:2D60FC82-0938-475A-AF4E-E2DBE24DlD8D

Diagnosis

Valves (Fig. 15A-B) high and rounded, with greatest height situated well in front of the middle; dorsal

margin with blunt angle towards the posterior side; ventral margin almost straight, without pronounced

mandibular curve. FV (Fig. 15A) with antero-ventral il well-developed in lower third of the valve, dorsal

il descending almost to ventral side, thus clearly overlapping with ventral il. RV (Fig. 15B) without

antero-ventral lapel (Fig. 15H-J). Carapace in dorsal and ventral views (Fig. 15D-F) with greatest width

in the middle, medium hirsute and slightly pitted, anteriorly with a hint of a rostrum; in lateral view (Fig.

15C, G) anteriorly with a clear FV>RV overlap.

Male unknown.

Etymology

This species is named after Dr Ivana Karanovic (South Korea), in recognition of her contributions to the

knowledge of the subterranean candonids of the Pilbara area.

Measurements (all measurements in pm - see Table 1 for measurements of all specimens illustrated

with SEM)

Holotype $ (WAMC52280): RV: F = 1475, H = 915. FV: F = 1533, H = 948.

Type locality

Rock pools (gnammas) on Holland Rocks, WA. Approximate coordinates: 33°2r35.66” S,

118°44’48.55” E (WGS 84) (sample DJC/02). All specimens collected and handpicked by D.J. Cale on

30 Aug. 2011.

Type material

Holotype

5 valves stored dry (WAMC52280).

Allotype

As males are unknown, no allotype is designated.

Paratypes

Four 5? with valves or carapaces stored dry (WAMC52281-52282; OC.3326-3327); ca. 25 $$ stored

dry as bulk in one micropalaeontological slide.

33

European Journal of Taxonomy 66 : 1-59 ( 2013 )

Other material investigated

? 5$ valves stored dry (WAMC52284) from YanneymooningRocks (OSTR013F).

Differential diagnosis

Bennelongia ivanae sp. nov. differs from all other WA species within the B. barangaroo lineage in the

total absence of an antero-ventral lapel on the RV and in the fact that the antero-dorsal il in the LV runs

almost entirely to the ventral margin. Bennelongia mckenziei Shearn et al., 2012 from Queensland also

lacks the antero-ventral lapel on the RV completely, but it has a notably different valve and carapace

Fig. 15. Bennelongia ivanae sp. nov., type specimens from type locality (Holland Rocks, DJC/02).

A. $ holotype, LVi (WAMC52280). B. $ holotype, RVi (idem). C. $ paratype, CpRL (OC.3327).

D. j paratype CpD (WAMC52281). E. $ paratype, CpV (WAMC52282). F. § paratype, CpV, detail

anterior (idem). G. $ paratype, CpRL, detail anterior (OC.3327). H. $ holotype, RVi, detail anterior

(WAMC52280). I. $ holotype, RVi, detail anterior, tilted (idem). J. $ paratype, RVi, detail anterior,

tilted (OC.3326). Scales: A-E = 1 mm; F = 400 pm; G-J = 200 pm.

34

MARTENS K., HALSE S. & SCHON I., The Bennelongia barangaroo lineage in Western Australia

shape, with a more pointed caudal margin and an evenly sloping dorsal margin in both valves, a shorter

antero-dorsal il in the LV and an anterior LV>RV overlap in a carapace in right lateral view which is

twice as large as in B. ivanae sp. nov.

Remark

Bennelongia ivanae sp. nov. and B. mckenziei have pronounced molecular differences, when the present

sequences of B. ivanae sp. nov. are compared with those of Shearn et al. (2012). Because of the shorter

lengths of the sequences provided by Shearn et al. (2012), the alignment of COI sequences from B.

ivanae sp. nov. and B. mckenziei are not shown in the present paper.

Ecology and distribution

The species has thus far been found with certainty only from rockpools at Holland Rocks. Two

tentatively identified females also originated from rock pools on another outcrop.

Bennelongia sp. nov. F2

Material investigated

Two $$ (KMWA.905, 906) in toto used for molecular screening, one $ with soft parts used for

molecular screening and with broken RV stored dry in micropalaeontological cavity slide (WAMC52285

= KMWA.904).

Locality

Lilian Stokes Rocks (eastern Wheatbelt-BVT/10/09), coordinates: 33 o 4’06” S, 120°05 , 49”E. Collected

on 25 Aug. 2010 by B.V. T imm s

Remarks

The specimens of the F2-group cluster close to those of B. ivanae sp. nov. in the phylogenetic tree (Fig.

2), but still constitute a separate genetic species (Table 2). A broken RV could be saved from only one

of these specimens (KMWA.904), and it could be ascertained that the antero-ventral lapel on the RV is

also fully absent (not shown). It is possible that cluster F2 will turn out to be a cryptic species within B.

ivanae sp. nov., but this remains to be tested with new material. Note that the same sample (BVT/10/09)

also contained one putative female of B. hirsuta sp. nov. (see above).

Bennelongia mcraeae sp. nov.

Figs 16-17

urn:lsid:zoobank.org:act:DFDlA720-E602-46CC-AABF-C6EDD63CD70E

Diagnosis

Valves (Fig. 16A-B, E-F) high and rounded, with greatest height situated on or close to the middle,

dorsal margins almost evenly rounded; ventral margin weakly sinuous. LV (Fig. 16A, E) with antero-

ventral il large, reaching over half of the anterior margin; dorsal il descending along ca. 4/5 of anterior

margin, both lists thus clearly overlapping. RV (Fig. 16B, F) with antero-ventral lapel fairly ventrally

inserted, large and pronounced, with rounded, weakly crenulated margin (Fig. 16D, I-L). Carapace in

lateral view (Fig. 16C- D) pitted, especially along anterior and posterior margins, and set with few setae.

Hemipenes (Fig. 17A, F) largely symmetrical. Is with broad base, ventrally bluntly beak-shaped, only

protruding significantly beyond ventral tip of ms. Right prehensile palp (Fig. 17C, E) with distal segment

stout, but of somewhat variable shape, either strongly subquadrate or with rounded dorsal and anterior

35

European Journal of Taxonomy 66 : 1-59 ( 2013 )

Fig. 16. Bennelongia mcraeae sp. nov., type specimens from type locality (Arro Lake, OSTR014). A. $

allotype, LVi (WAMC52287). B. $ allotype, RVi (idem). C. ? paratype, CpRL (WAMC52289). D. $

paratype, CpRL, detail anterior (idem). E. <5 holotype, LVi (WAMC52286). F. 3 holotype, RVi (Idem).

G. 3 holotype, LVe (idem). H. 3 holotype, RVe (idem). I. 3 holotype, RVe, detail anterior (idem). J.

3 holotype, RVi, detail anterior (idem). K. 3 holotype, RVi, detail anterior, tilted (idem). L. § allotype,

RVi, detail anterior, tilted (WAMC52287). Scales = 1 mm unless otherwise indicated.

36

MARTENS K., HALSE S. & SCHON L, The Bennelongia barangaroo lineage in Western Australia

Fig. 17. Bennelongia mcraeae sp. nov., type males from type locality (Arro Lake, OSTR014B). —A-C.

Holotype $ (WAMC52286). A. Hemipenis. B. Left prehensile palp. C. Right prehensile palp. — D-F.

Paratype $ (OC.3328). D. Left prehensile palp. E. Right prehensile palp. F. Hemipenis. Scale: A-F =

92 pm.

37

European Journal of Taxonomy 66: 1-59 (2013)

margins. Left prehensile palp (Fig. 17B, D) with distal segment long and slender, reaching beyond

ventro-apical margin of proximal segment with at least half of its length.

Etymology

The species is named in honour of Jane McRae (Perth, WA) in acknowledgement of her vast knowledge

of the taxonomy and morphology of many invertebrate groups of Western Australia. She also collected

the type material of the present species and has unrelentingly provided technical help towards the present

revision of Bennelongia since 2006.

Measurements (all measurements in pm - see Table 1 for measurements of all specimens illustrated

with SEM)

Holotype S (WAMC52286): RV: L = 1480, H = 878. LV: L = 1560, H = 896.

Allotype $ (WAMC52287): RV: L = 1608, H = 1002. LV: L = 1663, H = 1033.

Type locality

Arro Lake, ca. 11 km NW of Eneabba, WA. Approximate coordinates: 29°44’ 11” S, 115°09 , 58” E (WGS

84). All specimens collected by J. McRae & A. Pinder (voucher OST14B; locality code SPS182) on

23 Sep. 1999 with a sweep net. Arro Lake is an open lake with a Melaleuca/Casuarina fringe. Water

chemistry at the time of collecting: Salinity = 0.15 g/1, pH = 7.32. Nutrient levels were fairly high: total

N = 1700 pg/1; total P = 220 pg/1. The milky-white colour of the water equates to a high level of turbidity

(2200 NTU).

Type material

Holotype

S (WAMC52286), with soft parts dissected in a sealed slide and valves stored dry in a micropalaeon-

tological slide.

Allotype

5 (WAMC52287), with soft parts dissected in a sealed slide, and valves stored dry in a micro-

palaeontological slide.

Paratypes

Two SS (dissection: OC.3328; valves: WAMC52288) and one $ carapace (WAMC52289). Three

females in bulk in EtOH (WAMC52290).

Differential diagnosis

The species is characterised especially by the large and stout lapel on the RV but also by the large antero-

ventral il on the LV.

Ecology and distribution

Lake Arro (ca. 300 km N of Perth) is a large flat-bottomed body of water with episodic inflow that holds

water for about 4-24 months after inflow. The lake has a clay base and sediment-driven turbidity. This

species is known only from the type locality. From the same sample as the one that yielded B. macraeae

sp. nov., Timms (2002) described a new species of Anostraca, Branchinella complexidigitata Timms,

2002 .

38

MARTENS K., HALSE S. & SCHON I., The Bennelongia barangaroo lineage in Western Australia

Bennelongia scanloni sp. nov.

Figs 18-20

urn:lsid:zoobank.org:act:B41BF127-BEEC-47CE-A687-FACD6CBCF028

Diagnosis

Valves (Fig. 18A-B, D-E) high, with greatest height situated close to the middle, dorsal margin evenly

sloping towards the posterior side; ventral margin sinuous. FV (Fig. 18A, D) with antero-ventral il of

medium size, covering lower third of valve, antero-dorsal il descending to about halfway along antero-

ventral il. RV (Fig. 18B, E) with antero-ventral lapel tear-shaped in untilted lateral view; in tilted view,

lapel subtriangular with rounded ventral point (Fig. 18K-N). Carapace in dorsal and ventral views (Fig.

18G-J) with greatest width in the middle, hirsute and pitted, anteriorly with a clear rostrum; in right

lateral view (Fig. 18C, F) with large anterior FV>RV overlap, anterior margins of RV and FV not parallel.

Hemipenes (holotype: Fig. 20F) asymmetrical. Is with broad base, ventrally bluntly pointed (more so

in one hemipenis than in the other), largely protruding beyond ventral tip of ms. Right prehensile palp

(holotype: Fig. 20D) with distal segment stout and subquadrate, with sharp angle between anterior

and dorsal margins, both of these margins almost straight. Feft prehensile palp (holotype: Fig. 20E)

with distal segment sickle-shaped and of intermediate length, reaching beyond ventro-apical margin of

proximal segment with about half of its length.

Etymology

The species is named in honour of Mike Scanlon (Perth, WA) in acknowledgement of his unrelenting

technical help since 2006 towards the present revision of Bennelongia.

Measurements (all measurements in pm - see Table 1 for measurements of all specimens illustrated

with SEM)

Holotype $ (WAMC52291): RV: F = 1223, H = 694. FV: F = 1294, H = 714.

Allotype ? (WAMC52292): RV: F = 1263, H =752. FV: F = 1356, H = 775.

Type locality

One Tree Hill Creek, ca. 62 km SE of Dongara, WA. Approximate coordinates: 29°35 , 19.0” S,

115°26’31.0” E (WGS 84). All specimens (sample DJC/11; locality code SPS180) collected by D.J.

Cale on 10 Sep. 2011 with a sweep net. Water chemistry at time of collecting: K25 5.62 mS/cm, pH 6.68,

water temperature 20.8 °C.

Type material

Holotype

S (WAMC52291), with soft parts dissected in a sealed slide and valves stored dry in a micropalaeon-

tological slide.

Allotype

9 (WAMC52292) valves stored dry in a micropalaeontological slide.

Paratypes

Numerous males and females either as dissection, or as valves or carapaces stored dry (WAMC52293-

52304; OC.3329-3331). Ca. 30 SS and ? ? stored as bulk in EtOH (WAMC52305). ’

39

European Journal of Taxonomy 66: 1-59 (2013)

Fig. 18. Bennelongia scanloni sp. nov., type specimens from type locality (One Tree Hill Creek, DJC/11).

A. $ allotype, LVi (WAMC52292). B. $ allotype, RVi (idem). C. $ paratype, CpRL (WAMC52299).

D. S paratype, LVi (OC.3329). E. S paratype, RVi (idem). F. S paratype, CpRL (WAMC52295). G. $

paratype, CpD (WAMC52297). H. $ paratype, CpV (WAMC52298). I. $ paratype, CpV (WAMC52295).

J. S paratype, CpD (WAMC52296). K. 5 allotype, RVi, detail anterior (WAMC52292). L. $ allotype,

RVi, detail anterior, tilted (idem). M. S paratype, RVi, detail anterior (OC.3329). N. $ paratype, RVi,

detail anterior, tilted (idem). Scales: A-J = 1 mm; K-N = 200 pm.

40

MARTENS K., HALSE S. & SCHON I., The Bennelongia barangaroo lineage in Western Australia

Fig. 19. Bennelongia scanloni sp. nov., non-type specimens. — A-D. Pool at Latham-Coorow Rd

(BVT/10/10, ?, WAMC52307)). A. LVi. B. RVi. C. RVi, detail anterior. D. RVi, detail anterior, tilted.

— E-H. Tin Dog Creek (DJC/19, ?, OC.3335). E. LVi. F. RVi. G. RVi, detail anterior. H. RVi, detail

anterior, tilted. — I-O. Pool at Brookton Hwy (Warrine Park) (DJC/23, §, OC.3337). I. LVi. J. RVi. K.

RVi, detail anterior. L. RVi, detail anterior, tilted. M. LVe, detail anterior. N. RVe. O. RVe, detail anterior.

Scales: A-B, E-F, I-J, N = 1 mm; C-D, G-H, K-M, O = 200 pm.

41

European Journal of Taxonomy 66: 1-59 (2013)

Fig. 20. Bennelongia scanloni sp. nov., males. — A-C\ Non-type (OC.3338, Three Springs Tumulus

Stream - TST): A. Hemipenis (both hemipenes symmetrical in this specimen). B. Right prehensile

palp. C. Teft prehensile palp. C\ Idem, detail of ventro-apical part of first segment. — D-F. Holotype

(WAMC52291, One Tree Hill Creek - DJC/11): D. Right prehensile palp. E. Teft prehensile palp. F.

Hemipenis (both hemipenes symmetrical in this specimen). — G-J. Non-type (WAMC52327, from

OSTR013C): G. Hemipenis. H. Right prehensile palp. I. Teft prehensile palp. I’. Idem, detail of ventro-

apical part of first segment. J. Hemipenis. Scale: A-J = 92 pm; C’, T = 37 pm.

42

MARTENS K., HALSE S. & SCHON I., The Bennelongia barangaroo lineage in Western Australia

Other material investigated

One Tree Hill Creek. 29°35 , 19.0” S, 115°26’31.0” E, collected by S. Halse and A. Pinder on 11 Aug.

1999 (OSTR013C), see Fig. 20G-J (WAMC52327). Water chemistry at time of collecting: K25 3.12 mS/

cm, pH 7.65, water temperature 19.4 °C.

Pools at Latham-Coorow Rd. 29°5ES, 116°16’ E (sample BVT/10/10), collected by B.V. Timms on 10

Sep. 2010 (WAMC52306-52308; OC.3332-3333).

Dam at Solomons Well. Approximate coordinates: 31° 11 ’58.8” S, 116°21 ’47.7” E (sample DJC/04),

collected by D.J.Cale on 09 Sep. 2011 (4 $ $: WAMC52332-52334; OC.3343). Water chemistry at time

of collecting: K25 0.12 mS/cm, pH 7.85, water temperature 14.4 °C.

Three Pools along Eneabba-Carnamah Rd. Approximate coordinates: 34°18 , 32.80” S, 115°39’16.44”

E (sample DJC/09), collected by D.J. Cale on 10 Sep. 2011 (2 $$: WAM52310; OC.3334). Water

chemistry at time of collecting: K25 8.7 mS/cm, pH 7.2, water temperature 20.3 °C.

Second pool along Carnamah-Eneabba Road on south side (Eneabba Springs). Approximate

coordinates: 29°48 , 23.62” S, 115°25 , 6.11” E (sample DJC/10), collected by D.J. Cale on 10 Sep. 2011.

Water chemistry at time of collecting: K25 3.19 mS/cm, pH 6.6, water temperature 20.9 °C.

PetruderDam. Approximate coordinates: 30°25 , 20.87” S, 116°57 , 39.43”E (sample DJC/15), collected

by D.J. Cale on 11 Sep. 2011 (5 $ §: WAMC52329-52331, OC.3341-3342). Water chemistry at time of

collecting: K25 0.16 mS/cm, pH 7.85, water temperature 22.0 °C.

Tin Dog Creek. Approximate coordinates: 31°11 , 53.5” S, 117°01 , 41.4” E (sample DJC/19), collected

by D.J. Cale on 23 Sep. 2011 ($$: WAMC52311-52315, OC.3335-3336. Juveniles: WAMC523lb-

52318). Water chemistry at time of collecting: K25 1.56 mS/cm, pH 7.06, water temperature 24.2 °C.

Pools near Brookton Hwy (in Warrine Park). Approximate coordinates: 32°23 , 50.4” S, 116°48T)0.4”

E (sample DJC/23), collected by D.J. Cale on 01 Oct. 2011 (4 $$: WAMC52319-52321, OC.3337).

Water chemistry at time of collecting: K25 0.44 mS/cm, pH 8.1, water temperature 21.6 °C.

Three Springs Tumulus Stream. 29°35 , 31” S, 115°27 , 1” E, collected by A. Pinder on 29 Sep. 2010 (1

male: OC.3338; 2 $$: WAMC52322-52323).

East Lake Bryde. 33°21’ S, 118°49’ E (sample BRYDE7), collected by D.J. Cale on 22 Mar. 2006 (4

$ §: WAMC52324-52326; OC.3339). Water chemistry at time of collecting: K25 0.17 mS/cm, pH 6.91,

water temperature 25.8 °C.

Lake Cronin. Episodically filled waterbody with extensive shrub and Melaleuca fringe, collected

by S. Halse and A. Pinder on 25 Sep. 1997. Approximate coordinates: 32 o 23’02” S, 119°45 , 51” E.

Water chemistry at time of collecting: K25 0.23 mS/cm, pH 9.48, water temperature 18.0 °C. Material

investigated: one dissected $ (nr OS.544), with soft parts in a sealed slide and valves stored dry in

micropalaeontological cavity slide (illustrated: Fig. 24E-H).

Reserve Esperance 26140 near Munglinup. Seasonally filled lake with trees across most of flooded

area, collected 27 Oct. 1986 by S. Halse. Approximate coordinates: 33°26’24” S, 120°3E48” E. Water

chemistry: salinity 0.27 mg/L TDS, pH 6.93. Material investigated: one dissected $ (nr OS.604), with

soft parts in a sealed slide and valves stored dry in a micropalaeontological cavity slide (illustrated: Fig.

25I-L).

Remarks on the latter two localities, the lapels of both specimens are slightly larger than in most

specimens of B. scanloni sp. nov. and as (1) no males are at hand to check for the morphology of the

hemipenes and the prehensile palps in these populations and (2) no molecular data are available, the

identifications of these two specimens are tentative.

Additional notes on cryptic species

Specimens from the type locality (One Tree Hill Creek, sample DJC/11) all belong to cryptic species

El, which is thus the true B. scanloni sp. nov. s.s. Also specimens from sample DJC/23 (pools near

Brookton Hwy in Warrine Park) belong to this lineage (Fig. 191-0). Two specimens from cryptic species

43

European Journal of Taxonomy 66: 1-59 (2013)

B2 are also illustrated here, one female from pools beside Latham-Coorow Rd (western Wheatbelt)

(BVT/10/10) (Fig. 19A-D) and one female from Tin Dog Creek (DJC/19) (Fig. 19E-H). There are no

clear differences between the two cryptic species in valve morphology.

The soft parts of the male from Three Springs Tumulus Stream (TST) (Fig. 20A-C) are slightly different

from those of the holotype (Fig. 20D-F). Yet, within the molecular phylogeny the TST specimens cluster

closely with the cryptic species El, which is the same as for the type specimens. The hemipenis outline

and the prehensile palps of the male from OSTR013C (Fig. 20G-J) are almost identical to those of the

holotype, though no molecular data on this population are available, and it is thus also not clear to which

of the two cryptic species within B scanloni sp. nov. this specimen belongs.

Thus far, the two molecular species cannot be distinguished morphologically. Interestingly, these

two cryptic lineages occur sympatrically in no less than 4 localities (DJC/09, DJC/11, DJC/19 and

BVT/10/10).

Differential diagnosis

The drop-shaped lapel on the RV and the sharp angle on the distal margin of the terminal segment of

the right prehensile palp distinguish B. scanloni sp. nov. from all other species within the B. barangaroo

lineage.

Ecology and distribution

This is arguably the most common species in this lineage in the south-western part of WA. It typically

occurs in pools, dams and lakes with soft sediments.

Bennelongia calei sp. nov.

Figs 21-22

um:lsid:zoobank.org:act:DBD2498B-9E05-4E2A-9597-67844A85653E

Bennelongia barangaroo -De Deckker 1981a: 104, fig. 9 (partim).

Diagnosis (based on type specimens)

Valves (Fig. 21A-B, E-F) elongated, with greatest height situated close to the middle, dorsal margin

evenly sloping towards the posterior side; ventral margin almost straight. LV (Fig. 21 A, E) with antero-

ventral il large and reaching beyond middle of valve, antero-dorsal il descending to about halfway along

antero-ventral il. RV (Fig. 2 IB, F) with antero-ventral lapel large, elongated and wide, with crenulated

edge (Fig. 21C-D, G-H). Carapace in dorsal and ventral views (Fig. 21K-L) with greatest width in the

middle, most slender of all species described here, external surface rather smooth to weakly pitted, set

with only few short setae; anteriorly with a clearly delimited rostrum. Carapace in right lateral view

(Fig. 21I-J) with greatest height in the middle, dorsal margin evenly sloping to bluntly rounded posterior

margin; anteriorly with the widest LV>RV overlap of all species described here.

Males unknown.

Etymology

This species is named after D.J. Cale (Woodvale, WA) in honour of his longstanding contribution to the

knowledge of freshwater invertebrates in WA, including at Fraser Lake which is the type locality of the

present species (Cale et al. 2004), and also in recognition of the fact that he has collected so many of the

samples used for the present revision of the Bennelongia barangaroo lineage.

44

MARTENS K., HALSE S. & SCHON I., The Bennelongia barangaroo lineage in Western Australia

Measurements (all measurements in pm - see Table 1 for measurements of all specimens illustrated

with SEM)

Holotype ? (WAMC52335): RV: L = 1480, H = 857. LV: L = 1555, H = 865.

Type locality

Fraser Lake, ca. 8 km SE of Dowerin, WA. Approximate coordinates: 31 0 15’ 18.0” S, 117°4’22.0” E

(WGS 84). All material (sample code DJC/18) collected on 23 Sep. 2011 by D.J. Cale with a sweep net.

Water chemistry at time of collecting: K25 1.76 mS/cm, pH 8.1, water temperature 23.9 °C.

Type material

Holotype

9 (WAMC52335) valves stored dry in a micropalaeontological slide.

Allotype

As males are unknown, no allotype is designated.

Paratypes

Eight 55, either as dried valves or carapaces (WAMC52337-52341; OC.3344-3345). Ca. 45 females

stored as bulk in EtOH (WAMC52342).

Other material investigated

Fraser Lake (type locality). Four 55 valves and carapaces stored dry (WAMC52353-52356) from the

same locality, but collected on another date (sample nr SPM017B, collected by D.J. Cale, 24 Nov. 2000),

were also used during the present assessment of this species but are not considered as type material here.

Second pool along Carnamah-Eneabba Road on south side (Eneabba Springs). Approximate

coordinates: 29°48 , 23.62” S, 115°25 , 6.11” E (sample DJC/10), collected by D.J. Cale on 10 Sep.

2011 (one 5 WAMC52349). Water chemistry at time of collecting: K25 3.19 mS/cm, pH 6.6, water

temperature 20.9 °C.

Jerramungup West. Approximate coordinates: 33°59’16.03” S, 118°56 , 28.15” E (sample DJC/36),

collected by D.J. Cale on 21 Oct. 2011 (five 5 5 valves and carapaces stored dry WAMC52350-52352;

OC.3348-3349). Water chemistry at time of collecting: K25 0.73 mS/cm, pH 8.74, water temperature

25.8 °C

Oak Flat Wpitgnamma via Goomalling. Approximate coordinates 31°08'21” S, 116 0 52’46” E (sample

BVT/11/04), collected by B.V. Timms on 16 Aug. 2011 (four 55 valves and carapaces stored dry

WAMC52343-52344; OC.3346-3347).

Horse Collar gnamma, on Magee Rd via Kulin. Approximate coordinates: 32°48’04” S, 118°23’34” E

(sample BVT/11/05), collected by B.V. Timms on 4 Sep. 2011 (four 5 5 valves and carapaces stored dry

WAMC52345-52348).

Additional illustrations

Several other populations of this species were found and for four of these (listed above), valves of

female specimens are also illustrated here (Fig. 22). All of these specimens comply with the above

diagnosis, and where specimens were available for molecular analyses, they also fell into the B. calei sp.

nov. - cluster. No cryptic species were identified in this species.

Differential diagnosis

Bennelongia calei sp. nov. can easily be distinguished from all other species in the B. barangaroo

lineage by the elongated and stout antero-ventral lapel on the RV, which is slightly rounded and has a

45

European Journal of Taxonomy 66: 1-59 (2013)

Fig. 21. Bennelongia calei sp. nov. — A-H, J-L. Fraser Lake (type locality, females, DJC/18): A.

Paratype, LVi (OC.3344). B. Paratype, RVi (idem). C. Paratype, RVI, detail anterior (idem). D. Paratype,

RVI, detail anterior, tilted (idem). E. Holotype, LVi (WAMC52335). F. Holotype, RVI (idem). G.

Holotype, RVi, detail anterior (idem). H. Holotype, RVi, detail anterior, tilted (idem). J. Paratype, CpRL

(OC.3345). K. Paratype, CpD (WAMC52338). L. Paratype, CpV (WAMC52337). — I Oak Flat W pit

gnamma, via Goomalling (BVT/11/04, non-type female, WAMC52343). CpRL. Scales: A-B, E-F, I-L =

1 in in : C-D, G-H = 200 pm.

46

MARTENS K., HALSE S. & SCHON I., The Bennelongia barangaroo lineage in Western Australia

Fig. 22. Bennelongia calei sp. nov., non-type specimens. —A-D. OakFlat Wpit gnamma, via Goomalling

(BVT/11/04, $, OC.3346). A. LVi. B. RVi. C. RVi, detail anterior. D. RVi, detail anterior, tilted. — EH

Horse Collar gnamma, on Magee Rd via Kulin (BVT/11/05, ?, WAMC52345). E. LVi. F. RVi. G. RVi,

detail anterior. H. RVi, detail anterior, tilted. — I-L. Second pool along Carnamah-Eneabba Road on

south side, Eneabba Springs (DJC/10, §, WAMC52349). I. LVi. J. RVi. K. RVi, detail anterior. L. RVi,

detail anterior, tilted. — M-O. Pool at Jerramungup West (DJC/36, §, OC.3348). M. LVi. N. RVi. O.

RVi, detail anterior, tilted. Scales: A-B, E-F, I-J, M-N = 1 mm; C-D, G-H, K-L, O = 200 pm.

47

European Journal of Taxonomy 66: 1-59 (2013)

crenulated distal margin, the large anterior LV>RV overlap, the stout antero-ventral il on the LV and the

clear anterior rostrum on the carapace in dorsal or ventral view.

Ecology and distribution

The species is most common in pools and lakes in the southwest of WA. However, B. calei sp. nov.

was also recovered from a totally different kind of habitat, namely pit gnammas near Goomalling

(BVT/11/04) and Kulin (BVT/11/05). Whereas the other species in this lineage apparently occur either

in rock pools or in soft bottomed pools and lakes, B. calei sp. nov. can apparently survive in both

(very different) types of habitats. Bennelongia calei sp. nov. is, together with B. timmsi sp. nov. and B.

scanloni sp. nov., one of the more common species in its area.

Bennelongia dedeckkeri Shearn et al. , 2012

Figs 23, 24I-L

Bennelongia dedeckkeri sp. nov. - Shearn et al. , 2012: 10-14, figs 4-5.

Material investigated

Dam atKylena Well (Pilbara). Approximate coordinates: 22 o 06 , 00 ,, S, 119°39 , 00 ,, E (sample KIES 10).

Collected on 23 Apr. 2006 by the authors.

Unnamed saline billabong N of Coolcalaya Rd (Murchinson, Gascoyne). Approximate coordinates:

27°48 , 28” S, E (sample SIKE2). Collected on 5 Jul. 2011 by the authors. Water chemistry

at time of collecting: K25 8.8 mS/cm, pH 8.8, water temperature 11.0 °C.

McNeil Claypan , Carnarvon (Murchinson, Gascoyne). Approximate coordinates: 24 o 52 , 06 ,, S,

113°42’56” E (sample SIKE9). Collected on 6 Jul. 2011 by the authors. Water chemistry at time of

collecting: K25 0.19 mS/cm, pH 9.4, water temperature 10.8 °C.

Roadside ditch 7, North-West Coastal Huy. Minilya Station (Murchinson, Gascoyne). Approximate

coordinates: 23 0 54 , 25 ,, S, 114°01 ’45” E (sample SIKE18). Collected on 7 Jul. 2011 by the authors.

Water chemistry at time of collecting: K25 0.66 mS/cm, pH 7.4, water temperature 17.3 °C.

Roadside ditch 2 , North-West Coastal Hwy, (Murchinson, Gascoyne). Approximate coordinates:

23 0 54’25” S, 114°0r47” E (sample SIKE19). Collected on 7 Jul. 2011 by the authors. Water chemistry

at time of collecting: K25 0.69 mS/cm, pH 7.3, water temperature 17.3 °C.

Lake Gregory , south of Halls Creek. Approximate coordinates: 20°12’ S, 127 0 27’ E. Collected by S.

Halse on 29 May 1991 in fresh water (see Halse et al. 1998). One dissected $ (nr OS.260), with soft

parts in a sealed slide and valves stored dry in micropalaeontological cavity slide (illustrated in Fig.

24I-L).

Brief redescription

Smallest of the species described here, with females being only slightly longer than 1 mm.

Valves (Fig. 23A-B, E-F) high, with greatest height situated well in front of the middle, dorsal margin

evenly sloping towards the posterior side; ventral margin nearly straight. LV (Fig. 23A, E) with antero-

ventral il of medium size in lower half of valve, antero-dorsal il descending to about halfway along the

antero-ventral il. RV (Fig. 23B, F) with antero-ventral lapel almost rectangular, but slightly skewed and

bluntly pointed towards the ventral side (Fig. 23C-D, G-H).

Carapace in dorsal and ventral views (Fig. 23K-L) with greatest width in the middle, hirsute and heavily

pitted, anteriorly without a rostrum.

Males not yet found in WA.

48

MARTENS K., HALSE S. & SCHON I., The Bennelongia barangaroo lineage in Western Australia

Measurements (all measurements in pm - see Table 1 for measurements of all specimens illustrated

with SEM)

$ (Pilbara, KIES10, WAMC52357): RV: L= 1110, H= 676. LV: L= 1190, H= 701.

$ (Murchinson/Gascoyne, SIKE9, OC.3351): RV: L= 1125, H= 672. LV: L= 1188, H= 715.

Fig. 23. Bennelongia dedeckkeri , non-type females. — A-D, I, K-L. Dam at Kijlena Well, Pilbara

(KIES/10). A. LVi (WAMC52357). B. RVi (idem). C. RVi, detail anterior (idem). D. RVi, detail

anterior, slightly tilted (idem). I CpRL (OC.3350). K. CpD (WAMC52360). L. CpV (WAMC52359).

— E-H, J. McNeil Claypan, Murchinson/Gascoyne (SIKE9). E. LVi (OC.3351). F. RVi (idem). G.

RVi, detail anterior (idem). H. RVi, detail anterior, tilted (idem). J. CpRL (OC.3352). Scales: A-B,

E-F, I-L = 1 mm; C-D, G-H = 200 pm.

49

European Journal of Taxonomy 66: 1-59 (2013)

Ecology and distribution

Bennelongia dedeckkeri Shearn et al, 2012 was first described from Queensland, from a sexual

population. In WA it is not uncommon, but thus far only asexual populations have been found. To date

B. dedeckkeri is the only species of the B. barangaroo group found in both the eastern and the western

parts of Australia.

Fig. 24. Bennelongia spp. (no museum nrs), females. —A-D. Bennelongia sp. XI. Crane Pan (OS.255).

A. LVi. B. RVi. C. RVi, detail anterior. D. RVi, detail anterior, tilted. — E-H. ? Bennelongia scanloni sp.

nov., Lake Cronin (OS.544). E. LVi. F. RVi. G. RVi, detail anterior. H. RVi, detail anterior, tilted. — I-L.

Bennelongia dedeckkeri , Lake Gregory (OS.260). I. LVi. J. RVi. K. RVi, detail anterior. L. RVi, detail

anterior, tilted. Scales: A-B, E-F, I-J = 1 mm; C-D, G-H, K-L = 200 pm.

50

MARTENS K., HALSE S. & SCHON I., The Bennelongia barangaroo lineage in Western Australia

Bennelongia sp. indet.

There are several single specimens, mostly female, from various (older) collections that could not be

identified with certainty. Some of these are here illustrated to allow for future reference, in case new

collections will become available. It is entirely possible that some of these specimens represent hybrid

clades within the B. barangaroo-gro up. As none of these specimens have as yet been identified with

certainty, no museum numbers have been allocated to them.

Bennelongia sp. XI

Fig. 24A-D

Material investigated

One dissected $ (nr OS.255), with soft parts in a sealed slide and valves stored dry in a micro-

palaeontological cavity slide.

Locality

Canegrass covered claypan beside North-West Coastal Highway on Wooramel Station (CB35a),

25°40 , 52” S, 114°13 , 14” E, collected by S. Halse and A. Clarke on 24 Aug. 1994.

Morphology and affinities

The specimen is characterised by (1) a large antero-ventral lapel on the RV, subquadrate and only wealdy

crenulated, i.e., a shape unlike that of any of the other species (re-) described in the present paper, (2) a

relatively small antero-ventral il on the LV and, (3) the bluntly pointed posterior margin of the LV, with

a pronounced flange. The latter character is unique within the B. barangaroo lineage. If this character is

stable and also occurs in other specimens, it could be indicative at a specific level.

Bennelongia sp. X2

Fig. 25A-D

Material investigated

One dissected $ (nr KMWA.917), with soft parts in a sealed slide and valves stored dry in a

micropalaeontological cavity slide.

Locality

Three Springs Tumulus stream, 29°35 , 31”S, 115 0 27 , 1” E, collected by A. Pinder on 29 Sep. 2010.

Morphology and affinities

The specimen has some affinity with B. scanloni sp. nov., but the antero-ventral il on the LV is smaller

and the antero-ventral lapel of the RV is of a different shape, being larger and almost rectangular.

Remarks

Several other specimens from the same sample belong to cryptic species El of B. scanloni sp. nov., as

shown by morphological and molecular evidence (see above).

51

European Journal of Taxonomy 66: 1-59 (2013)

Bennelongia sp. nov. E2

Fig. 25E-H

Material investigated

One dissected $ (nr KMWA.806), with soft parts used for molecular analysis and valves stored dry in a

micropalaeontological cavity slide.

Fig. 25. Bennelongia spp. (no museum nrs). — A-D. Bennelongia sp. X2, Three Springs Tumulus

Stream (§, KMWA.917). A. LVi. B. RVi. C. RVi, detail anterior. D. RVi, detail anterior, tilted. — E-H.

Bennelongia sp. nov. E2, BYK (§, KMWA.806) = Species E2 in Figs 2-3. E. LVi. F. RVi. G. RVi, detail

anterior. H. RVi, detail anterior, tilted. — I-L. ? Bennelongia scanloni sp. nov., pool near Esperance ($,

OS.004). I. LVi. J. RVi. K. RVi, detail anterior. L. RVi, detail anterior, tilted. Scales: A-B, E-F, I-J = 1

mm; C-D, G-H, K-L = 200 pm.

52

MARTENS K., HALSE S. & SCHON I., The Bennelongia barangaroo lineage in Western Australia

Locality

Yakabindie Claypan, approximately 27°34’ S, 120°3E E (sample LN3006), collected by Outback

Ecology on 24 Mar. 2011.

Morphology and affinities

Although general valve appearance again shows some affinities to B. scanloni sp. nov., the different shape

of the antero-ventral il on the LV (less pronounced and evenly rounded) and the large and subquadrate

anteroventral lapel on the RV distinguishes this specimen from all other species (re-)described here.

Molecular analysis of the soft parts of the same specimen shows that it clusters outside of the B. scanloni

sp. nov. group (including cryptic species B2 and El), and constitutes a different genetic species. Lack of

additional material and males prevents us from formally describing this new species here.

General Discussion

De Declcker (1981a) reported two species of Bennelongia from WA: B. australis (Brady, 1886) and B.

barangaroo De Deckker, 1981. Meanwhile, Martens et al. (2012) described nine new species from three

separate lineages within this genus from WA (the B. australis lineage, the B. cygnus lineage and the B.

pinpi lineage) and showed that B. australis is actually a species group with at least seven nominal species

and potentially more. Shearn et al. (2012) contributed to the knowledge of the genus Bennelongia in

eastern Australia by confirming the validity of B. pinpi De Deckker, 1981 and B. harpago De Declcker

& McKenzie, 1981, using genetics to identify the occurrence of a cryptic lineage within a species of

the B. australis lineage ( B. cuensis Martens et al ., 2012), and by describing two new species in the B.

barangaroo lineage ( B. dedeckkeri Shearn et al ,2012 and B. mckenziei Shearn et al. , 2012) and one new

species within the B. nimala lineage ( B. regina Shearn et al. , 2012). The previous work of De Deckker

(1981a,b, 1982) and De Deckker & McKenzie (1981) as well as these two new papers (Martens et al ,

2012; Shearn et al ., 2012) show that the genus Bennelongia has extensive radiations in both eastern and

western Australia. The present paper formally describes seven new species within the B. barangaroo

lineage, redescribes B. dedeckkeri and indicates the putative presence of several other species within the

B. barangaroo lineage, all from WA. With the seven new species described here, the genus Bennelongia

now comprises 25 nominal species (Table 3) but several more await formal description.

Morphological features

Once again, the size and shape, and in some cases the sheer presence or absence, of the antero-ventral

lapel on the RV has proven to be indispensable to characterize species of Bennelongia. The plasticity

of this feature within the B. barangaroo lineage is amazing and ranges from being fully absent through

being small and triangular, to large and drop-like and to a large, elongated and heavily serrated structure.

Whereas Martens et al. (2012) hypothesized that the function of the lapel is to lock the sulcus in the

LV when valves need to be closed tightly ( e.g ., when attacked by predators, or in cases where habitat

is rapidly desiccating), one could ask why such a wide range of morphologies of the lapel is necessary

for highly similar functions in the different lineages and species. When homologous structures have

widely different morphologies in closely related species, sexual selection is often invoked as causality.

Several authors have indeed already mentioned the potential of sexual selection in ostracod radiations

(for example Tsukagoshi 1988; Martens 2000). However, such structures must usually (1) display sexual

dimorphism and be most common in males, (2) be accessible during pre-copulation by putative partners,

i.e., females, to determine whether or not to accept the male as a partner for reproduction, and (3) occur

only in the adult stage, i.e., after the final moult. Only the latter of these three conditions is fulfilled in

Bennelongia because there is no apparent sexual dimorphism in lapel-shape and lapels are in general not

easily available for inspection by females during the pre-copulatory stage. It is therefore unlikely that

lapel morphology has evolved through sexual selection.

53

European Journal of Taxonomy 66 : 1-59 ( 2013 )

Table 3. Species presently described in Bennelongia and their distribution (species in bold are newly

described here). Only certain distributions, based on type localities and documented range extensions,

are given here. * indicates the type species.

Bennelongia australis (Brady, 1886): SA

Bennelongia barangaroo De Deckker, 1981: WA

Bennelongia bidgelangensis Martens et al. , 2012: WA, Gascoyne

Bennelongia calei sp. nov.: WA

Bennelongia coondinerensis Martens et al, 2012: WA, Pilbara

Bennelongia cuensis Martens et al. , 2012: WA, Yilgam

Bennelongia cygnus Martens et al., 2012: WA, Swan Valley

Bennelongia dedeckkeri Shearn et al, 2012: QLD, WA

Bennelongiafrumenta Martens et al., 2012: WA, Wheatbelt

Bennelongia gnamma sp. nov.: WA

Bennelongia gwelupensis Martens et al., 2012: WA, Perth, southwest coast

* Bennelongia harpago De Deckker & McKenzie, 1981: QLD

Bennelongia hirsuta sp. nov.: WA

Bennelongia ivanae sp. nov.: WA

Bennelongia kimberleyensis Martens et al., 2012: WA, Kimberley

Bennelongia lata Martens et al., 2012: WA, Gascoyne-Murchinson region

Bennelongia mckenziei Shearn et al., 2012: QLD

Bennelongia mcraeae sp. nov.: WA

Bennelongia nimala De Deckker, 1981: NT

Bennelongia pinpi De Deckker, 1981: QLD

Bennelongia regina Shearn et al., 2012: QLD

Bennelongia scanloni sp. nov.: WA

Bennelongia strellyensis Martens et al, 2012: WA, Pilbara

Bennelongia timmsi sp. nov.: WA

Bennelongia tunta De Deckker, 1982: QLD

This leaves the possibility that lapels have evolved by chance (not a very parsimonious solution) or that

natural selection is acting on the evolution of this morphological feature and that selection pressures are

quite stringent. The morphological differences between species living in similar environments (pools,

lakes) can be either substantial, as is the case for B. macraeae sp. nov., B. scanloni sp. nov. and B. calei

sp. nov., or almost non-existent as in the case of the rock pool dwelling species, which are either cryptic

species without morphological differences or have very small differences (B. timmsi sp. nov. with 3

cryptic species and B. gnamma sp. nov.).

Bennelongia timmsi sp. nov. as a biological reality?

Recent research has shown that cryptic species are not uncommon in non-marine ostracods, as Shearn et

al. (2012) found a genetically distinct eastern Australian lineage within the otherwise western Australian

B. cuensis Martens et al., 2012, and confirmed that both clades within this species are morphologically

indistinguishable. Schon et al. (2012) found several cryptic species within putative ancient asexual

darwinulid ostracods, while Bode et al. (2010) revealed no less than 40 cryptic species within the

Palaearctic ostracod species Eucypris virens (Jurine, 1820).

54

MARTENS K., HALSE S. & SCHON I., The Bennelongia barangaroo lineage in Western Australia

In the B. barangaroo lineage, both B. timmsi sp. nov. and B. scanloni sp. nov. comprise cryptic species

as identified by molecular phylogenies based on the mitochondrial COI gene (Fig. 2) and the 4 theta

rule (Table 2). Bennelongia scanloni sp. nov. comprises three clusters and three unconnected genetic

networks (Fig. 3) of which two are identified as separate genetic species by the 4 theta rule, and the two

clusters together form a monophyletic clade within the tree. The smaller TST-clade is phylogenetically

slightly separate from the El clade, forms an additional network, but does not constitute a separate

genetic species. The situation in B. scanloni sp. nov. is thus a classic case of a monophyletic species

consisting of diverged, but morphologically unrecognisable, clades.

In B. timmsi sp. nov., however, the situation is less straightforward. Five clades are recognised in the

phylogenetic tree (Fig. 2 - Al-5) and there are six unconnected networks (Fig. 3), of which three are

considered valid genetic species (Al, A3, A2+4+5). However, unlike the situation in B. scanloni sp. nov.,

B. timmsi sp. nov. does not appear as a monophyletic taxon in the phylogenetic analysis, because A1 and

A3 cluster together in a different clade than A2+4+5. Extensive morphological comparisons (Figs 4-11)

could not reveal any specific differences in valve or soft part morphology between the clusters (though

see below), excluding the possibility that even clades A1+A3 on the one hand and clades A2+4+5 on the

other could be described as different monophyletic species. We considered it of little use to describe two

different species when they cannot be identified, except with molecular techniques.

Adding complexity to the phylogenetic uncertainty described above, the morphology of the antero-

ventral lapel on the RV shows some variability in length and position on the valve among specimens

within the B. timmsi sp. nov. clades, but this limited variability could not be linked to the phylogenetic

position of the specimens. While the shape of the terminal segment of the right prehensile palp was

variable within what is assumed to be the same cryptic species (in clade A5), it differed little between

cryptic species. Whether or not the differences observed in clade A5 specimens are real or artefacts

remains to be seen. The morphologies of the hemipenis-outlines and of the left prehensile palp were

fairly constant across all five clades and three cryptic species.

The situation in B timmsi sp. nov. is the first case within the revision of the genus Bennelongia where

morphological and molecular data are incongruent. At this stage, we have chosen to follow the results

of the morphological analyses, as it seems that, within the tree, the nodes indicated by an * (Fig. 2)

are weakly or not statistically supported, and if those nodes are collapsed into a polytomy, the virtual

polyphyletic position of B. timmsi sp. nov. would disappear. As soon as we have described and screened

all new species of Bennelongia from our collections, the molecular phylogeny of the genus as a whole

will be reconstructed and it is hoped that this more complete analysis will shed light on the presumed

polyphyletic status of B. timmsi sp. nov. In the meantime, Bennelongia timmsi sp. nov. is proposed here

as a valid biological species.

Distribution and Ecology

The continental-scale distribution of Bennelongia as a whole and phylogeography of selected species

will be dealt with elsewhere. At this stage, however, it is useful to point out that almost all species in this

genus have fairly restricted distributions: this appears to be so for the species within the B. australis and B.

cygnus lineages (Martens et al. 2012), as well as in the B. pinpi and the B. barangaroo lineages (Martens et

al. loc. cit; Shearn et al. 2012). One notable exception appears to be B. dedeckkeri , which has meanwhile

been reported from Queensland in eastern Australia (Shearn et al. 2012) and from both northern and

central Western Australia (present paper - Fig. 1). Molecular screening has shown that specimens from

WA and from QFD even have (near to) identical COI-sequences, showing that this species must have

a very efficient means of dispersal compared to its congeners. So far, only parthenogenetic populations

have been found in WA, which would be one way to explain a potentially recent expansion from the east

(where sexual populations do appear to exist) to the west. Indeed, parthenogens are assumed to be more

55

European Journal of Taxonomy 66 : 1-59 ( 2013 )

efficient dispersers, as one egg is potentially enough to establish a viable population, whereas sexual

groups must have both genders colonising the same habitat, and these moreover must find each other

in a spatially and temporally diluted environment (Horne & Martens 1999). All screened specimens

from WAhad identical COI sequences (Fig. 3), and the apparent absence of genetic diversity between

localities more than 1000 km apart supports this hypothesis of parthenogens.

Bennelongia dedeckkeri was recovered from both ephemeral lakes and pans (e.g., the McNeil Claypan

in Carnarvon), as well as semi-permanent lakes (e.g., the remote Lake Gregory, in the Tanami Desert,

but see Halse et al. 1998), and thus seems to be able to thrive in different types of environments. The

species might very well have a General Purpose Genotype (Van Doninck et al. 2003). The remainder

of the species of the B. barangaroo lineage are distributed in more particular habitat types: B. timmsi

sp. nov., B. gnamma sp. nov., B. hirsuta sp. nov. and B. ivanae sp. nov. occur only in rock pools, and

this diversity within a single lineage confirms the rock pools on the granite outcrops in southwestern

Australia as foci of diversification of aquatic animals (Pinder et al. 2000). Bennelongia mcraeae sp.

nov. and B. scanloni sp. nov. were sampled in seasonal or episodic soft-sediment lakes and pools only.

Bennelongia calei sp. nov. occurs mainly in the latter types of habitats, but was also found in at least

two localities in pit-gnammas, which are deep and narrow rock pools. Populations from both types of

habitats were also shown to belong to one and the same genetic species (Fig. 2), and this species thus

has a puzzling autecology.

Acknowledgements

The authors gratefully acknowledge the financial support by an ABRS-grant (nr RF211-33: ‘Biodiversity

and taxonomy of Ostracoda (Crustacea) from temporary water bodies of inland Western Australia’)

and by an Edith Cowan University Industry Collaboration grant. KM & IS acknowledge the Western

Australian Department of Parks and Wildlife (2006) and Bennelongia Pty Ltd (2008, 2009, 2010) for

financial support during several scientific stays in Perth, as well as the financial contribution of the FWO

Vlaanderen (Fund for Scientific Research, Flanders) in the form of travel grants in 2010 (V4.172.10N &

V4.173.10N) and the projects 1.5.172.09 (Krediet aan Navorsers) and G.0118.03N (projectonderzoek).

KM and IS also thank the people of Bennelongia Pty Ltd (Perth, WA) for continuous logistic support

(lab space, use of microscopes, etc.) and for unfailing companionship in the lab and their son Emrys

for invaluable help with sorting living specimens in the field as well as in the lab. David J. Cale, A.

Clarke and Adrian Pinder (DEC, Woodvale, Australia), Brian V. Timms (Newcastle, Australia), Outback

Ecology (Perth, Australia) and Jane McRae (Bennelongia Pty Ltd, Perth, Australia) are acknowledged

for collecting most of the material described in the present paper.

Julien Cilis and Claudine Behen (RBINS, Brussels, Belgium) provided technical assistance with the

SEM micrographs and with the line drawings, respectively. Kristiaan Hoedemakers (RBINS, Brussels,

Belgium) produced the SEM plates and supervised the entire production of the manuscript. Mike

Scanlon (Bennelongia Pty Ltd, Perth, WA) produced the map. IS thanks Annette Koenders (Edith Cowan

University, Perth) and Mike Johnson (University of Western Australia, Perth) for providing laboratory

space for the molecular part of this research. Dr Finn Viehberg (Koln, Germany) and Dr David J. Horne

(London, U.K.) are thanked for their valuable comments as referees.

References

Altschul S.F., Gish W., Miller W., Myers E.W. & Lipman D.J. 1990. Basic local alignment search tool.

Journal of Molecular Biology 215: 403-410. http://dx.doi.org/10,1016/80022-2836(05180360-2

Birky Jr. C.W. 2011. Species detection and identification in sexual organisms using population genetic

theory and DNA sequences. PLoS ONE 8: e52544. http://dx.doi.org/10.1371/iournal.pone.0052544

56

MARTENS K., HALSE S. & SCHON I., The Bennelongia barangaroo lineage in Western Australia

Birky Jr. C.W. & Barraclough T.G. 2009. Asexual speciation. In: Schon I., Martens K. & van Dijk R

(eds) Lost Sex. Springer Scientific Publishers, Dordrecht: 201-216.

Birky Jr. C.W., Adams J., Gemmel M. & Perry J. 2010. Using Population Genetic Theory and DNA

sequences for species detection and identification in asexual organisms. PLoS ONE 5: el0609. http://

dx.doi.org/10.1371/iournal.pone. 0010609

Birky Jr. C. W., Ricci C., Melone G. & Fontaneto D. 2011. Integrating DNA and morphological taxonomy

to describe diversity in poorly studied microscopic animals: new species of the genus Abrochtha Bryce,

1910 (Rotifera: Bdelloidea: Philodinavidae). Zoological Journal of the Linnean Society 161: 723-734.

http://dx.doi.Org/10.llll/i.1096-3642.2010.00674.x

Bode S.N.S., Lamatsch D.K., Martins M.J.F., Schmit O., Vandekerkhove J., Mezquita F., Namiotko T.,

Rossetti G., Schon I., ButlinR.K. & Martens K. 2010. Exceptional cryptic diversity and multiple origins

of parthenogenesis in a freshwater ostracod. Molecular Phylogeny and Evolution 54: 542-552. http://

dx.doi.org/10.1016/i.vmpev.2009.08.022

Broodbakker N.W. & Danielopol D.F. 1982. The chaetotaxy of Cypridacea (Crustacea, Ostracoda)

limbs: proposals for a descriptive model. Bijdragen tot de Dierkunde 52: 103-120.

Cale D.J., Halse S.A. & Walker C.D. 2004. Wetland monitoring in the wheatbelt of south-west Western

Australia: site descriptions, waterbird, aquatic invertebrate and groundwater data. Conservation Science

Western Australia 5: 20-135.

Clement M., Posada D. & Crandall K. 2000. TCS: a computer program to estimate gene genealogies.

Molecular Ecology 9: 1657-1660. http://dx.doi.Org/10.1046/i.1365-294x.2000.01020.x

Darriba D., Taboada G.F., Doallo R. & Posada D. 2012. jModelTest 2: more models, new heuristics and

parallel computing. Nature Methods 9: 772. http://dx.doi.org/10.1038/nmeth.21Q9

De Deckker P. 1981a. Taxonomy and ecology notes of some ostracods from Australian inland waters.

Transactions of the Royal Society of South Australia 105 (3): 91-138.

De Deckker P. 1981b. Ostracoda from Australian inland waters - notes on taxonomy and ecology.

Transactions of the Royal Society of Victoria 93 (1): 43-85.

De Deckker P. 1982. On Bennelongia tunta De Deckker sp. nov. A Stereo-Atlas of Ostracod Shells 9

(21): 117-124.

De Deckker P. & McKenzie K.G. 1981. Bennelongia , anew cyprididid ostracod genus from Australasia.

Transactions of the Royal Society of South Australia 105 (2): 53-58.

De Deckker P. & Martens K. 2013. Extraordinary morphological changes in valve morphology during

the ontogeny of several species of the Australian ostracod genus Bennelongia (Crustacea, Ostracoda).

European Journal of Taxonomy 36: 1-37. http://dx.doi.org/10.5852/eit.2013.36

Folmer O., Black M., Hoeh W., Lutz R. & Vrijenhoek R. 1994. DNA primers for amplification of

mitochondrial cytochrome c oxidase subunit I from diverse metazoan invertebrates. Molecular Marine

Biology and Biotechnology 3: 294-299.

Fontaneto D., Herniou E.A., Boschetti C., Caprioli M., Melone G., Ricci C. & Barraclough T.G. 2007.

Independently evolving species in asexual bdelloid rotifers. PLoS ONE 5: e87. http://dx. doi.org/10.1371/

iournal.pbio.0050087

Fontaneto D., Kaya M., Herniou E.A. & Barraclough T.G. 2009. Extreme levels of hidden diversity in

microscopic animals (Rotifera) revealed by DNA taxonomy. Molecular Phylogenetics and Evolution

53: 182-189. http://dx.doi.Org/10.1016/j.ympev.2009.04.011

57

European Journal of Taxonomy 66: 1-59 (2013)

Guindon S. & Gascuel O. 2003. PhyML - a simple, fast, and accurate algorithm to estimate

large phylogenies by maximum likelihood. Systematic Biology 52: 696-704. http://dx.doi.

org/10,1080/10635150390235520

Hall T. 2007. BioEdit: Biological sequence alignment editor for Win95/98/NT/2K/XP [Online]. Website

last modified on June 27, 2007 (accessed on September 13, 2011). Available at http://www.mbio.ncsu.

edii/BioEdit/bioedit.html

Halse S.A. 2002. Diversity of Ostracoda (Crustacea) in inland waters of Western Australia. Verhand-

lungen der Internationalen Vereinigungfur theoretische und angewandte Limnologie 28: 914-918.

Halse S.A., Shiel R.J. & Williams W.D. 1998. Aquatic invertebrates of Lake Gregory, north-western

Australia, in relation to salinity and ionic composition. Hydrobiologia 381: 15-29. http://dx.doi.

org/10.1023/A: 1003263105122

Horne D.J. & Martens K. 1999. Geographical parthenogenesis in European non-marine ostracods: post¬

glacial invasion or Holocene stability? Hydrobiologia 391: 1-7. http://dx.doi.org/10.1023/

A: 1003508210166

Home D.J., Cohen A. & Martens K. 2002. Taxonomy, Morphology and Biology of Quaternary and

Living Ostracoda. In: Holmes J.A. & Chivas A.R. (eds) The Ostracoda: Application in Quaternary

Research. Geophysical Monograph 131. American Geophysical Union, Washington, DC: 5-36.

Larkin M.A., Blackshields G., Brown N.P, Chenna R., McGettigan P.A., McWilliam H., Valentin F.,

Wallace I.M., Wilm A., Lopez R., Thompson J.D., Gibson T.J. & Higgins D.G. 2007. Clustal W and

Clustal X version 2.0. Bioinformatics 23: 2947-2948. http://dx.doi.org/10.1093/bioinformatics/btm404

Martens K. 1987. Homology and functional morphology of the sexual dimorphism in the antenna of

Sclerocypris Sars, 1924 (Crustacea, Ostracoda, Megalocypridinae). Bijdragen tot de Dierkunde 57: 183-

190.

Martens K. 2000. Factors affecting the divergence of mate recognition systems in the Limnocytherinae

(Cmstacea, Ostracoda). In: Horne D.J. & Martens K. (eds) Proceedings of the XIII International Symposium

on Ostracoda. Hydrobiologia 419: 83-101. http://dx.doi.org/10.1023/A: 1003954513004

Martens K., Halse S. & Schon I. 2012. Nine new species of Bennelongia De Decklcer & McKenzie,

1981 (Cmstacea, Ostracoda) from Western Australia, with the description of a new subfamily. European

Journal of Taxonomy 8: 1-56. http://dx.doi.Org/10.5852/eit.2012.8

Pinder A.M., Halse S.A., Shiel R.J. & McRae J.M. 2000. Granite outcrop pools in south-western

Australia: foci of diversification and refugia for aquatic invertebrates. Journal of the Royal Society of

Western Australia 83: 149-161.

Ronquist F., Teslenko M., van der Mark P, Ayres D., Darling A., Hohna S., Larget B., Liu L., Suchard

M.A. & Huelsenbeck J.P 2012. MrBayes 3.2: Efficient Bayesian phylogenetic inference and model

choice across a large model space. Systematic Biology 61: 539-542. http://dx.doi.org/10.1093/svsbio/

svs029

Schon I., Pinto R.L., Halse S.A., Smith A.J., Martens K. & Birky Jr. C.W. 2012. Cryptic species in

putative ancient asexual darwinulids (Cmstacea, Ostracoda). PLOS ONE 7: e39844. http://dx.doi.

org/10.1371/joumal.pone. 0039844

Shearn R., Koenders A., Halse S., Schon I. & Martens K. 2012. A review of Bennelongia De Deckker &

Mckenzie, 1981 (Cmstacea, Ostracoda) species from eastern Australia, with the description of three new

species. European Journal of Taxonomy 25: 1-35. http://dx.doi.org/10.5852/ejt.2012.25

58

MARTENS K., HALSE S. & SCHON I., The Bennelongia barangaroo lineage in Western Australia

Tamura K., Peterson D., Peterson N., Stecher G., Nei M. & Kumar S. 2011. MEGA5: Molecular

Evolutionary Genetics Analysis using Likelihood, Distance, and Parsimony methods. Molecular Biology

and Evolution 28: 2731-2739. http://dx.doi.org/10.1093/molbev/msr 121

Timms B.V. 2002. The fairy shrimp Branchinella Sayce (Crustacea: Anostraca: Thamnocephalidae) in

Western Australia, including a description of four new species. Hydrobiologia 486: 71-89. http://dx.doi.

org/10,1023/A: 1021330230369

Tsukagoshi A. 1988. Reproductive character displacement in the ostracod genus Cythere. Journal of

Crustacean Biology 8: 563-575. http://dx.doi.org/10.2307/1548693

Van Doninclc K., Schon I., Maes F., De Bruyn L. & Martens K. 2003. Ecological strategies in the ancient

asexual animal group Darwinulidae (Crustacea, Ostracoda). Freshwater Biology 48: 1285-1294. http://

dx.doi.org/10.1046/i. 1365-2427.2003.01078.x

Manuscript received: 29 July 2013

Manuscript accepted: 9 October 2013

Published on: 25 November 2013

Topic editor: Rudy Jocque

Desk editor: Kristiaan Hoedemakers

Printed versions of all papers are also deposited in the libraries of the institutes that are members of

the EJT consortium: Museum National d’Histoire Naturelle, Paris, France; National Botanic Garden

of Belgium, Meise, Belgium; Royal Museum for Central Africa, Tervuren, Belgium; Natural History

Museum, London, United Kingdom; Royal Belgian Institute of Natural Sciences, Brussels, Belgium;

Natural History Museum of Denmark, Copenhagen, De nm ark

59