Reef formation versus solitariness in two New Zealand serpulids does not involve cryptic species

Page 1

AQUATIC BIOLOGYAquat Biol

Vol. 16: 97–103, 2012doi: 10.3354/ab00444

Published June 26

INTRODUCTION

Serpulid worms (Annelida: Polychaeta: Sabellida:Serpulidae) are benthic suspension-feeders with achar acteristic calcareous tube and a tentacularcrown. About 10% of known species can sometimesgrow in groups of tubes tangled together in aggrega-tions which can grow upwards to form reefs up to afew metres in height and several kilometres in length(ten Hove & van den Hurk 1993). Although serpulidreefs have been studied for more than a century (e.g.Nardo 1847, Mörch 1863), the factors promotingaggregation are not well understood (Bosence 1973,ten Hove 1979). All species that occur in aggrega-tions are also found as individuals (ten Hove & vanden Hurk 1993).

In southern New Zealand there are 2 commonaggregating serpulid worms (Fig. 1): Galeolaria hystrix Mörch, 1863 and Spirobranchus cariniferus

(Gray, 1843)1. G. hystrix occurs in intertidal to subti-dal waters (0 to 10 m) as solitary individuals (some onrocks and others in sand), and also in deeper waters(10 to 20 m) of Paterson Inlet, Stewart Island, inaggregated patch reefs (Smith et al. 2005). S.cariniferus occurs as individuals or in patchy or belt-like aggregations in the lower intertidal zone of NewZealand north of about 45° S latitude (Morton &Miller 1973, as Pomatoceros caeruleus).

Morphologically similar serpulids can exhibit dif-ferences in habit. For example, Bastida-Zavala & tenHove (2002) noted that 4 species of western Atlantic

© Inter-Research 2012 · www.int-res.com*Email: [email protected]

Reef formation versus solitariness in twoNew Zealand serpulids does not involve cryptic

species

Abigail M. Smith1,*, Zoe E. Henderson1, Martyn Kennedy2, Tania M. King2, Hamish G. Spencer2

1Department of Marine Science, University of Otago, Dunedin, New Zealand2Allan Wilson Centre for Molecular Ecology and Evolution, Department of Zoology, University of Otago, Dunedin, New Zealand

ABSTRACT: Tissue from both solitary and aggregated serpulids Galeolaria hystrix and Spiro-branchus cariniferus from southern New Zealand was sequenced using 18S, histone H3 andcytochrome b in order to determine whether these differences in ecology and lifestyle reflect theexistence of cryptic species. In both cases, all 3 phylogenetic trees unequivocally combined soli-tary and aggregated individuals into 2 monophyletic groups corresponding to the nominal spe-cies. Some combination of larval behaviour, adult attractants and biotic/physical environmentalfactors are likely to be the drivers of reef formation in these serpulid worms. A previouslysequenced Australian specimen of G. hystrix is not in the same clade as the New Zealand samplesand requires re-investigation.

KEY WORDS: Polychaeta · Serpulidae · Galeolaria · Spirobranchus · Temperate reefs

Resale or republication not permitted without written consent of the publisher

1Some authorities (e.g. ten Hove & Kupriyanova 2009) referto S. cariniferus as S. carinifer, being of the view that thisformation has the correct masculine ending to agree withthe masculine genus name. Our consultation with Latinscholars, however, suggests both carinifer and carini ferusare correctly masculine (see also Herbert 1993); hence, weuse Gray’s original form of the name

Page 2

Aquat Biol 16: 97–103, 2012

Hydroides (H. spongicola, H. elegantulus, H. flori-danus and H. alatalateralis) were ‘very similar’ anddistinguishable chiefly by minute morphological dif-ferences (the length of wings along opercular spinesand the numbers of radioles, raduli and spines), butthe first was apparently restricted to living in associ-ation with a particular species of sponge. Though itis commonly theorised that environmental factors in -duce subtidal aggregation (e.g. ten Hove 1979), weknow of no studies that have examined cryptic spe-cies diversity as a possible explanation of these eco-logical differences.

Here we address the hypothesis that differentmodes of life in 2 species of southern serpulid worms,Spirobranchus carinferus and Galeolaria hystrix, arein fact reflections of such diversity, i.e. the aggregat-ing population is a different cryptic species from thesolitary population in each of 2 common species. Wenote that the related Australian nominal taxon, G.caespitosa, has recently been found to consist of 2

morphologically indistinguishable but geneticallydistant species (Halt et al. 2009).

MATERIALS AND METHODS

Specimens of solitary (i.e. individuals not touchinganother conspecific tube) and aggregated Galeolariahystrix and Spirobranchus cariniferus as well as outgroup species from the serpulid genera Protulaand Salmacina and the subfamily Spirorbinae werecollected from southern New Zealand (Table 1)between November 2010 and March 2011. DNA wasextracted from either the branchial crown or a largerregion of the dorsal part of the body for smallerspirorbids, using a Purelink™ Genomic DNA MiniKit (Invitrogen).

A combination of mitochondrial (cytochrome b) andnuclear (histone H3 and 18S) genes were used tocheck for concordance between these types of mark-

98

Fig. 1. Galeolaria hystrix and Spirobranchus cariniferus. (a) Solitary G. hystrix, settling plate, Big Glory Bay, Stewart Island,New Zealand (NZ); (b) solitary S. cariniferus, rock, Otago Harbour, NZ; (c) aggregated G. hystrix, seafloor, Big Glory Bay;(d) aggregated S. cariniferus, intertidal rocks, Sumner Beach, NZ. Photos by: M. A. Riedi and A. M. Smith. Scale bars are all

1 cm

Page 3

Smith et al.: New Zealand serpulid reefs

ers. Each of these markers evolves at a different rateand thus provides a different level of resolution. Ap-proximately 1800 bp of 18S was amplified usingprimers TimA (Norén & Jondelius 1999) and 18s2R(Kupri yanova et al. 2006), ~300 bp of histone H3 usingH3aF and H3aR (Colgan et al. 1998) and ~350 bp ofcytochrome b using Cytb424F and cobr825 (Halt et al.2009). All reactions were carried out in 20 µl volumesand contained 0.5 µM of each primer and 0.5 UMyTaq Red DNA polymerase (Bioline). Amplificationswere performed using an Eppen dorf Mastercycler epGradient S thermal cycler with a cycling profile of94°C for 120 s, followed by 45 cycles of 94°C for 30 s,45°C for 30 s (increased to 48°C and 120 s for 18S) and72°C for 60 s and a final extension of 72°C for 240 s.PCR products were run on 1% agarose gels containingSYBR Safe DNA gel stain (Life Technologies) and visualized under blue light (Uvitec UVIdoc HD2/20Blue). Bands were either purified directly or gel pu-rified using an E.Z.N.A. Ultra-Sep Gel Extraction Kit(Omega Bio-Tek). Purified PCR products were quanti-fied using a ND-1000 spectrophotometer (Nanodrop)and sequenced by University of Otago Genetic Analy-sis Services using an ABI 3730xl DNA analyser usingeither the PCR primers or, for 18S, internal primers18s2F and 1100R2 (Kupriyanova et al. 2006).

Sequences were edited and aligned usingClustalW2 within Sequencher 4.10.1 (Gene Codes).A BLAST search confirmed the correct gene regionshad been amplified (Altschul et al. 1990) and the new

sequences were submitted to GenBank (accessionnumbers JX144795–JX144879).

The 3 separate genes are treated individually inthe following analysis as we wished to evaluate theinformation in the nuclear and mitochondrial markersseparately. Phylogenetic analyses were performedwith MrBayes (Huelsenbeck & Ronquist 2001, Ron-quist & Huelsenbeck 2003) for Markov chain MonteCarlo Bayesian analysis and posterior probabilitiesand PAUP* v.4b10 (Swofford 2002) for maximumparsi mony (MP) and neighbour-joining (NJ) boot-strap searches (Felsenstein 1985). For visualizationpurposes, and because of the different taxa in eachdata set, different taxa were defined as the outgrouptaxa for 18S (Spirorbinae) and histone H3 (Protulabispiralis and Salmacina sp.) data sets, withcytochrome b drawn following the results of the other2 data sets (i.e. with Galeolaria hystrix and Spiro-branchus cariniferus both monophyletic).

The models of nucleotide substitution for theBayesian analyses were selected using the Akaikeinformation criterion of Modeltest 3.7 (Posada &Crandall 1998). The models selected for each generegion were submodels of GTR+I+G with more than2 substitution types (i.e. TrN+G for 18S, TVM+I+G forhistone H3, and K81uf+G for cytochrome b); thus, it ismore appropriate to use 6, rather than 2, substitutiontypes with each gene. Bayesian analysis was per-formed using MrBayes v.3.1.2 with the maximumlikelihood model employing 6 substitution types

99

Species Location Growth form No. individuals sequenced

18S Hist Cyt b

Galeolaria Katiki Beach, North Otago, 45° 27’ S, 170° 48’ E, intertidal Solitary 1 1 1hystrix Bravo Island, Paterson Inlet, Stewart Island, 46° 57.96’ S, 168° 07.92’ E, 9 m Aggregated 2 2 2

Islet Bay, Port Pegasus, Stewart Island, 47° 12.16’ S, 167° 38.20’ E, 2 m Solitary 1 1 0Quarantine Island, Otago Harbour, 45° 50’ S, 170° 39’ E, 2 m Solitary 3 3 3

Spirobranchus Quarantine Island, Otago Harbour, 45° 50’ S, 170° 39’ E, intertidal Solitary 4 4 4cariniferus Katiki Beach, North Otago, 45° 27’ S, 170° 48’ E, intertidal Solitary 2 2 2

Katiki Beach, North Otago, 45° 27’ S, 170° 48’ E, intertidal Aggregated 2 2 2Cave Rock, Sumner, Christchurch, 43° 34’ S, 172° 45’ E, intertidal Aggregated 4 0 4Cave Rock, Sumner, Christchurch, 43° 34’ S, 172° 45’ E, intertidal Solitary 3 0 3Hatfields Beach, Auckland, 36° 34’ S, 174° 42’ E, intertidal Solitary 3 3 2

Spirobranchus Puysegur Bank, southwest of South Island, 46° 34.10’ S, 165° 56.88’ E, 146 m Solitary 2 2 1latiscapus Puysegur Point, southwest of South Island, 46° 13.32’ S, 166° 28.59’ E, 264 m Solitary 2 2 0

Otago Shelf, east of South Island, 45° 47.33’ S, 171° 04.44’ E, 138 m Solitary 2 1 0

Protula bispiralis Bravo Island, Paterson Inlet, Stewart Island, 46° 57.96’ S, 168° 07.92’ E, 9 m Solitary 2 0 0

Salmacina sp. Otago Shelf, east of South Island, 45° 47.33’ S, 171° 04.44’ E, 138 m Solitary 2 2 0

Spirorbinae Islet Bay, Port Pegasus, Stewart Island, 47° 12.16’ S, 167° 38.20’ E, 2 m Solitary 1 0 0(unidentified)

Table 1. Serpulid worm species collected for this study, location and water depth of collection, and mode of life (aggregated or solitary), along with the number of individuals sequenced per gene. Hist: histone; Cyt b: cytochrome b

Page 4

Aquat Biol 16: 97–103, 2012

(‘nst = 6’) for each gene. For 18S and cytochrome b,rate variation across sites was modelled using agamma distribution, for which none of the sites wereinvariant (‘rates = gamma’). For histone H3 rate vari-ation across sites was modelled using a gamma distri-bution, with a proportion of the sites being invariant(‘rates = invgamma’). For the analyses the branchlength priors were set to ‘Unconstrained:Exponen-tial(100)’ to account for potential branch length esti-mation problems (see Brown et al. 2010). The Markovchain Monte Carlo searches were run twice with4 chains for 5 000 000 generations, and trees weresampled every 100 generations. Convergence of theduplicate runs was assessed both in Tracer v.1.4(Rambaut & Drummond 2007) and via the averageSD of split frequencies. Following this assessment,the first 10 000 trees, i.e. 1 000 000 generations, werediscarded as ‘burnin’ in each of the analyses.

Congruence with other measures of support wasevaluated using NJ and MP bootstrap analyses. TheNJ bootstrap analyses consisted of 10 000 replicatesusing the model selected by Modeltest. The equallyweighted MP bootstrap analyses consisted of 10 000replicates using a heuristic search (with TBR branch-swapping).

RESULTS AND DISCUSSION

Our alignments resulted in a 1884 bp fragmentof 18S (43 taxa), a 313 bp fragment of histone H3(30 taxa) and a 383 bp fragment of cytochrome b(58 taxa). In the 1884 bp fragment of 18S, 1345 bpwere constant, 539 bp were variable and 409 of thevariable characters were parsimony informative. Ofthe 313 bp fragment of histone H3, 205 bp were constant, 108 bp were variable and 98 of the vari-able characters were parsimony informative. Of the383 bp of cytochrome b, 147 bp were constant, 236 bpwere variable and 222 of the variable characterswere parsimony informative.

The Bayesian phylogenies (Fig. 2) for each of thegenes are broadly concordant with one another(while showing that each marker evolves at a differ-ent rate and, thus, has different levels of resolution).In each case (18S, histone H3 and cytochrome b)trees unequivocally place all our specimens of Gale-olaria hystrix monophyletically together (the poste-rior probabilities and bootstrap support values showthat this conclusion is well supported) irrespective ofcollection locality or mode of life. Similarly, Spiro-branchus cariniferus all belong to a single species; allspecimens group together and there is no difference

among aggregated and solitary individuals (i.e. thereis good posterior probability and bootstrap supportfor the monophyly of S. cariniferus).

Our genetic results conclusively disprove the hypo -thesis that ecological and lifestyle differences inthe southern serpulid worms Galeolaria hystrix andSpiro branchus cariniferus result from cryptic species.The variation in bathymetric distribution and distinc-tive modes of life exhibited by these species must beexplained using physiological and/or ecological dri-vers acting within species.

Interestingly, however, both the cytochrome b and18S trees show the Galeolaria hystrix individual collected from Adelaide, Australia, is separate fromthe New Zealand clade. As the type locality for G.hystrix is in New Zealand (Mörch 1863), this resultsuggests that the Australian species may require re-examination and possibly re-naming. In Australia,sequencing has split morphologically indistinguish-able ‘G. caespitosa’ into an eastern clade (renamedG. geminoea) and a southwestern clade (real G. cae-spitosa) (Halt et al. 2009), whose cytochrome b se -quences differ by >19%. The Australian G. hystrixdiffers from our New Zealand clade by an average of26% for the same gene. We note also that G. caespi-tosa and G. geminoea are both gregarious whereasthe Australian G. hystrix is generally solitary (Haltet al. 2009).

Understanding the factors affecting serpulid reef-building is important far beyond serpulid systemat-ics. Serpulid aggregations provide hard substrateand refugia for a wide variety of organisms andenhance biodiversity in the temperate waters wherethey occur (Haines & Maurer 1980, Moore et al. 1998,Haanes & Gulliksen 2011), as well as locally affectingwater flow and sedimentation (Schwindt et al. 2004).Serpulid aggregations affect the local environment inother ways: they clean the water by filtering out par-ticles as the worms feed (Davies et al. 1989); they alsoremove calcium carbonate from seawater for tubeproduction, acting as a carbon sink (Medernach et al.2000). Gregarious serpulids can thus be considered‘ecosystem engineers’ (sensu Jones et al. 1994); un -der standing the causes of gregariousness is impor-tant for protection of their fragile and disappearing(see e.g. Hughes 2011) temperate ecosystems.

Physical environmental influences on reef distribu-tion and growth have often been cited. A study ofSerpula vermicularis reefs in Loch Creran, Scotland,showed that reefs were most common where suitablesubstrate (bivalve shells) was available, and that theydid not occur in areas of strong currents (Moore et al.1998). Similarly, the distribution of reefs formed by

100

Page 5

Smith et al.: New Zealand serpulid reefs 101b

Salm

acin

a sp

. 1

Salm

acin

a sp

. 2

Prot

ula

sp. D

Q77

9761

P. tu

bula

ria E

F192

934

G. g

emin

eoa*

AF1

8523

3

Qua

rant

ine

Is. S

1

Brav

o Is

. A2

Katik

i Bea

ch S

1

Brav

o Is

. A1

Qua

rant

ine

Is. S

2

Port

Pega

sus

S1

Qua

rant

ine

Is. S

6

S. la

tisca

pus

Puys

eger

Ban

k 1

S. la

tisca

pus

Puys

eger

Ban

k 2

S. li

ma

EF19

2930

S. tr

ique

ter E

F192

932

S. la

tisca

pus

Puys

eger

Poi

nt 2

S. la

tisca

pus

Ota

go S

helf

1

S. la

tisca

pus

Puys

eger

Poi

nt 1

Katik

i Bea

ch A

2

Katik

i Bea

ch A

1

Auck

land

S2

Katik

i Bea

ch S

1

Auck

land

S1

Qua

rant

ine

Is. S

1

Qua

rant

ine

Is. S

6

Auck

land

S3

Qua

rant

ine

Is. S

5

Katik

i Bea

ch S

2

Qua

rant

ine

Is. S

2

1.00

100

100

0.99

99

98

Galeolaria hystrix

0.1

subs

titut

ions

/site

1.00

94

99

Spirobranchus cariniferus

0.94

100

1.00

76

99

0.99

98

100

0.84

0.58

84 0.

91 1

00 1

00

0.52

76

73

0.58

90

86

c

S. la

tisca

pus

Puys

eger

Ban

k 1

Sum

ner S

3Q

uara

ntin

e Is

. S5

Katik

i Bea

ch A

1Ka

tiki B

each

S2

Sum

ner A

2Su

mne

r A4

Katik

i Bea

ch A

2Au

ckla

nd S

1Ka

tiki B

each

S1

Auck

land

S3

Qua

rant

ine

Is. S

2Q

uara

ntin

e Is

. S1

Qua

rant

ine

Is. S

6Su

mne

r A3

Sum

ner A

1Su

mne

r S1

Sum

ner S

2

Aust

ralia

EU

2004

41Q

uara

ntin

e Is

. S2

Qua

rant

ine

Is. S

6Br

avo

Is. A

2Q

uara

ntin

e Is

. S1

Katik

i Bea

ch S

1Br

avo

Is. A

1G

. gem

ineo

a FJ

6465

48G

. gem

ineo

a FJ

6465

33G

. gem

ineo

a FJ

6465

36G

. gem

ineo

a* E

U18

4045

G. g

emin

eoa

FJ64

6535

G. g

emin

eoa*

EU

1840

42G

. gem

ineo

a FJ

6465

32G

. gem

ineo

a* E

U18

4039

G. g

emin

eoa

FJ64

6534

G. g

emin

eoa*

EU

1840

47G

. gem

ineo

a* E

U18

4041

G. g

emin

eoa*

EU

1840

43G

. gem

ineo

a* E

U18

4046

G. g

emin

eoa*

EU

1840

40G

. gem

ineo

a* E

U18

4044

G. c

aesp

itosa

FJ6

4654

5G

. cae

spito

sa F

J646

546

G. c

aesp

itosa

FJ6

4654

4G

. cae

spito

sa F

J646

547

G. c

aesp

itosa

FJ6

4654

3G

. cae

spito

sa E

U18

4050

G. c

aesp

itosa

EU

1840

53G

. cae

spito

sa F

J646

539

G. c

aesp

itosa

FJ6

4654

1G

. cae

spito

sa F

J646

537

G. c

aesp

itosa

EU

1840

48G

. cae

spito

sa F

J646

538

G. c

aesp

itosa

EU

1840

55G

. cae

spito

sa E

U18

4051

G. c

aesp

itosa

FJ6

4654

0G

. cae

spito

sa E

U18

4049

G. c

aesp

itosa

FJ6

4654

2G

. cae

spito

sa E

U18

4052

1.00

100

100

0.1

subs

titut

ions

/site

Spirobranchus cariniferusGaleolaria hystrix

1.00

100

100

0.73

62

0.82

81

1.00

75

98

1.00

92

100

1.00

99

100

0.99

58

100

0.78

63

1.00

100

1000.

60

0.61

57

86 69

57

aSp

irorb

inae

Prot

ula

bisp

iralis

2

Prot

ula

bisp

iralis

1

Salm

acin

a sp

. 2Sa

lmac

ina

sp. 1

G. g

emin

eoa*

AB1

0625

7

Aust

ralia

DQ

3148

39

Qua

rant

ine

Is. S

1Br

avo

Is. A

1

Katik

i Bea

ch S

1

Brav

o Is

. A2

Qua

rant

ine

Is. S

2

Port

Pega

sus

S1

Qua

rant

ine

Is. S

6

S. c

orni

cula

tus D

Q14

0410

S. c

orni

cula

tus E

U19

5381

S. la

tisca

pus

Puys

eger

Ban

k 1

S. la

tisca

pus

Puys

eger

Ban

k 2

S. la

tisca

pus

Ota

go S

helf

2

S. la

tisca

pus

Puys

eger

Poi

nt 2

S. la

tisca

pus

Ota

go S

helf

1S.

latis

capu

s Pu

yseg

er P

oint

1

S. tr

ique

ter D

Q31

7121

S. la

mar

cki D

Q14

0404

S. li

ma

DQ31

7130

Qua

rant

ine

Is. S

6

Auck

land

S1

Auck

land

S3

Katik

i Bea

ch A

1Ka

tiki B

each

S1

Sum

ner A

1

Sum

ner A

2Q

uara

ntin

e Is

. S2

Sum

ner S

1

Sum

ner A

4

Sum

ner S

2Q

uara

ntin

e Is

. S5

Sum

ner A

3

Auck

land

S2

Katik

i Bea

ch S

2

Katik

i Bea

ch A

2

Qua

rant

ine

Is. S

1Su

mne

r S3

1.00

100

100 1.

00 1

00 1

00

1.00

100

99

1.00

100

100

1.00

100

100

Galeolaria hystrix

1.00

100

100

1.00

99

1.00

100

100

1.00

100

100 1.

00 9

9 99

0.95

50

0.89

1.00

80

1.00

64 94

0.99

53

0.99

53

1.00

100

100

0.1

subs

titut

ions

/site

Spirobranchus cariniferus

Fig

. 2. G

aleo

lari

a h

ystr

ixan

d S

pir

obra

nch

us

cari

nif

eru

s. B

ayes

ian

ph

ylog

ram

s fo

r (a

) 18

S, (

b)

his

ton

e H

3 an

d (

c) c

ytoc

hro

me

b. A

ll 3

tre

es s

how

th

at G

. hys

trix

(ora

ng

eb

ar) a

nd

S. c

arin

ifer

us

(blu

e b

ar) a

re m

onop

hyl

etic

(i.e

. dis

cret

e an

d c

oher

ent s

pec

ies)

. Acc

essi

on n

um

ber

s in

the

taxo

n la

bel

s in

dic

ate

the

seq

uen

ces

not

from

this

stu

dy,

obta

ined

fro

m G

enB

ank

. T

he

nu

mb

ers

asso

ciat

ed w

ith

th

e b

ran

ches

rep

rese

nt

pos

teri

or p

rob

abil

itie

s (≥

0.5)

an

d b

oots

trap

s va

lues

(≥5

0%

), w

ith

nei

gh

bou

r-jo

inin

g i

nb

old

tex

t an

d m

axim

um

par

sim

ony

in it

alic

s. F

or t

he

G. h

ystr

ixan

d S

. car

inif

eru

ssa

mp

les

we

seq

uen

ced

, th

e co

llec

tion

loca

tion

is s

how

n (

see

Tab

le 1

for

mor

e in

form

a-ti

on).

A (

in t

he

red

fon

t la

bel

ed t

axa)

mea

ns

agg

reg

ate

and

S (

kh

aki f

ont)

mea

ns

soli

tary

; th

e n

um

ber

s re

fer

to t

he

ind

ivid

ual

seq

uen

ced

an

d a

re r

efer

ence

d in

th

e G

en-

Ban

k a

cces

sion

s. G

. g

emin

eoa

show

n w

ith

an

ast

eris

k (

*) w

ere

pre

viou

sly

des

ign

ated

as

G.

caes

pit

osa

(e.g

. S

tyan

et

al.

2008

), w

ith

G.

gem

ineo

ab

ein

g s

ub

seq

uen

tly

erec

ted

an

d u

sed

for

th

e d

isti

nct

eas

tern

Au

stra

lian

tax

on (

Hal

t et

al.

200

9)

Page 6

Aquat Biol 16: 97–103, 2012

Ficopomatus enigmaticus in Mar Chiquita Lagoon,Argentina, appears to be affected by current speed,salinity and nutrient availability (Schwindt et al.2004). Human activity, particularly anchor and moor-ing damage, can dramatically re duce serpulid reefabundance (Schwindt et al. 2004, Hughes 2011).

Biotic controls are less frequently invoked. Forintertidal belt-forming (occupying a zone of the inter-tidal) species, ten Hove (1979) suggested that aggre-gation in serpulids might result from competition forfood and space, at least at the lower limit. We knowof no study that demonstrates biotic factors influenc-ing serpulid reef distribution.

While physico-chemical environmental factors mayinfluence growth of existing reefs, what exactlycauses serpulid reef-forming to begin with? The mostlikely explanation is gregarious behaviour in larvae,perhaps influenced by chemical or physical attrac-tion by adults (Kupriyanova et al. 2001). Serpulid lar-vae may remain in the water column up to 2 mo andare able to delay settlement under unsuitable condi-tions (ten Hove 1979). In some species, larvae maypreferentially settle in areas where adult populationsare dense, though only where tubes are whole; damaged tubes appear to repel larvae (summary inKupriyanova et al. 2001). It is likely that larval be -haviour, adult attractants and biotic or physical environmental factors combine to enable the onsetof reef-forming in serpulid worms, though we cannotdis count the possibility that genetic differenceswithin species are also involved, perhaps underpin-ning some of the above factors.

The hypothesis that the different modes of life inthese 2 species reflect the existence of cryptic speciesis conclusively disproved here. Some combination oflarval behaviour, adult attractants, and other biotic orphysical factors must be the drivers of reef formation,distribution and growth in these serpulid worms.

Acknowledgements. We acknowledge financial supportfrom University of Otago summer studentship scheme.Specimens were collected by M. A. Riedi, G. O’Sullivan andV. C. Davis, with assistance from B. Dickson, P. Meredith, P.Heseltine and C. Garden. Some photos were provided by M.A. Riedi; K. Miller assisted with figure design. R. Hankeyand H. A. ten Hove kindly advised us on matters of Latingrammar; D. Lee and B. Marshall helped with ICZN no -menclatural rules. We acknowledge the contribution andimprovement by anonymous reviewers.

LITERATURE CITED

Altschul SF, Gish W, Miller W, Myers EW, Lipman DJ (1990)Basic logical alignment search tool. J Mol Biol 215:403−410

Bastida-Zavala JR, ten Hove HA (2002) Revision of Hydro -ides Gunnerus, 1768 (Polychaeta: Serpulidae) from theWestern Atlantic region. Beaufortia 52:103−178

Bosence DWJ (1973) Recent serpulid reefs, Connemara,Eire. Nature 242:40−41

Brown JM, Hedtke SM, Lemmon AR, Lemmon EM (2010)When trees grow too long: investigating the causes ofhighly inaccurate Bayesian branch-length estimates. SystBiol 59:145−161

Colgan DJ, McLauchlan A, Wilson GDF, Livingston SP andothers (1998) Histone H3 and U2 snRNA DNA sequencesand arthropod molecular evolution. Aust J Zool 46:419−437

Davies BR, Stuart V, de Villiers M (1989) The filtration activ-ity of a serpulid polychaete population (Ficopomatusenigmaticus (Fauvel) and its effects on water quality in acoastal marina. Estuar Coast Shelf Sci 29:613−620

Felsenstein J (1985) Confidence limits on phylogenies: anapproach using the bootstrap. Evolution 39:783−791

Gray JE (1843) Chapter 10. Fauna of New Zealand. In: Dief-fenbach E (ed) Travels in New Zealand; with contributionsto the geography, geology, botany, and natural history ofthat country, Vol 2. John Murray, London, p 177−296

Haanes H, Gulliksen B (2011) A high local species richnessand biodiversity within high-latitude calcareous aggre-gates of tube-building polychaetes. Biodivers Conserv20:793−806

Haines JL, Maurer D (1980) Benthic invertebrates associ-ated with a serpulid polychaete assemblage in a temper-ate estuary. Int Rev Gesamten Hydrobiol Hydrograph65:643−656

Halt MN, Kupriyanova EK, Cooper SJB, Rouse GW (2009)Naming species with no morphological indicators: spe-cies status of Galeolaria caespitosa (Annelida: Serpuli-dae) inferred from nuclear and mitochondrial genesequences and morphology. Invertebr Syst 23:205−222

Herbert DG (1993) Revision of the Trochinae, tribe Trochini(Gastropoda: Trochidae) of southern Africa. Natal MusAnn 43:239−308

Huelsenbeck JP, Ronquist F (2001) MRBAYES: Bayesianinference of phylogeny. Bioinformatics 17:754−755

Hughes DJ (2011) Where’s the ‘reef’? A five year study ofserpulid tube bioerosion in a Scottish sea loch. Mar EcolProg Ser 430:273−280

Jones CG, Lawton JH, Shachak M (1994) Organisms asecosystem engineers. Oikos 69:373−386

Kupriyanova EK, Nishi E, ten Hove HA, Rzhavsky AV (2001)Life-history patterns in serpulimorph polychaetes: eco-logical and evolutionary perspectives. Oceanogr MarBiol Annu Rev 39:1−100

Kupriyanova EK, Macdonald TA, Rouse GW (2006) Phyloge-netic relationships within Serpulidae (Sabellida, Annel-ida) inferred from molecular and morphological data.Zool Scr 35:421−439

Medernach E, Jordana A, Grémare A, Nozais C, Charles F,Amouroux JM (2000) Population dynamics, secondaryproduction and calcification in a Mediterranean popula-tion of Ditrupa arietina (Annelida: Polychaeta). Mar EcolProg Ser 199:171−184

Moore CG, Saunders GR, Harries DB (1998) The status andecology of reefs of Serpula vermicularis L. (Polychaeta:Serpulidae) in Scotland. Aquat Conserv 8:645−656

Mörch OAL (1863) Revisio critica Serpulidarum. Et bidrag tilrørormenes naturhistorie. Naturhistorisk Tidsskrift Hen-rik Krøyer, København (Ser 3) 1:347−470

102

Page 7

Smith et al.: New Zealand serpulid reefs

Morton J, Miller M (1973) The New Zealand sea shore, 2ndedn. Collins, London

Nardo GD (1847) Prospetto della fauna marina volgare delVeneto Estuario con cenni sulle principali specie com-mestibili dell’Adriatico, sulle venete pesche, sulle valli,ecc. In: Venezia e le sue lagune. G Antonelli, Venezia,p 113−156 (p 1−45 in reprint)

Norén M, Jondelius U (1999) Phylogeny of Prolecithophora(Platyhelminthes) inferred from 18S rDNA sequences.Cladistics 15:103−112

Posada D, Crandall KA (1998) Modeltest: testing the modelof DNA substitution. Bioinformatics 14:817−818

Rambaut A, Drummond AJ (2007) Tracer, version 1.4. Uni-versity of Edinburgh

Ronquist F, Huelsenbeck JP (2003) MRBAYES 3: Bayesianphylogenetic inference under mixed models. Bioinfor-matics 19:1572−1574

Schwindt E, De Francesco CG, Iribarne OO (2004) Indi -vidual and reef growth of the invasive reef-buildingpolychaete Ficopomatus engimaticus in a south-westernAtlantic coastal lagoon. J Mar Biol Assoc UK 84:987−993

Smith AM, McGourty CR, Kregting L, Elliot A (2005) Sub -

tidal Galeolaria hystrix (Polychaeta: Serpulidae) reefs inPaterson Inlet, Stewart Island, New Zealand. N Z J MarFreshw Res 39:1297−1304

Styan CA, Kupriyanova E, Havenhand JN (2008) Barriers tocross-fertlization between populations of a widely dis-persed polychaete species are unlikely to have arisenthrough gametic compatibility arms-races. Evolution 62:3041–3055

Swofford DL (2002) PAUP*. Phylogenetic analysis using parsimony (*and other methods). Sinauer Associates,Sunderland, MA

ten Hove HA (1979) Different causes of mass occurrence inSerpulids. In: Larwood G, Rosen BR (eds) Biology andsystematics of colonial organisms. Systematics Asso -ciation Spec Vol 11. Academic Press, London, p 281−298

ten Hove HA, Kupriyanova EK (2009) Taxonomy of Serpuli-dae (Annelida, Polychaeta): the state of affairs. Zootaxa2036:1−126

ten Hove HA, van den Hurk P (1993) A review of recent andfossil serpulid ‘reefs’; actuopalaeontology and the ‘UpperMalm’ serpulid limestones in NW Germany. Geol Mijnb72:23−67

103

Editorial responsibility: Judith Grassle, New Brunswick, New Jersey, USA

Submitted: February 22, 2012; Accepted: May 4, 2012Proofs received from author(s): June 15, 2012