decoupling of short- and long-distance dispersal pathways in the endemic new zealand seaweed...
TRANSCRIPT
DECOUPLING OF SHORT- AND LONG-DISTANCE DISPERSAL PATHWAYS INTHE ENDEMIC NEW ZEALAND SEAWEED CARPOPHYLLUM MASCHALOCARPUM
(PHAEOPHYCEAE, FUCALES)1
Joe Buchanan2 and Giuseppe C. Zuccarello
School of Biological Sciences, Victoria University of Wellington, P. O. Box 600, Wellington 6140, New Zealand
The processes that produce and maintain geneticstructure in organisms operate at different time-scales and on different life-history stages. In marinemacroalgae, gene flow occurs through gamete ⁄ zygotedispersal and rafting by adult thalli. Populationgenetic patterns arise from this contemporary geneflow interacting with historical processes. We ana-lyzed spatial patterns of mitochondrial DNA varia-tion to investigate contemporary and historicaldispersal patterns in the New Zealand endemicfucalean brown alga Carpophyllum maschalocarpum(Turner) Grev. Populations bounded by habitat dis-continuities were often strongly differentiated fromadjoining populations over scales of tens of kilome-ters and intrapopulation diversity was generally low,except for one region of northeast New Zealand (theBay of Plenty). There was evidence of strong connec-tivity between the northern and eastern regions ofNew Zealand’s North Island and between the Northand South Islands of New Zealand and the ChathamIslands (separated by 650 km of open ocean). Mod-erate haplotypic diversity was found in ChathamIslands populations, while other southern popula-tions showed low diversity consistent with LastGlacial Maximum (LGM) retreat and subsequentrecolonization. We suggest that ocean current pat-terns and prevailing westerly winds facilitate long-distance dispersal by floating adult thalli, decouplinggenetic differentiation of Chatham Island popula-tions from dispersal potential at the gamete ⁄ zygotestage. This study highlights the importance ofencompassing the entire range of a species wheninferring dispersal patterns from genetic differentia-tion, as realized dispersal distances can be contin-gent on local or regional oceanographic andhistorical processes.
Key index words: Carpophyllum maschalocarpum;dispersal; floating algae; mitochondrial DNA;New Zealand; Phaeophyceae; phylogeography
Abbreviations: AMOVA, analysis of molecular var-iation; CTAB, cetyltrimethylammonium bromide;LGM, Last Glacial Maximum; SAMOVA, spatialanalysis of molecular variance; SSCP, single-
stranded conformational polymorphism; SST,sea surface temperature; TBE, tris-borate-EDTA
Multiple dispersal mechanisms are found in manyorganisms. In many plant species, most seed dispersalis limited to local populations, but alternativedispersal mechanisms connect distant populations(Higgins et al. 2003). Short dispersal distances are alsofound in many sessile or demersal marine species, butoccasional long-distance dispersal events connect pop-ulations across wide habitat disjunctions (Reed et al.1988, Pakker et al. 1996, Fraser et al. 2009). Dispersaldistance in marine organisms is often correlated withthe duration of a pelagic larval stage (Bohonak 1999),but many marine organisms have the potential to dis-perse as adults, either because of inherent buoyancyor by rafting on floating substrata (e.g., drift wood,algae, or pumice) (Thiel and Gutow 2005a).
Long-distance dispersal mechanisms differ fromshort-distance mechanisms in two ways: Firstly, long-distance dispersal mechanisms are often mediatedby large-scale processes (climate patterns, ocean cur-rents, seasonal migrations of animal vectors), ratherthan local conditions (turbulence, wave regimes,inshore currents). Secondly, long-distance dispersalevents are rare, stochastic, and will seldom beobserved directly, but will become evident in a spe-cies’ range or population genetic structure overlong timescales. Consequently, factors that mediatelong-distance dispersal are decoupled from factorsthat mediate most observed dispersal (Kinlan et al.2005, Thiel and Haye 2006).
Dispersal by floating thalli might be an importantdispersal mechanism for macroalgae (Deysher andNorton 1982, van den Hoek 1987, Thiel and Gutow2005b, Fraser et al. 2009). Floating macroalgal thalliare abundant and can release viable propagules fora considerable time (Kingsford 1992, Smith 2002,Macaya et al. 2005, Hernandez-Carmona et al. 2006,McKenzie and Bellgrove 2008), allowing dispersalbetween populations separated by habitat disjunc-tions such as soft sediment coasts or large stretchesof open water (Dethier et al. 2003). Genetic evi-dence for long-distance dispersal in macroalgaeacross the open ocean is accumulating but still rare(Tatarenkov et al. 2007, Fraser et al. 2009).
1Received 2 December 2009. Accepted 27 September 2011.2Author for correspondence: e-mail: [email protected].
J. Phycol. 48, 518–529 (2012)� 2012 Phycological Society of AmericaDOI: 10.1111/j.1529-8817.2012.01167.x
518
Fucalean algae have a monophasic life historywith dispersal from maternal thalli by nonmotile,negatively buoyant eggs or zygotes (Kendrick andWalker 1991, 1995, Norton 1992, Reed et al. 1992,Chapman 1995). Strong genetic population differ-entiation is predicted from the very limited dispersalpotential at this stage. Population genetic studies ofNorthern Hemisphere Fucaceae have shown popula-tion differentiation (Coyer et al. 2003, Coleman andBrawley 2005a, Hoarau et al. 2007, Tatarenkov et al.2007), but often at larger scales than predicted fromdirect measurements of propagule dispersal. Long-distance dispersal by detached, floating thalli isoften invoked to explain population connectivity atlarge scales (van den Hoek 1987, Coleman andBrawley 2005b, Thiel and Gutow 2005b), especiallywhere local population differentiation is high (Muhlinet al. 2008, Coleman and Kelaher 2009),
More generally, the use of molecular markers hasuncovered higher than expected genetic populationstructure in marine organisms. Currents, upwelling,habitat disjunctions, and larval behavior limit dis-persal even for organisms with long pelagic duration(Sotka et al. 2004). Historical events are also evidentin genetic structure. Marko (2004), for example,found different biogeographic histories in two spe-cies of the intertidal gastropod Nucella in the north-eastern Pacific, despite similar dispersal modes.Retreat from high latitudes following glacial cooling,followed by recolonization from either northern orsouthern refugia is a common pattern (Hoarauet al. 2007, Maggs et al. 2008), although marine spe-cies appear to show more variable responses to glo-bal cooling than terrestrial species (Wares andCunningham 2001, Marko et al. 2010). In NewZealand, genetic population structure has beenshown in marine species (Hickey et al. 2009, Rosset al. 2011), alongside current mediated dispersal.Historical signatures of southern glaciation duringthe LGM have been observed in terrestrial (Marskeet al. 2011) and marine species (Hickey et al. 2009).
C. maschalocarpum is a common fucalean alga thatis endemic to New Zealand. All Carpophyllum speciesare dioecious. Female thalli produce large (c.200 lm diameter) nonmotile eggs that are extrudedto the surface of receptacles but retained by a muci-laginous stalk during fertilization (Naylor 1954,Clayton 1992). Following fertilization, zygotesremain attached to the maternal thallus for up to28 d before release (Delf 1939, Dromgoole 1973).As eggs and zygotes are negatively buoyant, the spe-cies is expected to have low dispersal ability at theselife stages, but long-distance dispersal is possible asadult thalli usually possess pneumatocysts and arefrequently found floating on the sea surface (Kings-ford 1992). Floating female thalli bearing zygotes orgermlings could establish new populations in theabsence of fertile males.
C. maschalocarpum forms extensive stands in theshallow subtidal areas of New Zealand’s rocky coasts,
where it is the second most abundant algal speciesby biomass and an important structuring organism(Shears and Babcock 2007). The species is presentaround both main islands of New Zealand and theChatham Islands but is absent from the southeast ofthe South Island and the sub-Antarctic islands (Mor-ton and Miller 1968, Nelson et al. 1991, Schiel et al.1995, Shears and Babcock 2007).
C. maschalocarpum’s extensive range, potentialmultiple dispersal modes, ecological importance,and ubiquity in rocky habitats make it an idealorganism for investigating phylogeographic pro-cesses. The species’ range is interrupted by signifi-cant habitat breaks, including extensive softsediment coast around both main islands of NewZealand and stretches of deep water between theNorth and South Islands (Cook Strait), and betweenthe mainland and the Chatham Islands, 650 km eastof the mainland (Fig. 1) (Morton and Miller 1968,Nelson 1994, Shears et al. 2008). The factors medi-ating the two mechanisms of dispersal should be amajor determinant of genetic structure in C. masc-halocarpum. If dispersal by floating thalli is rare, weexpect high population differentiation, especially
EAUC
ECC
SC
C H A T MH A I SR E
Bay ofPlenty
North Cape
HawkeBay
CookStrait
NORTHIS
LAND
S OUT
HI SLAND
East Cape
ChathamIslands
Northland
Taranaki
175°E
40°
N
200 km
15°C
15°C LGM
15°C LGM
Fiordland
Banks Peninsula
15°C
ABEL
COOK
PORTLAND
RAGLAN
NORTH-EASTERN
BANKS
STEWART IS.
FIORDLAND
WESTLAND
BULLER
FIG. 1. New Zealand map showing major currents (Heath1985) and locations referred to in the text. Boundaries of biore-gions proposed by Shears et al. (2008) are indicated by dashedlines and named in italics. EAUC: East Auckland Current; ECC:East Cape Current; SC: Southland Current. Dashed lines showpresent day 15�C sea surface temperature isotherm and estimated15�C isotherm for the Last Glacial Maximum (LGM) (Barrowsand Juggins 2005).
DISPERSAL IN CARPOPHYLLUM 519
between populations separated by discontinuities inhabitat. If floating dispersal is common, we expectconnectivity mediated between current and windpatterns.
The aim of this study was to determine if the popu-lation genetic variation of C. maschalocarpum is con-gruent with dispersal modes. Specifically, we compareconnectivity between adjacent (tens of kilometers)and distant (hundreds of kilometers) populations,test for congruence between phylogeographicstructure and biogeographic provinces proposed byShears et al. (2008) (Fig. 1), and test for genetic isola-tion by distance and population expansion.
MATERIALS AND METHODS
Sampling and DNA extraction. Specimens identified in thefield as C. maschalocarpum were collected from 32 sites aroundNew Zealand (Table 1), representing the entire range ofC. maschalocarpum with the exception of Fiordland, where thespecies is relatively rare (Shears and Babcock 2007). Thalli werecollected haphazardly, at least 1 m apart. Tips of thalli for DNAextraction were rapidly dried and stored in silica gel. Inaddition, one or more typical thalli from each species ⁄ site wereprepared as voucher specimens. Five Carpophyllum angustifoliumJ. Agardh specimens and a number of putative C. angustifoliumX C. maschalocarpum hybrid thalli were included as outgroupspecimens or to assist in genetic identification of hybridspecimens. DNA was extracted from �2–5 mg of dry tissueusing a modified cetyltrimethylammonium bromide (CTAB)buffer procedure (Zuccarello and Lokhorst 2005), with theaddition of 1% polyvinylpolypyrrolidone to the extractionbuffer.
PCR conditions and primers. We amplified the 23S-tRNA Lysmitochondrial spacer (225 bp) from a total of 651 samples,using primers (forward: fuc2625F 5¢-GCTGTGAGGTTTTTAGCTGACC-3¢; reverse: fuc3141R 5¢-TCCACCACTCTAACCAACTGAG-3¢) designed from alignments of the Fucus, Laminaria, andDesmarestia mitochondrial genomes (Oudot-Le Secq et al. 2002,2006).
PCR amplifications used a touchdown routine with an initialdenaturation step of 94�C for 4 min, followed by 5 cycles of1 min at 94�C, 1 min at 55�C ()1�C ⁄ cycle) and 1 min at 72�C,followed by 30 cycles of 1 min at 94�C, 1 min at 50�C and 1 minat 72�C, with a final extension of 72�C for 10 min. The PCRmix contained 1 lL genomic DNA, 0.5 U Taq DNA polymerase(New England Biolabs, Beverly, MA, USA) 1· ThermoPolreaction buffer (NEB), 7.5 pmoles of each primer, 200 nmolesdNTP, 5% DMSO, and 0.01% BSA. Amplified products werechecked for length and yield on 1% agarose gels stained withethidium bromide.
Populations were screened for variable haplotypes by single-stranded conformational polymorphism (SSCP) (Zuccarelloet al. 1999, Sunnucks et al. 2000). A quantity of 3–4mL of PCRproduct was mixed with 9 lL 98% formamide, 10 mM NaOH,0.025% bromophenol blue, and 0.025% xylene cyanol. Sam-ples were denatured at 95�C–100�C for 5 min, then snapcooled on ice before loading. Gels (225 mm long and 0.75 mmthick, D-Code System, BioRad, Hercules, CA, USA) contained9% 37.5:1 acrylamide ⁄ bis-acrylamide (Sigma Aldrich, St. Louis,MO, USA) in 0.5· Tris-borate-EDTA (TBE) buffer with theaddition of 10% glycerol. Electrophoresis was carried out for18–20 h at 7W in 0.5· TBE buffer at 4�C. After electrophoresis,gels were silver stained following protocols in Bassam et al.(1991), and banding patterns scored by eye. One or more ofeach haplotype indicated by SSCP was sequenced from eachpopulation, as well as any samples giving ambiguous SSCP
profiles. Amplification for sequencing used the same PCRconditions and reverse primer as above but used a novel forwardprimer (fuc2512F: 5¢-CCGGGTAGCTACATCGAGAA-3¢) thatanneals �100 bp further toward the 5¢ end of the fragment,to ensure accurate reads of the 5¢ end of the spacer.
Generally, SSCP banding and sequencing results wereconcordant. In very few cases, haplotypes were not easilydistinguished by SSCP, and all samples were sequenced. PCRproducts were cleaned with ExoSAP-IT (USB, Cleveland, OH,USA) enzymes and sequenced commercially (Macrogen Inc.,Seoul, South Korea).
Data analysis. Sequences from 210 samples were aligned inMEGA 3.1 (Kumar et al. 2004) and checked by eye. Alignmentwas straightforward, requiring insertion of a single gap toaccommodate outgroups. Haplotype networks were generatedin TCS 1.21 (Clement et al. 2000). Arlequin 3.11 (Excoffieret al. 2005) was used to generate summary population geneticstatistics and to perform neutrality tests (Tajima’s D, Fu andLi’s F * and D*, and Fu’s Fs). Neutrality indices were testedagainst 2,000 coalescence simulations assuming demographicstability and neutrality.
Pair-wise Fst values were calculated in Arlequin 3.11 for Bayof Plenty populations (in which shared haplotypes werereasonably frequent) and tested for significance by 1,023permutations. Mantel tests, also implemented in Arlequin 3.11,were used to determine isolation by distance, first for allpopulations, then for mainland populations (excluding Chat-ham Island populations). A matrix of shortest over-watergeographic distances between populations was assembled andtested for correlation with pair-wise Fst, with significance testedagainst 2,000 permutations. A second test was made using onlythe north–south component of these distances.
Analyses of molecular variation (AMOVAs), implemented inArlequin 3.11 were used to compare population partitionsfound in C. maschalocarpum with proposed marine biogeo-graphic regions (Shears et al. 2008). These were compared togenetic structure indicated by spatial analysis of molecularvariance (SAMOVA) (Dupanloup et al. 2002) implemented inSAMOVA 1.0. This uses a simulated annealing process to groupadjoining populations into K-groups so as to maximize geneticdistance between groups (maximum Fct), while minimizingdistance within groups (minimum Fsc). K was iterated insuccessive runs until a maximum Fct was calculated, andsignificance of fixation indexes tested by 1,000 permutations.
Mismatch distribution analysis (Rogers and Harpending1992), implemented in Arlequin 3.11, was used to test forsignatures of spatial expansion in frequency distributions ofpair-wise sequence differences. Analyses were applied to thetotal data and separately to populations partitioned into twogroups as indicated by SAMOVA and tested against 5,000coalescent simulations assuming spatial expansion. Resultswere interpreted in the context of simulations using stepping-stone and infinite-island models. Here, unimodal frequencydistributions of pair-wise mutational distances are regarded asevidence for past population expansions, but this signature isweakened or removed where populations are subdivided withlow migration rates (Ray et al. 2003, Excoffier 2004).
RESULTS
Haplotypic diversity and population structure. SSCPand subsequent sequencing identified 67 haplotypes(GenBank numbers HM070070 to HM070166).While SSCP might underestimate haplotypic diver-sity, sequencing of a large proportion of samplesshould reduce the likelihood of significant underes-timation of diversity. A statistical parsimony network
520 JOE BUCHANAN AND GIUSEPPE C. ZUCCARELLO
Table
1.
Geo
grap
hic
loca
tio
ns,
nu
mb
ers
of
spec
imen
ssa
mp
led
,mtD
NA
spac
erh
aplo
typ
ed
esig
nat
ion
s,an
dp
op
ula
tio
nd
iver
sity
ind
ices
for
Car
poph
yllu
mm
asch
aloc
arpu
m.
Hyp
hen
sin
gen
ed
iver
sity
and
nu
cleo
tid
ed
iver
sity
colu
mn
sin
dic
ate
po
pu
lati
on
sw
her
esu
bst
anti
alin
tro
gres
sio
nb
yC
.an
gust
ifol
ium
was
det
ecte
dan
dva
lues
wer
en
ot
cal-
cula
ted
.
Map
IDSi
ten
Lat
itu
de
Lo
ngi
tud
eH
erb
ariu
mn
um
ber
of
exem
pla
rsH
aplo
typ
esp
rese
nt
(n)
Gen
ed
iver
sity
h±
SDN
ucl
eoti
de
div
ersi
typ
±SD
1H
oo
per
Po
int,
No
rth
lan
d47
34�2
4¢56
¢¢S
172�
51¢1
3¢¢E
WE
LT
A03
1841
21(1
0),
34(3
4),
65(1
),66
(1),
67(1
)–
–
2N
ort
hC
ape,
No
rth
lan
d10
34�2
4¢57
¢¢S
173�
03¢0
9¢¢E
WE
LT
A03
1664
30(8
),31
(1),
32(1
)–
–3
Mat
aiB
ay,
No
rth
lan
d18
34�4
9¢37
¢¢S
173�
24¢5
4¢¢E
WE
LT
A03
1840
68(1
8)0
04
Ah
ipar
a,N
ort
hla
nd
2535
�10¢
10¢¢
S17
3�06
¢30¢
¢EW
EL
TA
0318
3940
(1),
52(2
4)0.
0800
±0.
0722
0.00
04±
0.00
085
Wek
aru
a,N
ort
hla
nd
1934
�56¢
42¢¢
S17
3�39
¢12¢
¢EW
EL
TA
0316
6530
(7),
33(1
),47
(7),
50(4
)–
6C
ape
Wiw
iki,
No
rth
lan
d21
35�0
9¢22
¢¢S
174�
07¢2
0¢¢E
WE
LT
A03
1835
21(2
1)0
07
Bla
nd
Bay
,B
ayo
fIs
lan
ds
2035
�20¢
31¢¢
S17
4�21
¢58¢
¢EW
EL
TA
SG29
013
(20)
00
8P
oo
rK
nig
hts
Isla
nd
s10
35�2
7¢40
¢¢S
174�
44¢2
0¢¢E
WE
LT
A03
1311
,A
0313
17,
A03
1319
21(1
0)0
0
9W
ater
fall
Ree
f,L
eigh
,N
ort
hla
nd
1136
�16¢
08¢¢
S17
4�47
¢53¢
¢EW
EL
TA
0313
18,
A03
1320
01(2
),13
(8),
36(1
)0.
4727
±0.
1617
0.00
67±
0.00
49
10M
ath
eso
nB
ay,
No
rth
lan
d34
36�1
8¢05
¢¢S
174�
47¢5
8¢¢E
WE
LT
A03
1932
01(2
),13
(27)
,55
(1),
56(1
),57
(2),
66(1
)0.
3708
±0.
1042
0.00
63±
0.00
44
11Sa
ilo
rsG
rave
,B
ayo
fP
len
ty20
36�5
7¢37
¢¢S
175�
50¢4
3¢¢E
WE
LT
A03
1312
–16,
A03
1331
20(1
),21
(1),
25(2
),27
(1),
28(1
3),
29(1
),70
(1)
0.58
42±
0.12
700.
0084
±0.
0056
12O
tan
ga,
Bay
of
Ple
nty
2537
�32¢
59¢¢
S17
8�09
¢41¢
¢EW
EL
TA
0318
30–3
114
(3),
16(3
),21
(2),
22(1
),26
(3),
28(9
),39
(3),
43(1
)
0.83
67±
0.05
390.
0152
±0.
0090
13M
arae
hak
oB
ay,
Bay
of
Ple
nty
2537
�40¢
20¢¢
S17
7�47
¢48¢
¢EW
EL
TA
0318
32–3
3,W
EL
TA
0318
1814
(15)
,15
(3),
16(2
),17
(1),
18(1
),19
(2),
23(2
),38
(2),
71(1
),72
(1)
0.74
25±
0.08
130.
0065
±0.
0045
14M
aket
u,
Bay
of
Ple
nty
2437
�44¢
41¢¢
S17
6�28
¢13¢
¢EW
EL
TA
0313
3358
(21)
,59
(1),
60(2
)0.
2355
±0.
1093
0.00
17±
0.00
1915
Op
ape,
Bay
of
Ple
nty
2137
�58¢
20¢¢
S17
7�25
¢15¢
¢EW
EL
TA
0318
3401
(1),
14(1
9),
69(1
),0.
1857
±0.
1102
0.00
21±
0.00
2116
Ho
roer
a,E
ast
Cap
e24
37�3
8¢17
¢¢S
178�
28¢2
9¢¢E
WE
LT
A03
1310
21(1
9),
52(1
),61
(1),
62(1
),64
(2)
0.37
68±
0.12
240.
0050
±0.
0038
17W
hal
eB
ay,
Wai
kato
2037
�49¢
22¢¢
S17
4�48
¢02¢
¢EW
EL
TA
0313
2701
(17)
,06
(1),
35(1
),37
(1)
0.28
42±
0.12
840.
0013
±0.
0016
18K
irit
eher
e,W
aika
to20
38�1
9¢31
¢¢S
174�
42¢0
7¢¢E
WE
LT
A02
4176
01(1
7),
02(1
),04
(1),
05(1
)0.
2842
±0.
1284
0.00
22±
0.00
22
19K
aiti
Bea
ch,
Gis
bo
rne
2338
�40¢
56¢¢
S17
8�01
¢46¢
¢EW
EL
TA
0313
3221
(21)
,63
(1),
67(1
)0.
1700
±0.
1025
0.00
54±
0.00
4020
Wai
pat
iki,
Haw
kes
Bay
2339
�18¢
19¢¢
S17
6�58
¢29¢
¢EW
EL
TA
0313
24,
A03
1325
08(2
1),
11(2
)0.
1660
±0.
0976
0.00
15±
0.00
17
21A
hu
riri
St,
Po
rtN
apie
r15
39�2
8¢42
¢¢S
176�
54¢1
3¢¢E
WE
LT
A03
1323
08(1
5)0
022
Pih
ama,
Sou
thT
aran
aki
2139
�31¢
17¢¢
S17
3�54
¢55¢
¢EW
EL
TA
0313
21,
A03
1322
,A
0313
3001
(18)
,07
(3)
0.25
71±
0.11
040.
0011
±0.
0015
23P
uke
rua
Bay
,W
elli
ngt
on
2041
�01¢
29¢¢
S17
4�54
¢06¢
¢EW
EL
TA
0313
0901
(20)
00
24R
iver
sdal
e,W
aira
rap
a15
41�0
6¢25
¢¢S
176�
04¢1
1¢¢E
WE
LT
A03
1308
01(1
2),
12(3
)0.
3429
±0.
1278
0.00
46±
0.00
3625
Wh
aran
ui,
Kai
kou
ra23
41�5
5¢20
¢¢S
174�
06¢0
0¢¢E
WE
LT
A03
1822
01(2
3)0
026
Wai
pap
aP
oin
t,K
aiko
ura
1742
�12¢
41¢¢
S17
3�52
¢29¢
¢EW
EL
TA
0318
2301
(17)
00
27R
ipap
aIs
lan
d,
Ban
ksP
enin
sula
2443
�37¢
07¢¢
S17
2�45
¢13¢
¢EW
EL
TA
0318
2401
(24)
00
28C
ape
Th
ree
Po
ints
,B
anks
Pen
insu
la16
43�4
9¢50
¢¢S
172�
54¢3
0¢¢E
WE
LT
A03
1821
01(1
6)0
0
DISPERSAL IN CARPOPHYLLUM 521
(Fig. 2b) grouped haplotypes into three clusters,each separated by five or more steps. The largestcluster (labeled [1] in Fig. 2b) joined 52 haplotypes.These are considered to arise from the C. maschalo-carpum mitochondrial lineage. A second cluster(labeled [2] in Fig. 2b) joined four closely relatedhaplotypes from North Cape (pop. 2 in Fig. 2a) andWekarua (pop. 5). This cluster was five stepsremoved from the nearest C. maschalocarpum haplo-type and six steps removed from the nearest haplo-type of the sister species C. angustifolium. It isunclear whether these haplotypes arise from theC. angustifolium or C. maschalocarpum mitochondriallineage. A third cluster (labeled [3] in Fig. 2b)joined 11 haplotypes from C. angustifolium and puta-tive hybrid specimens and was six steps removedfrom the nearest C. maschalocarpum haplotype. Hapl-otypes from C. angustifolium and any specimens con-sidered to have hybrid origin (including all of pops.1, 2, and 5) are not included in population geneticanalyses. Three haplotypes (15, 23, and 36) werefound only in specimens identified morphologicallyas either C. angustifolium or hybrids but differedfrom C. maschalocarpum haplotypes by only onemutational step. These are considered to havearisen in the C. maschalocarpum mitochondrial line-age and are included in subsequent analyses.
Geographically restricted haplotypes were com-mon (Fig. 2a). Forty-one C. maschalocarpum haplo-types were private (sampled from a singlepopulation only). At two sites, Matai Bay (pop. 3 inFig. 2a) and Maketu (pop. 14), all haplotypes sam-pled were private. At Ahipara (pop. 4), both haplo-types sampled were private aside from a singlespecimen of haplotype 52 also sampled at Horoera(pop. 16). Haplotypes from Bay of Plenty andHawke Bay populations were not shared with neigh-boring regions, apart from one Hawke Bay haplo-type that was also sampled in the Chatham Islands.Four haplotypes were endemic to the ChathamIslands. Population differentiation was moderate tostrong between populations with shared haplotypes(Table 2).
A small number of haplotypes were widely distrib-uted (Fig. 2a). Haplotype 01 was sampled in North-land (pops. 9 and 10), Bay of Plenty (pop. 15),Riversdale (pop. 24), the west coast of the NorthIsland (all populations except 4), the ChathamIslands (pops. 31 and 30), and in all South Islandpopulations. Haplotype 01 was the only haplotypesampled from the South Island. Three haplotypeswere shared between Northland and Portland popu-lations. Haplotype 21 was sampled frequently inboth regions. Also shared were haplotype 52 (pop. 4and a single sample from pop. 16) and haplotype67 (pops. 1 and 19). Haplotype 67 appears to bederived from the C. angustifolium lineage and is14 mutational steps removed from the closestC. maschalocarpum haplotype. Three Chatham Islandhaplotypes were shared with mainland New Zealand:T
able
1.
Co
nti
nu
ed.
Map
IDSi
ten
Lat
itu
de
Lo
ngi
tud
eH
erb
ariu
mn
um
ber
of
exem
pla
rsH
aplo
typ
esp
rese
nt
(n)
Gen
ed
iver
sity
h±
SDN
ucl
eoti
de
div
ersi
typ
±SD
29W
har
ekau
ri,
Ch
ath
amIs
lan
ds
1043
�42¢
18¢¢
S17
6�3
4¢55
¢¢W
WE
LT
ASG
257
24(1
0)0
030
Po
int
Do
rset
,C
hat
ham
Isla
nd
s24
43�4
9¢34
¢¢S
176
�42¢
21¢¢
WW
EL
TA
0313
2801
(16)
,03
(6),
24(2
)0.
5072
±0.
0929
0.00
60±
0.00
43
31W
aita
ngi
,C
hat
ham
Isla
nd
s22
43�5
6¢42
¢¢S
176
�33¢
42¢¢
WW
EL
TA
0313
2901
(6),
08(1
),09
(3),
10(5
),24
(4),
39(3
)0.
8398
±0.
0368
0.02
06±
0.01
17
32M
anu
kau
Ree
f,C
hat
ham
Isla
nd
s5
37�4
4¢41
¢¢S
176
�28¢
13¢¢
WW
EL
TA
0242
1308
(5)
00
522 JOE BUCHANAN AND GIUSEPPE C. ZUCCARELLO
haplotype 39 was sampled in Otanga (pop. 12) andWaitangi (pop. 31) only; haplotype 08 was sharedwith two Hawke Bay sites (pops. 20 and 21); and thecommon southern haplotype 01 was sampled at twoChatham Islands sites (pops. 30 and 31).
Haplotype and nucleotide diversity (Table 1) washighest in three Bay of Plenty populations (SailorsGrave, Otanga, and Maraehako Bay; pops. 11, 12,and 13, respectively), and in two Chatham Islandpopulations (Waitangi and Point Dorset; pops. 31
NORTHIS
LAND
35°S
40°
175°E 180°
S OU
TH
SLAND
200 km
a
1 n=2 n=3 n=5 n=15
(1)
(3)(2)
n=5
n=10
n=20
n=30
b
I
FIG. 2. (a) Distribution of mitochondrial haplotypes of Carpophyllum maschalocarpum. Area of pie charts is proportional to sample size.Locations are described in Table 1. (b) Statistical parsimony network inferred by TCS software (95% confidence limits). Area of circles isproportional to number of haplotypes sampled; small black circles represent extinct or unsampled haplotypes. Dashed line indicates analternative but unlikely connection.
DISPERSAL IN CARPOPHYLLUM 523
and 30). Diversity was lowest in the South Island (asingle haplotype sampled in four populations).
Biogeographic regions. SAMOVA analyses reachedmaximum Fct at K = 14 (Fct = 0.811425). At K = 14populations were partitioned into five groups andnine singletons (Fig. 3). The largest group of 10populations, dominated by the common haplotype01, included all South Island sites, Riversdale (pop.24), Point Dorset (pop. 30), and all west coastNorth Island sites except Ahipara (pop. 4). Thisgroup was unchanged for all iterations of K > 2.
Chatham Island populations did not grouptogether despite their proximity. Only Waitangi andManukau Reef (pops. 31 and 32) were retained
in the same group above K = 3, and these twopopulations separated into singletons at K = 13 andabove. Wharekauri (pop. 29) grouped with North-land populations and two Portland populations,then formed a singleton at K = 8. Point Dorset(pop. 30) remained grouped with South Island andsouthern North Island populations.
AMOVA based on biogeographic regions andprovinces proposed by Shears et al. (2008) explainedconsiderably less between-group variance than themaximum obtained by SAMOVA (Table 3). The lowdiversity in our South Island and southern NorthIsland data precluded showing the North ⁄ South dis-junction found in other studies. A biogeographicbreak at East Cape was not shown in C. maschalocar-pum, rather the data showed connectivity betweentwo Portland populations (pops. 16 and 19) andNorthland.
Departures from neutrality. Neutrality tests compareestimates of h using different parameters. Data fromdemographically stable populations using neutralmarkers should estimate similar values of h. Discrep-ancies in estimates of h indicate departure fromneutrality, with different tests sensitive to differentunderlying factors (Fu 1997).
Tajima’s D value (Tajima 1989) was negative(D = )0.7109), which suggests population expan-sion, but not significant (P = 0.268). Fu and Li’s D*test (Fu and Li 1993) was significant (D* = )2.3973,P = 0.016). Fu and Li’s F * test (Fu and Li 1993) wasalso significant (F * = )2.1275, P = 0.024). Fu’s Fs
value, which is sensitive to population expansion(Fu 1997), was strongly negative and highly signifi-cant (Fs = )25.2996, P = 0.0002). Fs tests were alsocarried out on data split into two groups as indi-cated by SAMOVA. Group 1 (North Island westcoast ⁄ South Island ⁄ Riversdale ⁄ Port Dorset) andGroup 2 (all other samples) both gave negative andsignificant Fs values (Fs = )11.7910, P < 0.0001 andFs = )23.1464, P = 0.0002, respectively).
Isolation by distance. Correlation between geo-graphic distance and genetic distance was low but sig-nificant (r2 = 0.10388, P < 0.0001). Recalculatingdistances to exclude east-west distances gave similarresults (r2 = 0.10820, P < 0.0001). Stronger correla-tion (r2 = 0.20768, P < 0.0001) was obtained byexcluding Chatham Islands samples from the analysis.
Table 2. Pair-wise Fst estimates from mtDNA spacer data (below diagonal) and minimum round coast distances (km,above diagonal) for Bay of Plenty populations of Carpophyllum maschalocarpum. Significance shown as ** = P £ 0.01 and * =P £ 0.05 after sequential Bonferroni correction for multiple tests. N. S. = No significant difference. Population numbersrefer to Table 1 ⁄ Figure 2.
Sailors Grave Maketu Opape Maraehako Bay Otanga
Sailors Grave (pop. 11) – 103 179 190 215Maketu (pop. 14) 0.54931** – 87 117 151Opape (pop. 15) 0.46280** 0.84906** – 48 86Maraehako Bay (pop. 13) 0.42232** 0.68024** N. S. – 38Otanga (pop. 12) 0.12573* 0.26357** 0.11334* 0.12548** –
34
109
876
11
14 1513 12
16
192021
17
18
22
2324
2625
2728
293031 32
Group 1
Group 2
FIG. 3. Grouping of populations of Carpophyllum maschalocar-pum inferred by spatial analysis of molecular variance (SAMOVA)to maximize Fct. Shaded areas show groups inferred with K = 14.Division into two groups shown was used for mismatch distributionanalyses.
524 JOE BUCHANAN AND GIUSEPPE C. ZUCCARELLO
Mismatch analyses. Mismatch analyses showed anapproximation to the expected distribution under amodel of sudden expansion (Fig. 4). Separation ofdata into two groups as indicated by SAMOVAresulted in two distinct distributions: an L-shapeddistribution in Group 1 (North Island west coast ⁄South Island ⁄ Riversdale ⁄ Port Dorset) and a unimo-dal distribution in Group 2 (all other populations).No group showed significant departure from theexpected distribution under a model of suddenexpansion (Group 1: P = 0.355; Group 2: P = 0.881;combined populations: P = 0.624). M (the parame-ter for gene flow where M = 2Nm and N is the demesize and m is the migration rate) is low in Group 1(M = 0.035) and high in Group 2 (M = 12.156).
DISCUSSION
Patterns of gene flow that arise from multipledispersal modes are complex as dispersal patternsmight overlap, especially when each mode is facili-tated by different physical processes and influencedby separate historical events: In our data, threemain patterns are evident: (i) often high geneticdifferentiation between populations, including adja-cent populations; (ii) connectivity between somedistant populations, with only slight overall isola-tion by distance; and (iii) generally low southerndiversity, with the notable exception of the Chat-ham Islands. These patterns are consistent withfucalean morphological and life-history characteris-tics that limit dispersal at the gamete ⁄ zygote stage,but facilitate occasional long-distance dispersal(Thiel and Haye 2006). Historical climatic changesalso appear to have influenced the distribution ofC. maschalocarpum.
1. Genetic differentiation. Genetic differentiationbetween many populations of C. maschalocarpum wason the order of that expected for fucalean algae(Coleman and Brawley 2005b, Hoarau et al. 2007,Tatarenkov et al. 2007, Coleman and Kelaher 2009).Strong genetic differentiation was observed betweensome adjacent populations that were separated byhabitat discontinuities, even where distances wererelatively short (e.g., Cape Wiwiki [pop. 6] andBland Bay [pop. 7], 39 km apart but separated bythe estuarine Bay of Islands). These discontinuitieswould be expected to provide a barrier to gameticor zygotic dispersal, but not to dispersal of floatingthalli. Despite the abundance of floating thalli(Kingsford 1992), it appears that this mode of dis-persal seldom contributes to gene flow.
Explanations for limited gene flow by floatingthalli include dioecy, small target areas for immi-grant floating algae (e.g., Maketu–a short, 1 km,rocky peninsula flanked by long stretches of softsediment coast), and density blocking (Deysher andNorton 1982, Hewitt 1996, Austerlitz et al. 2000),where relatively rare immigrants seldom contributealleles to established stands of algae. High regionalhaplotype endemism might reflect isolation of watermasses, for example, limited current transport intoembayments such as Hawke Bay and Bay of Plenty(Ridgway 1960, 1962, Stanton et al. 1997). Directstudies of transport of floating algae have not beenundertaken in New Zealand. A high prevalence ofwesterly winds might mean surface floating speciesare usually exported and lost from nearshorehabitat.
2. Long-distance dispersal. Shared haplotypes inNorthland and Portland populations, and NorthIsland and Chatham Island populations, show somehaplotypes have dispersed over long distances. Hap-lotype 21 is common in Northland and Portlandpopulations, and it is improbable that this haplotypearose independently in two areas. Furthermore,
ycneuqerF
Number of Differences310 1211109876543 4121
5,000
10,000
15,000
20,000
25,000
30,000
35,000
0
All samplesGroup 1Group 2
FIG. 4. Mismatch distributions. Group 1 (South Island ⁄ NorthIsland West Coast, Riversdale and Point Dorset populations):black bars; Group 2 (all other populations): gray bars; combinedpopulations: open bars.
Table 3. Hierarchical partitioning of molecular varianceestimated by analysis of molecular variation (AMOVA)from Carpophyllum maschalocarpum mtDNA spacer data. (a)Variation explained by grouping in biogeographic regionsproposed by Shears et al. (2008); (b) groupings at maxi-mum Fct inferred by spatial analysis of molecular variance(SAMOVA).
Source of variation dfSum ofsquares Variance
Percentageof variation
a. Groupings by biogeographic regionsAmong groups 6 603.55 Va = 1.026 33.05Among populations 22 626.90 Vb = 1.402 45.14Within populations 552 373.84 Vc = 0.677 21.81
b. Groupings by SAMOVA (Fig. 3)Among groups 13 804.64 Va = 1.652 81.14Among populations 15 6.94 Vb = 0.004 0.21Within populations 550 208.93 Vc = 0.380 18.65
(a) Fixation indices: Fsc = 0.674, Fst = 0.782, Fct = 0.331.(b) Fixation indices: Fsc = 0.011, Fst = 0.814, Fct = 0.811.
DISPERSAL IN CARPOPHYLLUM 525
haplotype 67, which clusters with C. angustifolium,was only sampled at Hooper Point (pop. 1) and KaitiBeach (pop. 19). Kaiti is well outside the range ofC. angustifolium, which does not extend south ofEast Cape (Morton and Miller 1968 and our obser-vations). It appears this haplotype entered theC. maschalocarpum gene pool by hybridization(Hodge et al. 2010) and dispersed to southern popu-lations. This connection, and the connectionbetween the mainland and the Chatham Islands, iscongruent with contemporary currents and surfacewinds. Northland and Portland are connected by theEast Auckland ⁄ East Cape Current. This current doesnot extend to the inner area of the Bay of Plenty(Heath 1985, Stanton et al. 1997), and our data sug-gest Bay of Plenty populations occupy an isolatededdy. Despite their spatial separation, the outerpopulations, Sailors Grave (pop. 11) and Otanga(pop. 12), are more connected to each other thanto inner populations. Haplotype 28 occurs in highfrequencies in these populations but was absentfrom the other Bay of Plenty populations.
Four lineages are present in Chatham Island pop-ulations, with haplotypes in each lineage the sameor closely related to haplotypes found in North andSouth Island populations, suggesting at least fourdispersal events between mainland New Zealandand the Chatham Islands. High connectivitybetween the mainland and the Chatham Islands issurprising as many Chatham Island marine algaeare endemic (Nelson et al. 1991, Schiel et al. 1995),several common mainland species are absent (e.g.,the nonbuoyant laminarian species Ecklonia radiataand Lessonia variegata) (Nelson et al. 1991, Nelson1994), and studies of other organisms have showngenetic isolation in Chatham Island populations(Goldstien et al. 2009). Our data suggest a very dif-ferent pattern for algae capable of dispersal by float-ing, with prevailing westerly winds and currentsfacilitating eastward dispersal from the mainland tothe Chatham Islands (Chiswell and Booth 2008).Removing the east–west component of geographicdistance between populations did not affect correla-tion with genetic connectivity. We suggest wind andcurrents facilitate long-distance transport to theChatham Islands, while north to south movementproceeds mostly by stepping-stone dispersal.
The north ⁄ south disjunction reported in otherNew Zealand studies (Apte and Gardner 2002,Sponer and Roy 2002, Waters and Roy 2004, Ayresand Waters 2005, Goldstien et al. 2006) was evidentin the sudden reduction in haplotype diversity insouthern populations, but this occurs further norththan the break found in previous studies. A barrierto dispersal by upwelling or current transport ofpropagules offshore around the north of the SouthIsland has been suggested, but evidence for this isequivocal (Ross et al. 2009). We suggest that histori-cal climate change explains the pattern in our databetter than a dispersal barrier and discuss this
below. Bioregions proposed by Shears et al. (2008)are partly reflected in the distribution of C. masc-halocarpum haplotypes (Table 3). In particular, dis-junctions just south of Ahipara and south of HawkeBay (pops. 4 and 20 ⁄ 21, Fig. 3) in our data are con-gruent with boundaries of bioregions in Shearset al. (2008). Conversely, the connection betweenNorthland and Portland, and the relative isolationof the Bay of Plenty and Hawke Bay populations ofC. maschalocarpum, is not evident in data obtainedby Shears et al. (2008). It is not clear whether thisarises from sampling differences, the stochasticnature of floating dispersal, or factors unique toC. maschalocarpum.
This pattern of population differentiation andconnectivity is consistent with a species that is gener-ally restricted in dispersal, but with intermittentlong-distance dispersal by floating (Thiel and Haye2006). Leapfrog dispersal, where more distant popu-lations show greater connectivity than adjacent pop-ulations, is a predicted outcome of intermittentfloating dispersal (Thiel and Haye 2006) and hasbeen observed in a variety of organisms (Snyder andGooch 1973, Bockelmann et al. 2003, Colson andHughes 2004). In C. maschalocarpum, leapfrogdispersal appears to be overlaid on patterns of lowdispersal or stepping-stone dispersal.
3. Low southern diversity and post-LGM recolonization.The abundant-center model (Eckert et al. 2008)predicts greatest genetic diversity in the center of aspecies’ range, with a decline toward peripheralpopulations. Our data are broadly consistent withthis model. Diversity is highest in the Bay of Plentyin northeast New Zealand and generally declineswith distance from this region. But there are twodepartures from this pattern: (i) haplotypicallydiverse populations in the far north (pops. 1, 2, and5) and (ii) high haplotypic diversity in the ChathamIslands. We invoke both contemporary factors andhistorical factors to explain these departures.
Far-north diversity appears to have hybrid origin,with gene flow between C. angustifolium and C. masc-halocarpum. Far-north populations have been identi-fied as either species by different workers (e.g.,C. maschalocarpum by Shears and Babcock 2007,C. angustifolium by Hay and Grant 2003). Hybridshave been reported between Carpophyllum speciesbased on intermediate morphology (Lindauer et al.1961, Dromgoole 1973, Shears and Babcock 2007)and molecular data (Hodge et al. 2010), and hybrid-ization in fucalean species has been widely reported(Scott and Hardy 1994, Kim et al. 1997, Coyer et al.2002, Engel et al. 2005).
The pattern of high northern diversity and lowsouthern diversity suggests relatively recent south-ward expansion with warming of the ocean watersfollowing the LGM. Water temperature is a majordeterminant of species range in many macroalgae(Luning 1990, Adey and Steneck 2001). Low diver-sity at high latitudes is often reported in species that
526 JOE BUCHANAN AND GIUSEPPE C. ZUCCARELLO
have recolonized high latitudes during postglacialwarming (Coyer et al. 2003, Marko 2004, Hoarauet al. 2007, Maggs et al. 2008, Fraser et al. 2009).The present-day distribution of C. maschalocarpumextends south to Fiordland (46oS) on the SouthIsland’s west coast (Nelson et al. 2002), where warmcurrents push in from the Tasman Sea, but only asfar south as Banks Peninsula (43�45¢S) on the eastcoast, where cold currents push up from the south(Adams 1994). This southern limit is correlated withthe 15�C sea surface temperature (SST) isobar forthe warmest month of the year (Fig. 1). While thereare few direct data on thermal tolerances in C. masc-halocarpum, the correlation between southern limitand SST on both coasts suggests that this is a low-temperature limit. Records of specimens of C. masc-halocarpum from driftlines on Stewart Island andsub-Antarctic islands (Dromgoole 1973) show thalliare transported further south, but populations havenot established.
Climate reconstructions of the LGM (Barrowsand Juggins 2005) place the 15�C warmest monthSST isobar north of Taranaki on New Zealand’s westcoast and north of Hawke Bay on the east coast(Fig. 1). These temperatures would have restrictedC. maschalocarpum to the northern half of the NorthIsland during the LGM. Recolonization southwarddown the North Island west coast, through the reo-pened Cook Strait (Lewis et al. 1994) and down theeast coast of the South Island would explain the lowdiversity and the different patterns of populationexpansion indicated by mismatch distributions.
Genetic diversity in Northern Hemisphere algaehas been shown to retain the signature of past cli-mate events–in particular the last glaciation, whichremoved algae from higher latitudes (reviewed byMaggs et al. 2008). Similar patterns have beenshown for marine species in the Southern Hemi-sphere (Hickey et al. 2009) including algae (Fraseret al. 2009). The distribution of haplotype diversityin our data is strikingly similar to patterns of thesame marker in Fucus serratus from Europe (Hoarauet al. 2007) and is similar to distributions inter-preted as showing postglacial range expansion for anumber of species (Hewitt 1999, Provan et al. 2005,Maggs et al. 2008).
We offer two explanations for the presence offour endemic haplotypes in the Chatham Islands.Either this is further evidence for the ease ofdispersal to the Chatham Islands, with these haplo-types either extinct or not sampled on the mainislands of New Zealand, and with dispersal pathwaysfacilitating codispersal of closely related haplotypes(e.g., haplotypes 08, 09, and 10); or ChathamIsland populations of C. maschalocarpum have per-sisted for a considerable time, allowing the accumu-lation of unique mutations. At present, summerwater temperatures around the Chatham Islandsare elevated as seasonal north ⁄ south movement ofthe subtropical convergence is constrained by an
area of shallow bathymetry, the Chatham Rise(Stanton 1997). Paleoclimate studies indicate warmsummer water temperatures around the ChathamIslands might have persisted during the LGM (Fenneret al. 1992, Nelson et al. 2000). Hoarau et al. (2007)estimated a mutation rate of 2% to 3.4% ⁄ Myr for thismitochondrial spacer in Fucus. If mutation rates aresimilar in C. maschalocarpum, it is unlikely that fourendemic haplotypes could arise in Chatham Islandpopulations since recolonization after the LGM.
The pattern of local population differentiation,but with specific pathways for gene flow between dis-tant populations, overlaid on a historical pattern ofretreat and recolonization is consistent with theintermittent and stochastic dynamics of long-distance dispersal by floating (Thiel and Haye 2006)and might be a general pattern for buoyant algae(Fraser et al. 2009) and organisms rafting on thesealgae. An important finding is that connectivitybetween distant populations can be high, evenwhere local connectivity is often low. This highlightsthe importance of sampling across species’ rangeswhen investigating dispersal, as large-scale processesdriving long-distance dispersal might not be apparentin studies at local scales.
We thank Wendy Nelson, Kate Neill, Tracy Farr, Nick Shears,Peter Martin, Erasmo Macaya, Fiona Hodge, and ChathamIslands Department of Conservation staff for help with collec-tions, and Peter Ritchie for help with early drafts of the man-uscript. Dr. Martin Thiel provided many useful comments.We also thank the New Zealand Department of Conservationfor funding assistance.
Adams, N. M. 1994. Seaweeds of New Zealand. Canterbury UniversityPress, Christchurch, New Zealand, 360 pp.
Adey, W. H. & Steneck, R. S. 2001. Thermogeography over timecreates biogeographic regions: a temperature ⁄ space ⁄ time-integrated model and an abundance weighted test for benthicmarine algae. J. Phycol. 37:677–98.
Apte, S. & Gardner, J. P. A. 2002. Population genetic variation in theNew Zealand greenshell mussel, Perna canaliculus: SSCP andRFLP analyses of mitochondrial DNA. Mol. Ecol. 11:1617–28.
Austerlitz, F., Mariette, S., Machon, N. & Gouyon, P.-H. 2000.Effects of colonization processes on genetic diversity: differ-ences between annual plants and tree species. Genetics 154:1309–21.
Ayres, K. L. & Waters, J. M. 2005. Marine biogeographic disjunctionin central New Zealand. Mar. Biol. 147:1045–52.
Barrows, T. T. & Juggins, S. 2005. Sea-surface temperatures aroundthe Australian margin and Indian Ocean during the LastGlacial Maximum. Quat. Sci. Rev. 24:1017–47.
Bassam, B. J., Caetano-Anolles, G. & Gresshoff, P. M. 1991. Fast andsensitive silver staining of DNA in polyacrylamide gels. Anal.Biochem. 196:80–3.
Bockelmann, A. C., Reusch, T. B. H., Bijlsma, R. & Bakker, J. P.2003. Habitat differentiation vs. isolation by distance: thegenetic population structure of Elymus athericus in Europeansalt marshes. Mol. Ecol. 12:505–15.
Bohonak, A. J. 1999. Dispersal, gene flow and population structure.Q. Rev. Biol. 74:21–45.
Chapman, A. R. O. 1995. Functional ecology of fucoid algae:twenty-three years of progress. Phycologia 34:1–32.
Chiswell, S. H. & Booth, J. D. 2008. Sources and sinks of larvalsettlement in Jasus edwardsii around New Zealand: where dolarvae come from and where do they go? Mar. Ecol. Prog. Ser.354:201–17.
DISPERSAL IN CARPOPHYLLUM 527
Clayton, M. N. 1992. Propagules of marine macroalgae: structureand development. Eur. J. Phycol. 27:219–32.
Clement, M., Posada, D. & Crandall, K. A. 2000. TCS: acomputer program to estimate gene genealogies. Mol. Ecol.9:1657–9.
Coleman, M. A. & Brawley, S. H. 2005a. Spatial and temporal var-iability in dispersal and population genetic structure of arockpool alga. Mar. Ecol. Prog. Ser. 300:63–77.
Coleman, M. A. & Brawley, S. H. 2005b. Are life history character-istics good predictors of genetic diversity and structure? A casestudy of the intertidal alga Fucus spiralis (Heterokontophyta;Phaeophyceae). J. Phycol. 41:753–62.
Coleman, M. A. & Kelaher, B. P. 2009. Connectivity among frag-mented populations of a habitat-forming alga, Phyllosporacomosa (Phaeophyceae, Fucales) on an urbanised coast. Mar.Ecol. Prog. Ser. 381:63–70.
Colson, I. & Hughes, R. N. 2004. Rapid recovery of genetic diversityof dogwhelk (Nucella lapillus L.) populations after localextinction and recolonization contradicts predictions fromlife-history characteristics. Mol. Ecol. 13:2223–33.
Coyer, J. A., Peters, A. F., Hoarau, G., Stam, W. T. & Olsen, J. L.2002. Hybridization of the marine seaweeds Fucus serratus andFucus evanescens (Heterokontophyta: Phaeophyceae) in a 100-year old zone of secondary contact. Proc. R. Soc. Biol. Sci. Ser. B269:1829–34.
Coyer, J. A., Peters, A. F., Stam, W. T. & Olsen, J. L. 2003. Post ice-age recolonization and differentiation of Fucus serratus L.(Phaeophyceae; Fucaceae) populations in Northern Europe.Mol. Ecol. 12:1817–29.
Delf, E. M. 1939. Studies in the Fucales of New Zealand I. Thegenus Carpophyllum Grev. J. Bot. Lond. 77:129–38.
Dethier, M. N., McDonald, K. & Strathmann, R. R. 2003. Coloni-zation and connectivity of habitat patches for coastal marinespecies distant from source populations. Conserv. Biol.17:1024–35.
Deysher, L. & Norton, T. A. 1982. Dispersal and colonisation ofSargassum muticum (Yendo) Fensholt. J. Exp. Mar. Biol. Ecol.56:179–95.
Dromgoole, F. I. 1973. A contribution to the biology of the genusCarpophyllum Grev. PhD dissertation, University of Auckland,New Zealand, 361 pp.
Dupanloup, I., Schneider, S. & Excoffier, L. 2002. A simulatedannealing approach to define the genetic structure of popu-lations. Mol. Ecol. 11:2571–81.
Eckert, C. G., Samis, K. E. & Lougheed, S. C. 2008. Genetic varia-tion across species’ geographical ranges: the central-marginalhypothesis and beyond. Mol. Ecol. 17:1170–88.
Engel, C. R., Daguin, C. & Serrao, E. 2005. Genetic entitries andmating systems in hermaphroditic Fucus spiralis and its closedioecious relative F. vesiculosus (Fucaceae, Phaeophyceae).Mol. Ecol. 14:2033–46.
Excoffier, L. 2004. Patterns of DNA sequence diversity and geneticstructure after a range expansion: lessons from the infiniteisland model. Mol. Ecol. 13:853–64.
Excoffier, L., Laval, G. & Schneider, S. 2005. Arlequin, version 3.0:an integrated software package for population genetics dataanalysis. Evol. Bioinformatics Online 1:47–50.
Fenner, J., Carter, L. & Stewart, R. 1992. Late quaternary paleocli-mate and paleoceanographic change over northern ChathamRise, New Zealand. Mar. Geol. 108:383–404.
Fraser, C. L., Nikula, R., Spencer, H. G. & Waters, J. M. 2009. Kelpgenes reveal effects of subantarctic sea ice during the LastGlacial Maximum. Proc. Natl. Acad. Sci. U. S. A. 106:3249–53.
Fu, X. Y. 1997. Statistical tests of neutrality of mutations againstpopulation growth, hitchhiking and background selection.Genetics 147:915–25.
Fu, X. Y. & Li, W. H. 1993. Statistical tests of neutrality of mutations.Genetics 133:693–709.
Goldstien, S. J., Gemmell, N. J. & Schiel, D. R. 2006. Comparativephylogeography of coastal limpets across a marine disjunctionin New Zealand. Mol. Ecol. 15:3259–68.
Goldstien, S. J., Gemmell, N. J. & Schiel, D. R. 2009. Colonisationand connectivity by intertidal limpets among New Zealand,
Chatham and Sub-Antarctic Islands I. Genetic connections.Mar. Ecol. Prog. Ser. 388:111–9.
Hay, B. & Grant, C. 2003. Marine Resources in Tai Tokerau. JamesHenare Maori Research Centre, University of Auckland, NewZealand, 83 pp.
Heath, R. A. 1985. A review of the physical oceanography of the seasaround New Zealand. N. Z. J. Mar. Freshw. Res. 19:79–124.
Hernandez-Carmona, G., Hughes, B. & Graham, M. H.2006. Reproductive longetivity of drifting kelp Macrocystispyrifera (Phaeophyceae) in Monterey Bay, USA. J. Phycol.42:1199–207.
Hewitt, G. M. 1996. Some genetic consequences of ice ages, andtheir role in divergence and speciation. Biol. J. Linn. Soc.58:247–76.
Hewitt, G. M. 1999. Post-glacial re-colonisation of European biota.Biol. J. Linn. Soc. 68:87–112.
Hickey, A. J. R., Lavery, S. D., Hannan, D. A., Baker, C. S. &Clements, K. D. 2009. New Zealand triplefin fishes (familyTripterygiidae): contrasting population structure and mtDNAdiversity within a marine species flock. Mol. Ecol. 18:680–96.
Higgins, S. I., Nathan, R. & Cain, M. L. 2003. Are long distancedispersal events in plants usually caused by non-standardmeans of dispersal? Ecology 84:1945–56.
Hoarau, G., Coyer, J. A., Veldsink, J. H., Stam, W. T. & Olsen, J. L.2007. Glacial refugia and recolonization pathways in thebrown seaweed Fucus serratus. Mol. Ecol. 16:3606–16.
Hodge, F., Buchanan, J. & Zuccarello, G. C. 2010. Hybridizationbetween the endemic brown algae Carpophyllum maschalocar-pum and C. angustifolium (Fucales): genetic and morphologicalevidence. Phycol. Res. 58:239–47.
van den Hoek, C. 1987. The possible significance for long rangedispersal for the biogeography of seaweeds. Helgol. Wiss.Meeresunters. 41:261–72.
Kendrick, G. A. & Walker, D. I. 1991. Dispersal distances forpropagules of Sargassum spinuligerum (Sargassaceae, Pha-eophyta) measured directly by vital staining and venturi suc-tion sampling. Mar. Ecol. Prog. Ser. 79:133–8.
Kendrick, G. A. & Walker, D. I. 1995. Dispersal of propagules ofSargassum spp. (Sargassaceae: Phaeophyta): observations oflocal patterns and consequences for recruitment and popula-tion structure. J. Exp. Mar. Biol. Ecol. 192:273–88.
Kim, K. Y., O’Leary, S. J. & Garbary, D. J. 1997. Artificial hybrid-ization between Ascophyllum nodosum and Fucus vesiculosus(Phaeophyceae) from Nova Scotia, Canada. Can. J. Bot.75:1133–6.
Kingsford, M. J. 1992. Drift algae and small fish in coastal waters ofnortheastern New Zealand. Mar. Ecol. Prog. Ser. 80:41–55.
Kinlan, B. P., Gaines, S. D. & Lester, S. E. 2005. Propagule dispersaland the scales of marine community processes. Divers. Distrib.11:139–48.
Kumar, S., Tamura, K. & Nei, M. 2004. MEGA3: integrated softwarefor molecular evolutionary genetics analysis and sequencealignment. Briefings Bioinfomatics 5:150–63.
Lewis, K. B., Carter, L. & Davey, F. J. 1994. The opening of CookStrait: interglacial tidal scour and aligning basins at a sub-duction to transform plate edge. Mar. Geol. 116:293–312.
Lindauer, V. W., Chapman, V. J. & Aiken, M. 1961. The marinealgae of New Zealand II: Phaeophyceae. Nova Hedwigia3:129–350.
Luning, K. 1990. Seaweeds: Their Environment, Biogeography, and Eco-physiology. Wiley, New York, 527 pp.
Macaya, E., Boltana, S., Buschmann, A., Hinojosa, I., Macchiavello,J., Valdivia, N., Vasquez, N., Vasquez, J., Vega, J. & Thiel, M.2005. Presence of sporophylls in floating kelp rafts of Macro-cystis spp. (Phaeophyceae) along the Chilean Pacific Coast.J. Phycol. 41:913–22.
Maggs, C. A., Castilho, R., Foltz, D., Henzler, C., Jolly, M. T., Kelly,J., Olsen, J. et al. 2008. Evaluating signatures of glacial refugiafor North Atlantic benthic marine taxa. Ecology 89:S108–22.
Marko, P. B. 2004. ‘What’s larvae got to do with it?’ Disparatepatterns of post-glacial population structure in two benthicmarine gastropods with identical dispersal potential. Mol. Ecol.13:597–611.
528 JOE BUCHANAN AND GIUSEPPE C. ZUCCARELLO
Marko, P. B., Hoffman, J. M., Emme, S. A., McGovern, T. M.,Keever, C. C. & Cox, L. N. 2010. The ‘Expansion-Contraction’model of Pleistocene biogeography: rocky shores suffer a seachange? Mol. Ecol. 19:146–69.
Marske, K. A., Leschen, R. A. B. & Buckley, T. R. 2011. Reconcilingphylogeography and ecological niche models for New Zealandbeetles: looking beyond glacial refugia. Mol. Phylogenet. Evol.59:89–102.
McKenzie, P. F. & Bellgrove, A. 2008. Dispersal of Hormosira banksii(Phaeophyceae) via detached fragments: reproductive viabilityand longetivity. J. Phycol. 44:1108–15.
Morton, J. E. & Miller, M. C. 1968. The New Zealand Sea Shore.Collins, London, 638 pp.
Muhlin, J. F., Engel, C. R., Stessel, R., Weatherbee, R. A. & Brawley,S. H. 2008. The influence of coastal topography, circulationpatterns and rafting in structuring populations of an intertidalalga. Mol. Ecol. 17:1198–210.
Naylor, M. 1954. The female conceptacle and young sporeling ofthe four New Zealand species of Carpophyllum. Trans. Proc.R. Soc. N. Z. 82:1–6.
Nelson, W. A. 1994. Distribution of macroalgae in New Zealand –an archipelago in space and time. Bot. Mar. 37:221–33.
Nelson, W. A., Adams, N. M. & Hay, C. H. 1991. Marine Algae of theChatham Islands. Nat. Mus. N. Z. Misc. Ser. 23:1–58.
Nelson, C. S., Hendy, I. L., Neil, H. L., Hendy, C. H. & Weaver, P. P.E. 2000. Last glacial jetting of cold waters through the Sub-tropical Convergence zone in the Southwest Pacific off easternNew Zealand, and some geological implications. Palaeogr.Palaeoclimatol. Palaeoecol. 156:103–21.
Nelson, W. A., Villouta, E., Neill, K. F., Williams, G. C., Adams, N. M.& Slivsgaard, R.. 2002. Marine macroalgae of Fiordland, New
Zealand. Tuhinga 13:117–52.Norton, T. A. 1992. Dispersal by macroalgae. Br. Phycol. J. 27:
293–301.Oudot-Le Secq, M.-P., Kloareg, B. & Loiseaux-de Goer, S. 2002. The
mitochondrial genome of the brown alga Laminaria digitata: acomparative analysis. Eur. J. Phycol. 37:163–72.
Oudot-Le Secq, M.-P., Loiseaux-de Goer, S., Stam, W. T. & Olsen, J. L.2006. Complete mitochondrial genomes of the three brownalgae (Heterokonta: Phaeophyceae) Dictyota dichotoma, Fucusvesiculosus and Desmarestia viridis. Curr. Genet. 49:47–58.
Pakker, H., Breeman, A. M., Prud’homme van Reine, W. F., vanOppen, M. J. H. & van den Hoek, C. 1996. Temperatureresponses of tropical to warm-temperate Atlantic seaweeds.I. Absence of ecotypic differentiation in amphi-Atlantic tropi-cal Canary Islands species. Eur. J. Phycol. 31:123–32.
Provan, J., Wattier, R. A. & Maggs, C. A. 2005. Phylogeographicanalysis of the red seaweed Palmaria palmata reveals a Pleisto-cene marine glacial refugium in the English Channel. Mol.Ecol. 14:793–803.
Ray, N., Currat, M. & Excoffier, L. 2003. Intra-deme moleculardiversity in spatially expanding populations. Mol. Biol. Evol.20:76–86.
Reed, D. C., Amsler, C. D. & Ebeling, A. W. 1992. Dispersal in kelps– factors affecting spore swimming and competence. Ecology73:1577–85.
Reed, D. C., Laur, D. R. & Ebeling, A. W. 1988. Variation in algaldispersal and recruitment: the importance of episodic events.Ecol. Monogr. 58:321–35.
Ridgway, N. M. 1960. Surface water movements in Hawke Bay, NewZealand. N. Z. J. Geol. Geophys. 3:253–61.
Ridgway, N. M. 1962. Nearshore surface currents in southernHawke Bay, New Zealand. N. Z. J. Geol. Geophys. 5:545–66.
Rogers, A. R. & Harpending, H. 1992. Population growth makeswaves in the distribution of pairwise genetic differences. Mol.Biol. Evol. 9:552–69.
Ross, P. M., Hogg, I. D., Pilditch, C. A. & Lundquist, C. J. 2009.Phylogeography of New Zealand’s coastal benthos. N. Z. J.Mar. Freshw. Res. 43:1009–27.
Ross, P. M., Hogg, I. D., Pilditch, C. A., Lundquist, C. J. & Wilkins,R. J. 2011. Population genetic structure of the New Zealandestuarine clam Austrovenus stutchburyi (Bivalvia: Veneridae)reveals population subdivision and partial congruence withbiogeographic boundaries. Estuar. Coast. 35:143–54.
Schiel, D. R., Andrew, N. L. & Foster, M. S. 1995. The structure ofsubtidal algae and invertebrate assemblages at the ChathamIslands, New Zealand. Mar. Biol. 123:355–67.
Scott, G. W. & Hardy, F. G. 1994. Observations of the occurrence ofhybrids between two sympatric species of fucoid algae. Crypto-gam. Algol. 15:297–305.
Shears, N. T. & Babcock, R. C. 2007. Quantitative description ofNew Zealand’s shallow sub-tidal reef communities. Sci. Conserv.280:1–126.
Shears, N. T., Smith, F., Babcock, R. C., Duffy, C. A. & Villouta, E.2008. Evaluation of biogeographic classification schemes forconservation planning: application to New Zealand’s coastalmarine environment. Conserv. Biol. 22:467–81.
Smith, S. D. A. 2002. Kelp rafts in the Southern Ocean. Glob. Ecol.Biogeogr. 11:67–9.
Snyder, T. P. & Gooch, J. L. 1973. Genetic differentiation in Litto-rina saxatilis (Gastropoda). Mar. Biol. 22:177–82.
Sotka, E. E., Wares, J. P., Barth, J. A., Groseberg, R. K. & Palumbi, S. R.2004. Strong genetic clines and geographical variation in geneflow in the rocky intertidal barnacle Balanus glandula. Mol. Ecol.13:2143–56.
Sponer, R. & Roy, M. S. 2002. Phylogeographic analysis of thebrooding brittle star Amphipholis squamata (Echinodermata)along the coast of New Zealand reveals high cryptic genetic var-iation and cryptic dispersal potential. Evolution 56:1954–67.
Stanton, B. R. 1997. Sea level and sea surface temperature vari-ability at the Chatham Islands, New Zealand. N. Z. J. Mar.Freshw. Res. 31:525–36.
Stanton, B. R., Sutton, P. J. H. & Chiswell, S. M. 1997. The EastAuckland Current, 1994–95. N. Z. J. Mar. Freshw. Res. 31:537–49.
Sunnucks, P., Wilson, A. C. C., Beheregaray, L. B., Zenger, K.,French, J. & Taylor, A. C. 2000. SSCP is not so difficult: theapplication and utility of single-stranded conformation poly-morphism in evolutionary biology and molecular biology. Mol.Ecol. 9:1699–710.
Tajima, F. 1989. Statistical method for testing the neutral mutationhypothesis by DNA polymorphism. Genetics 123:585–95.
Tatarenkov, A., Jonsson, R. B., Kautsky, L. & Johannesson, K. 2007.Genetic structure in populations of Fucus vesiculosus (Phaeo-phyceae) over spatial scales from 10 m to 800 km. J. Phycol.43:675–85.
Thiel, M. & Gutow, L. 2005a. The ecology of rafting in the marineenvironment. II. The rafting organisms and community. Oce-anogr. Mar. Biol. 43:279–418.
Thiel, M. & Gutow, L. 2005b. The ecology of rafting in the marineenvironment. I. The floating substrata. Oceanogr. Mar. Biol.42:181–263.
Thiel, M. & Haye, P. A. 2006. The ecology of rafting in the marineenvironment. III. Biogeographical and evolutionary conse-quences. Oceanogr. Mar. Biol. 44:323–429.
Wares, J. P. & Cunningham, C. W. 2001. Phylogeography and his-torical ecology of the North Atlantic intertidal. Evolution55:2455–69.
Waters, J. M. & Roy, M. S. 2004. Phylogeography of a high dispersalNew Zealand sea-star: does upwelling block gene flow? Mol.Ecol. 13:2797–806.
Zuccarello, G. C. & Lokhorst, G. M. 2005. Molecular phylogeny ofthe genus Tribonema (Xanthophyceae) using rbcL genesequence data: monophyly of morphologically simple algalspecies. Phycologia 44:384–92.
Zuccarello, G. C., West, J. A., Kamiya, M. & King, R. J. 1999. A rapidmethod to score plastid haplotypes in red seaweeds and its usein determining parental inheritance of plastids in the red algaBostrychia (Ceramiales). Hydrobiologia 401:207–14.
DISPERSAL IN CARPOPHYLLUM 529