Content uploaded by Steven A Trewick
Author content
All content in this area was uploaded by Steven A Trewick on Oct 25, 2021
Content may be subject to copyright.
This article appeared in a journal published by Elsevier. The attached
copy is furnished to the author for internal non-commercial research
and education use, including for instruction at the authors institution
and sharing with colleagues.
Other uses, including reproduction and distribution, or selling or
licensing copies, or posting to personal, institutional or third party
websites are prohibited.
In most cases authors are permitted to post their version of the
article (e.g. in Word or Tex form) to their personal website or
institutional repository. Authors requiring further information
regarding Elsevier’s archiving and manuscript policies are
encouraged to visit:
http://www.elsevier.com/copyright
Author's personal copy
Are you my mother? Phylogenetic analysis reveals orphan hybrid stick insect
genus is part of a monophyletic New Zealand clade
Steve A. Trewick
a,*
, Mary Morgan-Richards
a
, Lesley J. Collins
b
a
Allan Wilson Centre for Molecular Ecology & Evolution, Ecology Group PN 624, Institute of Natural Resources, Massey University, Private Bag 11-222, Palmerston North, New Zealand
b
Allan Wilson Centre for Molecular Ecology & Evolution, Institute of Molecular Biosciences, Massey University, New Zealand
article info
Article history:
Received 23 February 2007
Revised 6 March 2008
Accepted 20 May 2008
Available online 27 May 2008
Keywords:
Hybrid genesis
Phasmid
Parthenogenesis
mtDNA
rDNA
BLAST search
Outgroup
abstract
The hybrid stick insect genus Acanthoxyla Uvarov 1944 is unusual for an obligate parthenogen, in the
extreme morphological diversity it exhibits that has led to eight species being recognised. The New Zea-
land sexual species Clitarchus hookeri [White, A. 1846. The zoology of the Voyage of H.M.S. Erebus and
Terror. In: 1 Insects of New Zealand. E.W. Janson, London.] is the putative parental species in the hybrid-
ization that gave rise to the hybrid lineage Acanthoxyla. In an effort to identify the maternal ancestor of
Acanthoxyla we sequenced nuclear 28S rDNA and/or mtDNA COI & COII of all nine endemic New Zealand
stick insect genera, representing 17 of the 22 described species. We also sequenced 28S from eight non-
New Zealand stick insects to supplement published 28S sequence data that provided a taxonomically and
geographically broad sampling of the phasmids. We applied a novel search algorithm (SeqSSi = Sequence
Similarity Sieve) to assist in selection of outgroup taxa for phylogenetic analysis prior to alignment. Phy-
logenetic reconstructions resolved an exclusively New Zealand clade to which the maternal lineage of
Acanthoxyla belonged, but did not support existing higher level taxonomy of stick insects. We did not find
a sexual maternal species for Acanthoxyla but phylogenetic relationships indicate that this species lived in
New Zealand and could be classified among the New Zealand Phasmatinae. Among the available taxa, the
nearest evolutionary neighbours to the New Zealand phasmid fauna as a whole were predominantly from
the New Zealand region (Fiji, Australia, New Guinea, New Caledonia and South America). As it appears to
be an orphan, it is interesting to speculate that a combination of parthenogenetic reproduction and/or
hybrid vigour in Acanthoxyla may have contributed to the extinction of its mother.
Ó2008 Elsevier Inc. All rights reserved.
1. Introduction
Hybrid speciation is rare in animals (Coyne and Orr, 2004)
because reproductive isolation from parental taxa requires a
combination of traits that are uncommon. However, stick insects
have such a flexible reproductive pathway that hybrid lineages
have arisen many times using diverse strategies to by-pass the rou-
tine recombination and gamete fusion of sexual diploids (Bullini,
1994). The adoption of parthenogenetic reproduction is common
in stick insects and especially well documented in the genera
Timema (Law and Crespi, 2002) and Bacillus (Scali et al., 2003).
The best-studied group from the perspective of hybrid speciation
are the Sicilian Bacillus complex of stick insects where a diverse
range of reproductive mechanisms is implicated in the emergence
and maintenance of hybrid species (Scali et al., 2003). Asexual spe-
cies of the North American genus Timema have arisen without
hybridisation, and some show (and some do not) colour and host
differentiation from their sexual ancestors (Crespi and Sandoval,
2000). However, in Bacillus where asexual species have arisen with
and without hybridisation there is little ecological or morphologi-
cal variations among parthenogentic species and their sexual
ancestors. The situation among a genus of New Zealand stick
insects is strikingly different (Morgan-Richards and Trewick, 2005).
The genus Acanthoxyla Uvarov 1944 is the most speciose of the
nine endemic New Zealand stick insect genera (Jewell and Brock,
2002). All eight species of Acanthoxyla are obligate parthenogens,
entirely lacking males but with females routinely producing ‘‘fertile”
eggs. The extent of morphological diversity within Acanthoxyla is
unusual and includes variation in colour, the number and size of
spines and development of abdominal flanges. These morphologi-
cally distinct parthenogenetic lineages appear to share a common
origin, whereas other hybrid animal species appear to result in sin-
gle morphological daughter lineages (Bullini, 1994). In comparison
to other taxa, the high diversity of cuticle colour and texture
expressed in Acanthoxyla, contrasts with low mitochondrial DNA
sequence variation (<2% over 1448 bp of COI and COII) within the
genus (Morgan-Richards and Trewick, 2005). Analysis of mitochon-
drial and nuclear DNA sequences, in addition to evidence from allo-
zyme data, karyology and flowcytometry (Morgan-Richards and
1055-7903/$ - see front matter Ó2008 Elsevier Inc. All rights reserved.
doi:10.1016/j.ympev.2008.05.025
* Corresponding author. Fax: +64 6 350 5623.
E-mail address: s.trewick@massey.ac.nz (S.A. Trewick).
Molecular Phylogenetics and Evolution 48 (2008) 799–808
Contents lists available at ScienceDirect
Molecular Phylogenetics and Evolution
journal homepage: www.elsevier.com/locate/ympev
Author's personal copy
Trewick, 2005) indicate that the Acanthoxyla lineage most probably
originated from one or few hybridization(s) between male Clitarchus
hookeri (White, 1846) and another, unknown, species (see Table 1).
Evidence from DNA sequence data could be interpreted as indi-
cating more than one hybridization event in Acanthoxyla but there
is no direct evidence for eight separate events to match the species
number. The presence of two distinct and size polymorphic classes
of ITS nuclear DNA sequences within Acanthoxyla demonstrates its
hybrid origin (Clades I and II; Fig. 1). Acanthoxyla individuals have
ITS DNA sequences belonging to one or both clades, and individuals
of the same morpho-species express these alternative conditions.
One clade of ITS sequences evidently originated in Clitarchus as the
identical sequence occurs in both Clitarchus hookeri individuals
and some Acanthoxyla individuals. PCR assays indicate the operation
of concerted evolution in homogenising the nuclear rDNA gene clus-
ter in some lineages (Morgan-Richards and Trewick, 2005). A combi-
nation of morphological (taxonomic), ITS (nuclear) and mtDNA
(maternal) data reveals a complex though apparently shallow his-
tory (Fig. 1). Significantly, these data cast doubt on the validity of
the current taxonomic treatment, and instead indicate that numer-
ous parthenogenetic lineages have independently converged on a
range of similar morphotypes. Identifying both parental species
would simplify estimation of the timing and number of hybridiza-
tion events and facilitate exploration of genomic processes that have
resulted in the morphological diversity, distinct karyotype and in-
creased DNA content that characterise this genus (Morgan-Richards
and Trewick, 2005; unpublished data). We would expect the mater-
nal ancestor to have ITS sequences identical (or almost identical) to
the ITS sequences that form the clade unique to Acanthoxyla (Clade
I), because the putative paternal ancestor (Clitarchus hookeri) and
Acanthoxyla share identical ITS sequences in Clade II (Fig. 1).
If the maternal ancestor of Acanthoxyla is extant, where should
we look for it? All New Zealand stick insects lack wings and New Zea-
land is separated by at least 1200 km of ocean from other major
landmasses. Isolation of the New Zealand continent (Zealandia) by
continental drift started about 80 MA and as a consequence it is
has been widely assumed that the biota of New Zealand is similarly
ancient and isolated (Goldberg et al., in press; Trewick and Morgan-
Richards, in press). However, recent molecular work has revealed
that many New Zealand endemic plants and animals share common
ancestors with Australian and Pacific taxa only a few million years
ago and are thus not monophyletic (Waters et al., 2000; Chambers
et al., 2001; Winkworth et al., 2002; Vink and Paterson, 2003; Arens-
burger et al., 2004; and see Trewick et al., 2007). In order to focus our
search for a maternal ancestor for Acanthoxyla we need to know if
the New Zealand phasmid fauna is monophyletic.
The wingless New Zealand fauna are thought to be related to Aus-
tralian phasmids. The Australian phasmid fauna comprises some
200 species placed within ten subfamilies, including the two (Pachy-
morphinae and Phasmatinae; Appendix 1) in which the New
Zealand species are currently placed, and the New Zealand mono-
typic genus Micrachus hystriculeus (Westwood, 1859) was formerly
placed within the Australian genus Pachymorpha (Salmon, 1991).
Hence, taxonomy implies that the New Zealand phasmid do not
form a monophyletic group. In general, higher level taxonomy of
stick insects is widely regarded as in need of work (Otte and Brock,
2003). Constraints of morphology imposed by crypsis (stick mim-
icry) have restricted cladistic resolution, and the solution has been
the identification of one or few key synapomorphies. Thus, we made
few assumptions as to the likely relatives of the New Zealand fauna.
The identification of appropriate outgroup taxa for phylogenetic
analysis is both important and problematic (Lyons-Weiler et al.,
1998; Milinkovitch and Lyons-Weiler, 1998; Holland et al., 2003).
Even where it is possible to obtain sequence data from a wide
diversity of potential outgroup taxa, the severest limitations on
computing efficiency time for phylogenetic analysis arise from tax-
on number. A compromise between increasing taxon sampling to
reduce phylogenetic error (Zwickl and Hillis, 2002) and minimising
taxon sampling to reduce computing time is sought. A further
problem with an unconstrained sampling scheme is that sequence
alignment is frequently more demanding and perhaps less reliable
where length-polymorphic genes are concerned (e.g. components
of the rDNA cluster). Sequence alignment can be critical for reliable
phylogenetic inference, therefore it is preferable to work with a
relatively small number of taxa that minimises the number of
likely alternative alignments (Ogden and Rosenberg, 2006). In
addition, sequence alignment should not be subject to circular rea-
soning by the assumption of evolutionary relationships. The New
Zealand stick insects provide an exemplary case where neither tax-
onomy nor biogeography can be relied upon for outgroup selec-
tion. Indeed it is possible that some putative outgroup taxa (i.e.
non-New Zealand species) could in fact be unidentified ingroups.
Therefore we devised an efficient and convenient method (SeqSSi)
to minimise the assumptions in outgroup identification. For our
study we were in the convenient position of being able to take
advantage of existing data representing stick insect diversity
worldwide (Whiting et al., 2003).
2. Materials and methods
2.1. Rationale
To investigate the evolutionary relationships within the New
Zealand fauna Morgan-Richards and Trewick (2005) sequenced
protein coding mitochondrial genes (COI and COII) and nuclear
internal transcribed spacers (ITS) from an extensive sample of indi-
viduals and species representing all New Zealand genera except
Pseudoclitarchus (Salmon, 1991). The distribution of genetic diver-
sity within the Acanthoxyla clade was not concordant with mor-
phological diversity implying that the currently recognised
obligate parthenogenetic species are not monophyletic lineages.
From analysis of nuclear (ITS) and mitochondrial (COI, COII) DNA
sequence data it was inferred that the mitochondrial genome of
Acanthoxyla came from a maternal bisexual ancestral species
(mother) and that two nuclear lineages (mother and father) remain
in the nuclear rDNA gene cluster of this genus (Morgan-Richards
and Trewick, 2005)(Fig. 2). These nuclear lineages are the signa-
tures of the two ancestral (maternal and paternal) species from
which Acanthoxyla originated.
However, because of the limited capacity of these mitochon-
drial protein coding genes to accumulate phylogenetic signal at
Table 1
Genomic anatomy of Acanthoxyla, the New Zealand obligate parthenogenetic stick
insect genus
Taxonomy Genus of eight species recognised on morphological grounds
a
Reproduction Obligate parthenogens, males never recorded, no females
observed in copula
b
mtDNA Distinct lineage with little sequence divergence within genus
c
Nuclear DNA
sequence
At least two nuclear sequence-lineages, one shared with
Clitarchus hookeri, the others are unique
Karyotype Diploid number same as Clitarchus hookeri but karyotype
dominated by metacentrics which are absent from C. hookeri
karyotype
c,d
Allozymes Fixed heterozygosity at 1 locus (Gpi) in some Acanthoxyla
lineages. All 8 alleles at 8 loci found in Clitarchus occur in
Acanthoxyla, but Acanthoxyla also has 3 alleles not found in
Clitarchus
e
a
Jewell and Brock (2002).
b
Salmon (1991).
c
Morgan-Richards and Trewick (2005), Buckley et al. (2008).
d
Parfitt (1980).
e
Buckley (1995).
800 S.A. Trewick et al. / Molecular Phylogenetics and Evolution 48 (2008) 799–808
Author's personal copy
deeper taxonomic levels (Szymura et al., 1996; Paton and Baker,
2006) we have here further utilised the nuclear rDNA gene cluster.
In their analysis of DNA sequences from stick insects and other in-
sect orders where 28S represented a little less than 50% of the se-
quence data, Whiting et al. (2003) found that it provided some 62%
of the phylogenetic signal and provided resolution throughout the
tree topology. Hence our present analysis focuses on this gene,
which provides a good chance of phylogenetic resolution, the
opportunity to incorporate an existing extensive resource of data,
and take advantage of the presence of both the maternal and pater-
nal lineages in Acanothoxyla already identified by ITS sequences
(Fig. 2).
2.2. Taxon sampling
Twenty-two species and nine genera of New Zealand stick
insect are recognised by Jewell and Brock (2002). The genera
(and number of species in brackets) for each subfamily are: Pachy-
morphinae—Micrarchus (1), Niveaphasma (1), Asteliaphasma (2),
Tectarchus (4), Spinotectarchus (1); Phasmatinae—Acanthoxyla (8),
0.001 substitutions/site
Ap.Khd-11
Ap.Bid-86
Ax.Bid-104
Ap.PN-2
Ax.Mak-142
Ap.Otr-30
PN.ac-1
Ap.Bal-1s
Ax.Opa-103s
Ap.Bal-32s
Khd-3
Ap.Khd-10
Dun-1
Ap.Dun-47
Stf-3s
Ap.Otr-31
Ap.Otr-29
Ax.Otk-146
Ap.Khd-14
Ax.Opa-137
Ax.Opa-136
Pse230
Pse231
Ap.Bal-1f
Ap.Bal-32f
Ax.Him-138
Ax.Omo-109
Ax.Omo-110
Ch.Wby-45
Ch.Rn-62
Ch.Khd-61
Ax.Bid-105
Ax.Opa-103f
Ax.Dun-148
Ot.Ch-1
Ap.Rn-1
Ap.PN-17
Ap.PN-16
Ap.Khd-15
Stf-3f
Arw-3
Ch.Bal-141
Ch.Bid-20
Ch.GB-37
Ch.GB-35
Ch.GB-36
Ch.GB-34
Hhh-2
Ch.Opa-67
Eintermedia
Mspeciosa
Mintermedia
Jintermedia
Bgeisovii
Dhuttoni
Kinermis
Bprasina
Aprasina
Bprasina
Cnr geisovii
Cgeisovii
Hgeisovii
Iprasina
Gprasina
Bnr geisovii
Fnr geisovii
Fintermedia
Fspeciosa
Lsuteri
Lprasina
Bprasina
Bprasina
Bprasina
inermis
Binermis
Ainermis
Aprasina
Bprasina
Ainermis
Ainermis
Nnr geisovii
inermis
Gprasina
Pseudoclitarchus
MtDNA
Clade I
Clade II
Fig. 1. Evidence from nuclear (rDNA) sequence that Clitarchus hookeri maybe one parent in hybrid origin of Acanthoxyla (Morgan-Richards and Trewick, 2005). Midpoint
rooted Neighbor Joining (NJ) tree of ITS1&2 (internal transcribed spacers) nuclear DNA sequences comprising three clades: Aconthoyxla (Clade I), Pseudoclitarchus (Pse), and
Clitarchus plus Acanthoxyla (Clade II). Acanthoxyla individuals are labelled with bold lettering, and annotated with species name (indicative of different morphologies) and
COI–COII mtDNA haplotype (bold uppercase letter). Acanthoxyla individuals found to have both classes of ITS sequence are indicated by star, and the two individuals (Khd-3,
Ap.Rn-1) used in other analyses presented here are indicated by grey boxes.
S.A. Trewick et al. / Molecular Phylogenetics and Evolution 48 (2008) 799–808 801
Author's personal copy
Argosarchus (2), Clitarchus (2), Pseudoclitarchus (1). Of these, we
have sampled 17 species for the present analysis, representing all
genera (Appendix 1).
The species not included in the present study are unlikely to
contribute to the current analysis, and comprise: (1) one species
each from two genera (Clitarchus and Argosarchus) that were
recently inferred from DNA sequence data to be monotypic (Tre-
wick et al., 2005). (2) Despite several attempts, nobody has found
Asteliaphasma naomi (Salmon, 1991) since its description from a
single specimen. However, we have collected Asteliaphasma jucun-
da (Salmon, 1991) from the type location of A. naomi.A. naomi may
have been an unusual morph of A. jucunda. (3) Three of the four
Tectarchus (Salmon, 1948) species (T. huttoni (Brunner, 1907),
T. ovobessus Salmon, 1948,T. salebrosus (Hutton, 1899)) are
included but not the rare T. semilobatus (Salmon, 1948). (4) The
present analyses include seven of the eight described species of
Acanthoxyla (Fig. 1). Individuals representing two Acanthoxyla
species (A. geisovii, A inermis) were selected for analysis of 28S as
representing the extremes of COI–COII (1.8%) and ITS sequence di-
verge (Morgan-Richards and Trewick, 2005; see summary in Fig. 1).
We include additional geographic isolates of Micrarchus hystric-
uleus and Niveaphasma annulata (the latter including a putative
new species identified by Tony Jewell) and the monotypic sexual
species Pseudoclitarchus sentus Salmon, 1991.Pseudoclitarchus sen-
tus is restricted to a single offshore island that is a New Zealand
Scientific Reserve (Great King Island, Three Kings Islands) and as
a consequence was previously not available for analysis (Morgan-
Richards and Trewick, 2005). The inclusion of Pseudoclitarchus in
the present analysis means that all New Zealand Phasmatinae
(the subfamily that includes Acanthoxyla) are now represented.
Although Salmon (1948) suggested Acanthoxyla was taxonomically
near Clitarchus and Pseudclitarchus,Acanthoxyla species combine a
suite of features including highly sculptured egg capsules, presence
of abdominal flanges, well-developed thoracic spines and a distinc-
tive opercular spine, and colours not seen in other stick insects in
New Zealand.
Non-New Zealand stick insects were sampled using published
sequences for a set of species reported by Whiting et al. (2003)
and available from GenBank, with additional taxa from the West
Pacific: Australia (3 species), New Caledonia (3 species), Fiji (2 spe-
cies) and the Solomon Islands (2 species) (see Appendix). Identifi-
cation of Pacific island and Australian stick insects was provided by
Paul D. Brock (Natural History Museum, London) and Geoff Mon-
teith (Brisbane Museum, Australia), respectively. Together these
taxa comprise representatives of a range of subfamilies including
those (Pachymorphinae, Phasmatinae) found in New Zealand
(Appendix 1).
2.3. DNA extraction, amplification and sequencing
In most instances muscle tissue from fresh, frozen, or alcohol
preserved specimens was removed from a leg for genomic DNA
extraction using a salting-out method (Sunnucks and Hale, 1996).
Tissue was macerated and incubated with 5
l
L of 10 mg/mL
Acanthoxyla
Acanthoxyla
Acanthoxyla
Clitarchus
Clitarchus
Pseudoclitarchus
Pseudoclitarchus
Argosarchus
Argosarchus
a
b
Fig. 2. Unrooted MP trees for (a) COI–COII mitochondrial, and (b) 28S and ITS nuclear DNA sequence data for representatives of the New Zealand ‘‘Phasmatinae”. A single
shortest tree was returned in each case. The grey arrow highlights the placement of Acanthoxyla’s two distinct 28S and ITS DNA sequence variants. Male and female symbols
indicate inferred paternal and maternal lineages that gave rise to the Acanthoxyla lineage.
802 S.A. Trewick et al. / Molecular Phylogenetics and Evolution 48 (2008) 799–808
Author's personal copy
Proteinase-K in 600
l
L of TNES buffer (20 mM EDTA, 50 mM Tris,
400 mM NaCl, 0.5% SDS) at 50 °C for 1–4 h. 10% 5 M NaCl was
added and the extractions shaken vigorously for 20 s followed by
spinning at 14,000 rpm for 5 min. The supernatant was removed
and precipitated with an equal volume of cold 100% ethanol.
DNA was collected by spinning and washed with 70% ethanol, then
dried and dissolved in water. In the case of pinned museum spec-
imens (from Australia and New Caledonia), DNA extraction used
incubation at 55 °C with Proteinase-K and a CTAB buffer (2% Hex-
adecyltrimethylammonium bromide, 100 mM Tris–HCl, pH 8.0,
1.4 M NaCl, 20 mM EDTA), followed by a combined phenol/chloro-
form/isoamyl alcohol (25:24:1) cleanup.
2.4. PCR
To confirm the relationship and expected consistency of ITS and
28S sequences within Acanothoxyla we sequenced these genes plus
the mitochondrial fragment comprising partial COI and COII genes
from representatives of Acanthoxyla species (morphotypes) and the
other New Zealand Phasmatinae (Argosarchus,Clitarchus,Pseudocl-
itarchus). We compiled COI–COII data from representatives of all
New Zealand genera to assess the level of genetic diversity among
them. All taxa were subjected to PCR and sequencing targeting the
nuclear rDNA gene 28S.
The mitochondrial fragment, comprising the 3
0
end of cyto-
chrome oxidase I (COI), tRNA-Leucine, and cytochrome oxidase II
(COII) was amplified using the primers C1-J-2195 and TK-N3785
(Simon et al., 1994). The internal transcribed spacers (ITS regions
1 and 2) and the intervening 5.8S gene were amplified using the
primers ITS4, ITS5 (White et al., 1990) and STITS5F (Morgan-Rich-
ards and Trewick, 2005). For the majority of samples, 28S
sequences were amplified using the primers 1.2a, 28Sb and 28Sa,
7.1b (Whiting et al., 1997), although older material required ampli-
fication and sequencing with additional internal primers (4.2, 4.8,
5b) targeting shorter fragments.
PCR used standard conditions (Trewick et al., 2000). Amplifica-
tion products were treated to Shrimp Alkaline Phosphotase/Exonu-
clease I digestion prior to sequencing. Cycle sequencing with the
PCR primers used Bigdye chemistry (PE) following the manufac-
turer’s protocols, with automated reading on an ABI3730. Consen-
sus sequences were obtained using Sequencher v4.1 (ABI, PE), and
aligned using SeAl v2.0a3 (Rambaut, 1996).
2.5. SeqSSi
We devised a targeted BLAST-based (Altschul et al., 1997)
search algorithm to identify suitable outgroup and previously
unidentified ingroup taxa for our analyses of the New Zealand stick
insects, instead of using phylogeny or taxonomy based approaches.
This is particularly useful for the analysis of DNA with INDELS as
outgroup selection is done before multiple DNA sequences are
aligned. Ribosomal genes frequently contain many INDELS and
are thus sensitive to alignment bias.
The SeqSSi (Sequence Similarity Sieve) program is a Perl script
developed to search a group of sequences for those that are the
most similar to a nominated ingroup and thus locate possible out-
groups for further phylogenetic analysis. SeqSSi uses the stand-
alone gapped BLAST package (Altschul et al., 1997) available from
NCBI (http://www.ncbi.nih.nlm.gov). Sequences are input in a sin-
gle FASTA formatted file with the nominated ingroup of sequences
in a block at the end of this file. SeqSSi firstly formats the input file
(calling the program formatdb.exe) as a database for subsequent
searching by the blastall.exe program, then searches all of the se-
quences against every other sequence in the input file. SeqSSi
can use nucleotide or protein sequences as sequence type (aa or
nt) as nominated by the user. Our study used the BLAST default
parameters with the input sequences in nucleotide format (blastn
program option). Importantly, instead of searching GenBank in its
entirety, the SeqSSi procedure uses prior information, so that
ingroup and potential outgroup DNA sequences are matched for
locus coverage and broad taxonomic relevance (e.g. in this case
28S gene, and phasmids).
The BLAST program is a widely used tool for searching protein
and nucleotide sequence databases for similar sequence (Altschul
et al., 1997). Scores are produced for each similarity hit to indicate
the level of similarity between the query (input from user) and
subject (from database) sequences. In this case the subject and
query sequences are homologous as they represented the full
sequence of a single gene. After the BLAST search, SeqSSi records
each score from the generated output file. For each sequence in
the input file, SeqSSi records the difference between the top score
(which is always the 100% match between this sequence and itself
found in the database) and the scores for all the other sequences.
Thus, similar sequences will have very low difference values and
dissimilar sequences higher difference scores.
For each DNA sequence in the ingroup, the difference scores are
compared at different ranges to produce a rating of the outgroup
sequences based on their similarity to ingroup sequences. For our
data we used the following category bounds: category A (1–500
difference score), category B (501–1000 difference score), category
C (1001–2000 difference score) and category D (above 2001 differ-
ence score). Outgroup sequences are then ranked according to this
rating and how many times each sequence appears in each cate-
gory. This creates a list of outgroup sequence codes, starting with
those that appear most frequently in category A, then B then C
(D should contain all the sequences and is used as a control). We
summarise this output as a table comprising differences scores
(expressed as A–C rank codes for simplicity) for comparisons of
ingroup species sequences versus outgroup sequences (Table 2).
The incidence of A and B category difference scores was used to
select sequences of suitable outgroup species that were aligned
with ingroup sequences for phylogenetic analysis.
Statistical analysis of SeqSSi results was performed with the R
statistics software V2.4.0 (Ihaka and Gentleman, 1996). Difference
scores from outgroups against ingroup taxa were plotted as density
distributions so that the histogram had a total area of 1. Difference
scores for individual outgroup sequences were plotted against the
original histograms to obtain ‘SeqSSi-plots’ that express the distri-
bution of difference scores for specific outgroup taxa in the context
of the overall score distribution. Outgroup taxa that showed a
higher proportion of SeqSSi score differences in the lower ranges
(SeqSSi score difference <1000, i.e. category A and B in Table 2)
were included in subsequent analysis (Fig. 3).
2.6. Phylogenetic analysis
We used PAUP*4.0b9 (Swofford, 2002) to implement phyloge-
netic analyses with Neighbor Joining (NJ), Maximum Parsimony
(MP), and Maximum Likelihood (ML) criteria, and MrBayes
v3.1.2. (Ronquist and Huelsenbeck, 2003) for Bayesian analysis.
To select among alternative models of DNA evolution for the ML
and NJ analyses we used Modeltest version 3.06 (Posada and
Crandall, 1998).
For MP we used unweighted data, while for ML and NJ analysis
we used Modeltest v3.06 to separately estimate parameters under
the GTR+I+
C
model selected for the 28S and COI–COII data sets,
respectively. Bayesian analyses of 28S data used a GTR model with
gamma-distributed rate variation across sites (with four catego-
ries), a proportion of invariant sites and default priors. We used
two independent, simultaneous runs with three heated chains
and a burnin of first 25% of trees.
S.A. Trewick et al. / Molecular Phylogenetics and Evolution 48 (2008) 799–808 803
Author's personal copy
Analysis of COI–COII sequence data was undertaken with tRNA
Leucine excluded. Analysis of 28S DNA sequences used outgroup
taxa selected by SeqSSi analysis that received a minimum of two
hits in the category B ranking (this is an arbitrary cut off chosen
to provide an optimal number of outgroup taxa for this study).
Then 28S sequences for these outgroup (non-New Zealand) taxa
were aligned with those representing all New Zealand taxa, either
manually using SeAl v2.0a3 (Rambaut, 1996) or employing the
computer programs Muscle v3.6 (Edgar, 2004a, 2004b) and Clu-
stalW v1.82 (Higgins et al., 1994). For comparison we also gener-
ated alignments of the 28S sequences for the complete taxon set
using Muscle v3.6 (Edgar, 2004a,b) and ClustalW v1.82 (Higgins
et al., 1994) with default settings as implemented by Geneious
v3.5.4 (Drummond et al., 2007). Output alignments were then sub-
jected to Bayesian analysis with no data excluded.
Tree outputs were manipulated for display using TreeViewX
v0.5 (Page, 1996).
3. Results
The aligned COI–COII, ITS including the 5.8S, and 28S sequences
were 1380, 1806, and 2260 bp in length, respectively. New data
reported here are deposited on GenBank (Appendix). Aligned
sequences from New Zealand Phasmatinae (Acanthoxyla,Clitarchus,
Pseudoclitarchus,Argosarchus) yielded a data set with 3966 bp of
nuclear (28S and ITS) and 1380 bp of mitochondrial (COI, COII)
DNA sequence. Pairwise genetic distances estimated using a
GTR+I+
C
model applied to the COI–COII mtDNA data for represen-
tatives of New Zealand taxa are high (max 0.45), and comparisons
with non-New Zealand stick insects yielded similar high distances
indicating saturation.
Genetic distances (COI–COII) among the New Zealand Phasmat-
inae (Argosarchus,Clitarchus,Pseudoclitarchus,Acanthoxyla) were
similar to between family distances (Phasmatinae–Pachymorphi-
nae) within New Zealand. Either (a) the systematics of these stick
Table 2
Summary of SeqSSi analysis (Sequence Similarity Sieve algorithm) of 28S DNA sequence data, to find suitable outgroup taxa for the New Zealand stick insects
Letters A, B and C indicate ranked similarity score (A > B > C) for each outgroup/ingroup taxon combination. Missing entries indicate instances where similarity ranking was
lower than category C. For clarity only the summarised results of the top 21 of 44 potential outgroup taxa and one representative of each ingroup species are shown. Shading
indicates highest ranked taxa used in subsequent phylogenetic analysis.
Fig. 3. Example ‘SeqSSi-plots’ of difference scores obtained with SeqSSi (Sequence Similarity Sieve). The background histogram plots the probability densities of the total
SeqSSi score differences and is identical for the two plots shown here. Solid lines represent the relative densities of taxa that were included as outgroups in our phylogenetic
analysis. The taxa represented by dotted lines had profiles typical of those not used in our subsequent analysis. Outgroup taxa selected had a higher proportion of SeqSSi score
differences in the lower ranges (A and B in Table 3) whereas excluded taxa had a higher proportion of their scores in the higher ranges (C and above). (A) The proportional
density of the difference scores obtained with Outgroup 5-Ctenomorphodes (solid line) compared with those obtained with Outgroup 12-Neohirasea (dotted line). (B) The
proportional density of the difference scores obtained with Outgroup 3-Eurycnema (solid line) compared with those obtained with Outgroup 22-Gratida (dotted line).
804 S.A. Trewick et al. / Molecular Phylogenetics and Evolution 48 (2008) 799–808
Author's personal copy
insects is inappropriate and/or (b) the information content of the
mtDNA data at this level is negligible. The lowest intergeneric dis-
tance observed was between Pseudoclitarchus and Clitarchus (0.08);
a relationship confirmed by phylogenetic analysis (Fig. 2a).
Interspecific genetic distances from spatially separated
samples were typical of those observed in New Zealand insects
(0.02–0.05, Table 3).
The high functional constraint operating on the protein coding
genes (COI–COII) imposes a major limitation on their phylogenetic
utility and precludes their use for analysis of the relationships of
stick insects at deeper taxonomic levels. Analysis of the 28S and
ITS sequences together and apart revealed the same signal for
two distinct nuclear lineages within Acanthoxyla (Fig. 2b). Given
the close physical and thus heritable association of these two genes
in the rDNA cluster this congruence is as expected. The addition of
28S sequence data confirms the previous observation of two
(parental) lineages in Acanthoxyla (one close to Clitarchus hookeri),
and the inclusion of new nuclear sequence data (28S and ITS) from
Pseudoclitarchus sentus reveals that this species was not the source
of either of these (Figs. 1 and 2).
SeqSSi results facilitated the selection of 11 putative outgroup/
unidentified ingroup species (from a set of 44) with a ranked sim-
ilarity score of A or B with at least one New Zealand stick insect
species (Table 2). Examples of SeqSSi-plots are shown in Fig. 3.In
these plots, two outgroups are plotted onto each graph to contrast
the difference between sequences that were included in our anal-
ysis and those which were excluded. The 11 outgroup species
selected with SeqSSi originated from Australia (5), Fiji (1), New
Caledonia (1), New Guinea (1), South America (2) and the Carib-
bean (1). Analyses of 28S with New Zealand taxa and the 11 species
selected by SeqSSi for the outgroup reveal two important features,
(a) the New Zealand stick insects form a monophyletic clade in
most cases (see below), and (b) Acanthoxyla forms a well-sup-
ported clade with Clitarchus and Pseudoclitarchus. The monophyly
of the New Zealand taxa is supported by Bayesian analysis imple-
mented by MrBayes and ML implemented by PAUP. In these
analyses Spinotectarchus acornutus is placed as basal to the other
New Zealand taxa. Bayesian posterior probabilities indicate strong
support for this monophyletic grouping (Fig. 4), as does bootstrap
resampling (10,000 replicates) using NJ distance criteria with a
GTR+I+
C
model. To save computing time, a reduced data set of
12 taxa comprising representatives of the New Zealand fauna plus
just 3 outgroup taxa, selected on the basis of previous analysis
were subjected to ML (PAUP) bootstrap analysis (1000 replicates)
employing a GTR+I+
C
model. This also revealed strong support
for the monophyly of the New Zealand taxa (Fig. 5).
Bayesian analyses of 28S sequences in both, the full taxon set
and the SeqSSi selection, each aligned using Muscle v3.6 and Clu-
stalW v1.82, were entirely consistent with the observation above.
There was high statistical support (posterior probabilities of 0.99
or 1.00) for monophyly of New Zealand stick insects, with, in all
cases Spinotectarchus being sister to all other New Zealand species.
These analyses also supported the grouping of Acanthoxyla,Clitar-
chus and Pseudoclitarchus (as above), although internal edge
lengths for this clade were disproportionately short in trees that
included all taxa (i.e. including very distant relatives in the
outgroup). It was clear that resolution of ingroup structure was
enhanced by exclusion of least-similar sequences.
However, NJ using HKY and simpler models of DNA evolution,
and MP using unweighted data, applied to the full 28S data set,
each returned trees with a topology that differed in the placement
of Spinotectarchus among the outgroup taxa. MP bootstrap resam-
pling (10,000) replications gave 84% for the New Zealand clade
without Spinotectarchus. These latter results most likely reflect
the high degree of variation among taxa (a tendency for long
branch attraction) and model misspecification. In contrast, the
clade comprising Acanthoxyla,Clitarchus and Pseudoclitarchus was
returned by all analyses.
4. Discussion
The SeqSSi method of outgroup selection was extremely useful
in reducing the number of taxa that might be a suitable outgroup
before alignment of the INDEL-rich rDNA sequences. Outgroup
selection prior to phylogenetic analysis can be difficult when there
are a large number of species to choose from, taxonomy is unreli-
able or incomplete and alignment error or bias may result in mis-
leading topologies or poor tree resolution. While BLAST results
should generally be treated with caution in any evolutionary anal-
ysis, the use of BLAST to select outgroup taxa in the targeted man-
ner performed by SeqSSi, has potential. Although SeqSSi is still
under development and is undergoing further testing and evalua-
tion, we note that in this case the ingroup identified using SeqSSi
was also revealed by Bayesian analysis of all available taxa. This
supports our inference that SeqSSi correctly identified the most-
suitable outgroup taxa (by excluding sequences with lowest
Table 3
Pairwise genetic distances using a GTR+I+
C
model for COI–COII mtDNA nucleotide sequence data (excluding tRNALeu), for representatives of the New Zealand stick insects
S.A. Trewick et al. / Molecular Phylogenetics and Evolution 48 (2008) 799–808 805
Author's personal copy
similarity). The value of SeqSSi is evident in enhanced phylogenetic
resolution of the ingroup in analysis of the SeqSSi sample, and this
probably reflects better alignments and enhanced model specifica-
tion obtained from a more appropriate sample set.
The present data do not support monophyly of the subfamilies
Phasmatinae and Pachymorphinae, neither within New Zealand
nor across the geographic/taxonomic sample as a whole. The rep-
resentatives of the two New Zealand subfamilies appear to be
polyphyletic, including poor support for the grouping of Argosar-
chus with the other New Zealand Phasmatinae (Acanthoxyla,
Clitarchus,Pseudoclitarchus). However, there is good evidence that
the New Zealand fauna as a whole form a monophyletic group.
The monotypic genus Spinotectarchus was consistently returned
as sister to the other New Zealand taxa and as such these data
support the classification of the genus Asteliaphasma Jewell and
Brock (2002), splitting A. jucundus from S. acornutus; species that
are clearly not sister taxa as implied by the previous classifica-
tion (Salmon, 1991). Interestingly, the only New Zealand genus
apart from Acanthoxyla with more than two described species
in this study (Tectarchus) consistently failed to form a monophy-
letic clade, emphasising the inherent difficulty of resolving rela-
tionships within a group where morphology is extremely
constrained for crypsis.
We cannot reject the hypothesis that the parthenogenetic genus
Acanthoxyla is part of an exclusively New Zealand clade. We have
included all described New Zealand genera and most species of
New Zealand stick insects. It is possible that species remain unde-
scribed, but it is less likely that genera remain unnoticed. Although
two new New Zealand stick insect genera were recently erected,
the species they comprise were already well characterised (Jewell
and Brock, 2002). We have failed to find a mother for Acanthoxyla,
but know where to look; New Zealand.
If a mother existed, we expected the unique Axanthoxyla ITS se-
quences of Clade I to be nested within its sexual maternal species
in much the same way as the Acanthoxyla Clade II sequences are
nested within Clitarchus hookeri diversity. In addition, mitochon-
drial DNA sequences from a maternal species would form part of
the existing Acanthoxyla mtDNA clade. Salmon (1991) proposed,
on the basis of morphological evidence that the genus ‘‘Pseudocl-
itarchus forms an evolutionary link between Clitarchus and Acanth-
oxyla”. Indeed, Pseudoclitarchus sentus was originally described by
Salmon (1948) as an Acanthoxyla species (Acanthoxyla senta).
Although a clade consisting of these three genera is supported by
all DNA data, all analyses show that the sexual taxon Pseudoclitar-
chus sentus is not the maternal ancestor of Acanthoxyla.(Buckley
et al., 2008). On the basis of mtDNA nucleotide data, Pseudoclitar-
chus is genetically closer to Clitarchus than Acanthoxyla.
The sexual mother of Acanthoxyla appears to have been sister to
Clitarchus and Pseudoclitarchus (Fig. 2). This relationship, indicated
by 28S sequence data (Figs. 4 and 5) is corroborated by analysis of
mtDNA sequence indicating that the Acanthoxyla maternal lineage
is genetically closer to these than any other taxa known. This sug-
gests that the parthenogenetic Acanthoxyla and, more specifically,
the half of the genome represented by the unique 28S/ITS se-
quences are all that is left of the sexual maternal species. An alter-
native explanation is that the presence of Clitarchus 28S/ITS
sequences in the genome of Acanthoxlya may be the result of cur-
rent, opportunistic mating between male Clitarchus and Acanth-
oxyla (Buckley et al., 2008). This would require that Acanothyxla
is a facultative parthenogen rather than an obligate parthenogen.
Cn.rac
AY125315
fiji21
AY125300
AY125326
AY125318
AY125294
Oz.Pac1
AY125319
AY125295
AY125296
BluS246
Otr1
Mi.Bid87
Niv.Pis126
nadun42
Ast.Opa72
Ast.LWk152
Ts.PH1
Khd2
Tec.Rg241
Otr5
Arg.Omo159
Khd3
ApRn1
HhH2
HhH1
Pse231
Pse230
Pse232
Sp.Opa70
S
p
.Co238
0.73
0.58
0.75
1.00
0.87
0.65
0.99
0.98
1.00
1.00
1.00
0.91
1.00
1.00
1.00
1.00
1.00
1.00
0.99
1.00
1.00
0.95
1.00
0.97
1.00
Acanthoxyla nr geisovii
Acanthoxyla inermis
Clitarchus hookeri
Pseudoclitarchus sentus
Spinotectarchus acornutus
Argosarchus horridus
Tectarchus huttoni
Tectarchus ovobessus
Tectarchus salebrosus
Asteliaphasma jacunda
Niveaphasma annulata
Niveaphasma sp.
Micrarchus hystriculeus
Micrarchus sp.
Graeffea crouanii
Pachymorpha sp.
Tropidoderus childrenii
Cnipsus rachis
Ctenomorphodes briareus
Extatosoma tiaratum
Eurycnema goliath
Eurycantha insularis
Lamponius guerini
Agathemera crassa
Paraphasma rufipes
New Zealand
Fig. 4. Consensus tree from Bayesian analysis of 28S rDNA stick insect data using GTR+I+
C
with outgroup taxa selected by SeqSSi. Posterior probabilities estimated after
removal of 25% burnin are given at nodes. The star indicates the base of clade comprising New Zealand stick insects, triangle indicates clade that includes Acanthoxyla.
806 S.A. Trewick et al. / Molecular Phylogenetics and Evolution 48 (2008) 799–808
Author's personal copy
An inter-specific reproductive strategy of this type (hybridogene-
sis—Bullini, 1994) is known among Sicillian stick insects (Bacillus).
In that case, successful egg production requires mating with males
of the paternal species each generation but the paternal genome is
discarded prior to gametogenesis, and the taxa involved (B. rossius-
grandii) are congenerics (Mantovani and Scali, 1992; Mantovani
et al., 2001). Karyological evidence for New Zealand stick insects
indicates such a situation is unlikely (Morgan-Richards and Tre-
wick, 2005; Unpublished data). The karotypes of some Acanthoxyla
species (lineages) are diploid and distinct from that of Clitarchus.
Furthermore, the karyotype of Acanthoxyla with a single 28S/ITS
sequence variant is the same as that of Acanthoxyla individuals
bearing both 28S/ITS sequences (i.e. including the Clitarchus type).
This would be an unlikely condition if hybridisation is ongoing.
Further evidence against this interpretation comes from extensive
field and captive observations; although sexual Clitarchus pairs are
a common sight, and Clitarchus and Acanthoxyla are often sympat-
ric, inter-generic coupling has never been observed.
Therefore, Acanthoxyla is probably an orphan, and it is interest-
ing to speculate that it may have competed with its bisexual
mother species and contributed to ‘her’ extinction. The success of
parthenogens is commonly associated with occupation of range
margins (geographic parthenogenesis) and has been explained by
avoidance of the homogenising effect of gene flow allowing adap-
tation to marginal habitats (Peck et al., 1998). In addition, the
numerical advantage of parthenogenetic reproduction during
range expansion associated with Pleistocene climate cycling may
have given Acanthoxyla an advantage over bisexual species. For
Acanthoxyla it will be difficult to distinguish pathenogenetic
advantage from hybrid advantage (Kearney, 2005), but either and
both conditions could have given the genus an adaptive advantage
over a sexual maternal species. It is therefore intriguing that
Acanthoxyla is widely sympatric with its paternal species, Clitar-
chus hookeri, which is sexual in northern New Zealand and fre-
quently asexual at its southern range limits. One expression of
hybrid vigour may be the food range exhibited by Acanthoxyla.
All Acanthoxyla species eat a very wide range of plant species; from
New Zealand endemics including rata (Myrtaceae) and totara
(Podocarpaceae) to introduced rose (Roseacea), pine (Pinaceae),
and cypress (Cupressaceae). This dietary plasticity has allowed
the successful invasion of Acanthoxyla into the United Kingdom
(Brock, 1987). In contrast, paternal Clitarchus has a much more lim-
ited range of host plants and a narrower species range (Trewick
and Morgan-Richards, 2005; Trewick, 2007).
Acknowledgments
Paul Brock assisted us with identification. Tony Jewell, Geoff
Monteith, Dinah & Scott Dunavan, Craig Morley, Rhys Richards,
Ted and Bee Trewick, Judy and Llyn Richards, Mary and Ralph Pow-
lesland, Richard Murray and family of Bluff Station, and Julia Gold-
berg, provided stick insects. Michael Whiting’s lab (Brigham Young
University) provided details of internal 28S primers. Matt Philips,
Martyn Kennedy, Barbara Holland and Klaus Schliep helped with
Oz.Pac1
AY125319
Mi.Bid87
nadun42
Ast.LWk152
Arg.Omo159
Tec.Rg241
Sp.Co238
Khd3
HhH2
Pse230
52
78
54
52
88
95
Acanthoxyla nr geisovii
Clitarchus hookeri
Pseudoclitarchus sentus
Spinotectarchus acornutus
Argosarchus horridus
Asteliaphasma jacunda
Niveaphasma annulata
Micrarchus hystriculeus
Pachymorpha sp.
Eurycnema goliath
Tropidoderus childrenii
Tectarchus ovobessus
New Zealand
AY125315
Fig. 5. Bootstrap consensus tree from Maximum Likelihood (ML) analysis of 28S rDNA stick insect data for a reduced taxon set, using GTR+I+
C
with 1000 replications.
Bootstrap values are given at nodes. The star indicates the base of clade comprising New Zealand stick insects, triangle indicates clade that includes Acanthoxyla.
S.A. Trewick et al. / Molecular Phylogenetics and Evolution 48 (2008) 799–808 807
Author's personal copy
analysis. The New Zealand Department of Conservation provided
permits and we are particularly grateful for the assistance of Tony
Beauchamp, Andrea Booth, Donna Stuthridge and Dave King. Spec-
imens of Pseudoclitarchus sentus were collected on the Three King’s
Islands Nature Reserve by DoC staff under permit NO-13868-FAU,
with permission of the tangata whenua, Ngati Kuri and Te Aupouri.
This work was partly funded by Massey University Research Fund
grant (to MMR; MURF 05/2044) and Marsden grant (to
SAT:GNS302).
Appendix A. Supplementary data
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.ympev.2008.05.025.
References
Altschul, S.F., Madden, T.L., Schaffer, A.A., Schnaffer, A.A., Zhang, J., Zhang, Z., Miller,
W., Lipman, D.J., 1997. Gapped BLAST and PSI-BLAST: a new generation of
protein database search programs. Nucleic Acids Res. 25, 3389–3402.
Arensburger, P., Buckley, T.R., Simon, C., Moulds, M., Holsinger, K.E., 2004.
Biogeography and phylogeny of the New Zealand cicada genera (Hemiptera:
Cicadidae) based on nuclear and mitochondrial DNA data. J. Biogeogr. 31,
557–569.
Brock, P.D., 1987. A third New Zealand stick-insect (Phasmatodea) established in
the British Isles, with notes on the other species, including a correction. In:
Mazzini, M., Scali, V. (Eds.), 1st International Symposium on Stick Insects.
University of Siena, pp. 125–132.
Buckley, T., 1995. An allozyme survey of stick insects (Insecta: Phasmida) from the
Wellington region. Unpublished Honours thesis, Victoria University of
Wellington.
Buckley, T.R., Attanayake, D., Park, D.-C., Ravindran, S., Jewell, T.R., Normak, B.B.,
2008. Investigating hybridization in the parthenogenetic New Zealand stick
insect Acanthoxyla (Phasmatodea) using single-copy nuclear loci. Mol. Phylogen.
Evol 48, 335–349.
Bullini, L., 1994. Origin and evolution of animal hybrid species. Trends Ecol. Evol. 9,
422–426.
Chambers, G.K., Boon, M.W., Buckley, T.R., Hitchmough, R.A., 2001. Using molecular
methods to understand the Gondwanan affinities of the New Zealand biota:
three case studies. Aust. J. Bot. 49, 377–387.
Coyne, J.A., Orr, H.A., 2004. Speciation. Sinauer Associates, Sunderland, MA.
Crespi, B.J., Sandoval, C.P., 2000. Phylogenetic evidence for the evolution of
ecological specialization in Timema walking-sticks. J. Evol. Biol. 13, 249–262.
Drummond, A.J., Ashton, B, Cheung, M., Heled, J., Kearse, M., Moir, R., Stones-Havas,
S., Thierer, T., Wilson, A., 2007. Geneious v3.5, Available from: <http://
www.geneious.com/>.
Edgar, R.C., 2004a. MUSCLE: multiple sequence alignment with high accuracy and
high throughput. Nucleic Acids Res. 32, 1792–1797.
Edgar, R.C., 2004b. MUSCLE: a multiple sequence alignment method with reduced
time and space complexity. BMC Bioinform. 5, 113.
Goldberg, J., Trewick, S.A., Paterson, A.M., in press. Evolution of New Zealand’s
terrestrial fauna: A review of molecular evidence. Philosophical Transcations of
the Royal Society, London.
Higgins, D., Thompson, J., Gibson, T., Thompson, J.D., Higgins, D.G., Gibson, T.J., 1994.
ClustalW: improving the sensitivity of progressive multiple sequence alignment
through sequence weighting, position-specific gap penalties, and weight matrix
choice. Nucleic Acids Res. 22, 4673–4680.
Holland, B.R., Penny, D., Hendy, M.D., 2003. Outgroup misplacement and
phylogenetic inaccuracy under a molecular clock—a simulation study. Syst.
Biol. 52, 229–238.
Ihaka, R., Gentleman, R., 1996. R: a language for data analysis and graphics, J. Comput.
Graph. Stat. 5, 299–314. See, R Development Core Team. 2006. R: A Language and
Environment for Statistical Computing, R Foundation for Statistical Computing,
Vienna, Austria. ISBN: 3-900051-07-0 <http://www.R-project.org>.
Jewell, T., Brock, P.D., 2002. A review of the New Zealand stick insects: new genera
and synonymy, keys, and a catalogue. J. Orthopt. Res. 11, 189–197.
Kearney, M., 2005. Hybridization, glaciation and geographical parthenogenesis.
Trends Ecol. Evol. 20, 495–502.
Law, J.H., Crespi, B.J., 2002. Recent and ancient asexuality in Timema walkingsticks.
Evolution 56, 1711–1717.
Lyons-Weiler, J., Hoelzer, G.A., Tausch, R.J., 1998. Optimal outgroup analysis. Biol. J.
Linn. Soc. 64, 493–511.
Mantovani, B., Scali, V., 1992. Hybridogenesis and androgenesis in the stick insect
Bacillus rossius-grandii benazzii (Insects, Phasmatodea). Evolution 46, 783–796.
Mantovani, B., Passamonti, M., Scali, V., 2001. The mitochondrial cytochrome
oxidase II gene in Bacillus stick insects: ancestry of hybrids, androgenesis, and
phylogenetic relationships. Mol. Phylogenet. Evol. 19, 157–163.
Milinkovitch, M.C., Lyons-Weiler, J., 1998. Finding optimal ingroup topologies and
convexities when the choice of outgroups is not obvious. Mol. Phylogenet. Evol.
9, 348–357.
Morgan-Richards, M., Trewick, S.A., 2005. Hybrid origin of a parthenogenetic genus?
Mol. Ecol. 14, 2133–2142.
Ogden, T.H., Rosenberg, M.S., 2006. Multiple sequence alignment and phylogenetic
inference. Syst. Biol. 55, 314–328.
Otte, D., Brock, P., 2003. Phasmida species file: a catalog of the stick and leaf insects
of the world. ISBN:1-929014-03-1.
Page, R.D.M., 1996. TREEVIEW: an application to display phylogenetic trees on
personal computers. Comput. Appl. Biosci. 12, 357–358. <http://
darwin.zoology.gla.ac.uk/~rpage/treeviewx/index.html>.
Parfitt, R.G., 1980. The cytology and feulgen-DNA microdensitometry of some New
Zealand stick insect species (Phasmatodea: Phasmatidae). MSc thesis, Victoria
University of Wellington.
Paton, T.A., Baker, A.J., 2006. Sequences from 14 mitochondrial genes provide a
well-supported phylogeny of the Charadriiform birds congruent with the
nuclear RAG-1 tree. Mol. Phylogen. Evol. 39, 657–667.
Peck, J.R., Yearsley, J.M., Waxman, D., 1998. Explaining the geographic distributions
of sexual and asexual populations. Nature 391, 889–892.
Posada, D., Crandall, K.A., 1998. Modeltest: testing the model of DNA substitution.
Bioinformatics 14, 817–818.
Rambaut, A., 1996. Se-Al: Sequence Alignment Editor. Available from: <http://
evolve.zoo.ox.ac.uk/>.
Ronquist, F., Huelsenbeck, J.P., 2003. MrBayes 3: Bayesian phylogenetic inference
under mixed models. Bioinformatics 19, 1572–1574.
Salmon, J.T., 1948. New genera, species and records of Orthoptera from the Three
Kings Islands New Zealand. Rec. Auckland Inst. Mus. 3, 291–311.
Salmon, J.T., 1991. The Stick Insects of New Zealand. Reed, Singapore.
Scali, V., Passamonti, M., Marescalchi, O., Mantovani, B., 2003. Linkage between
sexual and asexual lineages: genome evolution in Bacillus stick insects. Biol. J.
Linn. Soc. 79, 137–150.
Simon, C., Frati, F., Beckenbach, A., Crespi, B.J., Liu, H., Flook, P., 1994. Evolution,
weighting, and phylogenetic utility of mitochondrial gene sequences and a
compilation of conserved polymerase chain reaction primers. Ann. Entomol.
Soc. Am. 87, 651–701.
Sunnucks, P., Hale, D.F., 1996. Numerous transposed sequences of mitochondrial
cytochrome oxidase I–II in aphids of the genus Sitobion (Hemiptera:
Aphididae). Mol. Biol. Evol. 13, 510–524.
Swofford, D.L., 2002. PAUP*. Phylogenetic analysis using parsimony (*and other
methods). Version 4. Sinauer Associates, Sunderland, MA.
Szymura, J.M., Lunt, D.H., Hewitt, G.M., 1996. The sequence and structure of the
meadow grasshopper (Chorthippus parallelus) mitochondrial srRNA, ND2, COI,
COII, ATPase8 and 9 tRNA genes. Insect Mol. Biol. 5, 127–139.
Trewick, S., 2007. Stick insects, Te Ara – the Encyclopedia of New Zealand, updated
21-sep-2007. Available from: <http://www.TeAra.govt.nz/TheBush/
InsectsAndOtherInvertebrates/StickInsects/en>
Trewick, S.A., Morgan-Richards, M., 2005. New Zealand Wild: Stick Insects. Reed
Publishing, Hong Kong.
Trewick, S.A., Morgan-Richards, M., in press. New Zealand biology. In: Gillespie, R.,
Clague, E. (Eds.), Encyclopaedia of Islands. University of California Press,
Berkeley.
Trewick, S.A., Morgan-Richards, M., Wallis, G.P., 2000. Phylogeographic pattern
correlates with Pliocene mountain building in the alpine scree weta
(Orthoptera: Anostostomatidae). Mol. Ecol. 9, 657–666.
Trewick, S.A., Goldberg, J., Morgan-Richards, M., 2005. Fewer species of Argosarchus
and Clitarchus stick insects (Phasmida, Phasmatinae): evidence from nuclear
and mitochondrial DNA sequence data. Zool. Scr. 34, 483–491.
Trewick, S.A., Paterson, A.M., Campbell, H.J., 2007. Hello New Zealand. J. Biogeogr.
34, 1–6.
Vink, C.J., Paterson, A.M., 2003. Combined molecular and morphological
phylogenetic analyses of the New Zealand wolf spider genus Anoteropsis
(Araneae: Lycosidae). Mol. Phylogen. Evol. 28, 576–587.
Waters, J.M., Andres-Lopez, J., Wallis, G.P., 2000. Molecular phylogenetics and
biogeography of galaxiid fishes (Osteichthyes: Galaxiidae): dispersal,
vicariance, and the position of Lepidogalaxias salamandroides. Syst. Biol. 49,
777–795.
White, A., 1846. The zoology of the voyage of H.M.S. Erebus and Terror. In 1 Insects
of New Zealand. E.W. Janson, London.
White, T.J., Bruns, T., Lee, S., Taylor, J., 1990. Amplification and direct
sequencing of fungal ribosomal RNA genes for phylogenetics. In: Innis,
M.A., Gelfand, D.H., Sninsky, J.J., White, T.J. (Eds.), PCR Protocols: A Guide
to Methods and Applications. Academic Press, Inc., New York, NY, pp. 315–
322.
Whiting, M.F., Carpenter, J.C., Wheeler, Q.D., Wheeler, W.C., 1997. The Stresiptera
problem: phylogeny of the holometabolous insect orders inferred from 18S and
28S ribosomal DNA sequences and morphology. Syst. Biol. 46, 1–68.
Whiting, M.F., Bradler, S., Maxwell, T., 2003. Loss and recovery of wings in stick
insects. Nature 421, 264–267.
Winkworth, R.C., Wagstaff, S.J., Glenny, D., Lockhart, P.J., 2002. Plant dispersal NEWS
from New Zealand. Trends Ecol. Evol. 17, 514–520.
Zwickl, D.J., Hillis, D.M., 2002. Increased taxon sampling greatly reduces
phylogenetic error. Syst. Biol. 51, 588–598.
808 S.A. Trewick et al. / Molecular Phylogenetics and Evolution 48 (2008) 799–808