Copyright is owned by the Author of the thesis.  Permission is given for 
a copy to be downloaded by an individual for the purpose of research and 
private study only.  The thesis may not be reproduced elsewhere without 
the permission of the Author. 
 
 WELLINGTON GECKOS MEET WAIRARAPA 
GECKOS: 
Hybridisation between two genetically and morphologically distinct populations of the 
New Zealand common gecko complex (Hoplodactylus maculatus) 
 
A thesis presented in partial fulfilment of the requirements for the degree of  
Masters of Science in  
Zoology 
 
At Massey University, Palmerston North 
New Zealand 
 
Josephine Fitness 
2010 
 
          
I 
 
Abstract 
The purpose of this study was to use molecular techniques and morphological 
measurements to set out to find whether a hybrid zone exists between two coastal 
populations of the common gecko (Hoplodactylus maculatus), on the Wellington south 
coast. I collected geckos from five sites in a coastal transect from the population of 
small geckos to the large geckos. Using four genetic loci, one mitochondrial (16S) and 
three nuclear (Rag-1, Rag-2, C-mos), I was able to determine that the coastal 
populations do have geneflow, however each population maintains some unique alleles. 
Morphological evidence reveals a significant difference in gecko sizes from Turakirae 
Head and those caught at Ocean Beach, separated by just 15 km. Adult geckos at 
Turakirae Head are on average 10mm smaller (snout-to-vent) than adult geckos at 
Ocean Beach, representing almost a doubling in average weight.  The centre of the steep 
frequency clines of four characters is coincident and the widths are concordant. The 
narrower morphological clines indicate stronger selection on the size of the gecko, than 
on genetic loci.  
 
 
 
 
 
 
 
 
II 
 
 
 
 
 
 
 
 
 
DEDICATION 
Dedicated to my fiancé 
Adam Sullivan 
 For his wonderful support and encouragement 
 
 
 
 
 
 
 
 
 
III 
 
 
Acknowledgements 
There have been a number of people who have helped in some way on this project and I 
am thankful to them all. My fiancé, Adam who went out on most field trips gecko 
sampling through some of the extreme weather and disappointing weekends. Also to my 
field assistants Christina, Shane, Shaun, Kayla, and Lewis who also helped collect 
samples. Also thanks to Sophie for some PCR work in the Lab and Trish for her 
knowledge with primers. 
To Mary Morgan-Richards and Rod Hitchmough for their guidance and supervision of 
the project, which without would not have got so far.  
To Ecology staff that helped organise equipment for field work. 
 
 
 
 
 
 
 
 
 
 
 
IV 
 
 
Contents 
ABSTRACT          I 
ACKNOWLEDGEMENTS        III 
INTRODUCTION 
 Hybrid Zones         1 
 Species Status        3 
 New Zealand Geckos        5 
METHODS 
 Study Site        9 
 Morphological Sampling       11 
 Genetic techniques        12 
RESULTS 
 Morphometrics        14 
 Genetic         21 
 Cline analysis         32 
DISSCUSSION         34 
CONCLUSION        41 
REFERENCE         43 
APPENDIX          51 
V 
 
   
CONTENTS OF FIGURES  
FIGURE 1 Map of study site         9 
FIGURE 2 Scatterplot of PCA values                            14 
FIGURE 3 Scatterplot of weight vs. snout-vent length    15 
FIGURE 4 Box plot of female snout-vent length      18 
FIGURE 5 Box plot of male snout-vent length      19 
FIGURE 6 Bar graph of eye colour        19 
FIGURE 7 Bar graph of 4th toe lamellae       20 
FIGURE 8 16S network          24 
FIGURE 9 Rag-1 allele network       28 
FIGURE 10 Rag-2 allele network       28 
FIGURE 11 C-Mos allele network        29 
FIGURE 12 Line graph of allele frequency       31 
FIGURE 13 Hybrid cline                    33 
 
 
 
 
 
 
VI 
 
 
Contents of Tables 
TABLE 1 ANOVA results for females       17 
TABLE2 ANOVA results for males        18 
TABLE 3 Percentage of pattern       21 
TABLE 4 Summary of mitochondrial population statistics     23 
TABLE 5 Haplotypes for 16S        23 
TABLE 6 Summary of nuclear statistics       26 
TABLE 7 Rag-1 alleles         27 
TABLE 8 Rag-2 alleles         27 
TABLE 9 C-mos alleles         27 
TABLE 10 Genetic differentiation       30 
TABLE 11 Pairwise differences for nuclear loci     30 
TABLE 12 Percentage of Heterozygosity       31 
TABLE 13 Clines centres and widths       32
1 
 
Introduction 
Hybrid Zones 
Hybridisation is the mating and production of offspring between individuals from 
genetically distinct populations (Harrison 1993). Hybrid zones form between two 
populations and/or species at the point of contact of their ranges. Many studies have 
looked at hybrid zones and selection against hybrids within the zone (Barton & Hewitt 
1985; Moore 1977; Bronson et al 2003). The position and width of hybrid zones are 
usually stable over many generations, due to equilibrium between the ability of the 
taxon to disperse and the selective disadvantage suffered by the hybrid offspring 
(Barton & Hewitt 1985). Further stability is ensured when zones lie in density troughs 
(Barton 1979) or on ecotones (Moore 1977). Most hybrid zones involve secondary 
contact of populations that have diverged in isolation. Some hybrid zones effectively act 
as a barrier for speciation by allowing limited gene flow between the populations when 
hybrids backcross to parentals, and keeping populations genetically connected. 
However, separation of the two populations can be maintained by selection against the 
hybrids in a narrow contact zone and selection can maintain populations or species as 
distinct even when there is some gene flow (Alexandrino et al 2005). A narrow zone 
forms when hybrids are much less successful than the parental taxa (selective 
disadvantage against hybrids), for example if hybrids are sterile or suffer more from 
parasites. Alexandrino et al (2005) noted that there was a strong selection against 
hybrids in Ensatina (salamanders) at a hybrid zone, and that this was most likely due to 
a postzygotic mechanism; that reduce hybrid fitness compared to the parental fitness in 
this study. A wide zone forms where hybrids are as fit as parentals and this leads to the 
homogenising of populations and the eventual loss of a distinct zone of contact 
2 
 
(Johansson et al 2008).  
Hybrid zones can be studied with any characters that distinguish the populations or 
species such as size, colour, behaviour, flowering time, chromosomes, enzymes, or 
DNA. Molecular markers are particularly useful in measuring hybrid zones and to 
detect dispersal of hybrids (Berry et al 2004). Mitochondrial DNA (mtDNA) is good to 
use as it is maternally inherited and non-recombinant with a rapid rate of evolution 
(Godinho et al 2006). However, because mtDNA is transmitted only by the maternal 
line it only provides one-sided information (Godinho et al 2006). Nuclear loci are useful 
in obtaining evidence of biparental gene flow and many single copy nuclear genes can 
be amplified from reptiles using universal primers (Gamble e al 2008; Godinho 2006, 
Vidal & Hedges 2004). For example, C-mos is a single coding exon of approximately 2 
kb in the nuclear DNA that is present in all vertebrates (Godinho et al 2006). But any 
loci that show allele frequency differences among populations can be used to estimate 
gene flow.  The three nuclear loci used in this study were all single copy genes and only 
partial genes were used. 
There have been many studies looking at hybridisation in birds, toads and insects. In 
New Zealand there have been a few hybridisation studies, such as between the 
Auckland and Wellington tree weta, Grey and Mallard ducks, and Black and Pied stilts 
(Morgan-Richards et al 2009). It is common for evidence from morphology to tell a 
different story from that told by genetics. Sattler and Braun (2000) noted that previous 
morphological studies on a contact zone between Black-capped and Carolina chickadees 
found evidence of hybridisation while some studies found none. It was their study using 
genetic markers that showed hybridisation occurred frequently. This study showed the 
importance of using genetics in studies of species that may be cryptic to some extent. 
3 
 
The difference between morphology and genetics may be due to morphology usually 
being under selection, and many genetic differences being neutral. One method that is 
used to study hybrid zones is the use of hybrid clines (Leache and Cole, 2007). Often 
characters are scored using a hybrid index from 0-1 and/or principle component analysis 
or simple allele frequencies are used. This information can then be put into a hybrid 
cline that will indicate where the centre of the hybrid zone is located and estimate the 
zone width. Leache and Cole (2007) conducted a study on Fence lizards, Sceloporus 
undulatus complex, in the USA. They focused on a contact zone in eastern Arizona 
where two mtDNA lineages meet at an ecotone between grassland and woodland 
habitats. Northern and southern lineages differ in size, colouration, and squamation. 
Four characters were used for cline analysis including mtDNA, chromosomes, 
morphology and colour pattern. The results showed that three of four clines were 
coincident (Cline centre in the same place), but the mtDNA cline centre was shifted 
south from the other clines. The widths of these clines differed drastically and ranged 
from 1.5km-89km, thus they are nonconcordant. They suggested that the fence lizard 
contact zone has multiple processes working on it. 
Species status 
What is a species? This question is a dominant one in evolutionary biology of species 
formation. It has sparked great controversy and debate about the best way to 
differentiate species from each other. From the beginning species were considered based 
on their morphology alone, which would separate any morphological variant within a 
population into a different species as well as geographic populations (Bock 2004). As 
the understanding of genetics and speciation progressed it became clear that most 
populations and species had different varieties within populations as well as between 
4 
 
populations. One of the most common and popular concepts formed to try and define 
species was the biological species concept of Mayr (1942), which states that species are 
groups of individuals actually or potentially interbreeding natural populations which are 
reproductively isolated from other such groups. The understanding of species is 
important in understanding speciation including the hybridisation process and its 
consequences. Speciation is the formation of new species and hybridisation can have 
some positive or negative effects on this process.  Previously hybridisation was thought 
of as an evolutionary dead end where hybrids were selected against by natural selection 
to prevent the loss of diversity (Arnold 1997). The hybridisation of two different species 
may introduce new genes and characters into a population that are better adapted to the 
environment for survival and overall fecundity, called hybrid vigour. On the other hand 
the complete opposite could occur where hybridisation may cause a homogenising 
effect of the two breeding species reducing the diversity between them. 
It has been suggested by de Queiroz (1998) (quoted in Wiens & Servedio 2000) that the 
large number of species concepts agree on what a species is for sexual organisms, that is 
a species is a  lineage which is unified primarily by sexual reproduction or gene flow 
among its constituent parts. If we use this as grounds to determining a species then gene 
flow is an important component in recognising and delimitating species, and 
hybridisation (the successful mating between two individuals) results in gene flow. A 
full understanding of the degree of gene flow between distinctive populations is 
essential for making decisions about species status. As conservation efforts are directed 
towards maintaining biodiversity (usually at the species level) it is extremely important 
that taxonomy takes into account the level of hybridisation between potential species 
and its probable long term consequences.  
5 
 
At the contact zone of distinct population ranges a number of things can potentially 
happen. Two species may mate and produce fertile hybrids resulting in gene flow 
between species, or two species may not mate and just coexist in the same habitat. The 
biological species concept states that two species cannot interbreed (Mayr 1942). Some 
interesting questions arise at a hybrid zone 1) if two “species” meet and mate does this 
mean they are actually the same species? 2) If they were separate can they now merge 
into one? 3) How well do hybrids do in terms of fitness compare to the parent species? 
In order to decide whether the populations are distinct enough to be considered separate 
species a number of features can be examined. Selection against hybrids (endogenous 
selection) is expected to result in characteristic genetic patterns (Kruuk et al 1999). For 
example, linkage of loci would be detected, as the allele combinations in parental 
populations (under possible selection) would result in significant linkage disequilibrium 
even when loci are not physically linked (e.g. on the same chromosome).  Selection 
against hybrids would also be expected to result in a higher rate of heterozygosity in 
young geckos compared to adult geckos, and deviations from those expected by Hardy-
Weinberg equilibrium. In contrast a species may diverge from another by selection for 
adaptations to a particular environment, a process known as exogenous selection (Kruuk 
1999; Gee 2004). Ecological adaptation might result in ecotypes that differed only at 
loci under selection.    
New Zealand Geckos 
Geckos (Reptilia: Diplodactylidae) are a distinctive group of reptiles with a worldwide 
distribution (Robb 1980). The two genera of New Zealand geckos are monophyletic, 
with their closest relatives in Australia (Neilsen at al 2007). The species of Naultinus 
are predominantly diurnal, arboreal and usually bright green. The Hoplodactylus group 
6 
 
is usually nocturnal and found in a variety of habitats ranging from arboreal to ground 
dwelling and alpine to coast (Hitchmough 1997). Geckos are mostly small with some 
growing up to 35-40 cm (Robb 1980). There are about 17 described species of geckos in 
New Zealand (Hitchmough 1997), but this number continues to rise as genetic studies 
reveal cryptic species.  
The common gecko (Hoplodactylus maculatus) (Gray 1843) is grey/brown in colour 
with a variety of different markings that run across the body (Whitaker 1996). It is a 
small to medium sized gecko with size variation among populations, for example H. 
maculatus from coastal populations are often smaller then inland individuals 
(Hitchmough 1997). The common gecko, widespread throughout most of New Zealand, 
is now known to be a complex of at least 12 species (Hitchmough 1997). There is 
considerable morphological variation within the Hoplodactylus maculatus complex, but 
only the use of genetic information has allowed the identification of this species 
diversity. Hitchmough describes the Hoplodactylus maculatus gecko species in greater 
detail (Hitchmough In prep). My research will focus on two populations of 
Hoplodactylus maculatus from the south coast of the North Island that are recognized 
by Hitchmough (1997) as conspecific. 
Within the distribution of Hoplodactylus maculatus species complex are a series of 
monophyletic clades (Hitchmough 1997). The Northern group includes four species: H. 
maculatus and three tag species (H. “Mount Arthur”, H. “Kaikouras”, H. “Marlborough 
mini”). This Northern clade is found in the North Island and the top of the South Island 
around Nelson. The species H. maculatus (as referred to by Hitchmough 1997) is 
restricted to the North Island, including around Wellington and Wairarapa regions and 
lives in a variety of habitats including forest, scrub and grassland. Hoplodactylus 
7 
 
maculatus can be found under logs, rocks and debris and shingle banks (Robb 1980) 
and is more commonly found on the ground, being one of the less arboreal species 
(Whitaker 1996). Individuals from coastal areas may be bluish-grey in colour, with 
darker streaks on the dorsal surface (Robb 1980). They are nocturnal but may be seen 
sun-basking (Whitaker 1996). Coastal populations of the Hoplodactylus maculatus 
species complex are often smaller than inland sister groups, suggesting body size 
selection favours smaller animals in coastal habitats, where densities are often higher. 
Some coastal populations show sexual size dimorphism (with males larger than females; 
Whitaker 1982).  
There are external morphological differences between female and male Hoplodactylus 
maculatus individuals. Males have 1-2 enlarged blunt scales on each side of the base of 
the tail (Robb 1980); and a broad patch of preanal and femoral pores along the 
underside of the thighs (Whitaker 1996) which contain glands (Robb 1980). The young 
are born mid-late summer (Robb 1980). Importantly the tail of this species is 
regenerated (Robb 1980), which is a defence mechanism and useful for our study 
purposes.  
From previous study it has been observed by Whitaker (1987) that the geckos found on 
the coast are a lot smaller and more abundant than the geckos found further inland. This 
idea sparked this study looking at whether these different populations hybridise and 
therefore share genes and are still one species or whether they do not hybridise and are 
actually speciating into different species. To look at the differences between populations 
and seeing whether they are in the process of diverging we look at hybrid zones 
between the two populations that show the most difference.  
Initial genetic studies using nuclear markers and mtDNA sequencing (Hitchmough 
8 
 
1997; Nielsen et al 2007) have found relatively high levels of genetic diversity within 
Hoplodactylus maculatus. Populations on the south coast of Wellington were found to 
be genetically distinct from geckos in Wairarapa at just a few allozyme loci 
(Hitchmough 1997). Hybridisation is thought to occur between these two populations 
which are recognized as members of the same species; Hoplodactylus maculatus. 
Morphological differences between these southern North Island populations have also 
been noted (Rod Hitchmough per com). The Wairarapa geckos sampled were larger 
than the coastal Wellington geckos, and sexual size dimorphism was limited to the 
coastal locations (Whitaker 1982).  
My research will address five questions, which focus on the hypothesis that the 
Wellington coast and Wairarapa gecko populations might be cryptic species. 
1) Using morphological and DNA sequences can one distinguish Hoplodactylus 
maculatus individuals from the Wellington South coast from those in 
Wairarapa? 
2) Do these populations hybridise? 
3) How wide is the hybrid zone? 
4) Do the morphological and genetic (mtDNA and nuclear markers) zones have the 
same centres and similar widths? 
5) Is this hybrid zone an effective barrier allowing speciation to occur?   
 
 
9 
 
Methods 
Study site 
The study site was located on the coast of Wellington starting from Turakirae Head and 
heading South-West towards Ocean Beach in Wairarapa (Figure 1). The two extreme 
populations that were thought to be the most morphologically different are shown in the 
red circles on the map below. Three more locations were sampled between these two 
extreme populations, approximately 3km apart. For an inland representative, samples 
were provided by Rod Hitchmough from his study of geckos (PhD unpublished thesis 
1997) from Lake Pounui.  
 
Figure 1, Map of Study site where geckos were collected from. The red circles indicate the two 
extreme populations at Turakirae Head (TH) and Ocean Beach (OB). Yellow circles are other 
populations sampled; Barneys Whare (BW), Fishermans Rock (FR) and Gateway (GW). 
 
The locations were all within 20 meters of the shoreline, but they differed in habitat. 
Turakirae Head (TH) was fairly flat with large boulders and lots of small rock piles. The 
main vegetation was Muehlenbeckia complexa shrubs which are ideal for geckos to hide 
under (Whitaker 1982). The area was located close to Orongoronga station and there 
10 
 
were a number of sheep wandering around the area, as well as people. Barneys Whare 
(BW) is the next location and is a historic site. Around the cabin are lots of rocky places 
with M. complexa shrubs. The area is also close to some tall trees providing height and 
shelter from the sun. The third site is located at Fishermans rock (FR). This site is 
barren and the rock piles were within farmed grassland where sheep and cattle graze. 
The patch that housed the geckos caught for the study was small and located at the 
bottom of a hill. There were fewer shrubs there for protection. The next site is called 
Gateway (GW), and got its name due to a gate located across a path. This area was 
mostly steep hills with lots of rocks and almost scree like. The geckos were commonly 
found under reasonably sized rocks adjacent to the tall grass and the few shrubs. This 
site spanned the longest distance due to the sparsely populated spots ideal for geckos or 
gecko hunting. The last site is at Ocean beach (OB). This spot took the longest to find 
geckos suggesting that the population is less dense as they were able to elude capture 
due to the presence of human activity all year round. These geckos were found mostly 
in rocky areas in dried river beds. Later in the year they were found up to 20 cm 
underground in sandy and rocky areas next to shrubs or small trees. The coastal area 
around the study sites have been under some major geological influences. The 
particularly important geological activities that have altered the landscape are the 
southern North Island uplift and outwash from rivers. The Rimutaka Ranges which 
overshadow the study site are made up of Greywacke (Lee & Begg 2002). At the end of 
the Rimutakas at Turakirae Heads the area is dominated by beach and marginal marine 
terrace deposits consisting of marine gravels, sand, mud and the beach ridges (Lee & 
Begg 2002). Turakirae Heads is well known for its beach ridges which can be used to 
estimate dates of earthquakes that have raised the land. The latest beach ridge is about 
11 
 
2.7 meters from the current sea level. This rise was due to the 1855 Wairarapa fault 
earthquake that affected much of the region (McSaveney et al 2006). 
Morphological Sampling 
The common gecko was more easily found under medium sized rocks in rocky areas, or 
hiding under shrubs such as Muehlenbeckia that grows over rock piles. They were best 
caught when weather was still cool as they were less active due to their exothermic 
properties.  At each site I aimed to collect 30 geckos at random. The first 20 geckos 
were included in both the genetic and the morphological study; the last 10 geckos from 
each site were used only for morphology. At two locations the full sample of 30 geckos 
were not obtained (Ocean Beach (n=28) and Gateway (n=25)). All geckos were caught 
between October 2008 and June 2009, except for 13 geckos with code name RAH 
caught by Rodney Hitchmough, which were collected prior to 1997 (these 13 geckos 
were included in the genetic analysis but not the morphological study). 
 Each gecko caught was placed in a zip lock plastic bag for measuring, and all 
measurements were made by one person (JF). The geckos were weighed using mini 
scales (Tanita Model 1479V). Quantitative external morphological measurements taken 
were; snout-vent length (SV), head width (HW), snout-ear length (SE) and head height 
(HH). Morphological characteristics recorded were; sex, eye colour, whether they were 
an adult or juvenile and had a damaged tail at catching or not. Using a hand lens the 
number of lamellae on the 4th toe of the hind foot and external mites were counted and 
recorded. All lamellae, partial and whole, were counted from the base of the toe up to 
the tiny knob close to the toe nail. This knob was not actually included in the count 
however. Each gecko also had photos taken to categorise colour and pattern for 
frequency analysis.   
12 
 
Morphological measurements were studied using Principle component analysis (PCA) 
using Excel XLSTAT 2009 (An add-on that can be brought online). Using the PCA 
output, assessments of the population data were conducted. For the rest of the study 
only Snout-Vent length was used as this showed the most variation among populations, 
it is the measurement least prone to experimental error, and is the most commonly used 
measurement of reptiles. ANOVA and T-test were conducted using XLSTAT 2009. 
Genetic techniques 
Tissue for DNA extraction was collected as tail tips. The last 2-3 mm of the tail was 
removed using clean scissors and stored in 95% ethanol, with the procedure approved 
by the Massey Animal ethics committee. DNA was extracted using a Qiagen DNeasy 
tissue and blood kit. DNA was diluted to 1ng/µl for PCR reactions.  
Primer pairs, designed to amplify DNA from both mitochondrial and nuclear loci for 
other reptiles were tested on Hoplodactylus maculatus samples (See Appendix 2; list of 
all primers tested). One mitochondrial gene (16S) and three nuclear loci, (Rag-1, Rag-2 
and C-mos) were selected for this study.  Primers used were; 16Sd 
CTCCGGTCTGAACTCAGATCACGTAG, 16Sc 
GT[A/C]GGCCTAAAAGCAGCCAC (Reeder 1995) C-Mos: Mos-F  
CTCTGGKGGCTTTGGKKCTGTSTACAAGG, MOS-R 
GGTGATGGCAAANGAGTAGATGTCTGC (Godinho et al 2006), Rag-1: L2408 
TGCACTGTGACATTGGCAA, H2928 GACTGCYTGGCATTCATTTT (Vidal and 
Hedges 2004), Rag-2: PY1-F CCCTGAGTTTGGATGCTGTACTT, PY1-R 
AACTGCCTRTTGTCCCCTGGTAT (Gamble et al 2008). 
All PCR reactions were done in a thermocycler using the following protocol; 94°C 
2mins, 94°C for 20s, 50°C for 15s, 72°C for 130sec and repeated for 40 cycles then 
13 
 
72°C for a further 5mins. Following amplification, the PCR products were sequenced at 
the AWC genome sequencing centre, using one of the amplification primers: 16Sc, 
L2408, PY1-F or MOS-F. The preparation of the PCR products for sequencing was 
done by adding 1ml of the primer used, 2mls of PCR product and 12mls of water. 
Analysis of the DNA sequence used the following software; sequencer 4.7 to check the 
sequences by eye, SE-AL to align the data, check for stop codons and export as a 
NEXUS file to be used in McClade 4.0. In McClade the characters were compressed 
and the redundant taxa and characters removed. This left only the base pairs that 
differed and one sequence for each distinct allele. The data was used to draw minimum 
spanning trees of alleles for the nuclear loci. The 16S haplotype network was 
constructed using Automated Nested Clade Analysis v1.0.This program is freely 
downloaded from the website http://www.rubic.rdg.ac.uk/~mahesh/software.html and 
can be used to build a TCS haplotype tree using the nexus file of sequences. Hybrid 
clines were studied using the genetic analysis of all four genes, Analysis V1.3, 
www.biology.ed.ac.uk/research/institutes/evolution/software/Mac/Analyse/index.html.
Other analyses of genetic diversity were conducted using Arlequin 2.0 freely 
downloaded from http://lgb.unige.ch/arlequin/software/ and GENEPOP 4.0.10 from 
http://genepop.curtin.edu.au/.  
 
 
 
 
 
14 
 
Results 
Morphometrics  
143 geckos were caught and measured from five locations on the Wellington south 
coast (Appendix 1). At each location the random sample of 25-30 animals resulted in a 
mixture of adult and juvenile geckos, of both sexes. Juveniles were excluded from 
further morphological analysis, leaving 101 adult geckos.  
 
Figure 2 Morphological variation among adult geckos presented as PCA 1 against PCA 2. Individuals in 
the middle are of intermediate sizes. 
Five morphological characters were used in the Principle component analysis; weight, 
snout-vent length, snout-ear length, head width and head height (Figure 2). Snout-vent 
length has the highest contribution to PC1 (0.938) followed by weight. The first 
PCA case scores for adult geckos  
TH 
GW 
BW 
FR 
OB 
PC2 (6%) 
PC1 (89%) 
-0.3 
-0.5 
-0.8 
-1.1 
-1.4 
0.0 
0.3 
0.5 
0.8 
1.1 
1.4 
-0.3 -0.5 -0.8 -1.1 0.0 0.3 0.5 0.8 1.1 1.4 
15 
 
Principle component (PC1) accounted for 89% of variation and the second component 
explained only 6%. When adult males and females are combined PCA fails to clearly 
differentiate the geckos from each location; however the smallest adult geckos were 
caught at Turakirae Head and the largest at Ocean Beach (Figure 2). Although a two 
sample t-test of snout-vent length between males and females in the total sample found 
no significant difference (P-value > 0.05 at 0.356), male and female geckos were 
analysed separately due to sexual size dimorphism known to exist at Turakirae Head 
(Whitaker 1982).  
 
Figure 3 Scatterplot of snout-vent length against weight of males and females. Females were 
separated into pregnant versus non-pregnant to see if there is any difference between these females. 
 
Even though a number of measurements were initially taken in the field, only a few 
have been chosen for final assessment of the populations.  Weight has been rejected for 
analysis because of seasonal differences, such as collecting pregnant females during 
summer and non-pregnant females during winter (Figure 3). Variation in the season the 
geckos were caught would make analysis of weight an unreliable source of variation. 
0
2
4
6
8
10
12
30 40 50 60 70 80
W
ei
gh
t 
(g
ra
m
s)
Snout-vent length (mm)
Scatterplot of snout-vent length against 
Weight
Males
Females (Pregnant)
Female (Non-pregnant)
16 
 
Head height and head width were also excluded due to difficultly in being consistent in 
taking this measurement. This main difficultly in taking measurements was making sure 
the same amount of pressure was applied to each gecko. Although I tried to be careful 
and accurate sometimes I think I may have tightened the callipers more on some geckos 
than other making the reading less accurate.  
If there is a hybrid zone we expect to see a significant difference between the two end 
populations on the transit, and intermediate sized individuals somewhere in the middle. 
The average snout-vent length for adult geckos was compared among populations 
(Figures 4 and 5). The average snout-vent length between each population increases 
from one end of the coastal transect to the other. For female geckos this difference on 
average is about 12mm. This increase in size along the coastal transect will also 
increase the average weight from approximately 4 grams to 7.5 grams, therefore the 
overall mass of individual geckos almost doubles from Turakirae Head to Ocean Beach. 
Fishermans Rock individuals are intermediate in size. Male geckos also show an 
increase in size from small at Turakirae Heads to large (about 8mm difference), but 
Gateway not Ocean Beach has on average the largest male geckos (although ANOVA 
fails to show a significant difference between Ocean Beach and Gateway) (Table 2). 
The box plots of both male and female average snout-vent length show a steady increase 
in adult gecko size from one end of the transect to the other.   
The Tukey (HSD) test in an ANOVA analysis (Table 1) shows that there is a significant 
difference in snout-vent length between six of the ten pairwise comparisons for adult 
females. The four non-significant comparisons of average snout-vent length for females 
are adjacent locations with the exception of Gateway and Barney’s Whare. This pattern 
of differences is as expected of a cline for morphological size. For the males (Table 2) 
17 
 
however, there is a significant difference in snout-vent length between only four of the 
ten pairwise comparisons. The populations that differ significantly are Ocean Beach, 
Gateway, Turakirae Head and Barneys Whare, thus the further the populations are from 
each other the more different they are.  
Table 1 ANOVA results of female snout-vent lengths. It is clear that there is a significant difference 
between populations further apart.  
 
Contrast Difference 
Standardized 
difference 
Critical 
value Pr > Diff Significant 
OB vs TH 12.352 7.305 2.834 < 0.0001 Yes 
OB vs BW 6.103 3.342 2.834 0.013 Yes 
OB vs FR 5.575 3.125 2.834 0.024 Yes 
OB vs GW 1.882 0.971 2.834 0.867 No 
GW vs TH 10.471 5.785 2.834 < 0.0001 Yes 
GW vs BW 4.221 2.179 2.834 0.205 No 
GW vs FR 3.694 1.946 2.834 0.308 No 
FR vs TH 6.777 4.119 2.834 0.001 Yes 
FR vs BW 0.528 0.296 2.834 0.998 No 
BW vs TH 6.249 3.696 2.834 0.005 Yes 
Tukey's d critical value: 4.008   
18 
 
OB (S-V)fGW (S-V)fFR (S-V)fBW (S-V)fTH (S-V)f
70
65
60
55
50
D
a
t
a
Boxplot of Female Snout-Vent lengths at each location
Figure 4, Box plot of snout-vent length of female geckos at each location in the transect showing a 
cline in size from 53mm to 65mm over 15km 
 
Table 2 ANOVA results for male snout-vent lengths showing significant difference between 
populations that are further apart.  
Contrast Difference 
Standardized 
difference Critical value Pr > Diff Significant 
OB vs TH 8.584 3.486 2.847 0.010 Yes 
OB vs BW 6.101 3.532 2.847 0.008 Yes 
OB vs FR 3.297 1.740 2.847 0.421 No 
OB vs GW 0.673 0.374 2.847 0.996 No 
GW vs TH 7.911 3.213 2.847 0.020 Yes 
GW vs BW 5.428 3.142 2.847 0.024 Yes 
GW vs FR 2.625 1.385 2.847 0.640 No 
FR vs TH 5.286 2.086 2.847 0.245 No 
FR vs BW 2.804 1.533 2.847 0.547 No 
BW vs TH 2.483 1.030 2.847 0.840 No 
Tukey's d critical value: 4.026   
19 
 
OB(S-V)mGW (S-V)mFR (S-V)mBW (S-V)mTH (S-V)m
75
70
65
60
55
50
D
a
t
a
Boxplot of Male Snout-Vent lengths at each location
 
Figure 5 Box plot of snout-vent length of adult male geckos at each location in the transect showing a 
cline in size from 55mm to 62mm over. The * is an outlier of a small adult gecko in the GW population. 
 
 
Figure 6 Graph showing the percentage of the three different eye colours in each population. Green 
eyes are more common in the middle locations and towards Ocean beach but gold is more common at 
Turakirae Head. 
The eye colour of each gecko was scored as one of three colours (green, bronze or gold) 
and all three colours were recorded in every population (Figure 6). Green eyes were the 
0
10
20
30
40
50
60
70
TH BW FR GW OB
Percentage of Eye Colour in the 
different populations
Gold
Green
Bronze
20 
 
most common in four out of five populations. At Turakirae Head gold eyes were more 
common than anywhere else, and bronze eyes were fairly consistent across the five 
locations with the fewest at Fishermans Rock.   
 
Figure 7 Graph showing the percentage of numbers of 4th toe lamellae in different populations. The 
averages were the same at around 10 for each location suggesting no significant difference. 
As with eye colour, the number of 4th toe lamellae varies among individuals within each 
population. The range was between 9 and 12 lamellae per toe (Figure 7). The most 
common number of lamellae was 10 or 11, and the average lamellae number changed 
only slightly over the zone (9.8-10.5) and did not show a cline. Twelve lamellae per toe 
were more common from Fisherman’s Rock, Gateway and Ocean Beach than the other 
two populations.  
An arbitrary score for dorsal pattern was performed using photos of each gecko to see if 
populations had a higher frequency of a particular pattern (Table 3). There is no 
evidence to suggest any significant difference in patterns. However, Turakirae Head 
geckos had the highest percentage of a striped pattern (38%) than any other location.  In 
all other populations the horizontal bands were the common pattern found, with few 
0
10
20
30
40
50
60
70
80
TH BW FR GW OB
P
e
rc
e
n
ta
ge
Population
Percentage of the Number of 4th Toe 
lamellae in Different Populations
9
10
11
12
21 
 
having distinct spots and a few that had patterns that could not be distinguished as either 
strips, bands, or spots. Colour was not scored because geckos change colour depending 
on amount of sunlight (Hitchmough per. com.). It was noticed that more geckos caught 
in winter were darker than geckos caught in summer at the same location.  
Table 3 the percentage of geckos with a given score corresponding to pattern. 1 is a pure vivid striped 
patter and 5 is the pure vivid spotted pattern. The numbers in between are generally in between 
these two patterns, based on the clarity of pattern and the kind of pattern.  
Location 1 (%) 2 (%) 3 (%) 4 (%) 5 (%) 
TH 38 21 7 28 7 
BW 5 11 3 68 11 
FR 4 11 22 63 0 
GW 19 6 12 44 19 
OB 3 40 0 37 10 
 
Genetic analysis 
Mitochondrial DNA 
Four loci were used to analyse genetic differentiation within and between these gecko 
populations along the coast and an inland population. Mitochondrial gene 16S was 
amplified and sequenced from 132 geckos. The length of the fragment in this study was 
600 base pairs of which 6.67% (40bp) were variable. From 132 geckos from six 
locations 23 haplotypes were found. Between 3 and 8 haplotypes were found at each 
location (Table 4).  Nucleotide diversity ranges from a high of 0.014 (Ocean Beach) to 
22 
 
the low of 0.001 (Barney’s Whare). Haplotype A is the most common and was found 
everywhere except Lake Pounui (LP). Lake Pounui (LP) is 12km from Ocean Beach, 
being more inland and thus is expected to be more distinct from other locations. Each 
location has one or more unique haplotypes (Table 5). Using GENEPOP to detect 
population differentiation using the exact G-test for 16S, the only non-significant 
difference is between Gateway and Fisherman Rock (P>0.05). All other population 
pairs were significantly different from each other (P value ranging from 0.00 – 0.043). 
This is what we would expect considering the private alleles found at each location and 
the distribution of the alleles shared between adjacent populations.  
 A network of the 16S haplotypes shows that ten haplotypes are just one or two 
mutational steps away from the common haplotype A, which are four steps away from 
the next haplotype group (Figure 8). Using the haplotype network, and their 
distribution, the 23 haplotypes were coded into two clusters, split between haplotype G 
and S (Figure 8). Haplotype A and the ten haplotypes within one or two mutational 
steps from A are common in Turakirae Head, Barneys Whare and Fishermans Rock 
(referred to as the south-west group). The other group of haplotypes is more common in 
Gateway, Ocean beach and Lake Pounui geckos (North-east group). Although these two 
main haplotype groups change in frequency many populations have both haplotype 
group (Table 5). This coding of south-west versus north-east 16S haplotypes is used 
later in the clinal analysis.  
 
 
 
23 
 
Table 4 Genetic diversity found in geckos between Wellington and Wairarapa south coast using 
mitochondrial DNA sequences (600bp 16S) population statistics, n is the sample size.  
Population n Number of 
Haplotypes  
Gene 
Diversity 
Mean Number of 
Pairwise 
Difference 
Nucleotide 
Diversity 
TH 26 7 0.7600 2.270769 0.003797 
BW 28 3 0.3624 0.375661 0.000628 
FR 21 5 0.7762 5.200000 0.008696 
GW 23 8 0.8103 5.454545 0.009166 
OB 26 8 0.8667 8.656667 0.014428 
LP 8 6 0.9286 5.785714 0.005879 
 
Table 5 the number of individuals with a particular mtDNA haplotype (16S) at each location. 
Haplotype A is the most common, whereas all others are only present in a few individuals at a few 
locations.  Haplotypes are coded into two groups; red are predominantly south-west and blue is 
north-east (see Fig 8).  
Haplotype TH BW FR GW OB LP 
A 12 22 3 2 3  
G 3 1  3   
F  5     
H 4      
N 2      
M 2      
L 2      
R 1      
S   1    
E   4 1   
C   5 7 4  
B   8 7 7  
D    1 4  
T    1   
V    1   
U     1  
W     1  
J     2 2 
I     4 1 
K      2 
O      1 
P      1 
Q      1 
 
24 
 
 
Figure 8 Minimum network tree of mitochondrial haplotypes (based on 600bp of 16S) found in geckos 
between the Wellington and Wairarapa south coast along a transect. Colours indicate the two codes 
for the cline analysis: Red equal the small Wellington and blue are the large Wairarapa geckos. 
 
Nuclear DNA analysis 
Three nuclear loci were amplified and sequenced for 132 geckos. Genotypes were 
scored for each gecko using homozygote individuals to identify alleles and 
heterozygotes inferred from sequences with double peaks at variable nucleotide sites. 
The Rag-1 product was 327bp long of which 2.44% were variable, with 9 alleles. Rag-2 
fragment was 268bp long, 2.24% variable and this locus has 7 alleles. C-mos fragment 
was 294bp long and had 2.28% variability, and 9 alleles (Table 6). Deviations from 
Hardy-Weinberg equilibrium were tested for each locus at each location (Table 6). All 
three loci were in Hardy-Weinberg equilibrium in every population. Gateway 
population had both the highest frequency of heterozygotes (78%, Rag-1) and the 
lowest (9% at c-mos) (Table 12). Heterozygosity of adults was compared to that of the 
juveniles and found to be higher in adults than juveniles.  Pairwise linkage-
disequilibrium was tested for each population and no evidence was found that the loci 
are linked, with the exception of Rag-1 linked to C-mos in Turakirae Heads and C-mos 
linked to Rag-1 and Rag-2 at Fishermans Rock. However, with Bonferroni correction 
25 
 
for multiple test this is not significant (P>0.005). Linkage disequilibrium between 
mitochondria and nuclear loci is also non-significant (P>0.05) (tested in GENEPOP 
4.0.10).   
F-Statistics using Arlequin 2.0 gave an average FST over the three loci of 0.78 (which is 
significantly greater than 0: P>0.05). The AMOVA results indicated that only 7.76% of 
total variation is explained by differentiation between populations. However, 92.24% of 
the variation is within populations. 
The tables 7-9 list the number of geckos with a particular allele present in their 
genotype. The alleles are listed in order of those found in Turakirae Heads to those 
found in Lake Pounui, showing a cline-like structure to the alleles present in each 
population.   
Rag-1 
The most common Rag-1 allele (A) is between one and three mutation steps away from 
the other eight alleles observed (Figure 9). All locations had one or more alleles found 
nowhere else, with the exception of Barney’s Whare and Ocean Beach (Table 7). 
Although frequency of alleles varies among sample locations there was not a set of 
alleles (or a clade) restricted to either end of the zone.  
Rag-2 
For Rag-2 the most common allele (A) is between one and three mutation steps away 
from the other 6 alleles (Figure 10). Only Fisherman Rocks has a unique allele found 
nowhere else, where as all other locations share their alleles with one or more other 
populations (Table 8). As with Rag-1, the frequency of alleles varies but there is no set 
of alleles that are restricted to either ends of the zone.  
26 
 
C-mos 
The most common allele (A) is only a single mutation step away from 7 of the other 
alleles, with the exception of allele D which is two mutational steps away from allele A 
(Figure 11). Gateway, Ocean Beach and Lake Pounui have between one and two unique 
alleles not shared with other populations (Table 9). As can be seen from the tree there is 
no clade structure that restricts allele sets to one end of the zone from the other.  
Table 6 summary table for the three nuclear loci over coastal transect of geckos, n is the sample size.  
 
 
 
 
 
 
 Rag-1 Rag-2 C-Mos Rag-1 Rag-2 C-Mos 
Population n Gene Diversity 
Hardy-Weinberg 
Equilibrium: P-Value 
(Number of alleles) 
TH 26 0.565 0.774 0.111 
0.389 
(5) 
0.216 
(5) 
1.0 
(2) 
BW 28 0.642 0.667 0.143 
0.033 
(4) 
0.263 
(5) 
0.174 
(3) 
FR 21 0.524 0.703 0.511 
0.250 
(5) 
0.379 
(6) 
0.110 
(3) 
GW 22 0.644 0.548 0.086 
0.195 
(6) 
0.809 
(5) 
1.0 
(3) 
OB 26 0.540 0.608 0.217 
0.049 
(3) 
0.111 
(6) 
1.0 
(5) 
LP 8 0.714 0.241 0.339 
0.904 
(5) 
1.0 
(3) 
1.0 
(3) 
27 
 
Table 7, the number of gecko individuals that have a copy of each allele present in that population for 
the nuclear gene RAG-1 from Turakirae Head along the coast to Lake Pounui. 
Allele TH BW FR GW OB LP 
A 32 30 27 20 23 8 
B 12 14 11 19 26 3 
C 6 8 2 2 2  
E 1 3    3 
F 1      
G   1    
D   1 3   
I    1   
H      1 
 
 
Table 8 the number of alleles present in each population for the nuclear gene RAG-2 from Turakirae 
Head to Lake Pounui 
Allele TH BW FR GW OB LP 
A 12 30 15 30 31 14 
B 13 6 3 3 1 1 
D 7 9 18 4 8 1 
E 3 4 2 7 7  
C 17 7 3  4  
H   1    
G    2 1  
 
 
Table 9 the number of individuals in each population with a particular allele from the gene C-Mos 
from Turakirae Head to Lake Pounui 
Allele TH BW FR GW OB LP 
A 49 51 27 44 46 13 
B 3 4 1    
C   13  3  
E    1   
I    1   
D     1  
J     1  
H     1 2 
F      1 
28 
 
 
Figure 9, Network of Rag-1 alleles from Wellington and Wairarapa geckos. Red are alleles present in 
the Turakirae Head population whereas blue are alleles present in the Lake Pounui populations (the 
two extreme populations of the transect). 
 
Figure 10 Network of Rag-2 alleles from Wellington and Wairarapa geckos. Red are alleles present in 
the Turakirae Head population whereas blue are alleles present in the Lake Pounui populations (the 
two extreme populations of the transect). 
29 
 
 
Figure 11 Network of C-Mos alleles from Wellington and Wairarapa geckos. Red are alleles present in 
the Turakirae Head population whereas blue are alleles present in the Lake Pounui populations (the 
two extreme populations of the transect). 
As with the mitochondrial loci, GENEPOP was used to assess the population 
differentiation (exact G test) for the three nuclear loci (Table 10). Table 11 shows the 
FST values using a Pairwise distance method for all three nuclear loci pooled together. 
For Rag-1 the significant differences were greater in populations not adjacent to each 
other. Lake Pounui genotypes were significantly different from all other locations, a 
result expected considering the distance between Lake Pounui and the other locations. 
As with Rag-1 the other two loci also showed a trend of a greater difference between 
populations further apart. However, not all results were as expected, for example Ocean 
Beach and Lake Pounui were non-significant for Rag-2 and C-mos. Table 10 and 11 
work together to show that the further apart the populations are the greater the 
differences. With correction for multiple tests the significance level drops to P= 0.005, 
and the number of pairs that are now significant drops. 
 
30 
 
Table 10 shows the significance level (P = 0.05) for genetic differentiation between each location for 
the four loci; Rag-1, Rag-2, C-mos and 16S.  
Rag-1 Rag-2 C-mos 16S 
Location 
pair 
P-value Location 
Pair 
P-Value Location 
Pair 
P-value Location 
Pair 
P-value 
TH-BW 0.757 TH-BW 0.005 TH-BW 1.000 TH-BW 0.000 
TH-FR 0.524 TH-FR 0.001 TH-FR 0.000 TH-FR 0.000 
TH-GW 0.046 TH-GW 0.000 TH-GW 0.135 TH-GW 0.000 
TH-OB 0.012 TH-OB 0.000 TH-OB 0.040 TH-OB 0.000 
TH-LP 0.017 TH-LP 0.000 TH-LP 0.015 TH-LP 0.000 
BW-FR 0.120 BW-FR 0.070 BW-FR 0.000 BW-FR 0.000 
BW-GW 0.007 BW-GW 0.017 BW-GW 0.079 BW-GW 0000 
BW-OB 0.005 BW-OB 0.269 BW-OB 0.069 BW-OB 0.000 
BW-LP 0.025 BW-LP 0.090 BW-LP 0.013 BW-LP 0.000 
FR-GW 0.321 FR-GW 0.000 FR-GW 0.00 FR-GW 0.267 
FR-OB 0.064 FR-OB 0.022 FR-OB 0.006 FR-OB 0.011 
FR-LP 0.037 FR-LP 0.006 FR-LP 0.005 FR-LP 0.000 
GW-OB 0.258 GW-OB 0.188 GW-OB 0.222 GW-OB 0.043 
GW-LP 0.032 GW-LP 0.243 GW-LP 0.044 GW-LP 0.000 
OB-LP 0.001 OB-LP 0.129 OB-LP 0.186 OB-LP 0.004 
 
Table 11 Distance Method; Pairwise differences, Pooled FST data for three nuclear loci under diagonal 
matrix and p-values above diagonal. 
 TH BW FR GW OB LP 
TH 0 0.0175 0 0 0 0 
BW 0.035 0 0.0003 0.0149 0.0089 0.0542 
FR 0.099 0.07 0 0 0.0003 0.0009 
GW 0.13 0.033 0.132 0 0.4255 0.0146 
OB 0.122 0.035 0.097 -0.004 0 0.0148 
LP 0.174 0.046 0.13 0.043 0.053 0 
31 
 
 
Figure 12, Allele frequency of the most common allele at each of four loci  across the coast transect 
from Wellington to Wairarapa for six gecko populations. Frequency clines are shown by mtDNA and 
Rag-2. 
 
Table 12 Percentage of heterozygote’s (combined adults and juveniles) at each location over the three 
nuclear loci 
COMBINE Rag-1 Rag-2 C-Mos 
TH 54 62 12 
BW 54 64 11 
FR 62 62 67 
GW 78 61 9 
OB 73 62 23 
LP 75 25 38 
 
 
 
0
0.2
0.4
0.6
0.8
1
1.2
0 3 7 10 13 25
fr
e
q
u
e
n
cy
distance (km)
Allele frequency of common allele
Rag-1
Rag-2
Cmos
mtDNA
32 
 
Cline analysis 
For the three nuclear loci the change in the frequency of the commonest allele was 
examined over the transect (Figure 12).  Only the common Rag-2 allele shows a clinal 
change in frequency and was therefore used to examine the hybrid zone. Cline widths 
and centres were found using Analysis 1.3 (Baird 1996). For morphology, hybrid scores 
were snout-vent measurements scaled using the equation ((population mean – min. 
mean)/ (max mean – min. mean)) (Leache & Cole 2007). Table 13 shows the estimates 
of centres and widths for three characters: mitochondrial DNA, Rag-2 nuclear DNA and 
snout-vent lengths for males and females separately. Although the estimates of cline 
centres and widths are not identical they all overlap broadly, as expected of a hybrid 
zone. The widest cline estimate is for Rag-2 and the narrowest is for female snout-vent 
length (Table 13). The centre of the mitochondrial cline is significantly further north-
east than any of the other three clines.   
Table 12 Hybrid cline centres and widths for 4 characters.  
 Centre Width 
Mitochondrial (16S) 11.52 (11.21-11.97) 14.82 (14.83 – 16.07) 
Rag-2 7.39 (6.93 – 7.99) 38.56 (34.98 – 42.27) 
Male Snout-vent 
length 
5.32 (5.14- 5.58) 7.77 (7.47-8.76) 
Female Snout-vent 
length 
5.37 (5.28-5.62) 9.18 (8.63-9.62) 
 
33 
 
a) b)  
c) d)  
Figure 13 Clines of the 4 characters that were used in cline analysis. A) is the female snout-vent 
length, b) the male snout-vent length, c) Rag-2 nuclear cline and d) is the mitochondrial cline.  
 
 
 
 
 
 
 
34 
 
Discussion  
Adult geckos from the north-east end of the transect (Wairarapa) are longer and heavier 
than adult geckos from the south-west end of the transect (Wellington). Over the 15km 
sampled, the average adult female gecko size differed by 12mm and male geckos 
differed by approximately 8mm. The increase in mass approximately doubles for both 
males and females from Turakirae Head to Ocean Beach. Over the same 15km a change 
in allele frequency is seen in one nuclear locus and the mitochondrial 16S locus. This is 
strong evidence for a hybrid zone.  
The two ends of the cline are very different from each other based on morphological 
and genetic variation. The adult size of geckos from Turakirae Head does not overlap 
with adults from Ocean beach at the other end of the zone. Size of the geckos alone can 
be used to distinguish between the two populations (Figure 2). As well as the snout-vent 
length of geckos the overall bulk of the animal changes along this transect, with Ocean 
beach geckos getting considerably bulkier than Turakirae Head geckos. Although the 
weights were influenced by both season and tail loss, it is clear from the comparisons of 
adult males and field observations that there is an increase in mass.   
The significant genetic difference between the two end populations, Turakirae Head and 
Lake Pounui, is seen in the mitochondrial DNA locus. There are no shared haplotypes 
between these two populations (Table 5, Figure 8) despite the relatively short distance 
between them of only 25km (this includes the inland Pounui site). Among the three 
nuclear loci (Rag-1, Rag-2, and C-mos) there are alleles shared between the extreme 
populations, however, there are also alleles that are found in one population but not the 
other (Table 7-9; Figure 9-11). Turakirae Head and Ocean beach are very different from 
each other sharing only one haplotype. In addition, Ocean Beach and Lake Pounui are 
35 
 
different from each other sharing only two haplotypes. The geographic distance between 
Turakirae Head and Ocean beach is only 15km and between Ocean beach and Lake 
Pounui is 10 km. These changes occur over short distances. For all four loci the genetic 
differentiation between Turakirae Head and Lake Pounui is significant (p-value= 0.05) 
(Table 10). The FST pairwise differences for the three nuclear loci confirm that there is a 
significant difference between the two populations, Turakirae Head and Lake Pounui. 
This is also seen between Turakirae Head and Ocean Beach geckos (Table 11)  
Morphological variation is high, with polymorphism of toe lamellae number, eye 
colour, colour pattern and hue; however, none of these characters or combinations of 
characters can distinguish isolated populations. Genetic diversity is very high with all 
sampled populations having unique (private) alleles at some or all loci studied. This 
level of morphological and genetic diversity suggests both large populations and limited 
dispersal. 
An interesting observation in this study is the amount of genetic variation found for the  
mitochondria and three nuclear loci. Rato & Harris (2008) also noted high intraspecific 
variation in their study species of geckos. They noted that compared to their nuclear 
markers (Rag-1, C-Mos), their mtDNA had more variation, as found in this study with 
23 alleles for the mitochondria and between 7-9 alleles for the three nuclear alleles. 
However, the 7-9 alleles for the nuclear markers used in this study are high compared to 
many studies of reptile populations. Godinho et al (2006) in their study on the Iberian 
lizard found only 4 alleles for their C-mos gene and Arnold et al (2008) found six Rag-2 
alleles across ten islands and eight islets in the Cape Verde Islands, covering more than 
300kms. The variation in this study is as expected for a reptile species. For all loci used, 
mitochondria has the highest variation at 7.16% and the three nuclear loci < 3%. Note 
36 
 
that 7% is higher than the barcoding definition of species (Herbert et al 2003; Rubinoff 
et al 2006), which uses a more rapidly evolving gene (COI).  Leache and Cole (2007) 
found in their fence lizard hybrid zone a maximum of 10.63% divergence which is also 
significantly higher than the DNA barcode cut off.  
Hybrid fitness 
The width of a hybrid zone can be maintained by two types of selection, either by a 
balance of selection against hybrid individuals and the dispersal distance of individuals 
(endogenous) or by an environmental gradient (exogenous) (Endler 1977; Barton & 
Hewitt 1985; Barton & Gale 1993; Kruuk et al 1999; Fitzpatrick & Shaffer 2004; Gee 
2004). I found no evidence of reduced fitness of hybrids compared to the parents. There 
is a high level of heterozygosity in all populations at each location for two of the three 
nuclear genes (Rag-1, Rag-2, and Table 12). Fishermans rock population (near the 
centre of the clines) has for all three loci around 60% heterozygotes made up of adults 
and juveniles, however for the c-mos marker the other five populations either side of 
Fishermans Rock have a lower heterozygosity level compared to those same locations 
for Rag-1 and Rag-2. This may be the result of selection against hybrids at this locus 
with locus-specific selection acting on the c-mos gene but not the other two genes. 
However, c-mos did not have alleles specific to either end of the transect, but allele C 
was more common at Fishermans Rock than anywhere else sampled. A comparison 
between juveniles and adults show no significant lowering of heterozygosity as would 
be expected by hybrid disadvantage. In fact, heterozygosity in adults was higher than 
juveniles, suggesting heterozygotes may have an advantage (heterosis). However, 
juveniles showed a larger range among populations, from low to high heterozygosity. 
Heterozygosity at each locus was not significantly different from expected of Hardy-
37 
 
Weinberg. In addition, there is no evidence of linkage of loci as might be expected if 
parental genotypes are at a selective advantage. 
What may this mean for selection? It is often very difficult to determine which of the 
two kinds of selection is acting on the zone (Kruuk et al 1999; Johansson et al 2008) 
because both models produce similar shaped clines (Gifford 2008). If it was endogenous 
selection we would expect to see strong selection against hybrids and cline shapes that 
are coincident and concordant for unlinked markers (Marshall and Sites 2001). This 
kind of selection will form hybrid zones called “tension zone model” (Brumfield et al 
2001; Ruegg 2008). We see this concordant and coincident cline shape in our zone but 
we lack evidence of strong selection against hybrids. The other form of selection, 
exogenous, is where the populations’ ability to adapt to the local environment is 
selected against (Barton and Hewitt 1985; Kruuk et al 1999; Gee 2004). The conditions 
for this form of selection involve selection to act independently on each locus (Kruuk et 
al 1999; Durret et al 2000).  We may not see any effect on the genotypes present in the 
populations because those genotypes do okay in that location. The clines for different 
traits may also be disconcordant due to the selection pressures on each locus and/or trait 
(Durret et al 2000) and is often referred to as the “ecotone model” (Endler 1977).   
There have been a number of studies of hybrid zones showing each of the different 
selection types influencing hybrid zones. For example Bert & Arnold (1995) found 
evidence for both exogenous and endogenous selection in the hard clam Mercenaria 
species located in a polyhaline lagoon in east-central Florida. A narrow hybrid zone 
between chickadee species (Poecila atricapilla and P. carolinensis) appears to be 
maintained by endogenous selection (Bronson et al 2003), Phillips et al (2004) found 
endogenous selection acting on the hybrid zone of Carlia rubrigularis skinks. All 
38 
 
studies confirmed the endogenous selection based on the hybrid fitness. An example of 
exogenous selection is in a quail hybrid zone (Callipepla californica and C. gambelii) 
where the hybrid zone crosses an environmental gradient of precipitation and vegetation 
(Gee 2004). More examples consider exogenous as a contributing factor but not the 
dominating force for maintaining a hybrid zone (Johansson et al 2008). 
Further study looking at the populations in more depth to gather more information may 
better show the selection method acting on the hybrid zone and the populations.  
 Average FST from the three nuclear loci (0.78) indicates that there is considerable 
differentiation among the populations. This is what we would expect considering the 
movement of geckos throughout their life time based on mark-recapture studies. For 
example, in a study of H. maculatus in the South Island 90% of animals recaptured were 
less than 10m away from their original capture site (Lettink 2007).  So migration is 
limited and gene flow would be low. As in many other studies of population structuring, 
population differentiation is relatively high, such as in the Iberian lizards with a FST = 
0.46 found throughout their whole distribution of the species in the Iberian peninsula 
(Godinho et al 2006), Genets (type of cat) from South Africa had significant FST, but in 
contrast their South African distribution is much larger (but data not shown) (Gaubert et 
al 2003).   
The morphological data suggest that the hybrid zone is around the Fishermans Rock 
area where individuals are of intermediate size. Estimates of the cline centres for the 
genetic markers are coincidence, but only mtDNA cline is concordant (same width) as 
the morphology. Rag-2 has a significantly wider cline than the morphology and 
mtDNA. However, for the nuclear marker and the male and female snout-vent length 
the centres lie somewhere around the Fishermans rock location, 5km – 8km from 
39 
 
Turakirae Heads. The centre of the mitochondrial DNA is more to the North-East end 
closer to Ocean Beach site, but confidence intervals overlap with male snout-vent 
length. A shift in mtDNA cline has been observed in many studies including 
Hispaniolan lizard (Gifford 2008); Swainsons Thrush (26km difference between 
mtDNA and other markers) (Ruegg 2008) and in Fence lizards (7km difference between 
mtDNA and other markers) (Leache & Cole 2007). The reason for this may include that 
mtDNA is haploid (smaller population size) and being maternal it measures only the 
movement of females, who may move shorter distances than males (Sequeira et al 
2004).  Rag-2 and the snout-vent length for both males and females have centres close 
together but the difference in width may indicate weak selection on the different traits. 
Linkage disequilibrium does not exist between the three nuclear and mitochondrial loci 
suggesting that selection on any one locus does not affect other loci.  
The cline widths of both the male and female snout-vent lengths are relatively narrow 
compared to the nuclear DNA, this suggests that selection on morphology is greater 
than this nuclear marker. So it appears that although there is not a complete 
reproductive isolation of the geckos, selection based on the phenotype may favour 
smaller geckos in the south-west and larger geckos in the north-east, creating a narrow 
zone.  
Many studies look at selection factors for their hybrid zone, including climate (Ruegg 
2008), and habitat/ecotypes (Leache & Cole 2007). There is no major physical or 
climatic barrier along the coastal transect, so it seems unlikely that the environment is 
playing a major role in limiting geneflow between these geckos. However, rock type 
and vegetation do vary from one location to the next. Turakirae Head site is dominated 
by muehlenbeckia shrubs and rocks and Ocean beach dominated by river beds and 
40 
 
forest of Kanuka (Kunzea ericoides) trees and Tutu (Coriaria arborea). Ecological 
measures of habitat were not made so it is not known whether habitat affects the clines 
positions. If there is a transition in habitat it may have some influence on the size of the 
geckos creating a narrow hybrid zone for size. Genotypes are shared among the 
populations supporting the idea that geneflow and hybridisation occurs between 
adjacent populations. A study on two species of quails showed that geneflow was 
limited to the region of contact within the zone (Gee 2006) and therefore further 
sampling away from the zone would be informative. 
In many studies of species that form a steep transition along a transect indicate an 
environment barrier (Gifford 2008). In Gifford’s (2008) study on the Hispaniolan lizard 
the contact zone was on an ecotone characterised by a difference in temperature and 
precipitation. Gee (2006) had found clines of genetic, plumage and morphometric trait 
coincides with vegetation and climatic gradients. Leache and Cole (2007) study on 
fence lizards also meet at an ecotone of woodland and grassland.   
The age of this hybrid zone is unknown but based on the amount of mtDNA divergence 
between the two end populations it may be fairly old. The mitochondrial haplotype 
clade characteristic of Turakirae Head is considerably different from that at Lake 
Pounui.  
 
 
 
 
41 
 
 
Conclusion 
Using morphological and genetic techniques I am able to identify features that 
distinguish Turakirae Head geckos from Ocean beach geckos, or more in general 
distinguish coastal geckos from inland geckos. There is a significant difference in the 
average size of the adult geckos between the two populations at Turakirae Head and 
Ocean beach. Morphologically the trend is that Turakirae Head geckos are smaller than 
the Ocean beach geckos, based on an average, there is no overlap in the sizes 15km 
apart. However, between these two populations geckos are often of intermediate sizes. 
There is sharing of alleles between populations and some alleles are unique to a 
population but these are usually rare. Mitochondrial DNA may give more information 
into where a gecko may be from than nuclear DNA but this should not be used alone to 
distinguish populations, as there is still some minor overlapping variation. 
From the large amount of alleles that are shared between populations we can justify that 
there is a reasonable amount of hybridising between populations. There is enough of a 
difference between Turakirae Head and Ocean beach to suggest that gene flow is 
relatively slow. The movement of geckos is known to be limited in other locations and 
populations large and therefore we have a high differentiation between populations 
(Lettink 2007).  
The width of the nuclear cline (Rag-2) is wide compared to the narrow cline for the 
morphological characters. This suggests that there is greater selection for size of geckos 
than for genotype. It is very likely that size is important in how well a gecko does in a 
particular habitat and that this hybrid zone is evidence of environment-gradient 
42 
 
selection. Characterising the habitat change over the transect is an important step in 
future work to provide a better understanding of selection factors that may be occurring 
in the zone. 
Due to the amount of hybridising, sharing of alleles and widths of genetic markers it 
indicated that this zone is not an effective barrier to gene flow and unlikely to facilitate 
speciation. These populations are still overall similar to each other and therefore cannot 
be considered different species. Using the biological species concept there is no 
reproductive isolation between the two populations and that it would be wise to consider 
the populations as variants or subspecies. No evidence was found of assortative mating 
(mate choice based on size) to maintain the zone. But it is possible that selection 
pressures might result in increased frequency should such behaviour arise, and this 
could lead to the separate populations further differentiating into different species 
(Schluter 2009). The current divergence between the two end populations and the 
formation of a narrow hybrid zone with intermediate morphology for size is evidence 
that selection will form different species if the gene flow and habitat differences remain 
as they are now.  
 
 
 
 
 
 
43 
 
 
Reference 
Alexandrino J, Baird S, Lawson L, Macey R, and Moritz C. (2005). Strong selection 
against hybrids at a hybrid zone in the Ensatina ring species complex and its 
evolutionary implications. Evolution, 59(6):1334-1347  
Arevalo E., Davis S. and Sites J. (1994). Mitochondrial DNA sequence divergence and 
phylogenetic relationships among eight chromosome races of the Sceloporus 
grammicus complex (Phrynosomaidae) in central Mexico. Systematic biology,
 43(3):387-418 
Arnold E., Vasconcelos R., Harris D., Mateo J. and Carranza S. (2008). Systematics, 
biogeography and evolution of the endemic Hemidactylus geckos (Reptilia, 
Squamata, Gekkonidae) of the Cape Verde Islands: based on morphology and 
mitochondrial and nuclear DNA sequences. Zoologica Scripta, 37(6): 619-636     
Barton NH 1979. The dynamics of hybrid zones. Heredity 43: 341-359. 
Barton NH, Hewitt GM 1985. Analysis of hybrid zones. Annual Review of Ecology and
 Systematics 16:113-148. 
Barton N. and Baird S. (1996) ANALYSE 1.30 PPC. Available at
 www.biology.ed.ac.uk/research/institutes/evolution/software/Mac/Analyse/index.html 
Barton N. and Gale K. (1993) Genetic analysis of hybrid zones., pp 13-45 in Hybrid
 zones and the evolutionary process, edited by R.G. Harrison. Oxford University
 Press. Oxford. 
44 
 
Begg J., Johnston M. (Compilers) (2000) Geology of the Wellington area. Institute of
 Geological & Nuclear Sciences 1:250 000 geological map 10. 1 sheet + 64p.
 Lower Hutt, New Zealand: Institute of Geological & Nuclear Sciences Limited. 
Berry Os, Tocher M. And Sarre S. (2004). Can assignment tests measure dispersal?
 Molecular Ecology, 13:551-561 
Bert T. and Arnold W. (1995). An empirical test of predictions of two competing 
models for the maintenance and fate of hybrid zones: both models are supported 
in a Hard-Clam hybrid zone. Evolution, 49(2): 276-289 
Bock W. (2004). Species: the concept, category and taxon. Journal of zoological 
systematics and evolutionary research, 42: 178-190 
Bronson C, Grubb T, and Braun M. (2003). A test of the endogenous and exogenous
 selection hypotheses for the maintenance of a narrow avian hybrid zone.
 Evolution 57(3):630-637  
Brumfield R., Jernigan R., McDonald D. And Braun M. (2001). Evolutionary
 implications of divergent clines in an avian (Manacus: Aves) hybrid zone.
 Evolution, 55(10): 2070-2087 
Durret R., Buttel L. and Harrison R. (2000) Spatial models for hybrid zones. Heredity
 84:9-19 
Endler J. (1977) Geographic variation, speciation and clines. Princeton University, 
Press, Princeton, NJ 
Fitzpatrick B. and Shaffer H. (2004). Environment-dependent admixture dynamics in a 
tiger salamander hybrid zone. Evolution 58(6): 1282-1293 
45 
 
Gamble T., Bauer A., Greenbaum E. and Jackman T. (2008). Evidence for Gondwanan
 vicariance in an ancient clade of gecko lizards. Journal of Biogeography, 35: 88
 104 
Gamble T., Simons A., Colli G. and Vitt L. (2008). Tertiary climate change and the
 diversification of the Amazonian gecko genus Gonatodes (Sphaerodactylidae,
 Squamata). Molecular Phylogenetics and Evolution 46: 269-277 
Gaubert P., Taylor P., Fernandes C., Bruford M. and Veron G. (2005) Patterns of
 cryptic hybridisation revealed using an integrative approach: a case study on
 genets (Carnivora, Vivverridae, Genetta spp.) from the southern African
 subregion. Biological Journal of the Linnean Society 86:1-33 
Gee J. (2004) Gene flow across a climatic barrier between hybridising avian species,
 California and Gambel’s Quaill (Callipepla californica & C. gambelii).
 Evolution 58(5):1108-1121 
Gifford M. (2008) Divergent character clines across a recent secondary contact zone in
 a Hispaniolan lizard. Journal of Zoology 274:292-300 
Godinho R., Domingues V., Crespo E. and Ferrand N. (2006). Extensice intraspecific
 polymorphism detected by SSCP at the nuclear C-mos gene in the endemic
 Iberian lizard Lacerta schreiberi. Molecular Ecology, 15: 731-738 
Groth J. and Barrowclough G. (1999) Basal divergences in birds and phylogenetic utility
 of the nuclear RAG-1 gene. Molecular Phylogenetics and Evolution 12(2):115
 123 
Harrison RG 1993b.  Hybrids and hybrid zones: historical perspective. Pp. 3-12 in R. G.
46 
 
 Harrison, ed. Hybrid zones and the evolutionary process. Oxford University
 Press, New York. 
Hebert P., Cywinska A., Ball S. and deWaard J. (2003). Biological identification
 through DNA barcodes. Proc. R. Soc. Lond. B. 270: 313-321 
Hitchmough R. (1997). A systematic revision of the New Zealand Gekkonidae.
 Unpublished Ph.D  Thesis. Victoria University, Wellington, New Zealand  
Jennings W., Pianka E. and Donnellan S. (2003). Systematics of the lizard Pygopodidae
 with implications for the diversification of Australian temperate biotas.
 Systematic biology, 52(6):757-780  
Jesus J., Harria D. and Brehm A. (2007). Relationships of Afroablepharus Greer, 1074
 skinks from the Gulf of Guinea islands based on mitochondrial and nuclear
 DNA: Patters of colonisation and comments on taxonomy. Molecular
 Phylogenetics and Evolution, 45: 904-914 
Johansson H., Surget-Groba Y. and Thorpe R. (2008). The roles of allopatric divergence 
and natural selection in quantitative trait variation across a secondary contact 
zone in the lizard Anolis roquet. Molecular Ecology 17: 5146-5156 
Kocher T., Thomas W., Meyer A., Edwards S., Paabo S., Villablanca F. and Wilson A.
 (1989). Dynamics of mitochondrial DNA evolution in animals: Amplification
 and sequencing with conserved primers. Evolution, 86:6196-6200  
Kruuk L., Baird S., Gale K. and Barton N. (1999). A comparison of multilocus clines 
maintained by environmental adaptation or by selection against hybrids. 
Genetics, 153: 1959-1971 
47 
 
Leache A. and Cole C. (2007). Hybridisation between multiple fence lizard lineages in
 an ecotone: locally discordant variation in mitochondrial DNA, chromosomes,
 and morphology. Molecular Ecology 16:1035-1054 
Lettink M. (2007) Detectability, movements and apparent lack of homing in
 Hoplodactylus maculatus (Reptilia: Diplodactylidae) following translocation.
 New Zealand Journal of Ecology 31(1):111-116 
Macey J., Larson A., Ananjeva N., Fang Z. and Papenfuss T. (1997a). Two novel gene
 orders and the role of light-strand replication in rearrangement of the vertebrate
 mitochondrial genome. Molecular evolution biology, 14(1):91-104 
Macey J., Schulte J., Larson A., Ananjeva N., Wang Y., Pethiyagoda R., Rastegar
 Pouyani N. and Papenfuss T. (2000). Evaluating Trans-Tethys Migration: an
 example using acrodont lizard phylogenetics. Systematic biology, 49(2):233-256. 
Marshall J. and Sites J. (2001). A comparison of nuclear and mitochondrial cline shapes 
in a hybrid zone in the Sceloporus grammicus complex (Squamata; 
Phrynosomatidae). Molecular Ecology, 10: 435-449 
Mayr E. (1942) Systematics and the origin of species: From the viewpoint of a 
zoologist. New York: Columbia University Press 
McSaveney M., Graham I., Begg J., Beu A., Hull A., Kim K. And Zondervan A. (2006)
 Late Holocene uplift of beach ridges at Turakirae Head, south Wellington coast,
 New Zealand. New Zealand Journal of Geology and Geophysics 49:337-358 
Moore WS 1977. An evaluation of narrow hybrid zones in vertebrates. Quarterly
 Review of Biology 52: 263-277. 
48 
 
Morgan-Richards M., Smissen R., Shepherd L., Wallis G., Hayward J., Chan C.,
 Chambers G. and Chapman H. (2009). A review of genetic analyses of
 hybridisation in New Zealand. Journal of the Royal Society of New Zealand
 39(1):15-34 
Moritz C., Schneider C. and Wake D. (1992). Evolutionary relationships within the
 Ensatina eschscholtzii complex confirm the ring species interpretation.
 Systematic biology, 41(3):273 291 
Neilsen S., Bauer A. and Jackman T. (2007). Molecular Systematics of the  geckos of
 New Zealand. Poster at Evolution meeting, Christchurch. NZ. 
Oliver P., Hugall A., Adams M., Cooper S. and Hutchinson M. (2007). Genetic
 elucidation of cryptic and ancient diversity in a group of Australian
 diplodactyline geckos: The Diplodactylis vittalus complex. Molecular
 Phylogenetics and Evolution, 44: 77-88 
Pepper M., Doughty P. and Keogh J. (2006). Molecular phylogeny and Phylogeography
 of the Australian Diplodactylus stenodactylus (Gekkota: Reptilia) species-group
 based on mitochondrial and nuclear genes reveals an ancient split between
 Pilbara and non Pilbara D. stenodactylus. Molecular Phylogenetics and
 Evolution, 41: 539-555 
Phillips B, Baird S. and Mortizo C. (2004). When vicars meet: a narrow contact zone
 between morphologically cryptic phylogeographic lineages of the rainforest
 skink, Carlia rubrigularis. Evolution, 58(7):1536-1548 
Prychitko T. and Moore W. (1997). The utility of DNA sequences of an intron from the
 β Fibrinogen gene in phylogenetic analysis of Woodpeckers (Aves: Picidae).
49 
 
 Molecular phylogenetics and evolution, 8(2):193-204  
Rato C. and Harris D. (2008). Genetic variation within Saurodactylus and its
 phylogenetic relationships within the Gekkonoidea estimated from
 mitochondrial and nuclear DNA sequences. Amphibia-Reptilia, 29: 25-34 
Reeder T. (1995) Phylogenetic Relationships among Phrynosomatid Lizards as inferred
 from Mitochondrial Ribosomal DNA sequences: Substitutional bias and
 Information content of transitions relative to transversions. Molecular
 Phylogenetics and Evolution 4(2):203-222  
Robb J. (1980). New Zealand Amphibians and Reptiles. Auckland, New Zealand,
 William Collins Publishers Ltd 
Rubinoff D., Carmeron S. and Will K. (2006) A genomic perspective on the
 shortcomings of mitochondrial DNA for “barcoding” identification. Journal of
 Heredity 97(6): 581-594 
Ruegg K. (2008) Genetic, morphological, and ecological characterisation of a hybrid
 zone that spans a migratory divide. Evolution 62(2): 425-466 
Schluter D. (2009) Evidence for Ecological speciation and its alternative. Science
 323:737-741   
Sequeira F., Rocha A., Arntzen J. and Ferrand N. (2005). Genetic exchange across a
 hybrid zone within the berian endemic golden-striped salamander, Chioglossa
 lusitanica. Molecular Ecology 14: 245-254 
Vidal N. and Hedges S. (2004) Molecular evidence for a terrestrial origin of snakes.
 Proc. R. Soc. Lond. B 271: s226-s229 
50 
 
Whitaker T. (1996). New Zealand Frogs and Reptiles. Auckland, New Zealand, David
 Bateman Limited 
Whitaker A. (1982) Interim results from a study of Hoplodactylus maculatus
 (Boulenger) at Turakirae Head, Wellington. In: Newman DG ed. New Zealand
 herpetology: proceedings of a symposium held at Victoria University of
 Wellington 29-31 January 1980. New Zealand Wildlife service occasional
 publication 2: 363-374  
 
51 
 
Appendix 1 Data table with morphology and genetic data for each individual sampled from each location. 
Gecko Age Sex Eye colour Weight Snout-Vent Snout-Ear Head Width 
Head 
Height 
No. Of 4th toe 
lamellae Rag-1 Rag-2 cMos 16S 
TH001 A M Gold 4 55.65 14.27 13.09 4.83  AB AC AA A 
TH002 A M Gold 3 52.42 13.16 11.42 5.04 11 AA BB AA H 
TH003 J - Bronze 2 40.37 12.46 9.81 4.38  AC BC AA N 
TH004 J - Bronze 0.5 32.07 10.15 8.86 4.41  AA BD AA A 
TH005 J - Bronze 0.5 35.85 8.49 8.29 3.98  AA AC AA A 
TH006 A F Gold 3.4 47.5 13.4 11.19 5.48 10 AB AA AA A 
TH007 A F Green 4 55.5 16.67 13.01 5.82 11 AB CC AA R 
TH008 J - Green 2 45.65 10.69 9.23 3.15 11 AC AC AA H 
TH009 A F Green 4 51.47 13.6 12.67 6.13 11 AC DD AB G 
TH010 A F Bronze 3 52.61 15.13 12.06 5.53 11 AB BC AA A 
TH011 J - Bronze 0.5 27.77 6.51 6.52 2.82 10 AA AD AA G 
TH012 A F Bronze 5.3 55.52 12.96 12.84 5.1 11 AA CC AA A 
TH013 A F Green 4.4 54.51 13.57 12.07 5.21 11 AA BC AA A 
TH014 J - Gold 2.9 42.39 11.75 9.63 5.92 11 AA CC AA M 
TH015 A F Gold 6.1 54.83 13.94 13.23 6.51 10 BB BB AA A 
TH016 A F Gold 6 51.38 11.53 12.26 6.54 10 AA AC AA H 
TH017 J - Bronze 4 48.49 10.7 10.69 5.19 9 AA AB AA L 
TH018 J - Gold 2.4 40.69 10.19 9.82 4.12 11 AB BC AA A 
TH019 A F Green 5.7 53.51 12.95 12.05 5.26 11 AB BE AA N 
TH020 A F Green 5.2 50.26 12.46 11.81 5.47 9 BC CC AB L 
TH021 J - Gold 1.8 41.03 8.78 7.96 4.65 10 - - - - 
TH022 A F Gold 5.9 56.97 12.47 11.01 4.98 11 - - - - 
TH023 A F Gold 5.4 54.53 11.32 12.17 4.88 9 - - - - 
TH024 A M Gold 4.1 55.57 12.68 12.74 5.48 9 AA DD AA H 
TH025 J - Gold 1.8 40.04 8.76 9.33 3.71 10 - - - - 
TH026 A F Bronze 2.9 47.04 13.15 10.2 4.44 9 - - - - 
TH027 A M Green 7.5 57.17 15.96 14.83 5.54 10 AC AA AA M 
TH028 A F Bronze 6.3 50.15 13.85 12.8 5.86 9 CC AB AB A 
TH029 J - Green 4.7 46.63 11.89 9.67 4.48 11 - - - - 
TH030 J - Gold 0.9 38.62 11.21 9.78 4.93 10 - - - - 
52 
 
THRAN - - - - - - - - - AA BE AA A 
TH100 - - - - - - - - - BE AE AA A 
TH200 - - - - - - - - - BC CD AA G 
BW101 A F Bronze 7.4 60.49 14.49 14.02 6.7 10 AA AA AA A 
BW102 J - Gold 3.5 50.41 11.73 10.75 5.65 10 BB AE AA A 
BW103 J - Green 1.7 37.85 10.37 8.29 3.99 10 AA BC AB F 
BW104 A M Bronze 4.9 57.25 12.84 11.09 6.19 11 AC AD AA F 
BW105 J - Bronze 2 42.93 11.31 9.16 4.69 10 AA AC AC A 
BW106 A M Green 8.1 65.29 15.13 14.29 6.84 9 AA AC AB G 
BW107 A M Green 8.1 61.78 13.64 12.25 6.44 9 AC AC AA A 
BW108 J - Bronze 1.4 35.42 9.25 8.34 3.52 9 BE AD AA A 
BW109 A M Green 4.8 57.79 12.74 11.09 5.68 10 AB AE AA A 
BW110 A M Green 5.3 55.77 12.4 11.18 6.79 10 AB AA AA A 
BW111 A M Bronze 5.2 55.54 13.32 11.82 5.01 10 AA AA AA A 
BW112 A M Green 5.3 58.56 13.83 12.41 5.25 10 - - - - 
BW113 A F Gold 3.9 52.12 13.35 10.18 4.38 10 AB CD AA A 
BW114 A M Green 3.6 52.67 11.64 10.41 5.51 9 AC AD AA A 
BW115 J - Gold 2.3 48.6 12.46 11.63 5.54 10 AA EE AA A 
BW116 J - Green 2.3 50.15 12.04 9.81 5.01 9 AA AA AA A 
BW117 A F Green 7.1 59.95 13.91 12.21 6.06 9 AC CC AA A 
BW118 A F Green 4.5 57.93 12.57 10.92 6.09 10 AA AB AA A 
BW119 J - Bronze 2.1 47.79 11.36 9.88 5.09 10 AC AA AA A 
BW120 A M Green 1.5 52.58 12.84 9.96 5.25 10 AE AD AA F 
BW121 A M Green 0.4 52.46 13.25 10.78 6.05 10 - - - - 
BW122 A F Gold 3.6 57.55 12.07 10.79 5.02 10 BB AB AA A 
BW123 A F Green 5.6 61.07 12.24 12.23 6.02 9 AC AB AA A 
BW124 A M Green 5.1 56.86 12.09 11.19 5.02 10 AA AA AA A 
BW125 A F Green 4.7 58.1 13.56 11.31 5.48 10 AB AD AA A 
BW126 A M Green 5.3 61.21 13.52 11.19 5.83 9 - - - - 
BW127 A F Gold 1.8 54.97 12.51 10.22 5.5 10 AB AD AA A 
BW128 A F Green 1.6 62.42 13.87 12.1 5.99 11 BB AA AA F 
BW129 A F Green 8.4 63.45 14.45 11.97 6.49 9 - - - - 
BW130 A M Green 6.2 62.15 15.18 12.92 6.2 10 - - - - 
BW7091 - - - - - - - - - BB BB AA A 
BW7092 - - - - - - - - - AC AD BB A 
53 
 
BW7093 - - - - - - - - - CE AD AA F 
FR150 A M bronze 2.6 49.62 11.13 9.92 3.99 10 AB AB AC A 
FR151 A F bronze 2.2 51.04 12.27 10.83 5.4 10 AD BD BC E 
FR152 A M green 7.1 65.12 14.1 13.67 7.05 11 AB BE AC A 
FR153 J - bronze 0.8 32.65 8.35 6.95 3.56 10 AA AD AC C 
FR154 J - bronze 0.8 31.46 8.25 6.98 3.96 9 - - - - 
FR155 A M bronze 4.9 61.17 14.07 12.1 6.27 10 BB DD AA B 
FR156 A F green 6.2 64.29 14.93 12.26 6.2 10 AB AD AA A 
FR157 A F bronze 3.7 53.03 11.94 10.35 6 12 CG AA AA E 
FR158 A M green 5.9 57.45 14.18 12.17 6.74 11 AA DD AC C 
FR159 J - green 2.2 45.23 11.47 9.88 5.03 10 AA CE AC B 
FR160 A M green 6.1 63.84 14.67 13.42 6.73 10 AA AA AA B 
FR161 A F gold 6.9 64.21 14.5 13.12 7.02 10 AA DD AC B 
FR162 A M green 6.4 63.66 14.75 13.7 6.02 12 AA DD AC C 
FR163 A F green 5 53.03 13.14 11.43 5.9 10 AB AD AC C 
FR164 A F bronze 6 63.71 14.55 12.14 6.26 10 AB CD AC B 
FR165 J - green 1.7 41.24 10.39 8.55 4.32 10 - - - - 
FR166 J - bronze 0.7 28.9 8.23 6.85 2.97 10 AB AD AC B 
FR167 J - green 0.8 32.35 8.04 7.27 4.02 9 AB AH AA B 
FR168 A F green 5.5 58.16 13.12 11.62 5.79 10 AB AC AC C 
FR169 A F green 4.4 58.6 13.29 11.54 5.25 10 AB AD AA B 
FR170 A M green 4.9 57.81 14.13 12.19 5.75 10 AA DD AC E 
FR171 J - bronze 2.4 45.06 11.09 9.23 4.53 10 - - - - 
FR172 A F green 6.2 62.67 14.08 11.91 5.44 10 - - - - 
FR173 J - bronze 1.7 40.81 10.4 8.79 4.84 9 - - - - 
FR174 A F green 6.1 62.89 14.89 12.91 6.09 12 - - - - 
FR175 J - green 3.4 52.22 12.09 11.01 5.56 9 - - - - 
FR176 J - bronze 1.3 37.58 8.32 7.03 3.98 10 - - - - 
FR177 A M gold 6 61.96 14.78 13.85 6.57 10 - - - - 
FR178 A F green 4.9 61.03 12.64 10.94 6.05 10 - - - - 
FR179 A M green 6.6 63.77 13.29 12.49 6.65 11 AA AD AA E 
GW070 A F Bronze 6.6 58.04 14.94 13.3 5.16 10 AA AA AA A 
GW071 A F Dark green 9.1 65.69 17.22 14.14 6.84 11 AB AE AA C 
GW072 J - Bronze 2.7 43.26 10.42 8.85 4.73 11 AB AB AI A 
GW073 A F Bronze 8.9 66.45 14.27 11.93 5.99 11 BB AD AA B 
54 
 
GW074 J - Bronze 1.2 39.74 10.23 8.29 4.55 11 AB DE AA G 
GW075 A M green 7.6 65.13 15.9 14.02 6.83 12 BC AB AA B 
GW076 A F green 10.2 68.47 16.59 15.47 7 10 AC AA AA G 
GW077 A F green 7.1 60.59 13.92 12.51 5.61 12 BB AA AA B 
GW078 J - green 1.5 35.62 9.04 8.31 3.51 11 AB AD AA B 
GW079 A M green 8 61.94 16 13.67 6.63 10 BI AA AE B 
GW080 A M green 9 66.86 13.99 12.06 6.15 10 AB AE AA G 
GW081 J - Green 3.3 48.93 11.66 9.97 5.49 10 AB AA AA D 
GW082 A F Green 7.5 61.88 14.09 11.27 5.98 11 AA AD AA B 
GW083 A M Green 7.1 64.32 14.59 13.05 6.11 9 AB AE AA C 
GW084 A M Gold 6.5 60.24 15.31 13.12 6.41 10 AB AA AA C 
GW085 A M Green 7.5 66.66 15.36 13.53 6.63 9 AB AE AA C 
GW086 A M Gold 3.9 51.98 12.79 10.51 4.65 11 AB AB AA C 
GW087 J - Bronze 3.2 52.4 11.81 11.05 4.64 10 AB AA AA C 
GW088 A F Bronze 3.8 54.11 10.96 11.06 6.39 11 AB AA AA T 
GW089 A M green 7.2 63.87 15.3 14.1 6.57 10 AD AA AA E 
GW090 J - Bronze 2.1 45.73 11.15 9.17 5.07 9 AA AE AA B 
GW091 A M green 6 65.41 14.7 13.23 6 11 - - - - 
GW092 A M green 6.4 63.34 13.27 12.69 6.13 10 - - - - 
GW094 A F green 9.5 68.98 15.05 14.23 7.34 11 - - - - 
GW095 A M green 7.1 64.5 14.84 13.44 7.9 11 - - - - 
GWRAN - - - - - - - - - BD EG AA V 
GWRAN2 - - - - - - - - - DF AG AA C 
OB033 J - Green 5.3 50.85 13.89 12.44 5.46 11 AB AA AA D 
OB034 J - Gold 5.4 57.95 14.2 13.49 5.93 11 BB AD AC A 
OB035 A M Green 7.9 63.97 15.7 14.87 6.53 12 AB AA AA B 
OB036 A F Green 9.1 65.53 16.72 13.32 7.04 12 BB AD AA A 
OB037 A F Green 8 60.66 14.17 13.35 7.16 10 BB CG AA B 
OB038 A F Bronze 8.5 60.05 15.69 13.93 7.23 11 AB AC AH B 
OB039 A F Bronze 7.1 66.72 13.64 12.85 7.37 10 BB AD AA C 
OB040 A M Green 7.3 67.38 15.28 14.05 7.36 11 AB AE AA A 
OB041 A F Gold 6 70.11 16.39 13.08 6.25 10 - - - - 
OB042 A M Bronze 5.3 62.55 14.36 12.81 6.75 10 - - - - 
OB043 J - Bronze 8 42.21 9.94 8.76 4.1 10 AA EE AC J 
0B044 A F Bronze 6.9 65.06 13.31 12.61 7.47 10 AB AA AA B 
55 
 
OB045 J - Bronze 2 42.62 10.54 9.07 5.86 11 AB AA AD I 
OB046 J - Bronze 1.1 32.62 9.04 7.72 4.33 10 AB AA AA D 
OB047 A M Gold 5.8 62.88 14.15 13.03 6.16 10 AB AE AA U 
OB048 A M Bronze 6.6 60.13 14.65 12.82 6.52 10 AB AA AC W 
OB049 A F Bronze 5.5 58.53 14.76 11.79 6.13 9 AB BE AA B 
OB050 A F Green 7.4 68.16 13.63 11.7 6.16 10 AB AE AA J 
OB051 A F Gold 8.3 66.56 13.49 12.29 6.7 9 BB AC AA B 
OB052 J - Bronze 0.7 32.42 7.81 6.63 4.11 11 AA AA AA I 
OB053 A M Gold 4.9 59.3 14.78 11.68 6.41 10 AB AD AA B 
OB054 A M Bronze 5.7 58.18 13.79 12.02 6.18 9 AB AD AJ C 
OB055 A F Gold 8.8 67.7 15.57 12.9 5.75 11 - - - - 
OB056 A M Green 9.9 71.84 16.11 15.02 7.53 9 AB AA AA J 
OB057 A M Gold 6.5 60.16 13.66 11.91 5.69 11 - - - - 
OB058 J - Bronze 1.8 43.16 10.33 8.18 3.58 9 - - - - 
OB059 A M Green 7.3 66.01 13.9 13.06 6.26 11 - - - - 
OB060 A M Bronze 9.3 69.25 16.92 15.64 7.61 10 - - - - 
RAH33 - - - - - - - - - AB AA AF K 
RAH34 - - - - - - - - - AH AB AA K 
RAH146 - - - - - - - - - DE AA AA O 
RAH229 - - - - - - - - - AE AD AH P 
RAH230 - - - - - - - - - AB AA AA J 
RAH231 - - - - - - - - - AA AA AA Q 
RAH232 - - - - - - - - - AA AA AA I 
RAH345 - - - - - - - - - AC AD AA C 
RAH346 - - - - - - - - - AB AA AA D 
RAH347 - - - - - - - - - AB AD AA C 
RAH348 - - - - - - - - - AB AC AA I 
RAH349 - - - - - - - - - AC DE AA D 
RAH355 - - - - - - - - - AC AA AH S 
RAH498 - - - - - - - - - BE AA AH J 
56 
 
Appendix 2 Mitochondrial primers and their sequences 
MtDNA 
Gene 
Primer Primer Sequence Source 
Cytochrome 
b 
Cyt b1 
(L14841) 
AAAAAGCTTCCATCCAACATCTC
AGCATGATGA 
Kocher et al, 
1989 
 Cyt b2 
(H15149) 
AAACTGCAGCCCCTCAGAATGAT
ATTTGTCCTCA 
Kocher et al, 
1989 
With Cyt b2 P1 TGAGGACAAATATCATTYTGRGG Jesus et al, 
2007 
 MV216 AAATAGGAAG
ATATCAC
TTCTGGTT
TG
AAT 
Mortiz 
 MVZ04 GCAGCCCCTCAGAATGATATTTCC
T 
Phillips et al 
2004 
 Ph-1 GACCCCAATACGAAAAACCACCC Phillips et al 
2004 
ND2 ND2b GCCCATACCCCAAAAATGTYG Jennings et 
al, 2003 
 ND2c AACCAAACCCAAACACGAAARAT
YAT 
Jennings et 
al, 2003 
 ND2d AAACCAAGAGCCTTCAAAG Jennings et 
al, 2003 
 ND2f TGTRGTTATRTGDGATATYCG Jennings et 
al, 2003 
 ND2e GCGCGCTGGTTTGGGTDWTTAGY Jennings et 
57 
 
TGTTAA al, 2003 
 L(mod of 
Jennings 
GCCCATACCCCGAAAATSTTG Oliver et al, 
2007 
 H TTAGGGTRGTTATTTHGAYATKCG Oliver et al, 
2007 
 L4645 ACAGAAGCCGCAACAAAATA Macey et al, 
1997a 
 L4831 TGACTTCCAGAAGTAATACAAGG Macey et al, 
1997a 
 L4882 TGACAAAAACTAGCACC Macey et al, 
1997a 
 H4980 ATTTTTCGTAGTTGGGTTTGRTT Macey et al, 
1997a 
 L5002 AACCAAACCCAACTACGAAAAAT Macey et al, 
1997a 
 H4629 AAGTATTTTGTTGCGGCTTC Macey et al, 
2000 
 L4882mod CAACATGACAAAAATCGCCCC Pepper et al, 
2006 
 tRNA
Asn CTAAAATRTTRCGGGATCGAGGC
C 
Read et al, 
2001 
ND4 ND4 F CACCTATGACTACCAAAAGCTCA
TGTAGAAGC 
Arevalo et al 
1994 
 ND4 R 
(LEU) 
CATTACTTTTACTTGGATTTGCAC
CA 
Arevalo et al 
1994 
58 
 
16S 16SH CCGGTCTGAACTCAGATCAGGT  
 L2510 CGCCTGTTTATCAAAAACAT  
 H3056 CTCCGGTCTGAACTCAGATCACGT
AGG 
 
 16S-F CTAACCGTGCAAAGGTAGCGTAA
TCAC 
Gamble et al, 
2008 
 16Sc GT[A/C]GGCCTAAAAGCAGCCAC Reeder, 1995 
 16Sb GCGCTGTTATCCCTAGGGTAACTT
G 
Reeder, 1995 
12s rRNA 12Sa CTGGGATTAGATACCCCACTA Kocher et al 
1989 
 12Sb TGAGGAGGGTGACGGGCGCT Kocher et al 
1989 
 L1091 AAAAAGCTTCAAACTGGGATTAG
ATACCCCACTAT 
Kocher et al 
1989 
 H1478 TGACTGCAGAGGGTGACGGGCGG
TGTGT 
Kocher et al 
1989 
Control 
region 
L15926 TCAAAGCTTACACCAGTCTTGTAA
ACC 
Kocher et al 
1989 
 L16007 CCCAAAGCTAAAATTCTAA Kocher et al 
1989 
 H00651 TAACTGCAGAAGGCTAGGACCAA
ACCT 
Kocher et al 
1989 
 
59 
 
 
Appendix 3 Nuclear markers and their sequences tested. 
Gene Primers Primers Sequence Source 
RAG-1 L2408 TGCACTGTGACATTGGCAA Vidal & 
Hedges 2004 
 H2928 GACTGCYTGGCATTCATTTT Vidal & 
Hedges 2004 
 H2920 GCCATTCATTTTYCGAA Vidal & 
Hedges 2004 
 G396 TCTGAATGGAAATTCAAGCTGT
T 
Groth & 
Barrowclough 
1999 
 G397 AAAGGTGGCCGACCGAGGCAG
CATC 
Groth & 
Barrowclough 
1999 
 R18 GATGCTGCCTCGGTCGGCCACC
TTT 
Groth & 
Barrowclough 
1999 
 F700 GGAGACATGGACACAATCCAT
CCTA 
Bauer et al 
2007 cited in 
Gamble et al 
2008 
 R700 TTTGTACTGAGATGGATCTTTT
TGCA 
Bauer et al 
2007 cited in 
60 
 
Gamble et al 
2008 
 L-RAG-1b TTCCAGCCAYTGCATGCTCT  
 H-snRAG-1 ATTGCCAATGTCACAGTGCA  
RAG2 EM1-F TGGAACAGAGTGATYGACTGC
AT 
Gamble et al 
2008 
 EM1-R ATTTCCCATATCAYTCCCAAAC
C 
Gamble et al 
2008 
 PY1-F CCCTGAGTTTGGATGCTGTACT
T 
Gamble et al 
2008 
 PY1-R AACTGCCTRTTGTCCCCTGGTA
T 
Gamble et al 
2008 
C-mos G73 GCGGTAAAGCAGGTGAAGAAA Saint et al 
1998 
 G74 TGAGCATCCAAAGTCTCCAATC Saint et al 
1998 
 FU-F TTTGGTTCKGTCTACAAGGCTA
C 
Gamble et al 
2008 
 FU-R AGGGAACATCCAAAGTCTCCA
AT 
Gamble et al 
2008 
 Mos-F CTCTGGKGGCTTTGGKKCTGTS
TACAAGG 
Godinho et al 
2006 
 Mos-R GGTGATGGCAAANGAGTAGAT
GTCTGC 
Godinho et al 
2006 
 G78 AGRGTGATRGCAAAVGARTAR  
61 
 
ATG 
 G155 TGCTACTAWAGCYYTCCAGC  
 G303 AATTATGCCATCMCCTMTTCC  
Phosducin PHOF2 AGATGAGCATGCAGGAGTATG
A 
Bauer et al 
2007 cited in 
Gamble et al 
2008 
 PHOR1 TCCACATCCACAGCAAAAAAC
TCCT 
Bauer et al 
2007 cited in 
Gamble et al 
2008 
ACM4 tgF CAAGCCTGAGAGCAARAAGG Gamble et al 
2008 
 tgR ACYTGACTCCTGGCAATGCT Gamble et al 
2008 
 int-F TTTYCTGAAGAGCCCTCTGGTC Gamble et al 
2008 
 int-R CAAATTTCCTGGCAACATTRGC Gamble et al 
2008 
B-
fibrinogens 
FIB-B17U GGAGAAAACAGGACAATGACA
ATTCAC 
Prychitko & 
Moore 1997 
 FIB-B17L TCCCCAGTAGTATCTGCCATTA
GGGTT 
Prychitko & 
Moore 1997 
 FIB-B17U2 CATCCATGCAGTTCTGGCAATT
CCAAGT 
Prychitko & 
Moore 1997 
62 
 
 FIB-B17L2 TGGGAGGTGAAGCAGCTAAGA
AAAACAA 
Prychitko & 
Moore 1997 
 FIBL210 ACTTGGAATTGCCAGAACTGCA
T 
Prychitko & 
Moore 1997 
 FIBU330 CTAGGAACTGCAAGTAA Prychitko & 
Moore 1997 
Aldolase Ald 1 CRAAGAAGGATGGAGCTGACT
TTGC 
Phillips et al 
2004 
 Ad 2 CRGCCATTCTGTAACACAACAG
CCAA 
Phillips et al 
2004 
 Ald-1 TGTGCCCAGTATAAGGATCG Phillips et al 
2004 
 Ald-2 CCCATCAGGGAGAATTTCAGG
CTCCACAA 
Phillips et al 
2004 
Rhodopsin Rho3 CRCCTTGCCTGGACACCCTATG
CTG 
Phillips et al 
2004 
 Rho4 CRCTCTGGAATAAAGGAGAGG
GTCTCT 
Phillips et al 
2004 
Cyt P450 P450 UP1 TTGGAGACGGGCAGTGTGAT  
 P450 UP2 TCCGAGTACGTCCAGGC  
 P450 UP3 CATGTTCGAGGGTCACGACACC
ACG 
 
 P450 UP4 ACGCCTATGTACCCTTCAGC  
 P450 UP5 GCCTGGAATAACCTGGACGAC  
 P450 UP6 ACCAAGAAAATGACGCAACAA  
63 
 
AG 
 P450 DO1 GCCTGGACGTACTCGGA  
 P450 DO2 GGAGGGTCTTGAAGCAGGCAT
TA 
 
 P450 DO3 GTGTAGAGGACCCAGTTGAT  
 P450 DO4 CCAGCGCTGAAGGGTACATAG
GCGT 
 
 P450 DO5 GCGTCGTCGTCCAGGTTATT  
 P450 DO6 ATAAATAGCACCTTCCTCAGTA
G 
 
Myoglobin Myo2 GCCACCAAGCACAAGATCCC  
 Myo3 CGGAAGAGCTCCAGGGCCTT  
Gapd GapdH950 CATCAAGTCCACAACACGGTTG
CTGTA 
 
Gapd GapdL890 ACCTTTAATGCGGGTGCTGGCA
TTGC 
 
64