Content uploaded by Lachie Scarsbrook
Author content
All content in this area was uploaded by Lachie Scarsbrook on Oct 13, 2022
Content may be subject to copyright.
Zoological Journal of the Linnean Society, 2022, XX, 1–13. With 3 figures.
1
Ancient DNA from the extinct New Zealand grayling
(Prototroctes oxyrhynchus) reveals evidence for Miocene
marine dispersal
LACHIE SCARSBROOK1,2,*, KIEREN J. MITCHELL1, MATTHEW D. MCGEE3,
GERARD P. CLOSS4 and NICOLAS J. RAWLENCE1,*
1Otago Palaeogenetics Laboratory, Department of Zoology, University of Otago, Dunedin, New Zealand
2Palaeogenomics and Bio-Archaeology Research Network, School of Archaeology, University of Oxford,
Oxford, UK
3Behavioural Studies Group, School of Biological Sciences, Monash University, Melbourne, Victoria,
Australia
4Department of Zoology, University of Otago, Dunedin, New Zealand
Received 27 May 2022; revised 13 July 2022; accepted for publication 15 August 2022
The evolutionary history of Southern Hemisphere graylings (Retropinnidae) in New Zealand (NZ), including their
relationship to the Australian grayling, is poorly understood. The NZ grayling (Prototroctes oxyrhynchus) is the
only known fish in NZ to have gone extinct since human arrival there. Despite its historical abundance, only 23
wet and dried, formalin-fixed specimens exist in museums. We used high-throughput DNA sequencing to generate
mitogenomes from formalin-fixed P. oxyrhynchus specimens, and analysed these in a temporal phylogenetic
framework of retropinnids and osmerids. We recovered a strong sister-relationship between NZ and Australian
grayling (P. mareana), with a common ancestor ~13.8 Mya [95% highest posterior density (HPD): 6.1–23.2 Mya],
after the height of Oligocene marine inundation in NZ. Our temporal phylogenetic analysis suggests a single marine
dispersal between NZ and Australia, although the direction of dispersal is equivocal, followed by divergence into
genetically and morphologically distinguishable species through isolation by distance. This study provides further
insights into the possible extinction drivers of the NZ grayling, informs discussion regarding reintroduction of
Prototroctes to NZ and highlights how advances in palaeogenetics can be used to test evolutionary hypotheses in fish,
which, until relatively recently, have been comparatively neglected in ancient-DNA research.
ADDITIONAL KEYWORDS: extinction – formalin-fixed – Oligocene marine inundation – Prototroctidae
– Retropinnidae.
INTRODUCTION
The arrival of Polynesians in Aotearoa (later known
as New Zealand) around ad 1280 (Wilmshurst et al.,
2008), followed by Europeans (effectively in the late
1700s), resulted in widespread habitat modification
and many species extinctions (Tennyson & Martinson,
2007; McWethy et al., 2014; Rawlence et al., 2020).
Since human arrival, at least 70 species of birds have
gone extinct, in addition to one mammal (Rawlence
et al., 2016; Rawlence et al., 2020), one lizard (Worthy,
1987) and three frogs (Easton et al., 2021). The New
Zealand grayling Prototroctes oxyrhynchus Günther,
1870 (or upokororo) is the only freshwater fish
species known to have become extinct in New Zealand
(Fig. 1), with the last confirmed sighting in 1923 by
anthropologist Te Rangi Hīroa (Sir Peter Henry Buck),
who captured ‘over forty grayling in a funnel-shaped’
net in the Waipu River in northern New Zealand
(Hīroa, 1926). Later unverified reports (e.g. newspaper
articles) suggest survival into the 1950s, and species
distribution models imply persistence as late as the
1970s (McDowall, 1990; Lee & Perry, 2019). However,
despite their former abundance, the New Zealand
grayling is now only known from 23 formalin-fixed
specimens (both stuffed skins and wet-preserved) in
*Corresponding authors. E-mail: lachiescarsbrook@gmail.com;
nic.rawlence@otago.ac.nz
applyparastyle “fig//caption/p[1]” parastyle “FigCapt”
This is an Open Access article distributed under the terms of the Creative Commons Attribution License (https://
creativecommons.org/licenses/by/4.0/), which permits unrestricted reuse, distribution, and reproduction in any
medium, provided the original work is properly cited.
© 2022 The Linnean Society of London.
Editor's choice
Downloaded from https://academic.oup.com/zoolinnean/advance-article/doi/10.1093/zoolinnean/zlac077/6726151 by University College London user on 29 September 2022
2 L. SCARSBROOK ET AL.
© 2022 The Linnean Society of London, Zoological Journal of the Linnean Society, 2022, XX, 1–13
museums in New Zealand and the United Kingdom
(see: McDowall, 1976) – in-part due to the intensity of
20th century fishing practices (e.g. reports of cartloads
of graling being traded; Preston, 1986).
The life-history traits of the New Zealand grayling
are uncertain – their spawning habits are entirely
unknown, and they were reportedly cryptic, residing
in deep pools and feeding nocturnally (McDowall,
1976). Individuals commonly reached lengths of 270–
280 mm, and weighed up to 1.36 kg (McDowall, 1990).
Historically, this species inhabited both North and
South Island rivers and streams proximal to the ocean
(due to its amphidromous behaviour; McDowall, 1990),
and may have been one of ‘the most common freshwater
fish in many parts of New Zealand’ (Phillipps, 1923).
Ethnohistorical evidence suggests that the New
Zealand grayling was an important seasonal food
source for Māori (Leach, 2006; McDowall, 2011; Fyfe &
Bradshaw, 2020). However, confirmed skeletal remains
of the New Zealand grayling have not been found in
pre-European archaeological middens, although this
probably reflects a mixture of taphonomic (including
acidic soil chemistry) and excavation biases, and a lack
of comparative material [Leach, 2006; McDowall, 2011;
also see Seersholm et al. (2018) regarding galaxiids].
Overall, our understanding of the biology, ecology,
evolution and extinction of the New Zealand grayling
is incomplete and fragmentary.
The New Zealand grayling is generally considered
to be the sister-species to the endangered Australian
grayling Prototroctes maraena Günther, 1864, which
is presently restricted to south-eastern Australia
and Tasmania. Although the genus Prototroctes has
sometimes been attributed to a distinct family – the
Prototroctidae (McDowall, 1969; McDowall, 1976;
Nelson, 1994; Schwarzhans et al., 2021) – it is currently
considered to belong to a subfamily (Prototroctinae)
of Retropinnidae (Waters et al., 2002). Retropinnidae
(southern smelts and graylings) constitute a
monophyletic group sister to Osmeridae (northern
smelts; Li et al., 2010), sharing a common ancestor
~80 million years ago (Mya; Burridge et al., 2012).
Retropinnids are endemic to southern Australia and
New Zealand and constitute three genera: Retropinna
Gill, 1862 (southern smelts), Prototroctes Günther,
1864 (southern graylings) and the monotypic New
Zealand endemic Stokellia Stokell, 1941 (Johnson &
Patterson, 1996). Classification of the New Zealand and
Australian graylings as congeners is based exclusively
on morphological characteristics, with their division
into separate species based on higher counts of lateral
scale rows, vertebrae and gill rakers in P. oxyrhynchus
(McDowall, 1976). This hypothesis has not been tested
using molecular data.
The evolutionary history of graylings in New
Zealand – including the number of colonisation
events, the directionality and timing of dispersal,
and their relationship to Australian graylings – is
poorly understood. Otoliths (calcareous structures of
the inner ear) representing two species – P. modestus
Schwarzhans, 2011 and P. vertex Schwarzhans, 2011
– have been recovered from Early Miocene (18.7–
15.9 Mya) lacustrine deposits of the palaeolake
Manuherikia, located at St. Bathans (Schwarzhans
et al., 2011). Identification to Prototroctes was based
on morphological similarities to otoliths of P. maraena,
with their relationship to P. oxyrhynchus unknown
given the absence of known comparative material
(due to the dissolution of calcareous structures in
formalin and ethanol; see Supporting Information,
Fig. S1). In addition, two complete fossil fish have also
been recovered from a Late Pleistocene lacustrine
deposit on the north-eastern North Island (0.71–0.62
Mya), assigned to P. oxyrhynchus on the basis of
caudal fin morphology (McDowall et al., 2006; but see:
Schwarzhans et al., 2011). However, as the temporal
Figure 1. The extinct New Zealand grayling (Prototroctes oxyrhynchus). Artwork by Frank Edward Clarke. Annotations in
figure contain a nomen nudum. Museum of New Zealand Te Papa Tongarewa CC BY-NC-ND 4.0.
Downloaded from https://academic.oup.com/zoolinnean/advance-article/doi/10.1093/zoolinnean/zlac077/6726151 by University College London user on 29 September 2022
NEW ZEALAND GRAYLING PHYLOGENY 3
© 2022 The Linnean Society of London, Zoological Journal of the Linnean Society, 2022, XX, 1–13
origin of P. oxyrhynchus is unknown, it is not possible
to test whether these fossil taxa are likely to represent
either stem or crown members of Prototroctes – the
fossils from New Zealand could represent an ancient
(perhaps Gondwanan) and endemic lineage that
includes P. oxyrhynchus or, alternatively, an extinct
stem lineage, with P. oxyrhynchus descending from a
recent and independent dispersal (facilitated by the
juvenile marine phase of graylings).
Molecular data (i.e. ancient DNA) could help answer
several questions about the taxonomic affinities and
evolutionary origins of the New Zealand grayling.
Unfortunately, the success of ancient DNA studies
on historical formalin-fixed ‘wet’ (i.e. preserved in
fluid) and ‘dried’ (e.g. mounted) museum specimens
– which encompass most teleost specimens, New
Zealand grayling included – has historically been
highly variable (Friedman & Desalle, 2008; Garrigos
et al., 2013; Pierson et al., 2020; Pyron et al., 2022),
especially due to frequent uncertainties surrounding
early curatorial practices (e.g. changes to preservative
fluids, most commonly unbuffered formalin to
ethanol). While palaeogenetic studies focusing on fish
are becoming more common (for review see: Oosting
et al., 2019), the majority of studies to date have
focused on sequencing ancient DNA from bone (e.g.
Ferrari et al., 2021; Martinez-Garcia et al., 2021, 2022)
and not formalin-fixed tissue. Obtaining ancient DNA
from formalin-fixed specimens has been challenging,
as fixation in formalin rapidly degrades DNA (through
fragmentation and base pair modification) and
promotes cross-linkage both within and between DNA
molecules, as well as to cellular proteins; both of which
accumulate linearly with increased preservation
time (Hykin et al., 2015; Hahn et al., 2022). However,
methodological and technological advancements
– especially the advent of high-throughput DNA
sequencing – have dramatically improved success
rates of obtaining genetic data from such fluid-
preserved specimens. For example, Scarsbrook et al.
(2022) combined DNA extraction methods optimised
for ultra-short fragments (Dabney et al., 2013), single-
stranded library preparation (which breaks molecular
cross-links through heat-denaturation; Gansauge
et al., 2017) and hybridisation capture-enrichment
(González-Fortes & Paijmans, 2019) to successfully
obtain complete mitochondrial genomes from historical
‘wet’ preserved New Zealand geckos.
In this study, we used palaeogenetic techniques
to obtain mitochondrial genome sequences from
three New Zealand grayling specimens, which were
compared to a newly generated mitochondrial genome
from the Australian grayling, in addition to published
data from other retropinnid and osmerid species. These
data were used to test whether P. oxyrhynchus is most
closely related to P. maraena, to estimate the age of
the common ancestor of P. oxyrhynchus and its nearest
living relative, and to investigate genetic structuring
of P. oxyrhynchus populations within New Zealand.
We also discuss the implications of our findings for
(1) understanding the cause(s) of the extinction of the
New Zealand grayling, and (2) informing discussion
regarding the possible re-introduction of Prototroctes
to New Zealand.
MATERIAL AND METHODS
australian grayling mitochondrial genome
sequencing
Two live P. maraena individuals were provided via
Wayne Koster (Department of Environment, Land,
Water & Planning; Victoria, Australia), transported
to Monash University and euthanised with clove oil.
All procedures involving live fish were performed
under project AE17725 to co-author MDM at Monash
University. Genomic DNA was immediately extracted
from muscle tissue using a Qiagen Blood and Tissue
kit, following the manufacturer’s instructions for
animal tissue. DNA was sequenced by a commercial
provider (Deakin Biosciences) using a polymerase
chain reaction (PCR)-free protocol for 2 × 150 bp
(paired-end) sequencing on an Illumina NovaSeq S4
flow cell. The first 500 000 read pairs were extracted
using bbduk.sh in the BBTools package (http:/
sourceforge.net/projects/bbmap) and the mitochondrial
genome assembled using tadpole.sh at a kmer size of
25. Mitochondrial genes and RNAs (rRNAs/tRNAs)
were annotated using mitoFinder v.1.4 (Allio et al.,
2020) under default parameters, with the Retropinna
semoni (Weber, 1895) (NCBI KX421785) mitochondrial
genome used as a reference.
australian grayling cytochrome b sequencing
Genomic DNA was extracted in the modern
evolutionary genetics laboratory in the Department
of Zoology (University of Otago), from P. maraena
tissue samples (fin-clips) obtained from contemporary
populations in Victoria (Bunyip River; N = 5) and
Tasmania (N = 1). Specifically, tissue samples
(~2 mm2) were added to 400 μL resuspended ‘Chelex
Solution’ (2.5 g Chelex 100; 50 mL dH2O) and 2 μL
Proteinase-K (20 mg/mL) and incubated at 60 °C for
24 h at 600 rpm (on a thermoshaker). Tubes were then
heated at 90 °C for 8 min (on a heating block), followed
by centrifugation for 10 min at 13 000 rpm. A 1273 bp
fragment of the mitochondrial protein-coding gene
cytochrome b (Cytb) and associated tRNA-Thr was
amplified using the following primer combination –
forward: HYPSLA (5′ GTG GCT TGA AAA ACC ACC
GTT 3′; Thacker et al., 2007); reverse: Ret.Thr31 (5′
Downloaded from https://academic.oup.com/zoolinnean/advance-article/doi/10.1093/zoolinnean/zlac077/6726151 by University College London user on 29 September 2022
4 L. SCARSBROOK ET AL.
© 2022 The Linnean Society of London, Zoological Journal of the Linnean Society, 2022, XX, 1–13
CTC CAA CCT CCG ACT TAC AAG 3′; Page & Hughes,
2010). PCRs (10 μL) contained 1 × AmpliTaq Gold
Buffer, 0.75 mmol/L MgCl2, 0.25 mmol/L dNTPs, 0.5
U AmpliTaq Gold DNA polymerase (ThermoFisher),
1 μmol/L each primer (HYPSLA/Ret.Thr31) and 1 μL
template DNA. Amplification was performed on an
Eppendorf ProS Mastercycler (‘thermocycler’) under
the following conditions: an initial denaturation of
94 °C for 3 min; 40 cycles of 94 °C for 30 s, 52 °C for
1 min and 72 °C for 2 min, with a final extension of
72 °C for 10 min. Resulting amplicons were visualised
using gel electrophoresis (0.5 g agarose; 25 mL 1 × TAE
Buffer; 0.5 μL SYBRsafe) and purified using ExoSap
[1 μL Shrimp Alkaline Phosphatase and 1.5 μL diluted
(1:20) Exonuclease I; ThermoFisher], incubated at
37 °C for 30 min followed by 95 °C for 5 min on a
thermocycler. Purified amplicons were bidirectionally
Sanger sequenced on an ABI 3730xl DNA Analyzer
(University of Otago Genetic Analysis Service) using
BigDye terminator technology. Chromatogram trace
files were edited (i.e. primer and poor-quality base
trimming, ambiguous base-calling) in 4Peaks v.1.8
(https://nucleobytes.com/4peaks/index.html).
ancient dna analysis oF the extinct new
Zealand grayling
Palaeogenetic work was conducted in the dedicated
Otago Palaeogenetics Laboratory in the Department
of Zoology (University of Otago). Ancient DNA was
extracted from either ‘dried’ scales or ‘wet’ tissue from
formalin-fixed historical P. oxyrhynchus specimens
from New Zealand museum collections (N = 12; see
Supporting Information, Table S1) using the QIAamp
DNA FFPE Tissue Kit, following the manufacturer’s
instructions (excluding the xylene paraffin removal
step). This method utilises a heat denaturation step
(at 90 °C) following tissue lysis to remove molecular
cross-links, characteristic of DNA from formalin-fixed
tissues (Hykin et al., 2015). Single-stranded Illumina
sequencing libraries were generated from the ancient
DNA extracts (following an adapted protocol; Scarsbrook
et al., 2022) to enhance recovery of highly fragmented
and altered DNA cross-linked to proteins (Gansauge &
Meyer, 2013). Optimal cycle number (No) for the indexing
of single-stranded libraries was determined through
quantitative PCR (qPCR), to reduce clonality and limit
heteroduplex formation (Gansauge et al., 2020). qPCRs
(10 μL) were performed using 1 × Maxima SYBR Green
qPCR Master Mix (Thermo Scientific), 0.2 mol/L each of
IS7 and IS8 primers (IS7: 5′ ACA CTC TTT CCC TAC
ACG AC 3′; IS8: 5′ GTG ACT GGA GTT CAG ACG TGT
3′) and 1 μL of diluted (1:10) library on a QuantStudio 5
Real-Time PCR System as follows: 95 °C for 10 min; 40
cycles of 95 °C for 30 s, 60 °C for 30 s and 72 °C for 30 s.
Full-length P5/P7 adapters containing unique 7-mer
barcode combinations were added to libraries through
quadruplicate indexing PCRs (to maximise complexity).
Indexing PCRs (25 μL) were performed using 1 × High
Fidelity PCR Buffer (Invitrogen), 2 mmol/L MgSO4,
0.25 mmol/L dNTPs, 1.25 U Platinum Taq DNA
polymerase High Fidelity (Invitrogen), 0.5 mol/L P5/P7
indexing primers and 5 μL library on a thermocycler
as follows: 94 °C for 12 min; No cycles of 94 °C for 30 s,
60 °C for 30 s and 72 °C for 45 s; 72 °C for 10 min.
Replicate indexed libraries were pooled, purified
using AMPure XP (Agencourt) at 1.1 × bead:template
ratio (following the manufacturer’s instructions) and
quantified using a Qubit dsDNA High Sensitivity Assay
kit (ThermoFisher).
Endogenous mitochondrial DNA was selectively
enriched through in-solution hybridisation-capture
(following an adapted protocol; Scarsbrook et al.,
2022), with biotinylated baits generated from sheared
mitochondrial amplicons of the New Zealand smelt
Retropinna retropinna Richardson, 1848. Specifically,
high-molecular weight (HMW) DNA was extracted from
the high-quality tissue (liver) sample of a R. retropinna
using the MagMAXTM DNA Multi-Sample Ultra 2.0
Kit (ThermoFisher), following the manufacturer’s
instructions for saliva or whole blood. The complete
R. retropinna mitochondrial genome (~16.5 kb) was
amplified (in a single fragment) through long-range
‘shuttle’ PCR using the following primer combination:
forward: S-LA-16S-L (5′ CGA TTA AAG TCC TAC
GTG ATC TGA GTT CAG 3′; Miya & Nishida, 2000);
reverse: S-LA-16S-H (5′ TGC ACC ATT RGG ATG
TCC TGA TCC AAC ATC 3′; Miya & Nishida, 2000).
Long-range ‘shuttle’ PCR (50 μL) was performed
using 1 × PrimeSTAR GXL Buffer, 200 mol/L dNTPs,
1.25 U PrimeSTAR GXL DNA Polymerase (Takara),
0.2 mol/L each primer (S-LA-16S-L/S-LA-16S-H)
and ~100 ng HMW DNA extract, on a thermocycler
as follows: 30 cycles of 98 °C for 10 s and 68 °C for
16 min. The resulting amplicon was visualised using
gel electrophoresis (0.5 g agarose; 25 mL 1 × TAE
Buffer; 0.5 μL SYBRsafe) and sheared to 150–300 bp
using a Picoruptor (Diagenode) sonicator with 13
cycles of: 30 s on, 30 s off, at 4 °C. Sheared amplicons
were purified using AMPure XP magnetic beads
(Agencourt) at 1.8 × bead:template ratio (following the
manufacturer’s instructions), quantified using a Qubit
dsDNA Broad Range Assay kit (ThermoFisher) and
used to generate biotinylated baits (see above).
Post-capture re-amplification PCRs (25 μL)
were performed using 1 × AmpliTaq Gold Buffer
(ThermoFisher), 2.5 mmol/L MgCl2, 0.25 mmol/L dNTPs,
1.25 U AmpliTaq Gold DNA polymerase (ThermoFisher),
0.5 mol/L each primer (IS5: 5′ AAT GAT ACG GCG ACC
ACC GA 3′, IS6: 5′ CAA GCA GAA GAC GGC ATA CGA
3′) and 5 μL post-capture library on a thermocycler as
follows: 94 °C for 12 min; Nno cycles (as above) of 94 °C
Downloaded from https://academic.oup.com/zoolinnean/advance-article/doi/10.1093/zoolinnean/zlac077/6726151 by University College London user on 29 September 2022
NEW ZEALAND GRAYLING PHYLOGENY 5
© 2022 The Linnean Society of London, Zoological Journal of the Linnean Society, 2022, XX, 1–13
for 30 s, 60 °C for 30 s and 72 °C for 45 s; 72 °C for 10 min.
Reamplified libraries were purified and quantified (as
above) for equimolar pooling. Mean fragment size of the
captured libraries was measured on a QIAxcel Advanced
System using a 25 bp–10 kb alignment marker; with
raw data calibration and visualisation performed using
QIAxcel scre enge l v.1.6.0 (Qiagen). Each library
was diluted to 10 nmol/L and run on an Illumina
NextSeq (Garvan Institute of Medical Research) using
2 × 75bp (paired-end) sequencing chemistry and custom
sequencing primers (CL72: 5′ ACA CTC TTT CCC TAC
ACG ACG CTC TTC C 3′; G’stein: 5′ GGA AGA GCG
TCG TGT AGG GAA AGA GTG T 3′).
Reads were demultiplexed using saBre v.1.0 (https://
github.com/najoshi/sabre) with no mismatches allowed
(-m: 0). Adapter sequences were removed and paired-end
reads collapsed using adaPterremoval v.2.3.1 (Schubert
et al., 2016), with a mismatch rate of 0.33 (-mm: 3). Low-
quality bases (Phred Quality Score < 30) were trimmed
(--minquality: 30; --trimns) and collapsed reads shorter than
25 bp discarded (-minlength: 25). To ensure the effective
removal of adapters, read quality was visualised using
Fastqc v.0.11.9 (https://www.bioinformatics.babraham.
ac.uk/projects/fastqc/). Mitochondrial sequence alignments
were generated from collapsed reads using the BAM
pipeline implemented in PALEOMIX v. 1.2.14 (Schubert
et al., 2014). Briefly, collapsed reads were mapped against
an Australian grayling (Prototroctes maraena) reference
mitochondrial genome (this study) using Bwa v.0.7.17
(-n: 0.01; -o: 2; Li & Durbin, 2009). samtools v.0.1.19
(Li et al., 2009) was used to select reads with a mapping
Phred Quality Score > 25 (-q: 25), with duplicate reads
discarded in Picard v.2.1.0 (http://broadinstitute.github.io/
picard/) using the MarkDuplicates.jar tool. Misalignments
from reads overlapping indels were improved using
the IndelRealigner tool implemented in gatk v.4.1.4.1
(McKenna et al., 2010). Finally, to ensure ancient DNA
authenticity, characteristic damage patterns (i.e. nucleotide
mis-incorporation and DNA fragmentation;Supporting
Information, Fig. S2) were assessed using maPdamage
v.2.0.8 (Ginolhac et al., 2011; Jónsson et al., 2013). Majority
consensus sequences (75%) were generated for each BAM
alignment in geneious Prime v.2021.2.2, with bases
only called at sites covered by ≥ 3 reads (with IUPAC
ambiguities otherwise called). Mitochondrial genomes
were annotated using the MITOS Webserver (Bernt et al.,
2013), and manually verified, to determine the location
and size of protein-coding genes and tRNAs/rRNAs.
Phylogenetic analysis
Prototroctes oxyrhynchus and P. maraena mitochondrial
genomes were aligned against published retropinnid
and osmerid sequences (Supporting Information,
Table S2) using MUSCLE v.3.8.425 (using ‘default’
parameters; Edgar, 2004) implemented in geneious
Prime v.2021.2.2 (Biomatters; https://www.geneious.
com/). Cytochrome b (1142 bp) was extracted (in
geneious Prime ) from the mitochondrial genome
alignment and aligned against both modern P. maraena
samples (N = 7; GenBank accession numbers:
ON161129–ON161135) and published retropinnid
sequences (N = 128; Supporting Information, Table
S2). 16S rRNA (1642 bp) was similarly extracted
and aligned against published sequences (N = 15;
Supporting Information, Table S2). Median-joining
haplotype networks (Bandelt et al., 1999) were
constructed from both 16S and Cytb alignments in
PoPart (Leigh & Bryant, 2015).
Topology estimation and molecular dating were
performed under a Bayesian framework in BEAST
v.1.8.4 (Drummond et al., 2012) using the mitochondrial
genome alignment (subsampled to include only the
most complete P. oxyrhynchus sequence: ON220596).
Data were analysed as four partitions, using best-fitting
substitution models identified using the Bayesian
information criterion in PartitionFinder v.1.0.1
(Lanfear et al., 2012): first-codon positions of H-strand
encoded protein-coding genes (GTR+I+G), second-
codon positions of H-strand encoded protein-coding
genes (GTR+I), first-codon positions of ND6 (TrN+G),
and second-codon positions of ND6 (K81+I). We applied
a birth–death tree prior and uncorrelated log-normal
relaxed clock model (with rate multiplier parameters
for each partition). The continuous-time Markov chain
scale reference prior (Ferreira & Suchard, 2008) was
applied to the mean rate parameter. We constrained
the age of the time to most recent common ancestor
(TMRCA) of the Osmeridae according to a log-normal
distribution with a hard minimum at 57 Mya, mean
of 13.85 and standard deviation of 1.0 (such that 95%
of the prior distribution fell between 57 and 100.5
Mya). The minimum age was based on the oldest
known fossil osmerid Speirsaenigma lindoei Wilson &
Williams, 1991 from the Palaeocene of Canada (Wilson
& Williams, 1991). The maximum age was based on
the absence of crown osmerids from Late Cretaceous
fossil localities (Burridge et al., 2012). The mean and
standard deviation were chosen to reflect Burridge
et al.’s (2012) posterior estimate for the equivalent
node (based on a larger dataset with additional fossil
age constraints). Other priors were set to their default
values. We ran three Markov chain Monte Carlos
(MCMCs) for 50 million generations each, sampling
trees and parameters every 5000 generations. The
first 10% of each chain was removed as burn-in and
remaining data were combined using logcomBiner
v.1.8.4. TRACER v.1.7.1 (Rambaut et al., 2018) was used
to monitor parameter values, ensuring convergence
and effective sample sizes > 200. A maximum clade
credibility tree was generated in treeanno tat or
v.1.8.4 and visualised in Figtree v.1.4.4.
Downloaded from https://academic.oup.com/zoolinnean/advance-article/doi/10.1093/zoolinnean/zlac077/6726151 by University College London user on 29 September 2022
6 L. SCARSBROOK ET AL.
© 2022 The Linnean Society of London, Zoological Journal of the Linnean Society, 2022, XX, 1–13
RESULTS
ancient mitochondrial genomes
Mitochondrial genomes (16 591 bp) were successfully
recovered from three of the 12 formalin-fixed historical
New Zealand grayling (Prototroctes oxyrhynchus)
specimens – in each instance, samples consisted of
scales derived from dried specimens, which are an
excellent source of endogenous DNA (Pinsky et al.,
2021). Alignments produced one complete (ON220596)
and two partial (ON220594–5) sequences, containing
between 1310 and 3659 unique mapping reads covering
79.8–96.2% of the reference sequence to at least
three-fold depth of coverage (3.6–10.8×; Supporting
Information, Table S3). Characteristic short fragment
lengths (47.7–51.4 bp; Supporting Information,
Table S3) and damage patterns (i.e. increased C-to-T
substitution frequencies at read ends) authenticated
recovered DNA sequences as ancient (Supporting
Information, Fig. S2). The remaining, predominantly
wet-preserved (i.e. formalin-fixed) specimens yielded
few mapped reads (0–115; Supporting Information,
Table S3).
Phylogenetic analysis
Our time-calibrated Bayesian phylogeny (Fig. 2) was
well resolved, with the majority of branches receiving
unequivocal support (i.e. posterior probabilities = 1.0).
Divergence of the Retropinnidae and Osmeridae
occurred 129.3 Mya (95% HPD: 83.5–183.6 Mya), with
crown-ages of each family estimated at 70.4 Myr (95%
HPD: 40.3–105.7 Myr) and 67.0 Myr (95% HPD: 57.2–
85.9 Myr), respectively. Within Retropinnidae, the
genus Prototroctes was the sister-taxon to Retropinna,
with the extant P. maraena and extinct P. oxyrhynchus
diverging 13.8 Mya (95% HPD: 6.1–23.2 Mya).
Conversely, earlier lineage separation was inferred
within the genus Retropinna, with divergence between
R. retropinna and R. semoni estimated at 57.0 Mya
(95% HPD: 30.6–87.4 Mya).
Cytochrome b and 16S rRNA haplotype networks
(Fig. 3) revealed no evidence for population structure
(or cryptic taxonomic diversity) in P. oxyrhynchus, with
all individuals sharing the same haplotype. Conversely,
two distinct haplotypes were observed in P. maraena,
however these were not geographically structured,
with both haplotypes present in the same population
(for Cytb; Fig. 3A).
DISCUSSION
grayling Phylogeny and BiogeograPhy
Our phylogenetic results and node age estimates among
extant taxa are concordant with the results of previous
studies (e.g. Burridge et al., 2012; Straube et al., 2018).
However, our new data allow us to include the extinct
New Zealand grayling in a molecular phylogeny for the
first time – our results unequivocally support a sister-
lineage relationship between the New Zealand and
Australian graylings, and suggest that the common
ancestor of the two species occurred 6.1–23.2 Mya (Fig.
1). This divergence timing is comparable to the age of
splits between osmerid sister-pairs in our analysis
(see Fig. 1), and pre-dates the most-recent common
ancestor of the two R. semoni individuals included in
our alignment, which probably represent two different
species (with one or both belonging to an undiagnosed
cryptic species; Hammer
et al., 2007; Hughes et al., 2014;
Schmidt et al., 2016). Consequently, it is clear that the
New Zealand and Australian graylings represent two
distinct species, consistent with previously observed
morphological differences (McDowall, 1976). Further,
extensive genetic distance between Prototroctes
species and the remaining Retropinnidae (see Fig. 2)
is consistent with previous assertions of subfamily-
level (e.g. Prototroctinae) – but not family-level (e.g.
Prototroctidae) – distinction of Southern Hemisphere
graylings.
Extant retropinnids are restricted to south-eastern
Australia and New Zealand, but a possible fossil
retropinnid – with closest affinities to Prototroctes
– has also been described from the Miocene of Chile
(Navidadichthys miru Schwarshans & Nielsen, 2021;
Schwarzhans & Nielsen, 2021). This exclusively
Southern Hemisphere distribution suggests that
Gondwanan vicariance may have played a role in
determining the distribution of retropinnid lineages.
Our estimate for the age of the common ancestor of
R. semoni and R. retropinna (30.6–87.4 Mya) is broadly
consistent with this hypothesis – the distribution
of these two lineages in Australia and New Zealand
may have been driven by rift-related separation
of these respective continental blocks through the
opening of the Tasman Sea (52–85 Mya; Cooper &
Millener, 1993). However, our results for P. maraena
and P. oxyrhynchus suggest that marine dispersal,
not Gondwanan vicariance, was the driver of their
respective distributions in Australia and New Zealand.
While our results implicate marine dispersal
in Prototroctes – made highly plausible by the
amphidromous life cycle of graylings – they are
equivocal on the pre-Miocene distribution of the
grayling lineage. Dispersing individuals could have
originated from New Zealand, Australia or even South
America (since fossil data indicate that this lineage was
present in Chile during the Miocene). While P. modestus
and P. vertex are known from New Zealand’s St. Bathans
assemblage (Schwarzhans et al., 2011), the age of these
fossil deposits (18.7–15.9 Mya) overlaps with our
estimate for the divergence between P. maraena and
Downloaded from https://academic.oup.com/zoolinnean/advance-article/doi/10.1093/zoolinnean/zlac077/6726151 by University College London user on 29 September 2022
NEW ZEALAND GRAYLING PHYLOGENY 7
© 2022 The Linnean Society of London, Zoological Journal of the Linnean Society, 2022, XX, 1–13
P. oxyrhynchus. Consequently, P. modestus and P. vertex
could represent early members of the lineage leading
to P. oxyrhynchus, and do not necessarily indicate pre-
Miocene presence of graylings in New Zealand. If so, our
age estimates suggest that colonisation of New Zealand
by graylings might have occurred only once, after the
height of the Oligocene marine transgression (27–22
Mya) during which > 80% of continental New Zealand
(i.e. Zealandia) was submerged (Mildenhall et al., 2014).
The extent and distribution of freshwater habitats in
New Zealand would have been severely reduced (if not
entirely absent) during this time, possibly resulting in
Figure 2. Time-calibrated Bayesian phylogeny of the Osmeriformes (Retropinnidae and Osmeridae) constructed from first
and second codon positions of all 13 mitochondrial protein coding genes. 95% highest probability density (95% HPD) of the
age estimate for each node is indicated by horizontal grey bars, with Bayesian posterior probabilities of 1.0 represented
by white circles (otherwise values are reported). The x-axis represents time in millions of years before present (Mya). Red
circles denote phylogenetic position and age of described retropinnid fossil material: Prototroctes modestus and P. vertex
(18.7–15.9 Myr; Schwarzhans et al., 2011), Navidadichthys mirus (18–17 Myr; Schwarzhans et al., 2021), Prototroctes
oxyrhynchus (0.71–0.62 Myr; McDowall et al., 2006) – we also indicate Speirsaenigma lindoei (57 Myr; Wilson & Williams,
1991), which was used to constrain the minimum age of crown Osmeridae. Timing of relevant geological events, such as
rifting of the Zealandian and Australian continental blocks (Cooper & Millener, 1993) and maximum marine inundation (i.e.
‘Oligocene drowning’; Mildenhall et al., 2014) are highlighted. Continental distribution of retropinnid lineages are indicated
by outline maps. Extinct taxa are denoted †.
Downloaded from https://academic.oup.com/zoolinnean/advance-article/doi/10.1093/zoolinnean/zlac077/6726151 by University College London user on 29 September 2022
8 L. SCARSBROOK ET AL.
© 2022 The Linnean Society of London, Zoological Journal of the Linnean Society, 2022, XX, 1–13
the extinction and turnover of many freshwater animal
lineages. Concomitant expansion of freshwater habitats
in the Early Miocene (e.g. Lee et al., 2007; Schwarzhans
et al., 2011) might have provided opportunities for
colonisation and/or diversification that were previously
limited by density-dependent founder-takes-all
processes (Waters et al., 2013). Indeed, the respective
common ancestors of two clades of galaxiid fishes found
in New Zealand also date to the Late Oligocene or
Early Miocene, where similar processes are implicated
(Burridge et al., 2012; Burridge & Waters, 2020).
grayling extinction
The drivers of the extinction of the New Zealand
grayling are debated. Reductions in population size
and geographic distribution were identified shortly
after European arrival in New Zealand, with concerns
raised as early as the 1870s (Rutland, 1877; Allen,
1949). By the early 1920s, the New Zealand grayling
was rare and restricted to isolated rivers and streams
away from human settlement (Phillipps, 1923; Allen,
1949), but it did not receive government protection
until 1952, and was officially declared extinct in
2002 (Hitchmough, 2002). This decline and eventual
extinction has been variously attributed to: (1)
overfishing; (2) the introduction of trout, resulting
in direct predation or transfer of disease (e.g. fungus
and parasites), which has caused epidemics in
Australian grayling populations (e.g. Johnston, 1882;
Saville-Kent, 1887; Kaminskas, 2021); and/or (3)
environmental modification and habitat degradation
(Allen, 1949). Some authors have suggested that the
disappearance of grayling from pristine unmodified
rivers indicates that habitat degradation was not
a primary cause of extinction (McDowall, 1990), but
other authors note that source-sink dynamics could
cause indirect depletion of populations, even in these
pristine habitats, as individuals migrate to the lower
density degraded habitats (Lee & Perry, 2019).
Lee and Perry (2019) suggested that up to 30% of
grayling per generation could have been sustainably
harvested in the absence of source-sink dynamics,
reducing to only 5% when considered a single
panmictic meta-population in the presence of source-
sink dynamics. The presence or absence of source-sink
dynamics is theoretically testable using genetic data
– marked phylogeographic structure, as observed in
Australian smelt (Hughes et al., 2014) and suggested by
colour variation in New Zealand grayling across their
range (McDowall, 1990), would imply that source-sink
dynamics were not operating over large distances. We did
not observe deep divergences among our three samples
(Fig. 2), but these individuals were only drawn from two
South Island localities (the Clutha and Hokitika rivers
on opposite sides of the Southern Alps) – as attempts to
extract ancient DNA from other (mostly ‘wet’) grayling
specimens from throughout New Zealand were not
successful. However, our results suggest that additional
sampling, especially of North Island individuals (e.g.
Whanganui Regional Museum), combined with more
specialised molecular techniques – optimised for
Figure 3. Median-joining haplotype networks of the Retropinnidae constructed from A, cytochrome b (1363 comparable
sites) and B, 16s rRNA (575 comparable sites) alignments in PoPart (for GenBank Accession numbers, see Supporting
Information, Table S2). Haplotypes (circles) are proportional to frequency (numbers), with number of substitutions indicated
by hatches along branches. Black circles represent undetected intermediary haplotypes, with colours corresponding to
species.
Downloaded from https://academic.oup.com/zoolinnean/advance-article/doi/10.1093/zoolinnean/zlac077/6726151 by University College London user on 29 September 2022
NEW ZEALAND GRAYLING PHYLOGENY 9
© 2022 The Linnean Society of London, Zoological Journal of the Linnean Society, 2022, XX, 1–13
the recovery of highly degraded ancient DNA from
formalin-fixed specimens (especially for ‘wet’ specimens;
e.g. Campos & Gilbert, 2012; Dabney et al., 2013;
Straube et al., 2021; Hahn et al., 2022) or dried fish scales
– could yield additional mitochondrial (and potentially
nuclear) genomes from New Zealand grayling. These
data would allow the presence of source-sink dynamics
to be more effectively evaluated and provide insight
into the importance of different drivers in the extinction
of the New Zealand grayling.
restoring lost ecological Functions to rivers
By introducing australian grayling to new
Zealand
Historical reports suggest that New Zealand grayling
were omnivorous, feeding on small invertebrates and
algae (McDowall, 1990). As such, they probably occupied
an ecological feeding niche unlike any other current
extant native freshwater fish species (McDowall, 1990),
and hence their (likely) extinction has resulted in some
degree of lost ecological function in rivers where New
Zealand grayling were formerly abundant (McDowall,
1990). Where a species translocation to restore lost
ecological function is planned, the target species
for translocation should be both functionally and
genetically as close as possible to the species that was
originally lost (Armstrong & Seddon, 2008). Based on
our analysis, the Australian grayling may be the most
likely extant candidate species to fulfil this criterion,
pending further palaeoecological (e.g. stable dietary
isotopes; Durante et al., 2020; Welicky et al., 2021),
morphometric and functional palaeogenomic research.
That said, our work also confirms that New Zealand
and Australian grayling are distinct species; hence the
translocation of Australian grayling to replace the New
Zealand grayling is clearly a species introduction (i.e.
ecological surrogate), not a reintroduction (Armstrong
& Seddon, 2008). Whilst this point might seem to be
mere semantics, the distinction does serve to highlight
the potential challenges of such a management step.
Given the time since their last common ancestor, the
two species would have probably diverged in a variety
of key morphological and behavioural traits, thus
predicting how Australian grayling would respond to
New Zealand conditions is impossible, both in terms
of the viability of such an introduction, but also with
respect to the goal of restoring lost ecological function.
ACKNOWLEDGEMENTS
We thank Tom Trnski and Severine Hannam (Auckland
War Memorial Museum); Paul Scofield (Canterbury
Museum); Karen Cook (Fish and Game, Nelson); Andrew
Stewart, Jeremy Barker and Clive Roberts (Museum of
New Zealand Te Papa Tongarewa); Eimear Egan (National
Institute of Water and Atmospheric Research); and Kane
Fleury and Emma Burns (Otago Museum) for access
to P. oxyrhynchus material for genetic analysis; James
Maclaine (Natural History Museum, United Kingdom)
for generating X-ray images of NMUK P. oxyrhynchus
syntypes; Chris Burridge (University of Tasmania), Matt
Jarvis (University of Otago) and Wayne Koster (Arthur
Rylah Institute for Environmental Research) for access
to fin-clips and otoliths from contemporary Australian
grayling (P. maraena) populations; Jesse Wansbrough
(University of Otago) for providing the R. retropinna tissue
sample; and Alex Verry and Tania King (University of
Otago) for their assistance in optimising DNA extraction,
PCR and next-generation sequencing protocols. We
also acknowledge the use of New Zealand eScience
Infrastructure (NeSI) high-performance computing
facilities as part of this research. We acknowledge that
Māori, the indigenous people of Aotearoa New Zealand,
have kaitiakitanga (guardianship) over the organisms
from their rohe (tribal area).
FUNDING
Funding was provided by the University of Otago.
DATA AVAILABILITY
The data that support the findings of this study are
openly available, with consensus mitochondrial genomes
openly available in GenBank (Accession Numbers:
ON161129–ON161135; ON220594–ON220597).
REFERENCES
Allen KR. 1949. The New Zealand grayling – a vanishing
species. Tuatara 2: 22–26.
Allio R, Schomaker-Bastos A, Romiguier J, Nabholz B,
Delsuc F. 2020. MitoFinder: efficient automated large-
scale extraction of mitogenomic data in target enrichment
phylogenomics. Molecular Ecology Resources 20: 892–905.
Armstrong DP, Seddon PJ. 2008. Direction in reintroduction
biology. Trends in Ecology and Evolution 23: 20–25.
Bandelt HJ, Forster P, Röhl A. 1999. Median-joining
networks for inferring intraspecific phylogenies. Molecular
Biology and Evolution 16: 37–48.
Bernt M, Donath A, Jühling F, Externbrink F, Florentz C,
Fritzsch G, Pütz J, Middendorf M, Stadler PF. 2013.
MITOS: improved de novo metazoan mitochondrial genome
annotation. Molecular Phylogenetics and Evolution 69:
313–319.
Burridge CP, Waters JM. 2020. Does migration promote or
inhibit diversification? A case study involving the dominant
Downloaded from https://academic.oup.com/zoolinnean/advance-article/doi/10.1093/zoolinnean/zlac077/6726151 by University College London user on 29 September 2022
10 L. SCARSBROOK ET AL.
© 2022 The Linnean Society of London, Zoological Journal of the Linnean Society, 2022, XX, 1–13
radiation of temperate Southern Hemisphere freshwater
fishes. Evolution 74: 1954–1965.
Burridge CP, McDowall RM, Craw D, Wilson MV,
Waters JM. 2012. Marine dispersal as a pre-requisite for
Gondwanan vicariance among elements of the galaxiid fish
fauna. Journal of Biogeography 39: 306–321.
Campos PF, Gilbert TMP. 2012. DNA extraction from
formalin-fixed material. In: Shapiro B, Barlow A,
Heintzman PD, Hofreiter M, Paijmans JLA, Soares AER, eds.
Ancient DNA. New York: Humana Press, 81–85.
Cooper RA, Millener PR. 1993. The New Zealand biota:
historical background and new research. Trends in Ecology
and Evolution 8: 429–433.
Dabney J, Knapp M, Glocke I, Gansauge MT, Weihmann A,
Nickel B, Valdiosera C, García N, Pääbo S, Arsuaga JL,
Meyer M. 2013. Complete mitochondrial genome sequence
of a Middle Pleistocene cave bear reconstructed from
ultrashort DNA fragments. Proceedings of the National
Academy of Sciences of the USA 110: 15758–15763.
Drummond AJ, Suchard MA, Xie D, Rambaut A. 2012.
Bayesian phylogenetics with BEAUti and the BEAST 1.7.
Molecular Biology and Evolution 29: 1969–1973.
Durante LM, Sabadel AJM, Frew RD, Ingram T, Wing SR.
2020. Effects of fixatives on stable isotopes of fish muscle
tissue: implications for trophic studies on preserved
specimens. Ecological Applications 30: e02080.
Easton LJ, Tennyson AJ, Rawlence NJ. 2021. A new
species of Leiopelma frog (Amphibia: Anura: Leiopelmatidae)
from the late Pliocene of New Zealand. New Zealand Journal
of Zoology 49: 215–224. doi: 10.1080/03014223.2021.1979053
Edgar RC. 2004. MUSCLE: a multiple sequence alignment
method with reduced time and space complexity. BMC
Bioinformatics 5: 1–19.
Ferrari G, Cuevas A, Gondek-Wyrozemska AT,
Ballantyne R, Kersten O, Pálsdóttir AH, van der Jagt I,
Hufthammer AK, Ystgaard I, Wickler S, Bigelow GF,
Harland J, Nicholson R, Orton D, Clavel B,
Boessenkool S, Barrett JH, Star B. 2021. The preservation
of ancient DNA in archaeological fish bone. Journal of
Archaeological Science 126: 105317.
Ferreira MAR, Suchard MA. 2008. Bayesian analysis of
elapsed times in continuous-time Markov chains. Canadian
Journal of Statistics 36: 355–368.
Friedman M, DeSalle R. 2008. Mitochondrial DNA extraction
and sequencing of formalin-fixed archival snake tissue. DNA
Sequence 19: 433–437.
Fyfe R, Bradshaw J. 2020. A review of the role of diadromous
ikawai (freshwater fish) in the Māori economy and culture
of Te Wai Pounamu (South Island), Aotearoa New Zealand.
Records of the Canterbury Museum 34: 35–55.
Gansauge MT, Meyer M. 2013. Single-stranded DNA library
preparation for the sequencing of ancient or damaged DNA.
Nature Protocols 8: 737–748.
Gansauge MT, Gerber T, Glocke I, Korlević P, Lippik L,
Nagel S, Riehl LM, Schmidt A, Meyer M. 2017. Single-
stranded DNA library preparation from highly degraded
DNA using T4 DNA ligase. Nucleic Acids Research 45:
e79–e79.
Gansauge MT, Aximu-Petri A, Nagel S, Meyer M. 2020.
Manual and automated preparation of single-stranded DNA
libraries for the sequencing of DNA from ancient biological
remains and other sources of highly degraded DNA. Nature
Protocols 15: 2279–2300.
Garrigos YE, Hugueny B, Koerner K, Ibanez C, Bonillo C,
Pruvost P, Causse R, Cruaud C, Gaubert P. 2013. Non-
invasive ancient DNA protocol for fluid-preserved specimens
and phylogenetic systematics of the genus Orestias (Teleostei:
Cyprinodontidae). Zootaxa 3640: 373–394.
Ginolhac A, Rasmussen M, Gilbert MTP, Willerslev E,
Orlando L. 2011. mapDamage: testing for damage patterns
in ancient DNA sequences. Bioinformatics 27: 2153–2155.
González Fortes G, Paijmans JL. 2019. Whole-genome capture
of ancient DNA using homemade baits. In: Shapiro B, Barlow A,
Heintzman PD, Hofreiter M, Paijmans JLA, Soares AER, eds.
Ancient DNA. New York: Humana Press, 93–105.
Hahn EE, Alexander MR, Grealy A, Stiller J,
Gardiner DM, Holleley CE. 2022. Unlocking inaccessible
historical genomes preserved in formalin. Molecular Ecology
Resources 22: 2130–2147.
Hammer MP, Adams M, Unmack PJ, Walker KF. 2007. A
rethink on Retropinna: conservation implications of new taxa
and significant genetic sub-structure in Australian smelts
(Pisces: Retropinnidae). Marine and Freshwater Research
58: 327–341.
Hīroa TR. 1926. The Maori craft of netting. Transactions and
Proceedings of the New Zealand Institute 56: 597–646.
Hitchmough R, Bull L, Cromarty P. 2002. New Zealand
threat classification system lists. Threatened Species
Occasional Publication 23: 10.
Hughes JM, Schmidt DJ, Macdonald JI, Huey JA,
Crook DA. 2014. Low interbasin connectivity in a
facultatively diadromous fish: evidence from genetics and
otolith chemistry. Molecular Ecology 23: 1000–1013.
Hykin SM, Bi K, McGuire JA. 2015. Fixing formalin: a
method to recover genomic-scale DNA sequence data from
formalin-fixed museum specimens using high-throughput
sequencing. PLoS One 10: e0141579.
Johnson GD, Patterson C. 1996. Relationships of lower
euteleostean fishes. In: Stiassny MLJ, Parenti LR,
Johnson DG, eds. Interrelationships of fishes. London:
Academic Press, 251–332.
Johnston RM. 1882. General and critical observations on
the fishes of Tasmania; with a classified catalogue of all the
known species. Papers & Proceedings of the Royal Society of
Tasmania 51–170.
Jónsson H, Ginolhac A, Schubert M, Johnson PL,
Orlando L. 2013. mapDamage2.0: fast approximate
Bayesian estimates of ancient DNA damage parameters.
Bioinformatics 29: 1682–1684.
Kaminskas S. 2021. Alien pathogens and parasites impacting
native freshwater fish of southern Australia: a scientific and
historical review. Australian Zoologist 41: 696–730.
Lanfear R, Calcott B, Ho SY, Guindon S. 2012.
PartitionFinder: combined selection of partitioning schemes
and substitution models for phylogenetic analyses. Molecular
Biology and Evolution 29: 1695–1701.
Downloaded from https://academic.oup.com/zoolinnean/advance-article/doi/10.1093/zoolinnean/zlac077/6726151 by University College London user on 29 September 2022
NEW ZEALAND GRAYLING PHYLOGENY 11
© 2022 The Linnean Society of London, Zoological Journal of the Linnean Society, 2022, XX, 1–13
Leach F. 2006. Fishing in pre-european New Zealand.
Archaeofauna 15: 1–276.
Lee F, Perry GL. 2019. Assessing the role of off-take and
source–sink dynamics in the extinction of the amphidromous
New Zealand grayling (Prototroctes oxyrhynchus). Freshwater
Biology 64: 1747–1754.
Lee DE, McDowall RM, Lindqvist JK. 2007. Galaxias
fossils from Miocene lake deposits, Otago, New Zealand:
the earliest records of the Southern Hemisphere family
Galaxiidae (Teleostei). Journal of the Royal Society of New
Zealand 37: 109–130.
Leigh JW, Bryant D. 2015. POPART: full-feature software
for haplotype network construction. Methods in Ecology and
Evolution 6: 1110–1116.
Li H, Durbin R. 2009. Fast and accurate short read alignment
with Burrows–Wheeler transform. Bioinformatics 25:
1754–1760.
Li H, Handsaker B, Wysoker A, Fennell T, Ruan J,
Homer N, Marth G, Abecasis G, Durbin R. 2009.
The sequence alignment/map format and SAMtools.
Bioinformatics 25: 2078–2079.
Li J, Xia R, McDowall RM, López JA, Lei G, Fu C. 2010.
Phylogenetic position of the enigmatic Lepidogalaxias
salamandroides with comment on the orders of lower
euteleostean fishes. Molecular Phylogenetics and Evolution
57: 932–936.
Martínez-García L, Ferrari G, Oosting T, Ballantyne R,
Van der Jagt I, Ystgaard I, Harland J, Nicholson R,
Hamilton-Dyer S, Baalsrud HT, Brieuc MS, Atmore LM,
Burns F, Schmölcke U, Jakobsen KS, Jentoft S,
Orton D, Hufthammer AK, Barrett JH, Star B. 2021.
Historical demographic processes dominate genetic variation
in ancient Atlantic cod mitogenomes. Frontiers in Ecology
and Evolution 9: 671281.
Martinez-Garcia L, Ferrari G, Hufthammer AK,
Jakobsen KS, Jentoft S, Barrett JH, Star B. 2022.
Ancient DNA reveals a southern presence of the Northeast
Arctic cod during the Holocene. Biology Letters 18: 20220021.
McDowall RM. 1969. Relationships of galaxioid fishes with
a further discussion of salmoniform classification. Copeia
1969: 796–824.
McDowall RM. 1976. Fishes of the family Prototroctidae
(Salmoniformes). Marine and Freshwater Research 27:
641–659.
McDowall RM. 1990. When galaxiid and salmonid fishes
meet – a family reunion in New Zealand. Journal of Fish
Biology 37: 35–43.
McDowall RM. 2011. Ikawai: freshwater fishes in Māori culture
and economy. Christchurch: Canterbury University Press.
McDowall RM, Kennedy EM, Alloway BV. 2006. A fossil
southern grayling, genus Prototroctes, from the Pleistocene
of north-eastern New Zealand (Teleostei: Retropinnidae).
Journal of the Royal Society of New Zealand 36: 27–36.
McKenna A, Hanna M, Banks E, Sivachenko A,
Cibulskis K, Kernytsky A, Garimella K, Altshuler D,
Gabriel S, Daly M, DePristo MA. 2010. The genome
analysis toolkit: a mapreduce framework for analyzing
next-generation DNA sequencing data. Genome Research 20:
1297–1303.
McWethy DB, Wilmshurst JM, Whitlock C, Wood JR,
McGlone MS. 2014. A high-resolution chronology of rapid
forest transitions following Polynesian arrival in New
Zealand. PLoS One 9: e111328.
Mildenhall DC, Mortimer N, Bassett KN, Kennedy EM.
2014. Oligocene paleogeography of New Zealand: maximum
marine transgression. New Zealand Journal of Geology and
Geophysics 57: 107–109.
Miya M, Nishida M. 2000. Use of mitogenomic information in
teleostean molecular phylogenetics: a tree-based exploration
under the maximum-parsimony optimality criterion.
Molecular Phylogenetics and Evolution 17: 437–455.
Nelson JS. 1994. Fishes of the world. New York: Wiley.
Oosting T, Star B, Barrett JH, Wellenreuther M,
Ritchie PA, Rawlence NJ. 2019. Unlocking the potential
of ancient fish DNA in the genomic era. Evolutionary
Applications 12: 1513–1522.
Page TJ, Hughes JM. 2010. Comparing the performance of
multiple mitochondrial genes in the analysis of Australian
freshwater fishes. Journal of Fish Biology 77: 2093–2122.
Phillipps WJ. 1923. Life history of the New Zealand grayling,
Prototroctes oxyrhynchus. The New Zealand Journal of
Science and Technology 6: 5–17.
Pierson TW, Kieran TJ, Clause AG, Castleberry NL. 2020.
Preservation-induced morphological change in salamanders
and failed DNA extraction from a decades-old museum
specimen: implications for Plethodon ainsworthi. Journal of
Herpetology 54: 137–143.
Pinsky ML, Eikeset AM, Helmerson C, Bradbury IR,
Bentzen P, Morris C, Gondek-Wyrozemska AT,
Baalsrud HT, Brieuc MS, Kjesbu OS, Godiksen JA,
Barth JMI, Matschiner M, Stenseth NC, Jakobsen KS,
Jentoft S, Star B. 2021. Genomic stability through time
despite decades of exploitation in cod on both sides of the
Atlantic. Proceedings of the National Academy of Sciences of
the USA 118: e2025453118.
Preston C. 1986. From the archives. West Coast Acclimatisation
Society Annual Report 1986. West Coast Acclimatisation
Society, 16.
Pyron RA, Beamer DA, Holzheuser CR, Lemmon EM,
Lemmon AR, Wynn AH, O’Connell KA. 2022.
Contextualizing enigmatic extinctions using genomic DNA
from fluid-preserved museum specimens of Desmognathus
salamanders. Conservation Genetics 23: 375–386.
Rambaut A, Drummond AJ, Xie D, Baele G, Suchard MA.
2018. Posterior summarization in Bayesian phylogenetics
using Tracer 1.7. Systematic Biology 67: 901–904.
Rawlence NJ, Collins CJ, Anderson CN, Maxwell JJ,
Smith IW, Robertson BC, Knapp M, Horsburgh KA,
Stanton JA, Scofield RP, Tennyson AJ. Matisoo-
Smith EA, Waters JM. 2016. Human-mediated
extirpation of the unique Chatham Islands sea lion
and implications for the conservation management of
remaining New Zealand sea lion populations. Molecular
Ecology 25: 3950–3961.
Downloaded from https://academic.oup.com/zoolinnean/advance-article/doi/10.1093/zoolinnean/zlac077/6726151 by University College London user on 29 September 2022
12 L. SCARSBROOK ET AL.
© 2022 The Linnean Society of London, Zoological Journal of the Linnean Society, 2022, XX, 1–13
Rawlence NJ, Verry A, Greig K, Maxwell JJ, Shepherd LS,
Walter R. 2020. Reconstructing the impact of humans
on Aotearoa New Zealand’s biodiversity. In: Decocq G, ed.
Historical ecology. Paris: Springer.
Rutland J. 1877. On the habits of the New Zealand grayling.
Transactions and Proceedings of the New Zealand Institute
Art XXXII: 250–252.
Saville-Kent W. 1887. Observations on the acclimatisation
of the true salmon (Salmo salar), in Tasmanian waters,
and upon the reported salmon disease at the breeding
establishment on the River Plenty. Papers & Proceedings of
the Royal Society of Tasmania: 54–66.
Scarsbrook L, Verry AJ, Walton K, Hitchmough RA,
Rawlence NJ. 2022. Ancient mitochondrial genomes
recovered from small vertebrate bones through minimally
destructive DNA extraction: phylogeography of the New
Zealand gecko genus Hoplodactylus. Molecular Ecology 00:
1–21. Doi: 10.1111/mec.16434.
Schmidt DJ, Islam MRU, Hughes JM. 2016. Complete
mitogenomes for two lineages of the Australian smelt,
Retropinna semoni (Osmeriformes: Retropinnidae).
Mitochondrial DNA Part B 1: 615–616.
Schubert M, Ermini L, Sarkissian CD, Jónsson H,
Ginolhac A, Schaefer R, Martin MD, Fernández R,
Kircher M, McCue M, Willerslev E. 2014. Characterization
of ancient and modern genomes by SNP detection and
phylogenomic and metagenomic analysis using PALEOMIX.
Nature Protocols 9: 1056–1082.
Schubert M, Lindgreen S, Orlando L. 2016. AdapterRemoval
v.2: rapid adapter trimming, identification, and read merging.
BMC Research Notes 9: 1–7.
Schwarzhans WW, Nielsen SN. 2021. Fish otoliths from
the early Miocene of Chile: a window into the evolution of
marine bony fishes in the Southeast Pacific. Swiss Journal of
Palaeontology 140: 1–62.
Schwarzhans W, Scofield RP, Tennyson AJ, Worthy JP,
Worthy TH. 2011. Fish remains, mostly otoliths, from the
non-marine early Miocene of Otago, New Zealand. Acta
Palaeontologica Polonica 57: 319–350.
Seersholm FV, Cole TL, Grealy A, Rawlence NJ, Greig K,
Knapp M, Stat M, Hansen AJ, Easton LJ, Shepherd L,
Tennyson AJ. 2018. Subsistence practices, past biodiversity,
and anthropogenic impacts revealed by New Zealand-wide
ancient DNA survey. Proceedings of the National Academy of
Sciences of the USA 115: 7771–7776.
Straube N, Li C, Mertzen M, Yuan H, Moritz T. 2018. A
phylogenomic approach to reconstruct interrelationships of
main clupeocephalan lineages with a critical discussion of
morphological apomorphies. BMC Evolutionary Biology 18:
1–17.
Straube N, Lyra ML, Paigmans JLA, Preick M, Basler N,
Penner J, Rodel MO, Westbury MV, Haddad CFB,
Barlow A, Hofreiter M. 2021. Successful application of
ancient DNA extraction and library construction protocols
to museum wet collection specimens. Molecular Ecology
Resources 21: 2299–2315.
Tennyson A, Martinson P. 2007. Extinct birds of New
Zealand. Wellington: Te Papa Press.
Thacker CE, Unmack PJ, Matsui L, Rifenbark N. 2007.
Comparative phylogeography of five sympatric Hypseleotris
species (Teleostei: Eleotridae) in south-eastern Australia
reveals a complex pattern of drainage basin exchanges with
little congruence across species. Journal of Biogeography 34:
1518–1533.
Waters JM, Saruwatari T, Kobayashi T, Oohara I,
McDowall RM, Wallis GP. 2002. Phylogenetic placement
of retropinnid fishes: data set incongruence can be reduced
by using asymmetric character state transformation costs.
Systematic Biology 51: 432–449.
Waters JM, Fraser CI, Hewitt GM. 2013. Founder takes all:
density-dependent processes structure biodiversity. Trends
in Ecology & Evolution 28: 78–85.
Welicky RL, Rolfe T, Leazer K, Maslenikov KP,
Tornabene L, Holtgrieve GW, Wood CL. 2021. Fluid-
preserved fishes are one solution for assessing historical
change in fish tropic level. Ecology and Evolution 11: 415–426.
Wilmshurst JM, Anderson AJ, Higham TF, Worthy TH.
2008. Dating the late prehistoric dispersal of Polynesians to
New Zealand using the commensal Pacific rat. Proceedings of
the National Academy of Sciences of the USA 105: 7676–7680.
Wilson MV, Williams RR. 1991. New Paleocene genus and
species of smelt (Teleostei: Osmeridae) from freshwater
deposits of the Paskapoo Formation, Alberta, Canada, and
comments on osmerid phylogeny. Journal of Vertebrate
Paleontology 11: 434–451.
Worthy TH. 1987. Osteological observations on the larger
species of the skink Cyclodina and the subfossil occurrence
of these and the gecko Hoplodactylus duvaucelii in the North
Island, New Zealand. New Zealand Journal of Zoology 14:
219–229.
SUPPORTING INFORMATION
Additional Supporting Information may be found in the online version of this article at the publisher’s website.
Figure S1. X-ray images of New Zealand grayling (Prototroctes oxyrhynchus) specimens at the British National
History Museum (NHM), used to assess preservation of calcareous otoliths. Images were generated by James
Maclaine (NHM). Top to bottom: 1870.5.22.17 (syntype); 1870.5.22.18 (syntype); 1873.12.13.69; 1886.11.18.80;
1935.3.14.65.
Figure S2. MapDamage reports for mapped collapsed reads of historic New Zealand grayling (Prototroctes
oxyrhynchus) single-stranded libraries (NIWA1, NIWA2, OMVT2922). The left panels show characteristic high
Downloaded from https://academic.oup.com/zoolinnean/advance-article/doi/10.1093/zoolinnean/zlac077/6726151 by University College London user on 29 September 2022
NEW ZEALAND GRAYLING PHYLOGENY 13
© 2022 The Linnean Society of London, Zoological Journal of the Linnean Society, 2022, XX, 1–13
frequency of purines (A and G) at read termini (top) and accumulation of 5′ C to T (red curve) misincorporations
(bottom), which authenticate ancient DNA. The right panel shows characteristic short fragment length of single-
end mapped reads (top).
Table S1. Sample information for New Zealand grayling (Prototroctes oxyrhynchus) specimens including:
institution [Auckland War Memorial Museum (AWMM), Canterbury Museum (CM), Museum of New Zealand
Te Papa Tongarewa (NMNZ), Otago Museum (OM)], sample identification (ID), tissue type and locality. The last
confirmed sighting of New Zealand grayling was in 1923.
Table S2. GenBank Accession numbers for Retropinnidae (R) and Osmeridae (O) sequences used in phylogenetic
analyses. * denotes sequences derived through annotation extraction from complete mitochondrial genomes.
Bolded accession numbers indicate sequences generated in this study.
Table S3. Mitochondrial genome assembly statistics for historic New Zealand grayling (Prototroctes
oxyrhynchus) specimens (see Supporting Information, Table S1) with reads generated through single-stranded
library preparation. Summary statistics are not reported (denoted by ‘-’) for ‘GC content (%)’, ‘Ambiguous Sites’,
‘Ambiguous bases (%)’, ‘Reference sequence coverage (%)’ and ‘Contig length (bp)’ in the majority of individuals
given insufficient reference sequence coverage.
Downloaded from https://academic.oup.com/zoolinnean/advance-article/doi/10.1093/zoolinnean/zlac077/6726151 by University College London user on 29 September 2022
