Improving ecological surveys for the detection of cryptic, fossorial snakes using eDNA on and under artificial cover objects

Abstract

Performing ecological surveys for secretive, fossorial snakes is challenging. Traditional survey methods involve visual observation under artificial cover objects (ACOs); this is labor-intensive and requires multiple consistent surveys of suitable habitats. Detection of snake DNA deposited under ACOs represents an innovative method for species detection. However, for terrestrial species, common issues with soil-based methods include the challenges of adequately removing enzyme inhibitors that reduce environmental DNA (eDNA) detection and potential photodegradation of DNA taken from surface samples. These issues may be circumvented by obtaining swabs and soil samples directly from the underside of ACOs for eDNA analysis. We demonstrate the application of this method in surveys of sharp-tailed snake (Contia tenuis), an endangered species under the Canadian Species at Risk Act. We describe the design and validation of a new quantitative real-time polymerase chain reaction (qPCR)-based eDNA eCOTE3 assay with high specificity and sensitivity for sharp-tailed snake. We developed a practical and robust protocol for obtaining eDNA samples by swabbing the underside of ACOs and collecting soil samples under ACOs. Traditional surveys were conducted over two successive years (2018–19) on 220 paired ACOs at 110 sites monitored between 12 and 30 times each. Of the 6,060 ACO visits, only 24 resulted in sharp-tailed snake observations (0.4% success rate) illustrating the considerable difficulty in detecting these snakes. During this same time, 109 swabs were taken directly from the undersides of ACOs and 78 soil samples were collected from a subset of these ACOs. Of the 24 occurrences where sharp-tailed snakes were visually observed, 13 of 23 ACO swabs (57%) and nine of 20 soil samples (45%) tested positive for DNA. eDNA deposition is likely low because of the small size and behavior of this cryptic species, yet DNA was detected from soil exposed to captured snakes for only 10 min. Nevertheless, sharp-tailed snake eDNA was detected at eight sites (9%) from ACO swabs (n = 86) and seven sites (13%) from soil samples (n = 56) where snakes were not observed. This is an overall detection rate of 25% (14/56) for swab and soil samples testing positive in sites where both were tested, representing a substantial reduction in the effort required for detection of this species. Given the time-consuming nature of traditional surveys, eDNA holds great promise as a complementary survey tool for this terrestrial species. While further work is needed to delineate DNA deposition rates, this work represents a significant advance in monitoring a challenging species.

Full text

Ecological Indicators 131 (2021) 108187
1470-160X/© 2021 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
Improving ecological surveys for the detection of cryptic, fossorial snakes
using eDNA on and under articial cover objects
Laura Matthias
a
, Michael J. Allison
b
, Carrina Y. Maslovat
a
, Jared Hobbs
c
, Caren C. Helbing
b
,
*
a
Independent Consultant, Salt Spring Island, British Columbia, Canada
b
Department of Biochemistry and Microbiology, University of Victoria, Victoria, British Columbia, Canada
c
J Hobbs Ecological Consulting Ltd., Pender Island, British Columbia, Canada
ARTICLE INFO
Keywords:
Articial cover object
Endangered reptile
environmental DNA
Fossorial snake
Non-lethal sampling
Robust methods development
ABSTRACT
Performing ecological surveys for secretive, fossorial snakes is challenging. Traditional survey methods involve
visual observation under articial cover objects (ACOs); this is labor-intensive and requires multiple consistent
surveys of suitable habitats. Detection of snake DNA deposited under ACOs represents an innovative method for
species detection. However, for terrestrial species, common issues with soil-based methods include the challenges
of adequately removing enzyme inhibitors that reduce environmental DNA (eDNA) detection and potential
photodegradation of DNA taken from surface samples. These issues may be circumvented by obtaining swabs and
soil samples directly from the underside of ACOs for eDNA analysis. We demonstrate the application of this
method in surveys of sharp-tailed snake (Contia tenuis), an endangered species under the Canadian Species at Risk
Act. We describe the design and validation of a new quantitative real-time polymerase chain reaction (qPCR)-
based eDNA eCOTE3 assay with high specicity and sensitivity for sharp-tailed snake. We developed a practical
and robust protocol for obtaining eDNA samples by swabbing the underside of ACOs and collecting soil samples
under ACOs. Traditional surveys were conducted over two successive years (2018–19) on 220 paired ACOs at
110 sites monitored between 12 and 30 times each. Of the 6,060 ACO visits, only 24 resulted in sharp-tailed
snake observations (0.4% success rate) illustrating the considerable difculty in detecting these snakes. Dur-
ing this same time, 109 swabs were taken directly from the undersides of ACOs and 78 soil samples were
collected from a subset of these ACOs. Of the 24 occurrences where sharp-tailed snakes were visually observed,
13 of 23 ACO swabs (57%) and nine of 20 soil samples (45%) tested positive for DNA. eDNA deposition is likely
low because of the small size and behavior of this cryptic species, yet DNA was detected from soil exposed to
captured snakes for only 10 min. Nevertheless, sharp-tailed snake eDNA was detected at eight sites (9%) from
ACO swabs (n =86) and seven sites (13%) from soil samples (n =56) where snakes were not observed. This is an
overall detection rate of 25% (14/56) for swab and soil samples testing positive in sites where both were tested,
representing a substantial reduction in the effort required for detection of this species. Given the time-consuming
nature of traditional surveys, eDNA holds great promise as a complementary survey tool for this terrestrial
species. While further work is needed to delineate DNA deposition rates, this work represents a signicant
advance in monitoring a challenging species.
1. Introduction
World Wildlife Fund’s Living Planet Report notes that about half of
Canadian vertebrate species studied are in substantial decline with
population decreases averaging 83% among species in Canada (World
Wildlife Fund WWF Canada, 2018). Half of the world’s ecosystems are
already degraded or transformed (Pojar, 2010) and ongoing habitat loss
is the greatest causal factor identied in species loss (World Wildlife
Fund WWF Canada, 2018). Efforts to detect cryptic, rare species are
often essential for guiding habitat conservation efforts and maintaining
biodiversity.
The sharp-tailed snake (Contia tenuis) is a small (20–30 cm), slender,
reddish-brown, non-venomous snake that is endemic to western North
America, ranging from Canada continuing south through Washington,
Oregon and California (Environment and Climate Change Canada,
2017). In Canada, the species is listed as Endangered under Schedule 1
* Corresponding author.
E-mail address: chelbing@uvic.ca (C.C. Helbing).
Contents lists available at ScienceDirect
Ecological Indicators
journal homepage: www.elsevier.com/locate/ecolind
https://doi.org/10.1016/j.ecolind.2021.108187
Received 1 February 2021; Received in revised form 8 August 2021; Accepted 5 September 2021

Ecological Indicators 131 (2021) 108187
2
of the federal Species at Risk Act (SARA) due to its restricted and
discontinuous distribution in southwestern British Columbia, and due to
ongoing threats from residential development and other human activ-
ities (Environment and Climate Change Canada, 2017).
Its secretive, primarily fossorial nature makes it difcult to study,
resulting in signicant data gaps in population biology, range, and
habitat requirements (Environment and Climate Change Canada, 2017).
Virtually nothing is known of their underground habitat use. Eggs have
never been found in the wild in Canada and have only been observed
once in the United States (Sharp-tailed Snake Recovery Team, 2008).
Sharp-tailed snakes have restricted intra-seasonal movements (usually
less than 55 m during their lifespan) and they do not migrate (Sharp-
tailed Snake Recovery Team, 2008). A poor understanding of the extant
distribution further challenges effective conservation and habitat pro-
tection for this species (Environment and Climate Change Canada,
2017).
Sharp-tailed snakes are found in woodland and forest openings on
warm aspect slopes under loose talus, coarse woody debris and/or in
ssures in rock (Environment and Climate Change Canada, 2017).
Although some cryptic species are often more abundant than survey data
might suggest, this pattern is highly unlikely for sharp-tailed snake given
their highly specialized life history requirements and occupancy of an
imperiled ecosystem that has suffered signicant habitat loss in Canada.
They require natural cover objects (NCOs) such as rocks or bark slabs
that provide shelter yet NCOs need to be thin enough to warm up quickly
and transfer heat efciently to the snake (i.e., behavioural thermoreg-
ulation) (Environment and Climate Change Canada, 2017). Snakes will
also use articial cover objects (ACOs) such as small, asphalt shingles, to
thermoregulate, when ACOs are placed on the surface of the ground in
suitable habitat (Engelstoft and Ovaska, 1997; Sharp-tailed Snake Re-
covery Team, 2008). Surveyors can easily lift ACOs (optimal survey
timing is during spring and fall) to detect snakes without disturbing their
habitat (Engelstoft and Ovaska, 1997; Sharp-tailed Snake Recovery
Team, 2008).
Although ACO observation is an important survey technique, it re-
quires multiple consistent surveys for success, and it often takes years of
monitoring before there is a detection (Sharp-tailed Snake Recovery
Team, 2008; Environment and Climate Change Canada, 2017). As such,
development of reliable and less time-intensive survey methods are
required to more efciently detect sharp-tailed snakes.
Recent technical advances in quantitative polymerase chain reaction
(qPCR) have allowed detection of tiny amounts of species-specic bio-
logical material from environmental DNA (eDNA) (Jerde et al., 2011;
Goldberg et al., 2016). The term “eDNA” refers to any trace fragment of
exogenous DNA that is released by an organism into the environment
(Ficetola et al., 2008). Aquatic eDNA sampling has been used success-
fully to survey for rare vertebrates (e.g., Goldberg et al., 2011; Thomsen
et al., 2012; Sigsgaard et al., 2015; Fukumoto et al., 2015; Thomsen and
Willerslev, 2015; Lacoursi`
ere-Roussel et al., 2016; Eiler et al., 2018;
Helbing and Hobbs, 2019). Relatively few studies have explored
detection of semi-aquatic snake eDNA from (primarily) water samples
with mixed success in detecting snake DNA (Halstead et al., 2017; Jor-
dan and Ratsch, 2018; Hunter et al., 2019; Rose et al., 2019; Baker et al.,
2020; Ratsch et al., 2020). This may have been due to low rates of DNA
release, sampling methods, or low sensitivity of the eDNA detection
assays. Techniques have also been investigated for detecting terrestrial
species using eDNA in soil, and several challenges have been encoun-
tered with this type of substrate, particularly the increased presence of
enzyme inhibitors and an absence of DNA diffusion as may be the case in
an aquatic environment (Anderson et al., 2012; Thomsen and Willerslev,
2015; Schwartz et al., 2017; Walker et al., 2017; Kucherenko et al.,
2018; Mauvisseau et al., 2019; Leempoel et al., 2020; Baudry et al.,
2021).
The present study explored the possibility of sampling eDNA directly
from the underside of ACOs as an alternative substrate for the detection
of cryptic, fossorial snake species. eDNA sampling protocols were
developed and samples were collected during traditional surveys.
Samples were analyzed at the University of Victoria (UVic) using a
species-specic qPCR assay which was validated using fecal and cloacal
swabs from sharp-tailed snakes captured and sampled in the eld. eDNA
samples were collected by swabbing the underside of ACOs, and by
collecting soil beneath snakes, to determine if eDNA could be detected in
these samples. ACO swabs and soil samples were also collected in lo-
cations where no sharp-tailed snakes had been observed (during surveys
completed between 2018 and 2019) to further test the efcacy of the
eDNA sampling tool.
2. Materials and methods
2.1. eDNA assay design and validation
2.1.1. Assay development
Quantitative real-time polymerase (qPCR) primers and probes were
designed using mitochondrial gene sequences obtained from the Na-
tional Center for Biotechnology Information (NCBI) database
(https://www.ncbi.nlm.nih.gov). All publically available sequences for
the all native reptile species that occur in BC, including: sharp-tailed
snake, rubber boa (Charina bottae), western painted turtle (Chrysemys
picta), western yellow-bellied racer (Coluber constrictor Mormon), west-
ern rattlesnake (Crotalus oreganus), northern alligator lizard (Elgaria
coerulea), western skink (Eumeces skiltonianus), desert nightsnake
(Hypsiglena torquata), common wall lizard (Podarcis muralis), Great Basin
gopher snake (Pituophis catenifer deserticola), terrestrial gartersnake
(Thamnophis elegans), northwestern gartersnake (T. ordinoides), com-
mon gartersnake (T. sirtalis), plus corresponding genes for human and
dog were assembled and aligned using ClustalW (http://www.genome.
jp/tools-bin/clustalw). Candidate assay components were designed
and chosen using BioEdit (Ibis Biosciences, Carlsbad, CA, USA) and
Primer Premier Version 6 (Premier Biosoft, Palo Alto, CA, USA) based on
regions of the genes that were unique to the sharp-tailed snake. Special
care was taken to ensure the test would not amplify human DNA. Finally,
assay candidate sequences were input into NCBI Primer BLAST
(https://www.ncbi.nlm.nih.gov/tools/primer-blast/) to ensure the se-
quences would not amplify any sympatric species that were not
considered in the design phase.
2.1.2. Assay lab validation
Assay specicity was empirically determined using DNA isolated
from voucher tissues and swabs collected from specimens housed at the
Royal BC Museum and the University of Victoria laboratory under
Wildlife Permit #NA18-286900. Human total DNA was obtained from a
HEK293 cell line (American Type Culture Collection (ATCC) Manassas,
VA; Catalog number CRL-1573). All primers and the probe containing a
5′FAM reporter dye and 3′ZEN/Iowa Black FQ quencher were ordered
from Integrated DNA Technologies (IDT; Coralville, IA, USA). The
primers were rst tested for specicity against total DNA from sharp-
tailed snake and all potentially confounding (sympatric reptile) spe-
cies (Table 1) with two technical replicates using SYBR green (Invi-
trogen, Carlsbad, CA, USA) qPCR assay and agarose gel visualization of
the amplied product (amplicon).
Once the primers were conrmed to produce an amplicon of the
desired length, they were tested in combination with their correspond-
ing candidate Taqman hydrolysis probe in two technical replicates
containing 5 µg/L (equivalent to 10 picograms per reaction) gDNA based
on Nanodrop (Thermo Fisher Scientic, Waltham, MA, USA) A
260
spectrophometry readings. This quantity of gDNA is a mixture of mito-
chondrial and nuclear DNA with varying proportions of each. Despite
this, 10 picograms per reaction was sufcient to reliably expect 100%
amplication of all technical replicates from the target DNA sample. We
tested gDNA from seven individual sharp-tailed snake specimens plus
the other eight species in Table 1, and a no-template control (NTC). If
amplication was detected in a reaction within 50 cycles it was scored as
L. Matthias et al.

Ecological Indicators 131 (2021) 108187
3
a positive. The run conditions for each amplication reaction on a
CFX96 thermocycler (Bio-Rad Laboratories, Hercules, CA, USA) were as
follows: Two µL of DNA sample were run in a 15 µL total reaction volume
consisting of 10 mM Tris-HCl (pH 8.3 at 20 ◦C), 50 mM KCl, 3 mM
MgCl
2
, 0.01% Tween 20, 0.8% glycerol, 69.4 nM ROX (Life Technolo-
gies, Burlington, ON, Canada), 10.5 pmol of forward and reverse PCR
primer, 1.5 pmol of TaqMan hydrolysis probe, 200 µM dNTPs (Frogga-
Bio Inc., North York, ON, Canada), and one unit of Immolase DNA po-
lymerase (FroggaBio). DNA amplication reactions were subject to the
following thermocycle conditions: an initial activation step of 9 min at
95 ◦C followed by 50 cycles of 15 sec denaturation at 95 ◦C, 30 sec
annealing at 64 ◦C, and 30 sec polymerization at 72 ◦C.
Sharp-tailed snake DNA isolated from voucher tissue was also tested
in two replicates at ve different concentrations between 0.008 and 5
µg/L based on Nanodrop (Thermo Fisher Scientic, Waltham, MA, USA)
A
260
spectrophometry readings to determine initial sensitivity (Supple-
mentary Fig. 1). Candidate assays that exhibited specicity and sensi-
tivity were further tested to bring total technical replicates for six of the
species and all target specimen DNA concentrations to 25. The assay
eCOTE3, which amplies a 220 bp region of the NADH dehydrogenase
subunit 4 gene, was selected based on its superior sensitivity and
selectivity (Table 2).
As only part of the genomic DNA from a tissue sample is from the
mitochondria and the number of mitochondria and mitochondrial DNA
can vary widely from tissue to tissue and between individuals, we used
synthetic DNA as a template as a reproducible way to create a stan-
dardized means for expressing eDNA assay performance (Klymus et al.,
2017; Langlois et al., 2020). The eCOTE3 test efciency was further
empirically assessed using gBlocks® synthetic DNA from Integrated
DNA Technologies (Coralville, Iowa, United States) using a 5-fold serial
dilution according to protocols described in Hobbs et al. (2019). This
step allows for a standardized indicator of assay performance. Briey, a
10
7
copies/µL synthetic DNA stock was made containing 10 ng/µL tRNA
as a stabilizer (Sigma-Aldrich Canada Co., Oakville, ON, Canada). One
µL of this dilution was added to 31 µL of working tRNA solution to
produce a temporary stock containing 312,500 copies/µL. This stock
was then serially diluted ve-fold with 10 ng/µL tRNA in UltraPure
DNAse/RNAse-free distilled water (Thermo Fisher Scientic).
to produce a range of ten synthetic DNA concentrations from 31,250
copies/µL to 0.016 copies/µL. Two µL of each dilution were run in qPCR
reactions with eight technical replicates. Therefore the nal range tested
per reaction was 0.032 to 62,500 copies per reaction. These data were
used to calculate the limit of blank (LOB), limit of detection (LOD), and
limit of quantitation (LOQ) as dened by Lesperance et al. (2021) and
described in section 2.6.
2.1.3. Assay eld validation
Field validation was accomplished in two ways – by collecting swabs
directly from animals and by taking DNA samples from ACOs where
snakes were found. Cloacal DNA samples were collected from two sharp-
tailed snakes found in the eld in 2018. Cloacal swabs were taken by
rotating a wet (water-moistened), long-stem sterile swab around the
outside of the cloacal vent as described in Ford et al. (2017). An addi-
tional Q-swab was collected from fecal matter deposited on a surveyor’s
gloved hands by one of the snakes. A blank Q-swab was included as a
negative control. Swabs were stored frozen at −20 ◦C until the DNA was
isolated from them.
Two captured snakes were placed in ex situ soil to determine if short-
term deposition of eDNA could be detected (adapted from Kucherenko
et al., 2018). Each snake was placed in a separate small plastic 250 mL
container with 100 mL of ex situ soil for 10 min before being gently
placed back in the location where it was found. The soil was collected
from forested habitat well outside the habitat for sharp-tailed snakes. A
sample of forest soil that did not have a snake added was tested to assess
potential for false positives.
All live animals were treated with due consideration to alleviate
distress according to procedures and permits reviewed and approved by
the Animal Care Committee at the University of Victoria for compliance
under the Canadian Council on Animal Care guidelines (UVic Animal
Care Committee (ACC) Protocol #2018-010; Species at Risk Act Permit
#SARA-PYR-2018-0422; Wildlife Permit #NA18-286900).
2.2. Field surveys
The study location was on Salt Spring Island, British Columbia, one
of 15 known sharp-tailed snake subpopulations in Canada (Environment
and Climate Change Canada, 2017). Salt Spring Island has an area of
190 km
2
and is located 1 km east of Vancouver Island and 44 km west of
Vancouver (Fig. 1). Sharp-tailed snake were previously detected at one
site near the summit of a low mountain on the south end of Salt Spring
Island, and at another discrete site next to the ocean (1.4 km away); the
area between these two sites had not been surveyed.
The study site is located in a Garry Oak (Quercus garryana) ecosystem
in the Coastal Douglas-r (CDF) moist maritime biogeoclimatic zone
(Meidinger and Pojar, 1991). This biogeoclimatic zone, including the
Garry oak ecosystems within it, rank very high in conservation impor-
tance in British Columbia, supporting over 100 at-risk ora and fauna
found nowhere else provincially or nationally (Fuchs, 2001). Garry oak
ecosystems have <5% of their original area remaining in near-natural
condition (Lea, 2006), and the CDF biogeoclimatic zone has less than
1% remaining old-growth (>250 years) forested habitat (Madrone
Environmental Services, 2008).
In 2018, 126 (63 pairs of) ACOs were installed on three land tenures
(Crown, Ecological Reserve, and Transport Canada) on the upper
portion Mt. Tuam, Salt Spring Island, British Columbia, Canada. In 2018
Table 1
Names and abbreviations of species used for sharp-tailed snake eDNA test
validation. All species listed were initially validated with two technical repli-
cates, and those in bold were validated with 25 technical replicates.
Species Name Common Name Species
Abbreviation
Percent
Detection by
eCOTE3
Contia tenuis Sharp-tailed snake COTE 100%
Canis familiaris Dog CAFA 0%
Elgaria coerulea Northern Alligator
Lizard
ELCO 0%
Homo sapiens Human HOSA 0%
Lithobates (Rana)
catesbeiana
American Bullfrog LICA 0%
Podarcis muralis European Wall
Lizard
POMU 0%
Thamnophis
elegans
Western Terrestrial
Garter Snake
THEL 0%
Thamnophis
ordinoides
Northwestern
Garter Snake
THOR 0%
Thamnophis
sirtalis
Common Garter
Snake
THSI 0%
No Template
Control
No Template Control NTC 0%
Table 2
Nucleotide sequences for the qPCR-based eCOTE3 eDNA tool comprised of
primers and a probe for Contia tenuis detection. The amplicon sequence for the
creation of the synthetic DNA sequence is indicated.
Sequence Type Sequence
Forward Primer 5′CACATAGGCTTAGTCATTGC
Reverse Primer 5′TTATTAGGCTGGTTAGGAGTC
Probe 5′FAM- CTCCTCAGCACTCTTCTGCTTAGCCAACAC-ZEN/Iowa
Black FQ
Amplicon 5′CACATAGGCTTAGTCATTGCCGCAATCATTATTCAAACACAAT
GAAGCCTATCAGGGGCCATAGCCCTTATAATCGCTCACGGCTTC
ACCTCCTCAGCACTCTTCTGCTTAGCCAACACCACCTACGAACG
AACCACAACCCGAATTATAATTCTCACACGAGGTTTCCACAATA
TCCTACCAATAACTACAGCCTGATGACTCCTAACCAGCCTAATAA
L. Matthias et al.

Ecological Indicators 131 (2021) 108187
4
and 2019, these ACOs were regularly monitored along with 64 (32 pairs
of) ACOs that were previously installed from 2008 to 2015 on two land
tenures: Transport Canada land and private covenanted land (Fig. 1). In
2019, an additional 30 (15 pairs) ACOs installed in 2011 on Crown land
(on the lower elevation sites on the same mountain) were monitored.
Each ACO measured approximately 0.6 m ×0.6 m and was cut from a
roll of asphalt roong-shingle material. They were placed in areas of
suitable sharp-tailed snake habitat, near rotting woody debris or loose
talus, in areas with prolonged sun exposure (following protocols rec-
ommended in Engelstoft and Ovaska, 1997; Sharp-tailed Snake Recov-
ery Team, 2008). The ACOs were secured in place with small rocks or
pieces of woody debris and were visited and checked on warm sunny
days (ambient temperatures between 10 and 22 ◦C). The ACOs on the
higher area of the mountain were visited in the spring and fall during
peak snake activity (mid-day) 12 times from May 31, 2018 to March 20,
2019 and in the following spring and fall 18 times from April 4-October
23, 2019. The ACOs on the lower area of the mountain were visited 12
times between May 4, 2019 and October 9, 2019.
When a snake was found its relative age (hatchling, juvenile, sub-
adult, adult) and suspected sex was recorded. Individual snakes were
also distinguished by their unique throat patterns and distinctive
markings including ventral tail colouration, branching in their ventral
patterns or other small spots or distinctive markings.
2.3. eDNA sample collection protocols
eDNA samples were collected using specic protocols developed to
prevent contamination as adapted from provincial aquatic eDNA sam-
pling guidelines (Hobbs et al., 2017). New gloves were put on before
each new sample was collected. One surveyor was responsible for
handling snakes (if found) and a second surveyor collected the eDNA
samples to avoid cross-contamination. Hands were cleaned with alcohol
wipes after any snakes (regardless of species) were handled.
2.3.1. ACO swab collection
ACO swabs were collected from a subset of ACOs surveyed. In 2018,
ACO swab kits were prepared in the laboratory at UVic using Whatman 3
lter paper cut to the size of a 42.5 mm Whatman 1 Filter as a guide,
with a 5 mm tail to enable a rm grip on the lter while swabbing. Each
newly cut, tailed lter was placed in a paper coin envelope and then put
in a separate sealable plastic ziploc bag. This technique was modied in
year two of the project as the lter paper often fell apart while collecting
the swabs, thus complicating lab analysis. In 2019, a cotton nger cot,
placed on the index nger over top of the disposable gloves, was used for
swabbing: the nger cot increased the consistency of the area that was in
contact with the ACO by providing a more focused area of contact be-
tween the surveyor’s nger and the ACO (Fig. 2).
Swab samples were collected from the underside of an ACO that was
in contact with the substrate. The surveyor put on sterile gloves and
sprayed either the lter paper or nger cot with 70% isopropyl alcohol
(Figs. 2D and 3A) until it was damp. The back of the nger cot (i.e., over
the ngernail) was marked with an “x” using a permanent marker to
later enable easy identication of the part of the nger cot that was in
contact with the ACO (Fig. 3B). The following swab pattern (Fig. 3C) was
used on each ACO: 1) small square midway between the edge of the ACO
and the centre, starting at the top left; 2) an “X” pattern connecting each
corner; 3) a “+” pattern; and 4) along the outside edges of the ACO. After
swabbing, the lter paper samples were folded so the side that was in
Fig. 1. Locations of Articial Cover Objects (ACOs) monitored on Mt. Tuam, Salt Spring Island, British Columbia, Canada. ACOs are indicated as squares (installed
prior to 2018, white; installed 2018–19, yellow). (For interpretation of the references to colour in this gure legend, the reader is referred to the web version of
this article.)
L. Matthias et al.

Ecological Indicators 131 (2021) 108187
5
contact with the ACO was inside the fold, the tail was torn off the lter
and the lter was folded a second time and inserted into a paper coin
envelope. Finger cots were placed into a coin envelope and then into a
sealable plastic bag as they came off the glove (Fig. 3D–F). Field blanks
consisted of the lter paper or nger cot being sprayed with 70% iso-
propyl alcohol and then placed into the paper coin envelope.
Each coin envelope was labelled with the ACO site information,
location co-ordinates, date, surveyor name, and sample number. Self-
indicating silica dessicant beads (approximately one tablespoon) were
added to each sealable plastic bag and samples were handled as
described in section 2.3.3 below.
2.3.2. Soil sample collection (in situ)
Soil samples were collected from under a subset of ACOs for com-
parison with the ACO swabs. Fresh gloves were put on prior to collecting
each soil sample. The samples were collected by placing the open end of
a sterile 100 mL vial at the surface of the soil and scraping the surface
soil directly into the vial. Soil samples had variable mixes of small rocks
(talus), decayed wood, and organic matter. Care was taken not to touch
the soil, even with gloved hands, and to limit contact with the outside of
the container while the collection was being made.
2.3.3. Sample storage
All swab and soil samples were kept in a cooler with frozen ice packs
while in the eld, then transferred after the eld day and stored for up to
one month in a non self-defrosting −20 ◦C freezer until transported to
the University of Victoria. Once at the University of Victoria lab all
samples were stored in the dark at 4 ◦C until processing.
2.4. DNA isolation
All samples were randomized and assigned sample processing
numbers to reduce processing bias (Hobbs et al., 2019). The method of
DNA extraction was chosen based upon the sample material collected.
2.4.1. Tissue, Q-swab and ACO swabs
In a laminar ow hood, total DNA was recovered from each lter,
swab, or nger cot sample using the DNeasy Blood and Tissue Kit
(Qiagen Inc., Mississauga, ON, Canada; Cat# 69506) using methods
outlined in Hobbs et al. (2019). Tissue and swabs were extracted as per
the manufacturer’s protocol, and lter and nger cot samples were
extracted using the same methods with the following modications: 280
µL of Buffer ATL was used in the initial incubation, then following in-
cubation both buffer and lter were transferred to a QIAshredder
Fig. 2. Field site photos illustrating the terrain, ACO placement, and eld sampling technique. A-B) Examples of the diversity of terrain and placement of ACOs. C)
Demonstration of physical survey method. D) Demonstration of moistening a nger cot with 70% isopropyl alcohol in preparation for taking an ACO swab for eDNA.
E) ACO ipped so the underside is visible and the area covered is exposed to the left, F) Photo of a visual observation of a sharp-tailed snake under an ACO. The snake
is indicated by an arrow. Photos credit: Laura Matthias.
L. Matthias et al.

Ecological Indicators 131 (2021) 108187
6
Column (Qiagen; Cat# 79654) using forceps that had been bleach-
treated and rinsed thoroughly with deionized distilled water. The
liquid collected from the QIAshredder was incubated with 300 µL Buffer
AL for ten minutes at 56 ◦C, then vortexed and centrifuged with 300 µL
of 100% Ethanol prior to spinning through the DNeasy spin column.
DNA was eluted from the spin column with 150
μ
L of Buffer AE. DNA
samples were stored at −20 ◦C prior to use in the eDNA qPCR assay.
2.4.2. Soil samples
Early attempts to selectively liberate DNA from soil samples of suf-
cient quality using DNeasy PowerSoil kits (Qiagen; Cat# 12888) were
unsatisfactory as the kits could not adequately handle more than 0.25 g
at a time. Additional limitations were identied; PowerSoil kits failed to
produce positive target species detections, and phenol extraction
methods lack the reliability of column-based kits (Deiner et al., 2015).
Taking a subsample of the 100 mL soil samples was undesirable, so we
elected to develop a method that suspended and concentrated the DNA
prior to isolation.
All personnel in contact with the samples wore nitrile gloves and lab
coats, and replaced or sterilized gloves between samples. All benchtops
were wiped with a 10% bleach solution, followed by a 70% ethanol
solution. Two paper towels were laid out on the benchtop, overlapping
in the middle and soil samples were handled on these paper towels to
reduce the amount of soil that touched the bench. A clean scoop was
used to portion ~ 15 mL soil into a 50 mL Falcon tube. The Falcon tube
was lled to the 45 mL line with UltraPure™ DNAse/RNAse-free
distilled water (UltraPure-dH
2
O) (Invitrogen, Waltham, MA, USA).
The cap was secured and shaken vigorously. If needed, more Ultrapure-
dH
2
O was added to reach 45 mL line.
The tube was vortexed for 30 s, shaken in hand, then vortexed for
another 30 s. The sample was allowed to settle at 4 ◦C for 24 h. Paper
towels were changed and the previous steps repeated for the remaining
samples. Between handling separate soil samples, gloves were wiped
with ethanol or changed for a fresh pair. All implements were well
immersed in a 50% (v/v) bleach solution, then well-rinsed with distilled
water and wiped with a paper towel. The suspended material was
ltered using sterile cheese cloth as a barrier to large particulate matter
into a 250 mL single-use polypropylene lter funnel with a 0.45 µm pore
size cellulose nitrate membrane as described in Hobbs et al. (2020). The
DNA from one quarter of each lter was extracted with the DNeasy
Blood and Tissue kit using the same protocol as was used for lter and
swab extractions above.
Soil samples taken between September 28, 2019 and October 23,
2019 were isolated with a slightly modied protocol. Up to 100 mL of
soil was added to a clean plastic 500 mL bottle, and autoclaved double-
distilled water was added up to 450 mL. The sample was shaken
vigorously for 30 s and allowed to settle at 4 ◦C for 24 h prior to ltration
above.
2.5. qPCR analysis of eld samples
Due to the high sensitivity of the eCOTE3 assay, extreme rigour was
exercised in sample handling to reduce the possibility of cross contam-
ination between samples. Nitrile gloves and lab coats were worn at every
stage of sample analysis, and all surfaces were sterilized with a 10% (v/
v) bleach solution and 70% ethanol. All qPCR reagents and DNA isolates
were handled using sterile technique in a sterile laminar ow hood
located in a designated room. Amplicons were kept in a separate room
and personnel were required to apply separate lab coats and gloves to
access them.
All samples were analyzed through a two-tiered targeted eDNA
analysis approach (Veldhoen et al., 2016; Hobbs et al., 2019). Prior to
eCOTE3 testing, the eDNA sample was rst examined using the Integ-
ritE-DNA
TM
test (Veldhoen et al., 2016; Hobbs et al., 2019) based upon
the detection of an endogenous plant and algae chloroplast target to
evaluate the integrity of the eDNA sample.
For this test, four technical replicates of the IntegritE-DNA
TM
qPCR
assay were run as described in Veldhoen et al. (2016) and Hobbs et al.
(2019) using the same run conditions stated in section 2.1.2. If the
sample failed to amplify before 30 qPCR cycles, it was concluded that
there was either too much inhibition or the integrity of the DNA was
Fig. 3. Demonstration of the ACO swabbing technique and storage of the nger
cot. A) A cotton nger cot that has been moistened with 70% isopropyl alcohol
(see Fig. 3D) is run across the underside of an ACO. B) The top side of the nger
cot is clearly labelled with a permanent marker to aid in later DNA isolation. C)
The swabbing pattern used on the ACO follows the pattern indicated. D) The
nger cot is removed and E) carefully placed into a labelled coin envelope. F)
The coin envelope is placed into a plastic bag containing moisture indicating
silica beads for storage prior to DNA processing.
L. Matthias et al.

Ecological Indicators 131 (2021) 108187
7
compromised. A sample that failed initial IntegritE-DNA
TM
testing was
cleaned in a Zymo OneStep PCR Inhibitor Removal Kit (Cedarlane,
Burlington, ON, Canada; Cat# D6030S) and retested with the same
assay. Samples that showed positive chloroplast amplication were
tested for sharp-tailed snake DNA using the validated eCOTE3 assay
with eight technical replicates as described in section 2.1.2. All qPCR
tests included two positive DNA controls and eight NTCs per reaction
plate.
2.6. Statistical analysis
The limit of blank (LOB), limit of detection (LOD), and limit of
quantitation (LOQ) were determined using the eLowQuant R code
(https://github.com/mlespera/eLowQuant) that uses a modied
Binomial-Poisson distribution model (Lesperance et al., 2021). The
default settings were used such that the false positive and false negative
uncertainty was 0.05 and the coefcient of variation for the limit of
quantication was 0.2.
Estimated copy numbers per sample were calculated as follows. If the
qPCR reaction results produced 8/8 hits, then the equation in Fig. 4B
was applied to determine estimated copies per reaction from each C
t
value and the standard error of the mean was calculated from the eight
technical replicates. If the qPCR reaction results produced less than 8/8
hits, then the estimated copy number with standard error for that sample
were calculated using the eLowQuant script. Regardless of quantica-
tion method, swab reaction results were multiplied by a factor of 75
since 2 µL were run in a qPCR reaction from a total of 150 µL isolated
DNA per swab. For soil samples, a quarter of the lter was used to isolate
DNA from, so the multiplication factor was 4X75 =300 (Supplementary
Table 1).
3. Results
3.1. Validation of the eCOTE3 assay
After extensive validation, the eCOTE3 assay exhibits highly robust,
sensitive and selective amplication of sharp-tailed snake DNA.
Including qPCR from all validation phases and sample testing, over 700
NTCs were run with no amplication, setting the background false
positive rate at zero with a high degree of condence. Indeed, the
calculated LOB =0 (Lesperance et al., 2021). No off-target amplication
was recorded for any sympatric or confounding species DNA at con-
centrations of 5 µg/L (Table 1). The gBlocks® sensitivity validation
demonstrated that the assay is able to reliably amplify target total DNA
at extremely low concentrations (Figs. 4 and 5) and the mean copy es-
timates and proportion detects were computed using the Binomial-
Poisson maximum likelihood no intercept model (Lesperance et al.,
2021). The calculated LOD was 0.2 copies/sample (95% condence
interval =0.1–0.3) and LOQ of 0.7 copies/sample (95% condence in-
terval =0.5–1.2). Using this model and derived standard errors, we were
Fig. 4. (A) Percent of positive qPCR reactions containing known starting
amounts of gBlocks® synthetic DNA. At four copies per reaction, 100% of
technical replicates were positive so the C
t
values obtained from these reactions
and at higher concentrations (gray region) could be used to derive the standard
curve in “B”. Quantication of copy numbers below ~ 4 copies per reaction
required the application of a Binomial-Poisson model in Fig. 5.
Fig. 5. Maximum Likelihood (ML) Binomial-Poisson model of gBlocks® syn-
thetic DNA for the quantitation of low copy number DNA using eLowQuant. (A)
Unconstrained binomial estimate of the Poisson mean copy number and
transformation of exact 95% binomial condence intervals versus starting copy
number of target DNA per reaction. (B) The proportion of detected target and
the ML t on the probability scale relative to starting copy number of a sample.
The ML model t is depicted as a blue line and is used to calculate LOD, LOQ
and LOB using eLowQuant (Lesperance et al., 2021). (For interpretation of the
references to colour in this gure legend, the reader is referred to the web
version of this article.)
L. Matthias et al.

Ecological Indicators 131 (2021) 108187
8
able to estimate the copy numbers per reaction when there were less
than 100% hits out of eight technical replicates (Supplementary
Table 1). These values were then converted to copy number with stan-
dard error per sample depending upon the sample type (Supplementary
Table 1).
Initial eld validation of the eCOTE3 eDNA assay was accomplished
in two steps. First, two snakes were briey handled and cloacal swabs
taken. A fecal swab was taken from one snake. All three samples tested
positive for DNA integrity using the IntegritE-DNA
TM
assay (Table 3). No
sample inhibition was detected and all three samples returned strong
positive detection ranging from 1,060 ±113 to 49,678 ±4392 esti-
mated copies of eDNA per sample (Table 3). The swab blank containing
no DNA returned no hits for either assay, as expected (Table 3). Second,
two different snakes were placed onto forest soil for 10 min and the soil
was used to extract DNA for eDNA analysis. Both samples required in-
hibitor clean up to pass the IntegritE-DNA
TM
assay (Table 3). One soil
sample returned a comparatively weak positive signal (estimated 72 ±
45 eDNA copies; Table 3) while the other produced no detectable signal
(0/8, false negative; Table 3). The swab blank from forest soil passed the
IntegritE-DNA
TM
test and returned no hits for the eCOTE3 assay, as
expected (Table 3). The data suggest that while the eCOTE3 assay works
on multiple snakes, the deposition of DNA over a short time is not
extensive. Taken together, the eCOTE3 assay was validated to level 5
according to Thalinger et al. (2021).
3.2. Field sample analysis
Of the 109 ACO swab samples, 57 (52%) ACO swab-derived DNA
samples required inhibitor clean-up as determined by the IntegritE-
DNA
TM
assay. After inhibitor clean up, 100 (92%) samples passed the
IntegritE-DNA
TM
assay (Table 4 and Supplementary Table 2) and the
DNA from nine samples were deemed poor quality. Of the 78 soil sam-
ples that were paired with a subset of the ACO swab samples, 47 (60%)
required inhibitor clean-up as determined by the IntegritE-DNA
TM
assay.
After inhibitor clean up, 69 (88%) samples passed the IntegritE-DNA
TM
assay and the DNA from nine samples were deemed poor quality
(Table 4 and Supplementary Table 2).
Overall, 21 ACO swab and 16 soil samples had detectable sharp-
tailed snake eDNA for an overall detection rate of 21% for ACO swabs
(n =100) and 23% for soil samples (n =69; Table 4 and Fig. 6). When
both swab and soil samples were taken at an ACO on the same day and
the DNA was of sufcient quality (n =66), only ve sites (8%) had
positive detections (Supplementary Table 2). However, when swabs and
soil samples are considered together independent of ACO and sampling
event, the overall detection rate was 25% (14/56; Supplementary
Table 2). Locations of positive eDNA detections and where snakes were
found are indicated in Fig. 6. DNA extracted from soil using up to 100
mL of the sample exhibited roughly the same overall positive rate as that
extracted using 15 mL (19 and 21%, respectively), but if sites with no
snake observations are excluded, this comparison changes dramatically
to 19% for larger volumes (n =36) and 0% for smaller volumes (n =19)
attesting to an advantage of using greater amounts of soil for analysis to
ensure target DNA detection, if present.
There were 24 snake sightings under 17 different ACOs that included
repeat observations of two individual snakes on separate days, and one
occurrence with two snakes under the same ACO at the same time (Fig. 6
and Supplementary Table 2). This enabled determination of
false negative rates for the ACO swab and soil samples. Of the eDNA
samples taken where sharp-tailed snakes were visually observed during
the surveys, 13 out of 23 ACO swabs (57%) and 9 out of 22 soil samples
(41%) tested positive. As all ACO swabs related to snake sightings were
of sufcient quality, the true positive rate for the 23 samples was 57%
(13/23) (Table 5). Twenty-two corresponding soil samples were taken
(Supplementary Table 2). Of these, two soil samples were of poor
quality. Of the remaining 20 samples, nine (45%) were positive for
sharp-tailed snake eDNA (Table 5 and Supplementary Table 2). There-
fore the false negative level was 43% and 55% for swab and soil samples,
respectively (Table 5).
For sites tested for eDNA at which no sharp-tailed snake obervation
was recorded during the simultaneous physical surveying, sharp-tailed
snake eDNA was detected at eight sites for ACO swab (n =86) (9%)
and seven sites for soil samples (n =56) (13%) (Table 5 and Fig. 6). All
eld blanks were negative for sharp-tailed snake.
Examination of the relative amounts of estimated target snake eDNA
detected revealed a considerable range over three orders of magnitude
(5 ±5 to 5,607 ±348 copies from ACO swabs and 21 ±21 to 29,235 ±
1,793 copies from soil samples; Supplementary Table 2). No clear trend
in eDNA accumulation was evident in the limited cases where multiple
snakes were observed under the same ACO together or on different
sampling dates (Supplementary Table 2).
Table 3
Swab results from three individual snakes. All snakes were in Zone 10U.
Sample
#
Date Easting Northing Sample Type Snake
Found
Sex IntegritE-DNA
Frequency
Clean Up
Required
COTE
Frequency
Lab Call Estimated
Counts per
Sample
b
1 2018-
09-28
464,058 5,397,321 Q-swab
(cloacal)
a
Juvenile Female 4/4 N 8/8 Y 2,395 ±385
2 “ “ “ Q-swab
(fecal)
a
“ “ 4/4 N 8/8 Y 1,060 ±113
3 2018-
09-28
464,058 5,397,321 Q-swab
(cloacal)
Hatchling Female? 4/4 N 8/8 Y 49,678 ±4,392
4 2019-
03-18
464,059 5,397,283 Snake placed
on forest soil
Juvenile Male 4/4 Y 3/8 Y 72 ±45
5 2019-
03-18
464,019 5,397,253 Snake placed
on forest soil
Subadult Female 4/4 Y 0/8 False
negative
0 ±0
6 2018-
09-26
– – Swab Blank – – 0/4 Y 0/8 N 0 ±0
7 2019-
03-18
463,324.8 5,399,082 Forest soil
Blank
– – 4/4 Y 0/8 N 0 ±0
a
A cloacal and fecal swab was taken from the rst snake observed under this ACO.
b
For 8/8 hits, the mean ±standard error of the mean. For less than 8/8 hits, the estimated copy number ±standard error as calculated by eLowQuant based upon
binomial data is shown.
Table 4
Summary of the overall eDNA analysis results for sharp-tailed snake.
ACO Swab Soil
Total # of eDNA samples 109 78
# eDNA samples requiring inhibitor clean-up 57 (52%) 47 (60%)
# of sufcient quality samples after clean-up 100 (92%) 69 (88%)
# of eDNA detections 21 (21%) 16 (23%)
L. Matthias et al.

Ecological Indicators 131 (2021) 108187
9
4. Discussion
Surveys using eDNA are increasingly popular, especially when
applied to cryptic species that are difcult to conrm using traditional
survey methods. The development of a qPCR-based assay capable of
detection of sharp-tailed snake is an important new tool that can
improve the ability to conrm the distribution and range of this elusive
fossorial species.
In comparison with eDNA analysis of more typical aquatic samples,
soil and swab samples are much more likely to require a cleanup step.
While purication columns are effective at removing humic acids and
other qPCR inhibitors, the DNA concentration is still adversely affected
by these protocols (McKee et al., 2015). Optimization of soil and sedi-
ment sample extraction methods is therefore a crucial step toward
reliable eDNA analysis in terrestrial environments. Sample extractions
using the DNeasy Blood and Tissue kit yielded better eukaryotic DNA
signals than did those extracted with the PowerSoil kit, despite more
often requiring a cleanup step. Similar ndings were discussed by
Goldberg et al. (2011) regarding the UltraClean Soil DNA isolation kit
(MoBio Laboratories, Inc.), that may be due to the soil-based kits which
are developed primarily to target microbial organisms.
Although eDNA survey methods cannot currently determine abun-
dance or identify individuals, they potentially provide an alternative
more efcient method relative to the application of conventional
methods when applied to survey of this cryptic species. On Mt. Tuam,
where 220 ACOs were monitored between 12 and 30 times resulting in
6,060 sample events, only 24 snakes were visually detected using
traditional ACO survey methods. This demonstrates a positive detection
rate of only 0.4%. By comparison, eDNA sampling had a much higher
overall detection rate of over 20%. However, eDNA was not consistently
detected when snakes were visually detected. The false negative rate
was lower for ACO swab samples compared to soil samples likely due to
the nature of the inhibitors present. It is currently unclear why the
known-positive detection rate was still only 57% for the ACO swab
samples.
There are several factors that could affect eDNA detection rates.
Snakes under ACOs are likely sedentary (when coiled, juveniles can be
the size of a 25 cent US or Canadian coin), and the snake may have come
up from the substrate and not have been in contact with the ACO in the
area targeted by the swabbing pattern. Similarly, it may not have even
deposited DNA in the surrounding soils. The snake may have only been
under the ACO momentarily before it was detected, which may not have
been long enough for eDNA deposition to occur. In future surveys we
will attempt to improve eDNA detection rates by swabbing the entire
underside of ACOs with the more robust nger cots.
Detecting new occurrences of sharp-tailed snakes using eDNA will
Fig. 6. Map showing the eDNA results and where sharp-tailed snakes were found. A positive eDNA detection is indicated by the blue circle and a negative eDNA
detection is indicated by a white circle. A poor quality sample is indicated in grey. A black triangle inset into the circle indicates where a visual snake observation was
made at the time of sampling. (For interpretation of the references to colour in this gure legend, the reader is referred to the web version of this article.)
Table 5
Breakdown of eDNA analysis results for sharp-tailed snake according to success
of conventional surveying method (visual observation).
ACO
Swab
Soil
Total # of visual detections during sampling
a
23 20
b
# Positive eDNA hits (True positives) 13 (57%) 9
(45%)
Estimated false negative rate 43% 54%
# ACOs where snakes were not observed during the time of
sampling
86 56
# Positive eDNA hits 8 (9%) 7
(13%)
a
Two snakes observed under one ACO at the same time were counted as one visual
detection event.
b
Two eDNA samples remained of poor quality and were not included in the tally.
L. Matthias et al.

Ecological Indicators 131 (2021) 108187
10
enable researchers and landowners to more efciently identify sites
where they are present. This may encourage conservation effort for the
species. Increased detection will also provide insight into this species’
habitat needs. These sites can be afforded legal protection under the
federal Species at Risk Act via designation of Critical Habitat and iden-
tication of residence locations (Fig. 7). Habitat protection may also
benet over 100 rare species that occur in Garry Oak and associated
ecosystems (Fuchs, 2001).
The present work indicates that eDNA deposition rates for sharp-
tailed snake are relatively fast (within 10 min). Receiving even one
positive detection in the ex situ samples is noteworthy since it suggests
much faster deposition rates than those found by Kucherenko et al.
(2018), who detected eDNA from the much larger Corn Snakes (Pan-
therophis guttatus) placed on ex situ soil at the rst sampling interval of
ve hours. However, handling of the snakes may have made them prone
to defecation and may not be indicative of deposition under natural
conditions.
Future research is required to determine eDNA deposition rates and
persistence under a range of environmental conditions (e.g., seasonality,
thermal conditions, etc.). While many studies have sought to identify the
relative impacts of variables that contribute to DNA degradation be-
tween eDNA deposition and analysis (Strickler et al., 2015; Thomsen
and Willerslev, 2015; Andruszkiewicz et al., 2017; Walker et al., 2017;
Kucherenko et al., 2018), most of these experiments have been focused
on aquatic environments, and isolation of the variables for independent
analysis is rare. Since the sampling and extraction methods were opti-
mized throughout the course of the present study, separate experiments
are required to properly investigate the conditions surrounding depo-
sition and degradation of sharp-tailed snake eDNA in natural terrestrial
ecosystems. Further research is also required to explore best methods for
collection of swabs and soil samples under natural cover objects (NCOs)
in situ, such as rock and bark.
Successful detection of reptiles using eDNA methods has shown
variability in the literature and appears highly dependent on context and
methods. In a study with a similar approach to the present study, Ratsch
et al. (2020) reported greater success of Kirtland’s snake detection with
tradional methods, but suggested that eDNA reliability was likely
hampered by assay design issues. Rose et al. (2019) found that their
eDNA approach underperformed compared to traditional trapping
methods, though the majority of their eDNA samples were tested using
only one qPCR replicate due to budget constraints, which substantially
limits detection success and statistical condence (Lesperance et al.,
2021). In contrast, Hunter et al. (2019) used digital droplet PCR to
determine Burmese python presence in Florida wetlands, which vastly
improved upon the success of traditional monitoring methods despite
high concentrations of enzyme inhibitor present in their environmental
samples. Determination of a method’s efcacy in the context of species
monitoring should take into account the reliability, efciency, cost,
timeliness, and degree of invasiveness associated with the method. In
these terms, we credit the success of our approach to the collaborative
effort in targeting the specic life history of sharp-tailed snakes, as well
as both eld and laboratory methods that evolved over the course of the
present study to improve detection reliability.
There are numerous at-risk species, including sharp-tailed snake,
that occur in Garry oak ecosystems; all face many threats (Environment
and Climate Change Canada, 2017). The biggest threat is direct loss of
habitat to land conversion associated with land or resource develop-
ment. In addition, these ecosystems are adapted to frequent res as re
suppression leads to increased conifer and shrub encroachment. Woody
plant inlling by both native and non-native ora also degrades Garry
oak ecosystems by shading habitat particularly for species that require
open, sunny sites for behavioral thermoregulation and egg deposition.
Fig. 7. Compilation of eDNA (blue circles) and visual sharp-tailed snake detections (prior to March 2018, green; between 2018 and 19, orange) and indicated
proposed critical habitat (yellow hatched area) and extension (pink hatched area). (For interpretation of the references to colour in this gure legend, the reader is
referred to the web version of this article.)
L. Matthias et al.

Ecological Indicators 131 (2021) 108187
11
Finally, the impacts of future climate change on sharp-tailed snakes is
not understood. Identifying sharp-tailed snake occurrence locations is a
necessary step towards conservation of their habitat. While there are
obstacles to reliable detection of reptilian DNA in a terrestrial environ-
ment, this study demonstrates that eDNA techniques have promising and
benecial use for monitoring cryptic and endangered fossorial species.
CRediT authorship contribution statement
Laura Matthias: Conceptualization, Data curation, Formal analysis,
Funding acquisition, Investigation, Methodology, Project administra-
tion, Validation, Visualization, Writing – original draft, Writing - review
& editing. Michael J. Allison: Conceptualization, Formal analysis,
Investigation, Methodology, Validation, Visualization, Writing – orig-
inal draft, Writing - review & editing. Carrina Y. Maslovat: Concep-
tualization, Data curation, Formal analysis, Funding acquisition,
Investigation, Methodology, Validation, Visualization, Writing – orig-
inal draft, Writing - review & editing. Jared Hobbs: Conceptualization,
Methodology, Writing - review & editing. Caren C. Helbing: Concep-
tualization, Data curation, Formal analysis, Methodology, Project
administration, Resources, Validation, Visualization, Writing – original
draft, Writing - review & editing.
Declaration of Competing Interest
The authors declare that they have no known competing nancial
interests or personal relationships that could have appeared to inuence
the work reported in this paper.
Acknowledgements
We would like to recognize and acknowledge that the eld research
took place on the unceded territories of the Coast Salish Peoples, spe-
cically of the Hul’qumi’num and the SENCOTEN speaking peoples.
This project would not have been possible without the support of Karen
Hall (Environmental Advisor, Transport Canada), Eric Gross (Species at
Risk, Biologist Canadian Wildlife Service), and Erica McLaren (Conser-
vation Specialist, BC Parks). We also acknowledge the expert technical
assistance of Sandy Shen, Lauren Bergman, and Neha Acharya-Patel. We
appreciate the time and generosity of the following for sharing their
ideas, suggestions and experience: Dr. John Herman, Gregg Schumer,
and Dr. Kristiina Ovaska. We are grateful for eld assistance from
Suzanne L’Heureux, Tony McLeod, Chris Wood, and Kevin Neill. Thank
you to Gavin Hanke of the Royal British Columbia Museum for provision
of tissue samples for the development of the eCOTE3 assay.
Funding
This work was supported by the Habitat Conservation Trust Foun-
dation, Park Enhancement Fund; Canadian Wildlife Service; and
Transport Canada, Interdepartmental Recovery Fund.
Appendix A. Supplementary data
Supplementary data to this article can be found online at https://doi.
org/10.1016/j.ecolind.2021.108187.
References
Anderson, K., Bird, K.L., Rasmussen, M., Haile, J., Breuning-Madsen, H., Kjaer, K.H.,
Orlando, L., Gilbert, M.T.P., Willerslev, E., 2012. Meta-barcoding of “dirt” DNA from
soil reects vertebrate biodiversity. Mol. Ecol. 21 (8), 1966–1979. https://www.ncb
i.nlm.nih.gov/pubmed/21917035.
Andruszkiewicz, E.A., Sassoubre, L.M., Boehm, A.B., Doi, H., 2017. Persistence of marine
sh environmental DNA and the inuence of sunlight. PLoS ONE 12 (9), e0185043.
https://doi.org/10.1371/journal.pone.0185043.
Baker, S.J., Niemiller, M.L., Stites, A.J., Ash, K.T., Davis, M.A., Dreslik, M.J., Phillips, C.
A., 2020. Evaluation of environmental DNA to detect Sistrurus catenatus and
Ophidiomyces ophiodiicola in craysh burrows. Conserv. Genet. Resour. 12 (1),
13–15. https://doi.org/10.1007/s12686-018-1053-9.
Baudry, T., Mauvisseau, Q., Goût, J.-P., Arqu´
e, A., Delaunay, C., Smith-Ravin, J.,
Sweet, M., Grandjean, F., 2021. Mapping a super-invader in a biodiversity hotspot,
an eDNA-based success story. Ecol. Ind. 126, 107637. https://doi.org/10.1016/j.
ecolind.2021.107637.
Deiner, K., Walser, J.-C., M¨
achler, E., Altermatt, F., 2015. Choice of capture and
extraction methods affect detection of freshwater biodiversity from environmental
DNA. Biol. Conserv. 183, 53–63. https://doi.org/10.1016/j.biocon.2014.11.018.
Eiler, A., L¨
ofgren, A., Hjerne, O., Nord´
en, S., Saetre, P., 2018. Environmental DNA
(eDNA) detects the pool frog (Pelophylax lessonae) at times when traditional
monitoring methods are insensitive. Sci. Rep. 8 (1) https://doi.org/10.1038/s41598-
018-23740-5.
Engelstoft, C., Ovaska, K., 1997. Sharp-tailed snake inventory within the Coastal
Douglas-r biogeoclimatic zone, June-November 1996. Unpublished report prepared
by Alula Biological Consulting for the B.C. Min. Environ., Lands and Parks,
Vancouver Island Region, Nanaimo, BC.
Environment and Climate Change Canada, 2017. Recovery Strategy for the sharp-tailed
snake (Contia tenuis) in Canada [Proposed]. Species at Risk Recovery Strategy
Series. Environmental and Climate Change Canada, Ottawa, ON. 17pp +42pp.
https://www.registrelep-sararegistry.gc.ca/virtual_sara/les/plans/rs_sharp_tailed_
snake_e_proposed.pdf.
Ficetola, G.F., Miaud, C., Pompanon, F., Taberlet, P., 2008. Species detection using
environmental DNA from water samples. Biol. Lett. 4 (4), 423–425. https://doi.org/
10.1098/rsbl.2008.0118.
Ford, B., Govindarajulu, P., Larsen, K., Russello, M., 2017. Evaluating the efcacy of non-
invasive genetic sampling of the Northern Pacic rattlesnake with implications for
other venomous squamates. Conserv. Genet. Resour. 9 (1), 13–15. https://doi.org/
10.1007/s12686-016-0606-z.
Fuchs, M.A., 2001. Towards a recovery strategy for Garry oak and associated ecosystems
in Canada: ecological assessment and literature review. Technical Report GBEI/EC-
00-030. Environment Canada, Canadian Wildlife Service. Pacic and Yukon Region.
Fukumoto, S., Ushimaru, A., Minamoto, T., 2015. A basin-scale application of
environmental DNA assessment for rare endemic species and closely related exotic
species in rivers: a case study of giant salamanders in Japan. J. Appl. Ecol. 52,
358–365. https://doi.org/10.1111/1365-2664.12392.
Goldberg, C.S., Pilliod, D.S., Arkle, R.S., Waits, L.P., Gratwicke, B., 2011. Molecular
detection of vertebrates in stream water: a demonstration using Rocky Mountain
tailed frogs and Idaho giant salamanders. PLoS ONE 6 (7), e22746. https://doi.org/
10.1371/journal.pone.0022746.
Goldberg, C.S., Turner, C.R., Deiner, K., Klymus, K.E., Thomsen, P.F., Murphy, M.A.,
Spear, S.F., McKee, A., Oyler-McCance, S.J., Cornman, R.S., Laramie, M.B.,
Mahon, A.R., Lance, R.F., Pilliod, D.S., Strickler, K.M., Waits, L.P., Fremier, A.K.,
Takahara, T., Herder, J.E., Taberlet, P., 2016. Critical considerations for the
application of environmental DNA methods to detect aquatic species. Methods Ecol.
Evol. 7, 1299–1307.
Halstead, B.J., Wood, D.A., Bowen, L., Waters, S., Vandergast, A.G., Ersan, J.S.M.,
Skalos, S.M., Casazza, M.L., 2017. An evaluation of the efcacy of using
environmental DNA (eDNA) to detect giant gartersnakes (Thamnophis gigas). US
Geological survey Open-le report 2017–1123, 41p. https://doi.org/10.3133/
ofr20171123.
Helbing, C.C., Hobbs, J., 2019. Environmental DNA standardization needs for sh and
wildlife population assessments and monitoring. Canadian Standards Association,
Toronto, ON. 41 pp. https://www.csagroup.org/wp-content/uploads/CSA-Group-
Research-Environmental-DNA.pdf.
Hobbs, J., Adams, I.T., Round, J.M., Goldberg, C.S., Allison, M.J., Bergman, L.C.,
Mirabzadeh, A., Allen, H., Helbing, C.C., 2020. Revising the range of Rocky
Mountain tailed frog, Ascaphus montanus, in British Columbia, Canada, using
environmental DNA methods. Environ. DNA 2 (3), 350–361. https://doi.org/
10.1002/edn3.v2.310.1002/edn3.82.
Hobbs, J., Goldberg, C., Helbing, C.C., Veldhoen, N., Vincer, E., 2017. BC Ministry of
Environment: Environmental DNA protocol for freshwater aquatic ecosystems
Version 2.2 (File: 489-027.01; revised November 2017), 48 pp.
Hobbs, J., Round, J.M., Allison, M.J., Helbing, C.C., St¨
ock, M., 2019. Expansion of the
known distribution of the coastal tailed frog, Ascaphus truei, in British Columbia,
Canada, using robust eDNA detection methods. PLoS ONE 14 (3), e0213849. https://
doi.org/10.1371/journal.pone.0213849.
Hunter, M.E., Meigs-Friend, G., Ferrante, J.A., Smith, B.J., Hart, K.M., 2019. Efcacy of
eDNA as an early detection indicator for Burmese pythons in the ARM Loxahatchee
National Wildlife Refuge in the greater Everglades system. Ecol. Ind. 102, 617–622.
https://doi.org/10.1016/j.ecolind.2019.02.058.
Jerde, C.L., Mahon, A.R., Chadderton, W.L., Lodge, D.M., 2011. Sight-unseen” detection
of rare aquatic species using environmental DNA. Conserv. Lett. 4, 150–157.
Jordan, M., Ratsch, R., 2018. State Wildlife Grant- Indiana: Efcacy of Using
Environmental DNA (eDNA) to Detect Kirtland’s Snake. Second year report of a two
year project. https://www.in.gov/dnr/shwild/les/fw-2018wsr_KirtlandsSnake.
pdf [accessed September 14, 2020]. State Wildlife Grant Program (T7R20) and
Purdue University Fort Wayne. 4 pp.
Klymus, K.E., Marshall, N.T., Stepien, C.A., Doi, H., 2017. Environmental DNA (eDNA)
metabarcoding assays to detect invasive invertebrate species in the Great Lakes.
PLoS ONE 12 (5), e0177643. https://doi.org/10.1371/journal.pone.0177643.
Kucherenko, A., Herman, J.E., Everham, E.M., Urakawa, H., 2018. Terrestrial snake
environmental DNA accumulation and degradation dynamics environmental
application. Herpetologica 74 (1), 38–49. https://doi-org.ezproxy.library.uvic.
ca/10.1655/Herpetologica-D-16-00088.
L. Matthias et al.

Ecological Indicators 131 (2021) 108187
12
Lacoursi`
ere-Roussel, A., Dubois, Y., Normandeau, E., Bernatchez, L., 2016. Improving
herpetological surveys in eastern North America using the environmental DNA
method. Genome, online NRC Research Press. http://www.nrcresearchpress.com/
doi/abs/10.1139/gen-2015-0218.
Langlois, V.S., Allison, M.J., Bergman, L.C., To, T.A., Helbing, C.C., 2020. The need for
robust qPCR-based eDNA detection assays in environmental monitoring and species
inventories. Environ. DNA 3 (3), 519–527. https://onlinelibrary.wiley.com/doi/full/
10.1002/edn3.164.
Lea, T., 2006. Historical Garry Oak ecosystems of Vancouver Island, British Columbia,
pre-European contact to the present. Davisonia 17 (2), 34–50.
Leempoel, K., Hebert, T., Hadly, E.A., 2020. A comparison of eDNA to camera trapping
for assessment of terrestrial mammal diversity. Proc. R. Soc. B. 287 (1918),
20192353. https://doi.org/10.1098/rspb.2019.2353.
Lesperance, M.L., Allison, M.J., Bergman, L.C., Hocking, M.D., Helbing, C.C., 2021.
A statistical model for calibration and computation of detection and quantication
limits for low copy number environmental DNA samples. Environ. DNA 00, 1–12.
https://doi.org/10.1002/edn3.220.
Madrone Environmental Services, 2008. Terrestrial ecosystem mapping of the Coastal
Douglas-r Biogeoclimatic zone – Madrone Environmental Services. Unpublished
report prepared for Integrated Land Management Bureau. https://bvcentre.ca/les/
research_reports/07-17_TEM_4522_rpt_only.pdf.
Mauvisseau, Q., T¨
onges, S., Andriantsoa, R., Lyko, F., Sweet, M., 2019. Early detection of
an emerging invasive species: eDNA monitoring of a parthenogenetic craysh in
freshwater systems. Manage. Biol. Invasions 10 (3), 461–472. https://doi.org/
10.3391/mbi.2019.10.3.04.
McKee, A.M., Spear, S.F., Pierson, T.W., 2015. The effect of dilution and the use of a post-
extraction nucleic acid purication column on the accuracy, precision, and
inhibition of environmental DNA samples. Biol. Conserv. 183, 70–76. https://doi.
org/10.1016/j.biocon.2014.11.031.
Meidinger, D., Pojar, J., 1991. Ecosystems of BC. Victoria, BC, Canada. BC Ministry of
Forests Special Report Series, Research Branch, Ministry of Forests, Victoria, BC. 330
pp.
Pojar, J., 2010. A new climate for conservation: Nature, carbon and climate change in
British Columbia. The Working Group on Biodiversity, For. Clim. 99 pp. https://
climateactionnetwork.ca/wp-content/uploads/2014/02/NewClimate_report_DSF.
pdf.
Ratsch, R., Kingsbury, B.A., Jordan, M.A., 2020. Exploration of environmental DNA
(eDNA) to detect Kirtland’s snake (Clonophis kirtlandii). Animals 10, 1057. https://
doi.org/10.3390/ani10061057.
Rose, J.P., Wademan, C., Weir, S., Wood, J.S., Todd, B.D., Schmidt, B.R., 2019.
Traditional trapping methods outperform eDNA sampling for introduced semi-
aquatic snakes. PLoS ONE 14 (7), e0219244. https://doi.org/10.1371/journal.
pone.0219244.
Schwartz, M.K., Penaluna, B.E., Wilcox, T.M., 2017. Not just for sheries biologists
anymore: environmental DNA sampling makes strides in wildlife. Wildlife
Professional 11 (6), 47–51.
Sharp-tailed Snake Recovery Team, 2008. Recovery Strategy for the Sharp-tailed snake
(Contia tenuis) in British Columbia. Prepared for the B.C. Ministry of Environment,
Victoria, BC. 27 pp. http://www.env.gov.bc.ca/wld/documents/recovery/
rcvrystrat/sharp-tailed_snake_rcvry_strat_2008.pdf.
Sigsgaard, E.E., Carl, H., Møller, P.R., Thomsen, P.F., 2015. Monitoring the near-extinct
European weather loach in Denmark based on environmental DNA from water
samples. Biol. Conserv. 183, 46–52. https://doi.org/10.1016/j.biocon.2014.11.023.
Strickler, K.M., Fremier, A.K., Goldberg, C.S., 2015. Quantifying effects of UV-B,
temperature, and pH on eDNA degradation in aquatic microcosms. Biol. Conserv.
183, 85–92. https://pubag.nal.usda.gov/catalog/5361083.
Thalinger, B., Deiner, K., Harper, L.R., Rees, H.C., Blackman, R.C., Sint, D., Traugott, M.,
Goldberg, C.S., Bruce, K., 2021. A validation scale to determine the readiness of
environmental DNA assays for routine species monitoring. Environ. DNA 3 (4),
823–836. https://doi.org/10.1002/edn3.189.
Thomsen, P.F., Kielgast, J., Iversen, L.L., Wiuf, C., Rasmussen, M., Gilbert, M.T.,
Orlando, L., Willerslev, E., 2012. Monitoring endangered freshwater biodiversity
using environmental DNA. Mol. Ecol. 21, 2565–2573. https://doi.org/10.1111/
j.1365-294X.2011.05418.x.
Thomsen, P.F., Willerslev, E., 2015. Environmental DNA - an emerging tool in
conservation for monitoring past and present biodiversity. Biol. Conserv. 183, 4–18.
Veldhoen, N., Hobbs, J., Ikonomou, G., Hii, M., Lesperance, M., Helbing, C.C.,
Melcher, U., 2016. Implementation of Novel Design Features for qPCR-Based eDNA
Assessment. PLoS ONE 11 (11), e0164907. https://doi.org/10.1371/journal.
pone.0164907.
Walker, D.M., Leys, J.E., Dunham, K.E., Oliver, J.C., Schiller, E.E., Stephenson, K.S.,
Kimrey, J.T., Wooten, J., Rogers, M.W., 2017. Methodological considerations for
detection of terrestrial small-body salamander eDNA and implications for
biodiversity conservation. Mol. Ecol. Resour. 17 (6), 1223–1230.
World Wildlife Fund (WWF) Canada, 2018. Living Planet Report 2018. https://wwf.ca/l
iving-planet-report.
L. Matthias et al.