Threat assessment for Pacific sand lance (Ammodytes personatus) in the Salish Sea

Abstract

Like many forage fish species, Pacific sand lance (Ammodytes personatus) play a key role in nearshore marine ecosystems as an important prey source for a diverse array of predators in the northeastern Pacific. However, the primary threats to Pacific sand lance and their habitat are poorly defined due to a lack of systematic data. Crucial information needed to assess their population status is also lacking including basic knowledge of their local and regional abundance and distribution. Sand lance are currently listed as ‘not evaluated’ under the IUCN red list and they have not been assessed by US and Canadian agencies. This hampers management and policy efforts focused on their conservation. To address this knowledge gap, we conducted a three-part, structured expert elicitation to assess the vulnerability of Salish Sea sand lance populations. Experts were asked to list and rank key threats to Salish Sea sand lance and/or their habitat, to further quantify the vulnerability of sand lance to identified threats using a vulnerability matrix, and to predict the population trajectory in 25 years from today. Impacts associated with climate change (e.g. sea level rise, sea temperature rise, ocean acidification, and extreme weather) consistently ranked high as threats of concern in the ranking exercise and quantified vulnerability scores. Nearly every expert predicted the population will have declined from current levels in 25 years. These results suggest sand lance face numerous threats and may be in decline under current conditions. This research provides vital information about which threats pose the greatest risk to the long-term health of sand lance populations and their habitat. Managers can use this information to prioritize which threats to address. Future research to reliably quantify population size, better understand the roles of natural and anthropogenic impacts, and to identify the most cost-effective actions to mitigate multiple threats, is recommended.

Full text

Threat assessment for Pacific
sand lance (Ammodytes
personatus) in the Salish Sea
Jacqueline R. Huard
1,2
*, Victoria Hemming
1
,
Matthew R. Baker
3
, Jennifer Blancard
4
, Ian Bruce
5
,
Sarah Cook
6
, Gail K. Davoren
7
, Phillip Dionne
8
, Virginia East
2
,
J. Mark Hipfner
9
, Nicola R. Houtman
10
, Brian A. Koval
5,11
,
Dayv Lowry
12
, Rowen Monks
13
, Graham Nicholas
14
,
Beatrice Proudfoot
15
, Micah Quindazzi
16,17
, Timothy Quinn
8
,
Clifford L. K. Robinson
15
, Emily M. Rubidge
1,18
, Dianne Sanford
19
,
James R. Selleck
20
, Anne Shaffer
21
, Nikki Wright
6
,
Jennifer Yakimishyn
22
and Tara G. Martin
1
1
Conservation Decisions Lab, Department of Forest and Conservation Sciences, University of British
Columbia, Vancouver, BC, Canada,
2
Comox Valley Project Watershed, Courtenay, BC, Canada,
3
University of Washington, Friday Harbor Labs, Friday Harbor, WA, United States,
4
The Loon
Foundation, Madeirda Park, BC, Canada,
5
Peninsula Streams Society, North Saanich, BC, Canada,
6
SeaChange Marine Conservation Society, Victoria, BC, Canada,
7
Department of Biological Sciences,
University of Manitoba, Winnipeg, MB, Canada,
8
Washington Department of Fish and Wildlife,
Olympia, WA, United States,
9
Environment and Climate Change Canada, Wildlife Research Division,
Delta, BC, Canada,
10
Department of Geography, University of Victoria, Victoria, BC, Canada,
11
Canadian Wildlife Service, Environment and Climate Change Canada, Sackville, NB, Canada,
12
National Oceanic and Atmospheric Administration, National Marine Fisheries Service, West Coast
Region, Protected Resources Division, Lacey, WA, United States,
13
Department of Bioscience,
University of Oslo, Oslo, Norway,
14
Tsleil-Waututh Nation/təsəlilwətaɬxʷəlməxʷ, North Vancouver,
BC, Canada,
15
Fisheries and Oceans Canada, Pacific Biological Station, Nanaimo, BC, Canada,
16
Fisheries, Marine Ecology and Conservation Lab, Department of Biology, University of Victoria,
Victoria, BC, Canada,
17
Marine Science Program, Pacific Salmon Foundation, Vancouver, BC, Canada,
18
Fisheries and Oceans Canada, Institute of Ocean Sciences, Sidney, BC, Canada,
19
Sunshine Coast
Friends of Forage Fish, Sechelt, BC, Canada,
20
Natural Resources Consultants, Seattle, WA, United States,
21
Coastal Watershed Institute, Port Angeles, WA, United States,
22
Pacific Rim National Park Reserve, Ucluelet,
BC, Canada
Like many forage fish species, Pacific sand lance (Ammodytes personatus) play a
key role in nearshore marine ecosystems as an important prey source for a
diverse array of predators in the northeastern Pacific. However, the primary
threats to Pacific sand lance and their habitat are poorly defined due to a lack of
systematic data. Crucial information needed to assess their population status is
also lacking including basic knowledge of their local and regional abundance and
distribution. Sand lance are currently listed as ‘not evaluated’under the IUCN red
list and they have not been assessed by US and Canadian agencies. This hampers
management and policy efforts focused on their conservation. To address this
knowledge gap, we conducted a three-part, structured expert elicitation to
assess the vulnerability of Salish Sea sand lance populations. Experts were
asked to list and rank key threats to Salish Sea sand lance and/or their habitat,
to further quantify the vulnerability of sand lance to identified threats using a
vulnerability matrix, and to predict the population trajectory in 25 years from
today. Impacts associated with climate change (e.g. sea level rise, sea
temperature rise, ocean acidification, and extreme weather) consistently
Frontiers in Marine Science frontiersin.org01
OPEN ACCESS
EDITED BY
Correigh Greene,
National Oceanic and Atmospheric
Administration (NOAA), United States
REVIEWED BY
Anna Kagley,
NOAA Fisheries, United States
Joshua Chamberlin,
National Marine Fisheries Service (NOAA),
United States
*CORRESPONDENCE
Jacqueline R. Huard
jachuard@gmail.com
RECEIVED 07 June 2024
ACCEPTED 30 September 2024
PUBLISHED 30 October 2024
CITATION
Huard JR, Hemming V, Baker MR, Blancard J,
Bruce I, Cook S, Davoren GK, Dionne P,
East V, Hipfner JM, Houtman NR, Koval BA,
Lowry D, Monks R, Nicholas G, Proudfoot B,
Quindazzi M, Quinn T, Robinson CLK,
Rubidge EM, Sanford D, Selleck JR, Shaffer A,
Wright N, Yakimishyn J and Martin TG (2024)
Threat assessment for Pacific sand lance
(Ammodytes personatus) in the Salish Sea.
Front. Mar. Sci. 11:1445215.
doi: 10.3389/fmars.2024.1445215
COPYRIGHT
©2024Huard,Hemming,Baker,Blancard,
Bruce, Cook, Davoren, Dionne, East, Hipfner,
Houtman, Koval, Lowry, Monks, Nicholas,
Proudfoot, Quindazzi, Quinn, Robinson,
Rubidge, Sanford, Selleck, Shaffer, Wright,
Yakimishyn and Martin. This is an open-access
article distributed under the terms of the
Creative Commons Attribution License (CC BY).
The use, distribution or reproduction in other
forums is permitted, provided the original
author(s) and the copyright owner(s) are
credited and that the original publication in
this journal is cited, in accordance with
accepted academic practice. No use,
distribution or reproduction is permitted
which does not comply with these terms.
TYPE Original Research
PUBLISHED 30 October 2024
DOI 10.3389/fmars.2024.1445215

ranked high as threats of concern in the ranking exercise and quantified
vulnerability scores. Nearly every expert predicted the population will have
declined from current levels in 25 years. These results suggest sand lance face
numerous threats and may be in decline under current conditions. This research
provides vital information about which threats pose the greatest risk to the long-
term health of sand lance populations and their habitat. Managers can use this
information to prioritize which threats to address. Future research to reliably
quantify population size, better understand the roles of natural and
anthropogenic impacts, and to identify the most cost-effective actions to
mitigate multiple threats, is recommended.
KEYWORDS
expert elicitation, threats, conservation, marine food web, ecological risk assessment,
forage fish, nearshore habitat
Introduction
Pacific sand lance, Ammodytes personatus (hereafter, sand
lance), are small forage fish that play a significant role in the
nearshore ecosystem of the northeast Pacific Ocean. Sand lance
are known to comprise part of the diet for at least 100 predators
(Robards and Piatt, 1999;Penttila, 2007;Harvey et al., 2010;Alheit
and Peck, 2019;Staudinger, 2020;Scordino et al., 2022;Shaffer et al.,
2023). Examples include seabirds, especially Alcids (Robards and
Piatt, 1999;Zamon, 2000;Pastran et al., 2021); fish such as Chinook
(Oncorhynchus tshawytscha)andcohosalmon(Oncorhynchus
kisutch;Duguid, 2020), Pacificcod(Gadus macrocephalus;
Gunther et al., 2023), and lingcod (Ophiodon elongatus;
Beaudreau and Essington, 2007); and larger mammals like Steller
sea lions (Eumetopias jubatus;McKenzie and Wynne, 2008), harbor
seals (Phoca vitulina richardii;Lance et al., 2012), and baleen whales
such as humpback (Megaptera novaeangliae;Wright et al., 2016),
minke (Balaenoptera acutorostrata;Okamura et al., 2009;Towers
et al., 2019) and fin(Balaenoptera physalus;Moore et al., 2019).
They play a particularly important role during the breeding of
seabirds because of the high energy content, and slender bodies that
are easily transported to, and consumed by, chicks (Willson et al.,
1999;Bertram et al., 2001;Hedd et al., 2006;Beaubier and Hipfner,
2013). Sand lance range from California to Alaska and are one
species within the Ammodytes genus, the only species that occurs in
the Salish Sea, and one of two in the eastern Pacific(Robards and
Piatt, 1999;Orr et al., 2015). They are dependant on specific
spawning and burying habitats that must have coarse, silt-free,
well oxygenated sandy substrates (hereafter “habitat”)(Baker et al.,
2024). Declines in sand lance abundance could have serious rippling
impacts on coastal ecosystems given their important role in the food
web (Bertram et al., 2001;Robards et al., 2002;Piatt et al., 2020).
Piecemeal observations by researchers, including some of the
authors, Indigenous groups, and anecdotal reports from
recreational anglers in the northeastern Pacificoverthelast
decade have resulted in a growing concern that forage fish,
including sand lance populations, face numerous threats (Dethier
et al., 2016;Frick et al., 2022;Robinson et al., 2023). While some
research has been undertaken on their biology (Penttila, 2007;
Haynes et al., 2008;Haynes and Robinson, 2011;Hipfner and
Galbraith, 2013;Matta and Baker, 2020;Zhukova and Baker,
2022;Robinson et al., 2023), distribution and seasonal abundance
(Selleck et al., 2015), and habitat distribution (Robinson, 2013;
Baker et al., 2019;Robinson et al., 2021;Huard et al., 2022;Baker
et al., 2023;Gunther et al., 2023), there is a lack of comprehensive
empirically derived information to contextualize their population
status at a scale as wide as the Salish Sea.
Sand lance have an uneven distribution throughout their range,
often restricted to small, specific, and patchy habitats that are not well
mapped (Haynes et al., 2008;Robinson et al., 2013;Speed and Baker,
2016;Greene et al., 2020;Robinson et al., 2021;Huard et al., 2022).
Theiroccurrenceistemporallyvariable(Selleck et al., 2015)andthey
are dormant much of the winter (van Deurs et al., 2010;Haynes and
Robinson, 2011;Robinson et al., 2013;Greene et al., 2015;Baker et al.,
2019,2023). Accurate estimates employed in the North Sea require use
of both acoustics and nets at great cost with imperfect results
(Greenstreet et al., 2006,2010). There has been limited species-
specific population assessments at the Salish Sea scale, and one
projecting forward in the US side of the Salish Sea. In Washington
State, field surveys targeting the San Juan Channel sand wavefield
(approximately 0.6 km
2
) have occurred over the past 15 years
estimating that there is an average of 81 million sand lance (Blaine,
2006;Baker et al., 2019;Baker et al., 2024); however, these surveys are
limited to one area and do not speculate on population trends (Speed
and Baker, 2016;Greene et al., 2020,Greene et al., 2021). The National
Oceanic and Atmospheric Administration (NOAA) has conducted
research on abundance trend estimates for several forage fish within
Puget Sound and found sand lance populations increased in some
basins over the study time (1971-1985; 2002-2003) (Greene et al.,
2015). A handful of additional studies relating to Pacific sand lance in
Huard et al. 10.3389/fmars.2024.1445215
Frontiers in Marine Science frontiersin.org02

the Salish Sea exist, however they lack specifictemporal,spatial,or
species-specific assessments to be useful to understanding the sand
lance population trends (Penttila, 1995;Selleck et al., 2015)(Tomlin
et al., 2021)(Bertram and Kaiser, 1993;Hedd et al., 2006;Thayer et al.,
2008)(Gunther et al., 2023), (Duguid, 2020;Baker et al., 2021)
(Einoder, 2009). The historical lack of research conducted on sand
lance may have contributed to the lack of knowledge and research into
population status as it can be difficult to acquire research funding if
there is no/little demonstratable risk to the species or ecosystem.
There are multiple, potentially interacting threats on sand lance
in the Salish Sea (Krueger et al., 2010;Hipfner et al., 2018;Baker
et al., 2019;Buchanan et al., 2019;Liedtke and Conn, 2021;Selden
and Baker, 2023). The Salish Sea is a densely populated area on the
Northeastern Pacific coast where rapid anthropogenic growth and
development is exerting increasing pressure on regional biodiversity
(Gaydos and Pearson, 2011;Gaydos et al., 2015;Singh et al., 2017).
Anthropogenic activities in the Salish Sea have the potential to
negatively impact sand lance populations and habitat, which could
result in cascading effects (Staudinger, 2020). As a species with an
obligate association to very specific, rare habitats, sand lance are
particularly exposed to impacts from habitat loss and degradation
(Pearson et al., 1984;Quinn, 1999;Robinson et al., 2013;Bizzarro
et al., 2016;Huard et al., 2022;Smith and Liedtke, 2022).
While many threats have been documented individually,
including the vulnerability of sand lance to climate change (Hare
et al., 2016;Rovellini et al., 2024), there is no comprehensive list of
threats acting on sand lance. These data deficiencies make it
challenging to understand pressures on sand lance, and to
prioritize management actions to abate threats. Evaluating the key
threats would help to understand drivers of decline and inform
decisions to conserve and manage sand lance.
In this study, we sought to overcome these challenges through the
use of expert judgement to identify a comprehensive list of threats to
their persistence, including a ranking of the vulnerability of sand lance
to those threats, and to develop a hypothesis for Salish Sea sand lance
population trends. Structured expert elicitation is routinely applied in
conservation and decision-making contexts, when data are incomplete
or unavailable, and time and resources to collect such data are limited
(Wolfson et al., 1996;Harwood, 2000;Wilson et al., 2005;De Lange
et al., 2010;Martin et al., 2012;Burgman et al., 2015;The Salish Sea
Pacific Herring Assessment and Management Strategy Team, 2018).
Structured approachesare designed to facilitate the elicitation of expert
judgement in such a way that common biases are mitigated, and the
resulting data conform to the same level of empirical control and
transparency afforded to other forms of empirical data and have been
shown to improve the accuracy and calibration of results (Hemming
et al., 2018;Camaclang et al., 2021;Hanea et al., 2022).
Ranking the potential impact of threats requires a clear definition
of each and connecting such impacts with putative individual or
population-level effects across a range of exposure levels (O’Hagan,
2019). Direct ranking methods are often expedient, but can over-rely
on the opinions of experts, and may result in availability biases, with
threats that are more easily recalled, or more familiar to experts, being
listed higher than other threats (e.g., Donlan et al., 2010). Indirect
methods, by contrast, identify explicit criteria for assessing the
magnitude of any threat, and can be more time consuming, but may
guard against availability bias. In ecology, vulnerability is often
categorized into exposure, sensitivity, and resilience. However; there
are a number ofdifferentapproaches (Van Straalen, 1993;Turner et al.,
2003;Wilson et al., 2005;De Lange et al., 2010;Speirs-Bridge et al.,
2010;Beroya-Eitner, 2016;Berrouet et al., 2018;Hou et al., 2022)
requiring a choice to be made about the best criteria to represent
vulnerability for the species and ecosystems in question. We used both
indirect and direct ranking methods to help order the list of threats,
choosing to adapt the methods and vulnerability scores developed by
Halpern et a l. (2007) and Tecket al. (2010); and repeated by Grech et al.
(2011) and Kappel et al.(2012) as applied in the marinerealm to the top
ranked threats of concern. The focus of our study was to expand the
knowledge of sand lance and their habitats to support ecosystem
conservation. Specifically, we aimed to:
1. Create a list of threats to sand lance habitats and populations;
2. Rank the threats through a vulnerability assessment to
identify those with the highest potential to reduce
populations and/or their critically required habitat; and
3. Generate a hypothesis on the population trajectory of
sand lance.
Methods
We applied the IDEA protocol for structured expert elicitation
(Investigate, Discuss, Estimate, and Aggregate); (Hanea et al., 2018;
Hemming et al., 2018). We began our assessment of threats to sand
lance with a literature review to identify a list of threats. We then
supplemented this list through an expert elicitation. This aimed to
overcome biases in the literature due to publication biases, such as
demographic biases, as well as lags in the literature whereby
emerging threats may not have been documented or sufficiently
studied for publication (Baum and Martin, 2018).
This research was undertaken under the University of British
Columbia Human Ethics H19-01635 for the Salish Sea Cumulative
Threats project, led by Dr. Tara Martin. The expert elicitation took
place from January –April 2021, using remote elicitation. To address
our questions, we administered three sequential surveys (Surveys
were delivered through Qualtrics [https://www.qualtrics.com/]),
through following the workflow of:
•Recruitment of experts.
•Survey 1: Initial review of threats and definitions.
•Videoconference: Provide feedback on Survey 1 and
introduce process for Surveys 2 and 3.
•Survey 2: Initial estimate of population size and
vulnerability assessment of threats.
•Survey 3: Review of Survey 2 results and finalize estimates.
Recruitment of experts
Species and/or habitat experts were identified based on
literature searches (including grey literature), word-of-mouth, and
Huard et al. 10.3389/fmars.2024.1445215
Frontiers in Marine Science frontiersin.org03

personal experience of the authors. A snowball technique was also
used where invited participants were asked to forward the
participation request to other potential knowledge holders.
Anyone who self-identified as being knowledgeable on sand lance
or on habitats in the Salish Sea was invited to participate. Over 80
invitations were extended, invitations (via e-mail) can be found in
the Supplementary Material.
Participants were from federal government agencies (28.5%,
11), not-for-profit organizations or non-governmental
organizations (23%, 7), academic institutions (20.5%, 2),
consulting firms (7.8%, 1), First Nation governments (staff not
necessarily Indigenous identifying individuals) (2.6%, 1), First
Nation non-governmental organizations (staff not necessarily
Indigenous identifying individuals) (2.6%, 1), and Other (5.1%, 2).
Respondents identified as men (39.5%, 15), women (44.7%, 17), and
prefer not to answer (15.8%, 7). Participants ranged in age from 25
to 75 years old, with a majority falling into the 25 to 34, and 35 to 44
years old bands. Four participants [10.3%] chose not to respond to
the age demographic question. Respondents identified as White
(70.5%), Hispanic or Latino (2.6%), and prefer not to answer
(26.9%). When asked about their type of expertise, 71.8% of
respondents identified having knowledge of sand lance; 10%
identified having knowledge of habitats; 5.4% selected both of the
previous two categories; and 12.8% identified having related but
other knowledge (such as knowledge of forage fishes in general in
the Salish Sea, and/or knowledge of species that rely on sand lance).
When asked about geographical area of expertise, 64.1% identified
as being most familiar with the Canadian Salish Sea (i.e., British
Columbia), 25.6% identified as being most familiar with the
American Salish Sea (i.e., Washington State), 7.7% identified as
being familiar with the entire Salish Sea, and 2.6% identified as not
having familiarity with the Salish Sea but feeling they were able to
provide insight to threats regardless of location.
Survey 1 –threat identification
An initial list of 20 potential threats to sand lance and/or their
habitats was developed from a literature review (see Supplementary
Table S1 for initial threat list). Each threat was accompanied by a
definition of the threat. This list was reviewed by a small group of
five experts to further reduce linguistic ambiguity and compiled into
Survey 1 for review by 39 experts. In Survey 1 experts were first
asked to review the initial list of threats and, if desired, provide and
describe up to five additional threats. They were then asked to categorize
the threats into one of three broad tiers of impact to sand lance: Of
Concern,Least Concern,andUncertain. Within each of the three tiers,
experts were asked to directly rank threats in order of greatest to least risk,
with no ties permitted, to the persistence of sand lance and/or to the
persistence of their habitats in the Salish Sea. This was used to help
generate a Cumulative Rank Score (CRS) across experts and threats (See
Supplementary Material for calculation methods).
A 1-hour video conference was held on January 23, 2022.
During the video conference, participants were formally
introduced to the project and reviewed and discussed the 20
threats of concern. They were then informed of the process for
completing Surveys 2 and 3 and invited to raise any questions or
concerns. Following the video call, the list of threats to include in
Survey 2 was reorganized, reworded, and in some cases expanded or
collapsed based on the discussions.
Survey 2: vulnerability assessment and
expert population trajectory assessment
Survey 2 was split into two parts: i) an assessment of the
population trajectory; and ii) the relative quantification of
vulnerability scores for the top threats.
Vulnerability assessment
For the vulnerability assessment (Survey 2), we sought to reduce
the number of threats provided from Survey 1 to avoid expert
fatigue. To determine the list of threats to be assessed, we ranked the
42 threats that received at least one vote as being a threat Of
Concern by the CRS score (See Supplementary Table S3). Following
the survey, the top 20 threats categorized as being Of Concern were
utilized (see Supplementary Material for specific methods used
Supplementary Table S2). The final 20 threats under evaluation
were divided into two groups (A and B) each with 10 threats.
Participants were then assigned to either group based on the threats
they indicated had the most familiarity with as indicated in Survey
1. Group A had 22 participants assigned to it and Group B had 21. A
copy of Survey 2 can be found in the Supplementary Material S2.
Experts were also asked to characterize the level of certainty for
their responses to a given threat and allowed space to provide
caveats and qualifiers to contextualize their responses. The specific
questions asked in the survey are shown in Table 1.
As per Halpern et al. (2007) and Teck et al. (2010), experts were
asked to consider the vulnerability criteria for each threat on an
annual basis, and assign a relative value (0-4 or 0-6) using a
TABLE 1 Vulnerability measures assessed in Survey 2 in the expert
elicitation of threats to sand lance and marine sand habitats in the
Salish Sea.
Vulnerability
Measure
Specific Question asked
Frequency What is the Cumulative Occurrence?
How many times per year (on average) do events
associated with this threat occur?
Area
(Physical Scale)
What is the Spatial scale (km
2
) of each event?
How much area does each event (e.g., one sea wall, on
average) cover?
Duration What is the Duration?
How long does each event last? Including any possible
construction plus general existence (where applicable).
Resistance What is the Resistance?
Over the next 25 years will the habitat/population in the
Salish Sea resist changing from its ‘natural’state in
response to this threat?
Certainty Certainty (very high to none)
How much certainty do you have on the
previous questions?
Huard et al. 10.3389/fmars.2024.1445215
Frontiers in Marine Science frontiersin.org04

predefined drop-down list (Table 2). Data from expert responses
were used to develop a weighted average vulnerability score to
represent a relative measure of how vulnerable sand lance and their
habitats are to each threat. This approach to calculating relative
vulnerability of sand lance is mathematically represented as
Vulnerability(threat,i)= o
k=1,…4
WkSi,k
where S
i,k
is the value of threat iand a vulnerability measure k,
and W is the weight assigned to vulnerability measure k, such that
W
k≥0
,∑
k=1,…,4
W
k
=1. The weights are normalized so that they sum
to one and were empirically derived using a multicriteria decision
model. Resistance was estimated to explain 66.5% of the
vulnerability (Teck et al., 2010), when experts were assessing the
vulnerability of various scenarios with pre-determined values and,
therefore, we used 0.665 for the weighting of resistance. For the
weighting of the three scale-related criteria, we assumed that
frequency, area, and duration were of equal weights and used 1 –
0.665 = 0.334 divided by 3 to give each a weight of 0.112.
Frequency, area, and duration vulnerability measure results
were standardized to have the same 0 to 4 scale (by multiplying
the vulnerability value by 4/6) so they are comparable and were all
given equal weight to resistance. Then, for each expert’s
vulnerability score, the values were divided by the highest possible
score (4) to give a value between 0 and 1. Values closer to 1
represent higher estimated vulnerability to that threat and values
TABLE 2 Ranking system for vulnerability measures to assess how threats affect sand lance and/or coarse, silt- free sand, habitats.
Vulnerability Measure Category Rank - Description/examples
Cumulative Frequency
How many times, per year, (on
average) do these events occur?
(e.g., how many sea walls are
installed every year)?
Never occurs
Occurs, but rare
Annually
Occasional
Regular
Often
Persistent
0 - Never observed, or observed from a low probability chance even e.g., < 1 in 500-year event.
1 - Irregular and/or sporadic: Less than once/year
2 - At least once a year
3-2–10 times a year, could occur seasonally
4 - Frequent, could be seasonal: >10 times a year, or once per month
5 - Common: >120 times a year, at least 10 or more times a month
6 - Close to or over >300 times a year, could be daily/constant
NA - Not applicable to this threat
DNK - Do not personally know, can’t provide even a guess
Spatial scale (km
2
) of each
individual
threat event
(e.g., A single port project, not all
Port Developments together)
Single, small beach
Large beach
Several beaches
Region
Entire basin
Entire Salish Sea
0 - Does not physically occur
1 - <1 km
2
2-1–10 km
2
(of beaches/populations)
3-10–100 km
2
(beaches/populations)
4 - 100 –1,000 km
2
(e.g. Courtney/Comox, Burrard Inlet)
5 - 1,000 –10,000 km
2
(e.g., Strait of Georgia, Haro Strait)
6 - >10,000 km
2
NA - Not applicable to this threat
DNK - Do not personally know, can’t provide even a guess
Duration of
Impact by each individual
threat event
(e.g.,
sea wall on one
property)
0 - None, or near instantaneous
1 - <1 day
2-1day–1 month
3 - 1 month –1 year
4-1–10 years
5-10–20 years
6 - >20 years/Permanent
NA - Not applicable to this threat
DNK - Do not personally know, can’t provide even a guess
Resistance
(of habitat or population,
not individuals)
Complete
High
Moderate
Low
None
0 - Sand lance populations/individuals or habitat do not experience any change in the presence of this
threat
1 - Sand lance populations/individuals or habitat do not change very much in response to this threat
2 - Detectable negative changes occur that impact the persistence of Sand lance populations/individuals
or habitat
3 - Sand lance populations/individuals or habitat are sensitive to this threat and the slightest occurrence
will causes a significant change to the persistence of the habitat or species
4 - Sand lance populations/individuals or habitat experience significant major changes from this threat.
It could be all or nothing
NA - Not applicable to this threat
DNK - Do not personally know, can’t provide even a guess
Certainty of participant response Very High
High
Medium
Low
None
0 - Extensive empirical work/local knowledge exists or the participant has extensive personal experience
1 - Body of empirical work/local knowledge exists or the participant has direct personal experience
2 - Some empirical work/local knowledge exists or participant has some personal experience
3 - Very little empirical work/local knowledge exists
4 - No knowledge on this threat exists
NA - Not applicable to this threat
DNK - Do not personally know, can’t provide even a guess
Huard et al. 10.3389/fmars.2024.1445215
Frontiers in Marine Science frontiersin.org05

closer to 0, lower vulnerability. For each threat, the mean and
confidence intervals across the experts were calculated to provide a
single weighted vulnerability score. The mean and confidence
intervals were calculated using the DescTools package in R,
‘meanCI’function using bootstrapping methods. The confidence
intervals are the 2.5
th
and 97.5
th
centiles, or a 95% confidence
interval of an equi-tailed, two-sided, nonparametric interval using a
basic bootstrap interval.
Population trajectory assessment
There were concerns that quantifying the trajectory of sand
lance may be too onerous for the experts to accurately assess given
the species’cryptic nature, boom-and-bust episodic reproductive
cycles, and the lack of a stock assessment. To overcome these issues,
the team elicited population trajectory predictions using questions
of increasing resolution. Experts were asked to consider a
hypothetical survey program that had averaged 100 fish per
survey each year over the previous five years. Experts were then
asked to consider if the same population in the Salish Sea was
evaluated 25 years from present (year 2047) would there be an
increase, decrease, or no change relative to the current five-year
average. The timeframe of 25 years was used because it encompasses
multiple generations of the sand lance life cycle and is within the
realm of experience that can be reasonably predicted by expert
participants. This question was accompanied by a visual image of
potential changes in sand lance (Figure 1). The purpose of this
framing was to convey population changes of sand lance in a
hypothetically observable and meaningful quantity for experts (a
key criterion for structured expert elicitation questions [Hemming
et al., 2018]).
Finally, experts were asked to quantify their estimates by
providing their: (1) high estimate; (2) low estimate; and (3) best
estimate of the relative catch of sand lance from a monitoring
program in 25 years, as compared to today (Figure 1). Experts were
also asked to provide an estimate between 50% and 100% for the
confidence level associated with their intervals (Speirs-Bridge et al.,
FIGURE 1
Information given to participants to help estimate the relative change in sand lance population trajectory provided to experts in Survey 2.
Huard et al. 10.3389/fmars.2024.1445215
Frontiers in Marine Science frontiersin.org06

2010). To compare between experts, the lower and upper bounds
were standardized to 90% credible intervals (Hemming et al., 2018).
Survey 3: review
Survey 3 provided an opportunity to review the results of the 2
nd
Survey, including the comments and allow experts to finalize their
estimates, adjusting if needed. Twenty-one experts completed Survey 3.
We used the results of Survey 3 to calculate the final relative vulnerability
assessments and as the final expert estimate of the population trajectory.
Results
Survey 1 –threat identification
Experts recommended 0 to 5 additional threats each. After
accounting for duplicates, 34 additional threats were provided by
the experts (Supplementary Table S2). When asked to broadly
categorize the list of threats, participants placed as few as one and
up to 23 in the Of Concern category (average of 10), zero to nine in
the Least Concern category (average of three), and zero to 11 in the
Uncertain category (average of seven). Climate change, followed by
shoreline armoring were the most voted-for threats in the Of
Concern category, with 31 and 30 total votes, respectively
(Table 3). Recreational boating (14), commercial fishing (12), and
recreational sites (11) were listed the most for threats of Least
Concern. Under the Uncertain category, geoduck harvesting
received the most votes (21), followed closely by renewable
energy and construction (20), and commercial ship anchorages
(19). Other threats with considerable uncertainty votes included
major shipping and port developments (15), aquaculture (15), and
freshwater dams (15). There were 19 threats that only received one
vote, and most were additional threats provided by an expert and
not seen by the remaining experts. The only way such a potential
threat could have received more than one vote would be for two
experts to identify the same threat independently.
The Cumulative Rank Scores (CRS) typically led to the same
ranking as the number of votes a given threat received (Table 3). For
example, climate change, followed by shoreline armoring, again
scored the highest in the total CRS, with scores of 642 and 608,
respectively. Some exceptions were: accidental spills (received 27
votes with a CRS of 479), and dredging (received 29 votes with a
CRS of 473). However, there were only three threats with a lower
CRS rank compared to the direct rank.
In the comments section, many participants addressed their
lack of certainty with the threats and difficulty in ranking them.
Several commented on how the spatiotemporal nature of a
particular threat impacted their ranking choices, while others
noted how their familiarity with a particular threat may have
driven their rankings. Others commented they were not able to
rank the threats without additional information given differences in
scale among threats. Several participants with geology backgrounds
pointed out that natural impacts such as earthquakes or shifts in
geophysical processes may also have impacts on sand lance habitats.
Several participants commented on confusion or specific concerns
they had with certain threats, or how the threats were categorized.
These comments were used to refine the final threats list and
description of threats for Survey 2 (Tables 3,4).
Survey 2 and 3: vulnerability assessment
and expert population trajectory
Vulnerability assessment
The vulnerability assessment was completed by 30 experts;
however, 5 abstained from providing estimates. Group A had 18
experts, and Group B had 12. In Survey 3, one expert altered just
one of their responses to the vulnerability matrix table. This expert
changed their response regarding the duration of shoreline
armoring from (4) 1-10 years, to a higher value of (5) 10-20
years, changing the overall average from 5.78 to 5.8. The results
presented are the final judgements by participants. The average of
expert’s vulnerability score, the order of the relative vulnerability,
and the most imminent threats scores were estimated to be (in order
of highest to least): sea level rise, sea temperature rise, extreme
weather, ocean acidification, decreased sediment loads from
freshwater sources, and shoreline armoring (Figure 2).
Aquaculture of kelp, salmon, and shellfish had the lowest scores.
The vulnerability scores were then disaggregated into area,
duration, and occurrence (Figure 3); resistance (Figure 2); and
certainty (Figure 4) and ordered according to the vulnerability
scores, from highest to least (Figure 5). The resistance values
roughly approximate the vulnerability scores, likely because of the
higher weight this criterion was given relative to other criteria.
However, duration, area, and frequency contribute to variability in
these rankings. For example, climate change threats made up four of
the top five threats in the overall relative vulnerability scores
(Figure 5), but they were not always the top scoring resistance
values (Figure 2). On average, experts thought sand lance are
betterfrmoval) had the lowest median resistance value (3.2,
Figure 2). Dredging had relatively low occurrence values, at least
once per year (1.78, Figure 3A), meaning experts think sand lance
and sand habitat are poorly able to cope with the impacts of
dredging; however, they also thought this threat does not occur as
often as other threats. At the other end of the threats assessed, kelp
aquaculture had the lowest relative vulnerability score, and the
lowest resistance values (0.6, Figure 2), but the highest occurrence
value (6, Figure 3A), with moderate duration and area values.
Certainty values varied little among threats, approximately 2, and
made very little difference to the rankings of threats.
Population assessment
Only five participants (of 30) declined to estimate the population
trajectory (declining to answer each of three population questions).
The results of the remaining 25 experts suggest that, on average, sand
lance are expected to decline to 63% [95%CI: 20.1 –185.5] of today’s
population under a ‘business as usual’scenario (Figure 6). While the
upper and lower bound on thisestimate exhibited high uncertainty, we
note that only one participant (3%) estimated the population to
increase, and three (9%) estimated it would remain the same. Many
Huard et al. 10.3389/fmars.2024.1445215
Frontiers in Marine Science frontiersin.org07

experts (40%) commented that they based their population projection
estimate onassumptions that are supported in the literature, including
anecdotal observations that suitable shoreline habitats are increasingly
impacted or lost, threat intensity and occurrence frequencies appear to
be increasing, extirpation and extinctions of other species in the Salish
Sea are occurring at a rapid rate, and that ecosystem processes are
decaying (Pimm et al., 2014;Dı

az et al., 2019;Chase et al., 2020;Arimitsu
et al., 2021;Laubenstein et al., 2023). The single expert that estimated a
population increase cited that sand lance are a resistant species despite
the habitat damage and loss, and changing climate conditions.
Three participants made changes in the follow-up survey after
discussing initial results with other experts. There were two
participants that changed their population trajectory assessment
and widened their bounds (decreased their certainty), one also
lowered their best guess. These changes made the overall average
slightly lower and the bounds slightly wider.
Discussion
We present the first Salish Sea-wide evaluation of expert-based
sand lance population trajectory, as well as the most comprehensive
threat and vulnerability assessment conducted for the species
to date.
TABLE 3 Collated results of Survey 1 threat ranking showing top 20 threats identified by experts as of concern including cumulative rank score, total
votes given (number of times listed under ‘threat of concern’category), number of times listed as Least concern, and number of times listed
as Uncertain.
Threats Cumulative
Rank
Score (CRS)
No. of votes
for Rank 1
No. of times Listed as
‘Threat of concern’
No. of times Listed
as Least Concern
No. of times
Listed
as Uncertain
Climate change 642 13 31 0 6
Shoreline armoring 608 9 30 0 4
Pollution, Accidental spills 479 4 27 2 3
Dredging (Sediment Removal
and dumping)
473 1 29 2 2
Increase in sediment loads 401 1 23 1 9
Pollution, Wastewater 362 1 21 2 8
Major shipping port
& developments
356 2 22 4 5
Pollution, Microplastics 222 0 14 0 15
Recreational sites 190 0 12 11 10
Aquaculture (salmon, shellfish,
& kelp)
182 0 12 3 15
Shipping, Commercial
ship anchorages
106 0 8 6 19
Recreational boating 133 0 11 14 7
Freshwater dams (Decrease
in sediment)
84 0 6 11 15
Aquaculture,
Geoduck harvesting
95 0 7 5 21
Loss of riparian zone 74 0 4 0 0
Shipping, Increased ship traffic47 0 3 0 0
Climate change-related shifts in
prey phenology*
41 0 3 0 1
Shoreline development* 21 0 1 0 0
Deoxygenation specificto
Climate change*
21 0 1 0 0
Pollution, Water quality* 20 0 1 0 0
*Indicates a threat was provided as an additional threat by a participant.
Huard et al. 10.3389/fmars.2024.1445215
Frontiers in Marine Science frontiersin.org08

TABLE 4 Threat list developed through results of Survey 1 and the definitions provided to participants for Survey 2.
Threat List
1 Aquaculture: Kelp –The breeding, rearing, and harvesting of kelp typically for commercial purposes within marine waters usually from floating lines. A single event
includes a single operation/farm.
2 Aquaculture: Salmon –The breeding, rearing, and harvesting of Atlantic salmon typically for commercial purposes in open net pens within marine waters usually
from floats anchored to shore and to the seafloor. A single event includes a single operation/farm.
3 Aquaculture: Shellfish –The breeding, rearing, and harvesting of shellfish typically for commercial purposes such as oysters, scallops, and mussels within marine
waters usually from floating rafts and lines. A single event includes a single operation/farm.
4 Beach Recreation –People spending time at beaches, walking dogs, or swimming, usually for recreational reasons. Includes all land ownership categories such as
parks, reserves, Indigenous lands, private land, and public lands. A single event includes any recreational use (at an intensity) at a single beach.
5 Climate Change: Extreme weather –Extreme precipitation, or lack of it and/or wind events or major shifts in seasonality of storms as a result of long-term climatic
changes. Includes thunderstorms, droughts, atmospheric rivers, tornados, hailstorms, ice storms or blizzards, dust storms, and erosion of beaches during storms. A
single event includes a single above average storm, flood, or other weather event.
6 Climate Change: Sea level rise –The rise of average sea level, resulting in ‘‘coastal squeeze’as a result of long-term climatic changes that may be linked to climate
change and other severe climatic or weather events outside the natural range of variation. A single event includes the incremental rise of average water level 1mm or
more that remains after a year i.e., not tidal changes.
7 Climate Change: Sea temperature rise –The global increase in sea surface temperatures at a rate of approximately 1.1C from 1971 to 2010, and similar results have
been observed in the Salish Sea Region. It is related to increasing atmospheric temperatures, and increased carbon dioxide and other gases and has cascading effects
including those on oxygen availability, and trophic relationships (prey availability). A single event includes heat waves, major oceanic temperature changes (e.g. The
Blob), that are outside the normal range (encompassing El Nino event), and/or the incremental rise of average ocean water temperature 1°C or more that remains
after a year.
8 Climate Change: Ocean acidification –The ongoing decrease in the pH value of the Earth’s oceans, caused by the uptake of carbon dioxide (CO
2
) from the
atmosphere. A single event includes the incremental decrease of average ocean pH of 0.001 per year.
9 Dredging: Sediment dumping –The release of dredge material (silt, sediments, and other benthic material), often in the tons, typically at designated deep-water sites.
A single event includes the dumping of accumulated dredged sediments.
10 Dredging: Sediment removal –The removal of silt, sediments and other benthic material from the bottom of bodies of marine or estuarine waters for any purpose
but often occurring at marinas, in high traffic channels, ports, and under overhead features like bridges. A single event includes any single occurrence/project to
dredge a specific area (e.g. channels, marinas, ports etc.).
11 Pollution: Accidental spills –The unintentional release of any substance, naturally occurring or otherwise. Includes any liquid, solid, or gas such as chemicals, oil,
crude, petroleum product, etc. from ships, pipelines, and any other location. A single event includes the accidental release, spill, or dumping of a product/products
from one location and one time.
12 Pollution: Microplastics –Particles 1 to 5,000 mm, including spheres, fragments, and fibers resulting from the deterioration of larger plastics fragment into ever-
smaller debris over time, eventually becoming nanoplastics (<1 mm), as well as intentionally manufactured microbeads. A single event includes the incremental
annual increase in average microplastics content contributed each year (estimated at 4.8 to 12.7 million tons per year and expected to increase tenfold by 2025).
13 Pollution: Non-point source –Water or snowmelt that moves over or through the landscape picking up pollutants, eventually depositing them into the marine
environment. Examples include fertilizers and nutrients from lawns, golf courses, and agriculture, oil or sediment from roads, and contaminated sediments. A single
event for includes any event or action that leads to release of non-point source pollutants into the marine environment.
14 Pollution: Urban wastewater –Discharge from municipal waste treatment plants, leaking septic systems, untreated sewage, road salt, and the effluent from industrial
and commercial facilities. Water-borne sewage and nonpoint runoff from housing and urban areas that include nutrients, toxic chemicals and/or sediments. A single
event includes any urban site (private, public, commercial, or industrial), occurrence, project, or site releasing waste products into the environment that enter the
marine environment.
15 Port activities and developments –The construction, presence, and daily activities of shipping terminals including legacy sites (existing active or inactive ports) as
well as any proposed/in construction (e.g., Robert’s Bank Terminal 2). Includes any industrial marine foreshore facility/activity in the Salish Sea including large
major ports, such as the Port of Metro Vancouver, Port of Everett, Port of Seattle, Port Angelas, Port of Nanaimo, as well as smaller ports such as the Portof
Bellingham, Squamish Terminal, Ogden Point Terminal, and so on. A single event includes a single port site or proposed site.
16 Riparian area loss or removal –Removal of vegetation immediately along the foreshore that often provides shade and erosion protection, and sometimes overhangs
the intertidal zone. A single event includes the loss/remove of vegetation from one property.
17 Sediment decrease/reduction –A decrease in sediment delivery to coastal systems from human actions such as freshwater dams, and river diversion. A single event
for includes any event that leads to a decrease in sediments available.
18 Sediment Increase –An increase in sediment delivery to coastal systems from human actions such as land-clearing and deforestation. This threat could also be
related to climate change as terrestrial temperatures, precipitation, glacial melting, slide frequencies, forest fire, snow depth, etc. patterns change resulting in changes
to flooding/freshet patterns. A single event includes any event that leads to the accumulation of sediments.
19 Shipping –The presence of, daily activities associated with, and movement of large vessels usually for commercial/industrial purposed but may also include very
large private vessels. This includes all anchoring activities, noise, light pollution, and wake associated with these vessels. This does not include any pollutants,
contaminants, or spills. A single event includes the presence of any one large vessel in the Salish Sea.
(Continued)
Huard et al. 10.3389/fmars.2024.1445215
Frontiers in Marine Science frontiersin.org09

Threats
The vulnerability assessment gathered experts from a wide
range for backgrounds to carefully garner (1) a comprehensive list
of threats to Salish Sea sand lance (Supplementary Material S1),
(2) a list of the most concerning threats, and (3) prioritization of
these top threats (Figures 2–6). The results indicate that sand lance
are considered most vulnerable to climate change, which was
thought to be a major threat to sandy beach habitats since the
early 2000s (Brown and McLachlan, 2002). There is some existing
research that specifically supports the notion that climate change
poses a significant threat to sand lance populations. For example, a
number of studies assessing sand lance mean body weight, size,
condition, and fat content all negatively correlate with periods of
increased temperatures (e.g., marine heat waves, Robards et al.,
2002;Hipfner et al., 2018;von Biela et al., 2019;Arimitsu et al.,
2021;Robinson et al., 2023). In addition, lab research on ocean
acidification impacts to Ammodytes dubius, a species within the
same genus occurring in the Northwest Atlantic, showed A. dubius
eggs are highly sensitive to changes in CO
2
at levels within the range
of expected climate change values (Murray et al., 2019;Baumann
et al., 2022). Finally, sea level rise is also projected to influence tidal
currents, which may lead to the erosion and loss of sand,
endangering rare shoreline and subtidal sand lance habitats
(Healy, 1996;Greene et al., 2017; Greene et al., 2021).
While mitigation measures for climate change require national
and global scale efforts, other threats (e.g., sediment loads, shoreline
armoring and pollution) can be abated through local regulatory
and/pr policy decisions of the Salish Sea (e.g., those approving
permits for development or dredging). Shoreline armoring was
among the top-ranked threats in the initial survey and has been of
high concern amongst conservation groups since the early 2000s (de
Graaf, 2010;Hart, 2010;de Graaf, 2014,2017). Of the top threats,
changes to sediment loads (e.g. such as that from dams or shoreline
development) and shoreline armoring may be easier to address
relative to the multifaceted, global nature of climate change
(Hornsey and Fielding, 2019;Habel et al., 2020;Toft et al., 2021).
Addressing local impacts, such as damage or loss of crucial
spawning habitat, would contribute to offsetting complex threats
such as sea level rise, sea temperature rise, pollution, and marine
riparian area loss. There is mounting evidence of the benefits of dam
removal which returns natural sediment regimes (Frick et al., 2022;
Shaffer et al., 2023), avoiding shoreline armoring in favor of
restoring and maintaining the natural capacity of nature to buffer
adverse impacts using natural shoreline designs (Brown and
McLachlan, 2002;Gittman et al., 2016;Martin and Watson,
2016), and avoiding dredging (Wenger et al., 2017).
While our assessment focused on threats to sand lance, we note
that many of the top-ranked threats have been previously
highlighted as key issues for other species and across ecosystems
(Halpern et al., 2007;Crain et al., 2009;Defeo et al., 2009;Teck et al.,
2010;Gaydos et al., 2015). They are also likely to be applicable to
species that use the same habitats as sand lance, and for the
predators that depend on them (Beaudreau and Essington, 2007;
Defeo et al., 2009;Page et al., 2011;The Salish Sea Pacific Herring
Assessment and Management Strategy Team, 2018;Smith and
Liedtke, 2022). Therefore, applying a threat-based approach to
management of sand lance habitats, such as a priority threat
management plan (Martin et al., 2018), could provide an efficient
means of safeguarding multiple species facing the same threats.
The threat assessment revealed that there is still much
uncertainty about many threats, with 28 threats grouped as
uncertain, and differences among experts in how they ranked
threats. This is not surprising given the lack of research on this
species and habitat. The uncertainty in the vulnerability matrix and
comments provided by experts identifies collective gaps for future
research (Supplementary Materials, Survey 2 Results Summary).
While the threat assessment revealed many potential threats acting
TABLE 4 Continued
Threat List
20 Shoreline armoring –The installation of any hard structure at or below the high tide line that interacts with natural sediment movement into or through the
intertidal zone. Examples include retaining walls, groynes, road armoring, rip rap, dykes, pipeline outfalls, placing of fill, pilings, dock installations, and seawalls. This
does not include pollution/contaminants, noise, or light. A single event includes a single sea wall built on one property, or for one project.
Threats are listed in alphabetical order.
FIGURE 2
Box plots of vulnerability matrix estimates for sand lance and sand lance habitat resistance to evaluated threats. The median is represented by a line,
the mean is displayed as a black diamond and the value of the mean is given. The color of each threat is consistent throughout, and the order of the
x-axis is based on the overall vulnerability score.
Huard et al. 10.3389/fmars.2024.1445215
Frontiers in Marine Science frontiersin.org10

on sand lance, the average resistance value was ‘moderate’,
suggesting sand lance may be resilient in the face of many threats.
This result was echoed by several expert comments. This may
explain the more optimistic predictions for population trajectory
from some experts. Perhaps the secret to sand lance resilience is
buried in their life history. Sand lance are more dormant
throughout the later fall and winter months buried in sandy
substrates, and, during this period, they regrow gonads in
preparation for annual spawning (van Deurs et al., 2010;Zhukova
and Baker, 2022). Mature sand eels in the North Sea are known to
balance the increased metabolic costs of warm years with reduced
gonad mass, which may buffer the effects of poor environmental
conditions (Wright et al., 2017). Throughout marine heat wave
years, when sea temperatures were above average for extended
periods and zooplankton communities (prey) were altered, sand
lance in Alaska (A. personatus) responded by burying more
(Arimitsu et al., 2021). Researchers suggested that these sand
lance may have adjusted daily and seasonal dormancy periods to
reduce metabolic costs associated with warmer water temperatures,
and subsequently thrived (Arimitsu et al., 2021). While sand lance
may display resistance, their body condition, and thus nutritional
value for predators, significantly declines during periods of warm
ocean conditions, and/or when prey communities are altered
(Litzow, 2000;Robards et al., 2002;von Biela et al., 2019;
Robinson et al., 2023). This double impact of reduced abundance
and nutritional value of sand lance and other forage fishes (e.g.,
FIGURE 3
Box plots of vulnerability matrix estimates for (A) area, (B) duration, and (C) occurrence frequency. The median is represented by a line, the mean is
displayed as a black diamond and the value of the mean is given. The color of each threat is consistent throughout, and the order of the x-axis is
based on the overall vulnerability score.
FIGURE 4
Results of vulnerability matrix estimates for certainty. The median is represented by a line, the mean is displayed as a black diamond and the value of
the mean is given. The color of each threat is consistent throughout, and the order of the x-axis is based on the overall vulnerability score.
Huard et al. 10.3389/fmars.2024.1445215
Frontiers in Marine Science frontiersin.org11

capelin, Pacific herring, Pacific sardine, and northern anchovy)
brought on by the 2014-2016 marine heat wave led to cascading
impacts, shifting distributions, and resulting in large-scale mortality
events of marine predators such as seabirds, marine mammals, and
groundfish (Piatt et al., 2020;Arimitsu et al., 2021). Although sand
lance populations may be resilient to climate change, it is not
without any impact to the wider community.
Population
The experts interviewed in this study collectively suggest that
sand lance populations in the Salish Sea could be 63% of what they
are today (a 37% decline) in 25 years under current management
scenarios. We acknowledge that there is much uncertainty around
this population’s future status, which can be attributed to inherent
biological attributes of forage fishes, a lack of wide scale, regularly
occurring population surveys for sand lance, and a dearth of
information about the direct impacts of specific threats on sand
lance at both the individual and population levels. The expert
projection contradicts the positive trend seen in the forage fish
study across subbasins of Puget Sound using integrated trawl data
(Greene et al., 2015). The authors of that study suggest that the
differences in subbasin forage fish populations may be attributed to
anthropogenic influences.
Our results highlight the need to better understand sand lance
populations. A decline of the magnitude predicted by experts here
would have ecosystem-wide implications particularly to predators
that depend on sand lance during critical times in their life history
(i.e., sea bird breeding) (Bertram and Kaiser, 1993;Bertram et al.,
2001;Beaudreau and Essington, 2007;Gutowsky et al., 2009;
Hipfner et al., 2018;Duguid, 2020). Given this species importance
to the ecosystem and our collective interest in the economic and
cultural values provided by top predators (e.g. salmon, orcas,
lingcod, sea birds, etc.), we recommend the development and
initiation of long-term, cross-border coordinated field-based
population monitoring programs to address the hypothesis
developed here that sand lance populations are in decline.
Method review
An additional benefit from this study is the progressive
application of a method for assessing populations and threats
together, and that these assessments may be applied to other
cryptic species. The process applied here was collaborative,
transboundary, and enabled information to be gathered from
experts at a regional scale. The elicitation was relatively expedient,
with results for 20 threats being obtained in three online surveys
within 10 months. Experts provided 34 new threats that were not
initially identified in the literature. While we did not quantify all
FIGURE 6
Population status estimate results from participants in Survey 2
(N=30, including edits in Survey 2 follow up). The upper and lower
bounds were extrapolated to represent 90% credible intervals.
0%=complete crash, 50%= 50% as many as today, 90% = 90% as
many as today, 100%= the same as today, 110%= 10% more than
today, 200%=twice as many as today, and 500%=five times as many
as today. The average estimate is 63% [95%CI: 20.1 –185.5].
FIGURE 5
Weighted relative vulnerability mean scores of threats assessed to sand lance and their habitats in the Salish Sea, and their associated bootstrapped
95 percent confidence intervals.
Huard et al. 10.3389/fmars.2024.1445215
Frontiers in Marine Science frontiersin.org12

threats with a vulnerability score, our list contains information on
these threats. The framing of the elicitation question for the
population status seemed to enable most experts (25 out of 30) to
provide a best estimate of the expected trajectory of sand lance, and
accompanying credible intervals, a response rate that surprised even
the authors.
There are aspects of the approach applied here that might be
improved for subsequent case studies. There were many decisions
in the method development process, and we deliberated frequently
over the best, or least-biased approach to acquire a truly relative
comparison of the threats. On reflection, an additional initial survey
to gather all the possible threats first, and then in a subsequent and
separate survey to conduct a threat ranking activity would have
allowed all experts to see the additional threats suggested by their
peers. This would take additional time but may have reduced
variability among experts.
Another improvement could be to hold a workshop to elicit
which vulnerability criteria to use in the assessment, which would
help better define criteria. This may lead to a more context-specific
assessment process in which participants are more engaged
(Stelzenmüller et al., 2018). Reassessing criteria weights used in
the vulnerability calculation may also be warranted to have the best
possible weighting specific to this context, area, and species and
provide further expert engagement. Providing additional time for
experts to spend thinking about the vulnerability criteria and the
best threat weighting methodology may lead to better
understanding of this complex concept and their application
when accessing threats.
Defining terms was a difficult and time-demanding task. Even
though attention was given to providing careful definitions, experts
may have interpreted the threats differently, which may have
contributed to some of the highly variable vulnerability scores.
Linguistic uncertainty is a common issue in expert elicitation
(Hemming et al., 2018) and underscores the importance of
allowing experts to review and update their judgements, which
can help to reveal and resolve uncertainties caused by ambiguity.
We faced a trade-off in mental load (how many threats we asked
experts to evaluate) and the quality of their answers (e.g., was 10
threats too many or sufficient)?. One expert objected to leaving a
particular threat out that they thought was important. Increasing
collaboration and opportunities to review and change the list of
threats may have addressed these concerns. To guard against a high
non-response rate, we deliberately had large groups. Given the high
response by experts, it may be possible to have assessed
vulnerability for more threats, or to have split the experts into
smaller groups of 5-8 experts to assess more of the threats.
Finally, many of the top threats identified (Table 4) occur over
long time scales, and have both direct impacts on individuals as well
as indirect impacts on sand lance habitat or food webs. Almost
certainly, these threats interact, overlap, and amplify one another
(Crain et al., 2008;Laubenstein et al., 2023). Although investigating
the interactions was beyond the scope of this study, understanding
if and how threats individually drive population responses and
interact (e.g., acidification and warming; [Crain et al., 2008)] would
help interpret and predict cumulative interactions.
Conclusion
Using expert elicitation, we gathered information on an
understudied and difficult to study species, Pacific sand lance, in
an efficient and low-cost manner. Expert opinion predict a decline
in sand lance abundance in the Salish Sea under a business-as-usual
scenario, highlighting the need for further investigation into Salish
Sea sand lance populations and their persistence. A decline in sand
lance abundance would have major cascading effects throughout the
Pacific coastal ecosystem. The top threats identified include climate
change, nearshore development, and pollution. Addressing
knowledge gaps identified here to improve conservation decisions
is one step toward a more sustainable, sand lance-abundant coast.
Data availability statement
The raw data supporting the conclusions of this article will be
made available by the authors, without undue reservation.
Ethics statement
The studies involving humans were approved by University of
British Columbia Human Ethics H19-01635 for the Salish Sea
Cumulative Threats project. The studies were conducted in
accordance with the local legislation and institutional
requirements. The participants provided their written informed
consent to participate in this study.
Author contributions
JRH: Conceptualization, Data curation, Formal analysis,
Funding acquisition, Investigation, Methodology, Project
administration, Resources, Software, Validation, Visualization,
Writing –original draft, Writing –review & editing. VH:
Conceptualization, Data curation, Methodology, Project
administration, Supervision, Writing –original draft, Writing –
review & editing. MB: Writing –review & editing. JB: Writing –
review & editing. IB: Writing –review & editing. SC: Writing –review
& editing. GD: Writing –review & editing. PD: Writing –review &
editing. VE: Writing –review & editing. JMH: Writing –review
& editing. NH: Writing –review & editing. BK: Writing –review &
editing. DL: Writing –review & editing. RM: Writing –review &
editing. GN: Writing –review & editing. BP: Writing –original draft,
Writing –review & editing. MQ: Writing –review & editing. TQ:
Writing –review & editing. CR: Conceptualization, Data curation,
Investigation, Methodology, Supervision, Writing –original draft,
Huard et al. 10.3389/fmars.2024.1445215
Frontiers in Marine Science frontiersin.org13

Writing –review & editing. ER: Conceptualization, Writing –review
& editing. DS: Writing –review & editing. JS: Writing –review &
editing. AS: Writing –review & editing. NW: Writing –review &
editing. JY: Writing –review & editing. TM: Conceptualization, Data
curation, Formal analysis, Funding acquisition, Investigation,
Methodology, Project administration, Resources, Supervision,
Validation, Writing –original draft, Writing –review & editing.
Funding
The author(s) declare financial support was received for the
research, authorship, and/or publication of this article. This
research was supported by funding from a Mitacs Fellowship
(JRH), NSERC Discovery Grant (TM), Liber Ero Chair in
Conservation (TM) and Reid and Laura Carter.
Acknowledgments
This research is part of a Master’s thesis work by first author,
JRH. The thesis can be found at https://open.library.ubc.ca/soa/
cIRcle/collections/ubctheses/24/items/1.0434260 (DOI: 10.14288/
1.0434260). We would like to thank all the experts interviewed
which included the co-authors and in addition, Theresa Liedtke,
Jenna Cragg, Alanna Vivani, Jennifer Boldt, William D.P. Duguid,
Correigh M. Greene, Haley Tomlin, John Harper, Tanya Prinzing,
Kirk Krueger, Jeremy Maynard, Karen Douglas and Tark Hamilton.
Conflict of interest
Author JS was employed by the company Natural
Resources Consultants.
The remaining authors declare that the research was conducted
in the absence of any commercial or financial relationships that
could be construed as a potential conflict of interest.
Generative AI statement
The author(s) declare that no Generative AI was used in the
creation of this manuscript.
Publisher’s note
All claims expressed in this article are solely those of the authors
and do not necessarily represent those of their affiliated organizations,
or those of the publisher, the editors and the reviewers. Any product
that may be evaluated in this article, or claim that may be made by its
manufacturer, is not guaranteed or endorsed by the publisher.
Supplementary material
The Supplementary Material for this article can be found online
at: https://www.frontiersin.org/articles/10.3389/fmars.2024.1445215/
full#supplementary-material
References
Alheit,J.,andPeck,M.(2019).Driversofdynamicsofsmallpelagicfish resources: biology,
management and human factors. Mar. Ecol. Prog. Ser. 617–618, 1–6.doi: 10.3354/meps12985
Arimitsu, M. L., Piatt, J. F., Hatch, S., Suryan, R. M., Batten, S., Bishop, M. A., et al.
(2021). Heatwave-induced synchrony within forage fish portfolio disrupts energy flow
to top pelagic predators. Global Change Biol. 27, 1859–1878. doi: 10.1111/gcb.15556
Baker, M. R., Greene, H. G., Aschoff, J., Aitoro, E., Bates, E., Hesselroth, D., et al.
(2024). Atlas of benthic habitat for sand lance –application of multibeam acoustics and
directed sampling to identify viable subtidal substrates. Mar. Environ. Res. 202, 106778.
doi: 10.2139/ssrn.4896945
Baker, M. R., Williams, K., Greene, G. H., Aschoff, J., Greufe, C., Lopes, H., and
Towler, R. (2021). Use of manned submersible and autonomous stereo-camera array to
assess forage fish and associated subtidal habitat. Fish. Res. 243, 106067. doi: 10.1016/
j.fishres.2021.106067
Baker, M., Matta, M., Beaulieu, M., Paris, N., Huber, S., Graham, O., et al. (2019).
Intra-seasonal and inter-annual patterns in the demographics of sand lance and
response to environmental drivers in the North Pacific. Mar. Ecol. Prog. Ser. 617–
618, 221–244. doi: 10.3354/meps12897
Baker, M. R., Smeltz, S., Williams, K., Greufe, C., Ewing, M., Chapman, J., et al.
(2023). Diel vertical migration in a pelagic forage fish associated with benthic
substrates. ICES J. Mar. Sci. 80, 1758–1772. doi: 10.1093/icesjms/fsad106
Baum, J. K., and Martin, T. G. (2018). It is time to overcome unconscious bias in
ecology. Nat. Ecol. Evol. 2, 201. doi: 10.1038/s41559-017-0441-y
Baumann, H., Jones, L. F., Murray, C. S., Siedlecki, S. A., Alexander, M., and Cross, E.
L. (2022). Impaired hatching exacerbates the high CO2 sensitivity of embryonic sand
lance Ammodytes dubius. Mar. Ecol.-Prog. Ser. 687, 147–162. doi: 10.3354/meps14010
Beaubier, J., and Hipfner, J. M. (2013). Proximate composition and energy density of
forage fish delivered to rhinoceros auklet cerorhinca monocerata nestlings at Triangle
Island, British Columbia. Mar. Ornithol. 41, 35–39. doi: 10.1111/jfb.12181
Beaudreau, A. H., and Essington, T. E. (2007). Spatial, temporal, and ontogenetic
patterns of predation on rockfishes by lingcod. Trans. Am. Fish. Soc. 136, 1438–1452.
doi: 10.1577/T06-236.1
Beroya-Eitner, M. A. (2016). Ecological vulnerability indicators. Ecol. Indic. 60, 329–
334. doi: 10.1016/j.ecolind.2015.07.001
Berrouet, L. M., MaChado, J., and Villegas-Palacio, C. (2018). Vulnerability of socio
—ecological systems: A conceptual Framework. Ecol. Indic. 84, 632–647. doi: 10.1016/
j.ecolind.2017.07.051
Bertram,D.F.,andKaiser,G.W.(1993).RhinocerosAuklet(Cerorhinca
monocerata) Nestling Diet May Gauge Pacific Sand Lance (Ammodytes hexapterus)
Recruitment. Can. J. Fish. Aquat. Sci. 50, 1908–1915. doi: 10.1139/f93-213
Bertram, D. F., Mackas, D. L., and McKinnell, S. M. (2001). The seasonal cycle
revisited: interannual variation and ecosystem consequences. Prog. Oceanogr. 49, 283–
307. doi: 10.1016/S0079-6611(01)00027-1
Bizzarro, J. J., Peterson, A. N., Blaine, J. M., Balaban, J. P., Greene, H. G., and
Summers, A. P. (2016). Burrowing behavior, habitat, and functional morphology of the
Pacific sand lance (Ammodytes personatus). Fishery Bulletin 114, 445–460. doi: 10.7755/
FB.114.4.7
Blaine, J. (2006). Pacific sand lance (Ammodytes hexapterus) present in the sandwave
field of central San Juan Channel, WA: Abundance, density, maturity, and sediment
association (Friday Harbor Laboratories, University of Washington), 23pp.
Brown, A. C., and McLachlan, A. (2002). Sandy shore ecosystems and the threats
facing them: some predictions for the year 2025. Envir. Conserv. 29, 62–77.
doi: 10.1017/S037689290200005X
Buchanan, M., Lesperance, A., McArdle, A. J., Sandborn, C., and Curran, D. (2019).
Saving Orcas by Protecting Fish Spawning Beaches (Victoria, British Columbia: World
Wildlife Fund). Available online at: http://www.elc.uvic.ca/wordpress/wp-content/
uploads/2019/11/2019-01-11-Saving-Orcas-by-Protecting-Fish-Spawning-Beaches.pdf
(Accessed February 7 2020).
Burgman, M., Jarrad, F., and Main, E. (2015). Decreasing geographic bias in
Conservation Biology. Conserv. Biol. 29, 1255–1256. doi: 10.1111/cobi.12589
Camaclang, A. E., Currie, J., Giles, E., Forbes, G. J., Edge, C. B., Monk, W. A., et al.
(2021). Prioritizing threat management across terrestrial and freshwater realms for
species conservation and recovery. Conserv. Sci. Pract. 3, e300. doi: 10.1111/csp2.300
Huard et al. 10.3389/fmars.2024.1445215
Frontiers in Marine Science frontiersin.org14

Chase, J. M., Blowes, S. A., Knight, T. M., Gerstner, K., and May, F. (2020). Ecosystem
decay exacerbates biodiversity loss with habitat loss. Nature 584, 238–243. doi: 10.1038/
s41586-020-2531-2
Crain, C. M., Halpern, B. S., Beck, M. W., and Kappel, C. V. (2009). Understanding
and managing human threats to the coastal marine environment. Ann. New York Acad.
Sci. 1162, 39–62. doi: 10.1111/j.1749-6632.2009.04496.x
Crain, C. M., Kroeker, K., and Halpern, B. S. (2008). Interactive and cumulative
effects of multiple human stressors in marine systems. Ecol. Lett. 11, 1304–1315.
doi: 10.1111/j.1461-0248.2008.01253.x
Defeo, O., McLachlan, A., Schoeman, D. S., Schlacher, T. A., Dugan, J., Jones, A.,
et al. (2009). Threats to sandy beach ecosystems: A review. Estuarine Coast. Shelf Sci.
81, 1–12. doi: 10.1016/j.ecss.2008.09.022
de Graaf, R. (2010). Preliminary Habitat Assessment for suitability of intertidally
spawning forage fish species, Pacific sand lance (Ammodytes hexapterus) and surf smelt
(Hypomesus pretiosus) Esquimalt Lagoon, Colwood, British Columbia (Victoria, British
Columbia: Capital Regional District). Available online at: https://www.crd.bc.ca/docs/
default-source/es-harbours-pdf/esqu imalt-lagoon/esquimaltlagoon-foragefishsurvey-
degraffe-2010.pdf?sfvrsn=2 (Accessed May 22 2020).
de Graaf, R. (2014). Bowen Island Surf smelt and Pacific sand lance Spawning Habitat
Suitability Assessments (Bowen Island, BC: Bowen Island Municipality, Islands Trust,
Bowen Island Local Trust Committee).
de Graaf, R. (2017). Galiano Island Surf smelt and Pacific sand lance Beach Spawning
Habitat Suitability Assessments (Islands Trust). Available online at: https://islandstrust.
bc.ca/document/galiano-forage-fish-report/ (Accessed March 1 2023).
De Lange, H. J., Sala, S., Vighi, M., and Faber, J. H. (2010). Ecological vulnerability in
risk assessment —A review and perspectives. Sci. Total Environ. 408, 3871–3879.
doi: 10.1016/j.scitotenv.2009.11.009
Dethier, M. N., Raymond, W. W., McBride, A. N., Toft, J. D., Cordell, J. R., Ogston,
A. S., et al. (2016). Multiscale impacts of armoring on Salish Sea shorelines: Evidence
for cumulative and threshold effects. Estuarine Coast. Shelf Sci. 175, 106–117.
doi: 10.1016/j.ecss.2016.03.033
Dı

az, S., Settele, J., Brondı

zio, E. S., Ngo, H. T., Agard, J., Arneth, A., et al. (2019).
Pervasive human-driven decline of life on Earth points to the need for transformative
change. Science 366, eaax3100. doi: 10.1126/science.aax3100
Donlan, C. J., Wingfield, D. K., Crowder, L. B., and Wilcox, C. (2010). Using expert
opinion surveys to rank threats to endangered species: a case study with sea turtles.
Conserv. Biol. 24, 1586–1595. doi: 10.1111/j.1523-1739.2010.01541.x
Duguid, W. (2020). British Columbia Adult Chinook and Coho Salmon Diet Program:
2019 Data Summary Report (Victoria, British Columbia: University of Vi ctoria
Fisheries Ecology and Marine Conservation Laboratory).
Einoder, L. D. (2009). A review of the use of seabirds as indicators in fisheries and
ecosystem management. Fish. Res. 95, 6–13. doi: 10.1016/j.fishres.2008.09.024
Frick, K. E., Kagley, A. N., Fresh, K. L., Samhouri, J. F., Ward, L. S., Stapleton, J. T.,
et al. (2022). Spatiotemporal variation in distribution, size, and relative abundance
within a Salish sea nearshore forage fish community. Mar. Coast. Fish. 14, e10202.
doi: 10.1002/mcf2.10202
Gaydos, J. K., and Pearson, S. F. (2011). Birds and mammals that depend on the
Salish sea: A compilation. Northwest. Nat. 92, 79–94. doi: 10.1898/10-04.1
Gaydos,J.K.,Thixton,S.,andDonatuto,J.(2015).Evaluatingthreatsin
multinational marine ecosystems: A coast Salish first nations and tribal perspective.
PloS One 10, e0144861. doi: 10.1371/journal.pone.0144861
Gittman, R. K., Scyphers, S. B., Smith, C. S., Neylan, I. P., and Grabowski, J. H.
(2016). Ecological consequences of shoreline hardening: A meta-analysis. BioScience
66, 763–773. doi: 10.1093/biosci/biw091
Grech, A., Coles, R., and Marsh, H. (2011). A broad-scale assessment of the risk to
coastal seagrasses from cumulative threats. Mar. Policy 35, 560–567. doi: 10.1016/
j.marpol.2011.03.003
Greene, H. G., Baker, M., and Aschoff, J. (2020). “Chapter 14 - A dynamic bedforms
habitat for the forage fish Pacific sand lance, San Juan Islands, WA, United States,”in
eafloor Geomorphology as Benthic Habitat,2nd ed., vol. S . Eds. P. T. Harris and E.
Baker (Mount Vernon, WA: Elsevier), 267–279. doi: 10.1016/B978-0-12-814960-
7.00014-2
Greene, H. G., Cacchione, D. A., and Hampton, M. A. (2017). Characteristics and
dynamics of a large sub-tidal sand wave field—Habitat for pacificsandlance
(Ammodytes personatus), Salish sea, Washington, USA. Geosciences 7, 107.
doi: 10.3390/geosciences7040107
Greene, C., Kuehne, L., Rice, C., Fresh, K., and Penttila, D. (2015). Forty years of
change in forage fish and jellyfish abundance across greater Puget Sound, Washington
(USA): anthropogenic and climate associations. Mar. Ecol. Prog. Ser. 525, 153–170.
doi: 10.3354/meps11251
Greene, H. G., Baker, M.R., Aschoff, J., and Pacunski, R. (2021). Hazards evaluation
of a valuable vulnerable sand wave field forage fish habitat in the marginal Central
Salish Sea using a submersible. Oceanologia. 65, 1–9. doi: 10.1016/oceano.2121.06.002
Greenstreet, S. P. R., Armstrong, E., Mosegaard, H., Jensen, H., Gibb, I. M., Fraser, H.
M., et al. (2006). Variation in the abundance of sandeels Ammodytes marinus off
southeast Scotland: an evaluation of area-closure fisheries management and stock
abundance assessment methods. ICES J. Mar. Sci. 63, 1530–1550. doi: 10.1016/
j.icesjms.2006.05.009
Greenstreet, S. P. R., Holland, G. J., Guirey, E. J., Armstrong, E., Fraser, H. M., and
Gibb, I. M. (2010). Combining hydroacoustic seabed survey an d grab sampling
techniques to assess “local”sandeel population abundance. ICES J. Mar. Sci. 67, 971–
984. doi: 10.1093/icesjms/fsp292
Gunther, K. M., Baker, M. R., and Aydin, K. Y. (2023). Using predator diets to infer
forage fish distribution and assess responses to climate variability in the eastern Bering
Sea. Mar. Ecol. Prog. Ser. SPF2. 741, 71-99. doi: 10.3354/meps14389
Gutowsky, S., Janssen, M., Arcese, P., Kyser, T., Ethier, D., Wunder, M., et al. (2009).
Concurrent declines in nestling diet quality and reproductive success of a threatened
seabird over 150 years. Endang. Species. Res. 9, 247–254. doi: 10.3354/esr00225
Habel, M., Mechkin, K., Podgorska, K., Saunes, M., Babinski, Z., Chalov, S., et al.
(2020). Dam and reservoir removal projects: a mix of social-ecological trends and cost-
cutting attitudes. Sci. Rep. 10, 19210. doi: 10.1038/s41598-020-76158-3
Halpern, B. S., Selkoe, K. A., Micheli, F., and Kappel, C. V. (2007). Evaluating and
ranking the vulnerability of global marine ecosystems to anthropogenic threats.
Conserv. Biol. 21, 1301–1315. doi: 10.1111/j.1523-1739.2007.00752.x
Hanea, A. M., Hemming, V., and Nane, G. F. (2022). Uncertainty quantification with
experts: present status and research needs. Risk Anal. 42, 254–263. doi: 10.1111/
risa.13718
Hanea, A. M., McBride, M. F., Burgman, M. A., and Wintle, B. C. (2018). The value
of performance weights and discussion in aggregated expert judgments. Risk Anal. 38,
1781–1794. doi: 10.1111/risa.12992
Hare, J. A., Morrison, W. E., Nelson, M. W., Stachura, M. M., Teeters, E. J., Griffis, R.
B., et al. (2016). A vulnerability assessment of fish and invertebrates to climate change
on the northeast U.S. Continental shelf. PloS One 11, e0146756. doi: 10.1371/
journal.pone.0146756
Hart, C. (2010). Marine Shoreline Armoring and Puget Sound; Shorelands and
Environmental Assistance Program. In: Frequently Asked Questions: 8 (Department of
Ecology State of Washington). Available online at: https://apps.ecology.wa.gov/
publications/documents/1006004.pdf (Accessed January 27 2024).
Harvey, C. J., Bartz, J., Davies, T. B., Francis, T. P., Good, A. D., Guerry, B., et al.
(2010). A Mass-balance Model for Evaluating Food Web Structure and Community-
scale Indicators in the Central Basin of Puget Sound (NOAA Technical Memorandum,
NOAA). Available online at: https://repository.library.noaa.gov/view/noaa/3751
(Accessed January 27 2024).
Harwood, J. (2000). Risk assessment and decision analysis in conservation. Biol.
Conserv. 95, 219–226. doi: 10.1016/S0006-3207(00)00036-7
Haynes, T., and Robinson, C. (2011). Re-use of shallow sediment patches by Pacific
sand lance (Ammodytes hexapterus) in Barkley Sound, British Columbia, Canada.
Environ. Biol. Fishes 92, 1–12. doi: 10.1007/s10641-011-9809-z
Haynes, T. B., Robinson, C. K. L., and Dearden, P. (2008). Modelling nearshore
intertidal habitat use of young-of-the-year Pacific sand lance (Ammodytes