Content uploaded by Jean-Noel Druon
Author content
All content in this area was uploaded by Jean-Noel Druon on Apr 21, 2022
Content may be subject to copyright.
Global-Scale Environmental Niche
and Habitat of Blue Shark (Prionace
glauca) by Size and Sex: A Pivotal
Step to Improving Stock
Management
Jean-Noël Druon
1
*, Steven Campana
2
,FredericVandeperre
3,4
,Fa
´bio H. V. Hazin
5
,
Heather Bowlby
6
,RuiCoelho
7,8
, Nuno Queiroz
9,10
, Fabrizio Serena
11
,FranciscoAbascal
12
,
Dimitrios Damalas
13
, Michael Musyl
14
,JonLopez
15
,BarbaraBlock
16
,PedroAfonso
3,4
,
Heidi Dewar
17
, Philippe S. Sabarros
18,19
,BrittanyFinucci
20
, Antonella Zanzi
1
,
Pascal Bach
18,19
,InnaSenina
21
, Fulvio Garibaldi
22
, David W. Sims
23,24
, Joan Navarro
25
,
Pablo Cermeño
26
,AgostinoLeone
18,27
,Guzma
´nDiez
28
,Marı
´
aTeresaCarreo
´n Zapiain
29
,
Michele Deflorio
30
,EvgenyV.Romanov
31
, Armelle Jung
32
,MatthieuLapinski
33
,
Malcolm P. Francis
20
,HumbertoHazin
34
and Paulo Travassos
35
1
Joint Research Centre (JRC), European Commission, Ispra, Italy,
2
Life and Environmental Sciences, University of Iceland,
Reykjavik, Iceland,
3
Okeanos, University of the Azores, Horta, Portugal,
4
Institute of Marine Research (IMAR), University of
the Azores, Horta, Portugal,
5
Departamento de Pesca, Laborato
´rio de Oceanografia Pesqueira, Universidade Federal Rural
de Pernambuco, Recife, Brazil,
6
Fisheries and Oceans Canada, Bedford Institute of Oceanography, Dartmouth, NS, Canada,
7
Department of Sea and Marine Resources, Portuguese Institute for the Ocean and Atmosphere, I.P. (IPMA, IP), Olhão,
Portugal,
8
CCMAR - Center of Marine Sciences of the Algarve, University of Algarve, Faro, Portugal,
9
Centro de
Investigac¸ão em Biodiversidade e Recursos Gene
´ticos (CIBIO), InBIO Laborato
´rio Associado, Universidade do Porto, Vairão,
Portugal,
10
BIOPOLIS Program in Genomics, Biodiversity and Land Planning, Centro de Investigac¸ão em Biodiversidade e
Recursos Gene
´ticos (CIBIO), Vairão, Portugal,
11
National Research Council –Institute of Marine Biological Resources and
Biotechnologies (CNR IRBIM), Mazara del Vallo, Italy,
12
Spanish National Research Council, Spanish Institute of
Oceanography, Santa Cruz de Tenerife, Spain,
13
Hellenic Centre for Marine Research, Institute of Marine Biological
Resources and Inland Waters, Heraklion, Greece,
14
Pelagic Research Group LLC, Honolulu, HI, United States,
15
Ecosystem
and Bycatch Program, Inter-American Tropical Tuna Commission, La Jolla, CA, United States,
16
Hopkins Marine Station,
Stanford University, Monterey, CA, United States,
17
Southwest Fisheries Science Center, National Oceanic and Atmospheric
Administration (NOAA), La Jolla, CA, United States,
18
MARBEC, Univ Montpellier, CNRS, Ifremer, IRD, Sète, France,
19
Institut de Recherche pour le De
´veloppement (IRD), Observatoire des Ecosystèmes Pe
´lagiques Tropicaux Exploite
´s (Ob7),
Sète, France,
20
National Institute of Water and Atmospheric Research (NIWA), Wellington, New Zealand,
21
Marine
Ecosystems Modeling and Monitoring by Satellites (MEMMS), CLS, Space Oceanography Division, Ramonville, France,
22
Department of Earth Sciences, Environmental and Life, University of Genova, Genova, Italy,
23
Marine Biological
Association of the United Kingdom, Plymouth, United Kingdom,
24
Ocean and Earth Science, National Oceanography Centre
Southampton, University of Southampton, Southampton, United Kingdom,
25
Institut de Ciències el Mar (ICM), Consejo
Superior de Investigaciones Científicas (CSIC), Barcelona, Spain,
26
Research and in situ Conservation Department,
Barcelona Zoo, Barcelona, Spain,
27
Department of Biological, Geological and Environmental Sciences, Laboratory of
Genetics and Genomics of Marine Resources and Environment, University of Bologna, Ravenna, Italy,
28
Arrantzuarekiko
Zientzia eta Teknologia Iraskundea (AZTI), Marine Research, Basque Research and Technology Alliance (BRTA), Sukarrieta,
Spain,
29
Laboratorio de Ecologı
´a Pesquera, Unidad B. Facultad de Ciencias Biolo
´gicas, Universidad Auto
´noma de Nuevo
Leo
´n, San Nicola
´s de los Garza, Mexico,
30
Department of Veterinary Medicine, University of Bari Aldo Moro, Valenzano, Italy,
31
Centre technique de recherche et de valorisation des milieux aquatiques (CITEB), Le Port, I
ˆle de La Re
´union, France,
32
Des Requins et Des Hommes (DRDH), Technopole Brest-Iroise, Plouzane
´, France,
33
Association AILERONS, Universite
´
Montpellier 2, Montpellier, France,
34
DCN- Department of Animal Science, Federal Rural University of Semiarid, Mossoro
´,
Brazil,
35
Dept. de Pesca e Aquicultura, Universidade Federal Rural de Pernambuco (UFRPE), Recife, Brazil
Frontiers in Marine Science | www.frontiersin.org April 2022 | Volume 9 | Article 8284121
Edited by:
Michele Thums,
Australian Institute of Marine Science
(AIMS), Australia
Reviewed by:
K. David Hyrenbach,
Hawaii Pacific University,
United States
Charlie Huveneers,
Flinders University, Australia
*Correspondence:
Jean-Noël Druon
jean-noel.druon@ec.europa.eu
Specialty section:
This article was submitted to
Marine Megafauna,
a section of the journal
Frontiers in Marine Science
Received: 03 December 2021
Accepted: 11 March 2022
Published: 21 April 2022
Citation:
Druon J-N, Campana S,
Vandeperre F, Hazin FHV, Bowlby H,
Coelho R, Queiroz N, Serena F,
Abascal F, Damalas D, Musyl M,
Lopez J, Block B, Afonso P, Dewar H,
Sabarros PS, Finucci B, Zanzi A,
Bach P, Senina I, Garibaldi F,
Sims DW, Navarro J, Cermeño P,
Leone A, Diez G, Zapiain MTC,
Deflorio M, Romanov EV, Jung A,
Lapinski M, Francis MP, Hazin H and
Travassos P (2022) Global-Scale
Environmental Niche and Habitat
of Blue Shark (Prionace glauca) by
Size and Sex: A Pivotal Step to
Improving Stock Management.
Front. Mar. Sci. 9:828412.
doi: 10.3389/fmars.2022.828412
ORIGINAL RESEARCH
published: 21 April 2022
doi: 10.3389/fmars.2022.828412
Blue shark (Prionace glauca) is amongst the most abundant shark species in international
trade, however this highly migratory species has little effective management and the need
for spatio-temporal strategies increases, possibly involving the most vulnerable stage or
sex classes. We combined 265,595 blue shark observations (capture or satellite tag) with
environmental data to present the first global-scale analysis of species’habitat
preferences for five size and sex classes (small juveniles, large juvenile males and
females, adult males and females). We leveraged the understanding of blue shark biotic
environmental associations to develop two indicators of foraging location: productivity
fronts in mesotrophic areas and mesopelagic micronekton in oligotrophic environments.
Temperature (at surface and mixed layer depth plus 100 m) and sea surface height
anomaly were used to exclude unsuitable abiotic environments. To capture the horizontal
and vertical extent of thermal habitat for the blue shark, we defined the temperature niche
relative to both sea surface temperature (SST) and the temperature 100 m below the
mixed layer depth (T
mld+100
). We show that the lifetime foraging niche incorporates highly
diverse biotic and abiotic conditions: the blue shark tends to shift from mesotrophic and
temperate surface waters during juvenile stages to more oligotrophic and warm surface
waters for adults. However, low productivity limits all classes of blue shark habitat in the
tropical western North Atlantic, and both low productivity and warm temperatures limit
habitat in most of the equatorial Indian Ocean (except for the adult males) and tropical
eastern Pacific. Large females tend to have greater habitat overlap with small juveniles
than large males, more defined by temperature than productivity preferences. In
particular, large juvenile females tend to extend their range into higher latitudes than
large males, likely due to greater tolerance to relatively cold waters. Large juvenile and
adult females also seem to avoid areas with intermediate SST (~21.7-24.0°C), resulting in
separation from large males mostly in the tropical and temperate latitudes in the cold and
warm seasons, respectively. The habitat requirements of sensitive size- and sex-specific
stages to blue shark population dynamics are essential in management to improve
conservation of this near-threatened species.
Keywords: foraging habitat, habitat niche, chlorophyll-a gradient, marine productivity, mesotrophic, oligotrophic,
mesopelagic micronekton, water temperature
INTRODUCTION
Rapid declines in the abundance of many elasmobranch
populations in the last decades are largely attributed to
vulnerability from exploitation (Dulvy et al., 2021).
Elasmobranchs’life history characteristics do not allow them
to withstand elevated fishing mortalities for extended periods
(Holden, 1974;Musick et al., 2000), and past accounts suggest
that intensive fisheries can be followed by a rapid decline in catch
rates or even a complete collapse of the fishery. In such a case, it
has been documented that shark stocks may need several decades
to recover (Castro et al., 1999). Shark declines can have strong
ecological consequences. Except in coral reef ecosystems
(Desbiens et al., 2021), the absence of sharks can indirectly
alter predation pressure on different fish species via behavioral
responses of meso-consumers released from predator
intimidation (Frid et al., 2008), altering the total fish
assemblage through trophic interactions and shaping marine
communities over large spatial and temporal scales (Stevens
et al., 2000;Ferretti et al., 2010). The manner in which sharks
structure communities may also be shifting with climate
change, where the recent increase of seawater temperatures
has been linked to a poleward shift in shark distributions
(Dolgov et al., 2005;Tanaka et al., 2021)andapotential
reduction of dive depths due to increased deoxygenation
(Vedor et al., 2021b).
The blue shark, Prionace glauca, is among the most abundant,
widespread, fecund and fast-growing of elasmobranchs, making
it somewhat more resilient to exploitation than other shark
species (Castro et al., 1999;Aires-da-Silva and Gallucci, 2007).
It is also one of the most heavily fished sharks in the world;
annual fishing mortality (mainly as bycatch) is estimated at 10.74
million.year
−1
(95% PI, 4.64–15.76 million.year
−1
;Clarke et al.,
2006), representing 5% of global shark landings in the 1990s,
peaking at almost 18% (137,973 mt) in 2013 before declining to
16% in 2017 (103,528 mt) (Okes and Sant, 2019).
Druon et al. Global-Scale Blue Shark Habitat
Frontiers in Marine Science | www.frontiersin.org April 2022 | Volume 9 | Article 8284122
As a bycatch species, the catch of blue sharks is intrinsically
linked to abundance but also to global market demand, both
being relatively complex drivers. Blue shark is amongst the most
abundant shark species in international trade (Okes and Sant,
2019) for the meat and/or fins (e.g., dominant shark species for
meat in Japan, Spain, Taiwan, and Uruguay: Okes and Sant,
2019; Brazil: Cruz et al., 2021; Italy: Serena and Silvestri, 2018;
Mancusi et al., 2020). Overall, the importance of blue shark in the
fin trade highlights its economic importance and the driver of the
demand (Porcher et al., 2021). Although some fleets release blue
shark bycatch (Campana, 2016), this species is a major
component of retained incidental catch of longline and driftnet
fisheries, particularly from nations with high-seas fleets (Okes
and Sant, 2019), often seasonally complementing the catches of
other pelagic species. For example, the blue shark is frequently
the dominant bycatch species in swordfish and tuna longlining
(Megalofonou et al., 2000;Campana et al., 2016;Carpentieri
et al., 2021). Global spatial analyses reported blue shark to be at
significantly greater risk of exposure to fishing due to its high
distributional overlap with important fished areas (Queiroz et al.,
2019). However blue shark populations have declined at the
global scale less than other shark species in the last five decades
(Pacoureau et al., 2021) probably due to its relatively higher
reproductive and growth rates compared to other oceanic sharks
(Pacoureau et al., 2021) and even though fishing mortality was
close to or possibly exceeding the maximum sustainable yield
levels (Clarke et al., 2006). Tuna Regional Fisheries Management
Organizations (RFMOs) conduct several shark stock, but often
no consensus on current status is found (e.g., ICCAT, 2015). If
several shark-specific regulations exist (Simpfendorfer and
Dulvy, 2017), one RFMO only recently started to agree on
catch limits for blue shark (e.g., allocated total allowable
catches (TAC) in the North Atlantic and unallocated TAC in
the South Atlantic by ICCAT in 2019, (European Union, 2020;
European Union, 2021).
Highly mobile in nature, blue sharks are known to make
seasonal reproductive migrations following changes in water
temperature and currents (Nakano, 1994;Stevens, 1999). The
links between blue shark distribution and oceanographic features
are an important focal point of current research. An improved
knowledge of blue shark habitat is needed to better understand
the ecology of a species that is distributed globally, but for which
only regional patterns have been studied in detail (e.g., Bigelow
et al., 1999;Carvalho et al., 2011;Adams et al., 2016;Vandeperre
et al., 2016). The provision of robust regional patterns of size- or
sex-specific habitat could improve the management of blue shark
populations by providing relevant spatio-temporal information
for potential fisheries mitigation measures. Focus should be given
to the life history stages that are most vulnerable to fishing to
possibly reduce the overall impact of its catch in global fisheries
as target and bycatch species.
In this study, we investigate the relationship between global
blue shark distribution and selected environmental variables to
define the species environmental niche. We frame our results in
the context of sustainable exploitation and conservation of this
species, to provide new perspectives for research and
management of blue shark populations. We justify the use of
two environmental proxies for food availability (chlorophyll-a
gradient and upper mesopelagic micronekton) as indicators of
preferred sex- and size-specific foraging habitat and identify the
unfavorable physical variables limiting habitat use. We then
build a habitat model for each sex and size classes using these
recognized environmental variables and clustering-based
parametrizations. The predicted habitat distribution is
compared with published knowledge, and model performance
is quantitatively compared against an independent validation
dataset of blue shark presence. The Supplementary Material
provides substantial details for a deeper insight of the analysis
and notably seven regional habitat animations by size and sex
classes with the overlay of blue shark presence data (observer and
electronic tagging data) for interpreting the seasonal movements.
MATERIALS AND METHODS
Blue Shark Presence Data
We collected extensive blue shark presence data throughout its
global distribution, mostly from observer programs of longline
fisheries but also purse seine fisheries, and from electronic
tagging programs. We used presence-only data from all
collected datasets for consistency. The full dataset included
589,450 observations, but only 496,080 observations
overlapped with the availability of environmental data
(chlorophyll-a data from MODIS-Aqua satellite sensor and the
other variables extracted from the EU-Copernicus Marine
Environment Monitoring Service, respectively). Thus, we
restricted all data to the time period from July 2002 to
December 2018. This included fishery-independent 234 tracks
from electronically tagged blue sharks in the Pacific (95), Atlantic
(132), and Indian Oceans (7) using different types of electronic
tags (Argos satellite transmitters and pop-up satellite-linked
archival transmitters, PSATs) on which we applied filtering
criteria for geolocation (see Supplementary Material, hereafter
SM). For electronic tags that did not have precise location
information (i.e., light- and temperature-based PSATs tags),
the majority of the likelihood surface for position estimates
tended to fall within a 50 km radius. This level of precision
was considered to be similar to the longline data since we
assumed sets had a maximum length of 100 km and we kept
the midpoint as the geographic position. Redundancy of all
presence data was avoided by filtering out data of the same
size (and sex for large individuals) when closer than 2.3 km on
the same day. This distance represented about half of the pixel
size of the habitat model grid, which was determined by the
resolution of the satellite remote sensing data of chlorophyll-a (1/
24° by 1/24° resolution). However, this redundancy filter mostly
removed eventual duplicates in the observer data since several
longline sets generally do not occur in the same location and day.
It was ineffective for electronic tagging data since only one high-
precision location was kept per day (see SM).
The presence data were partitioned by size and sex classes
following information available in the literature. While there are
Druon et al. Global-Scale Blue Shark Habitat
Frontiers in Marine Science | www.frontiersin.org April 2022 | Volume 9 | Article 8284123
differences in size thresholds among ocean regions, we followed
Vandeperre et al. (2016) and Megalofonou et al. (2009) for small
and large juveniles, and we averaged values from the following
sources: Castro and Mejuto (1995);Castro et al. (1999);
Campana et al. (2006);Megalofonou et al. (2009);Calich and
Campana (2015);Coelho et al. (2018) for adults of both sexes.
From the 496,907 observations of blue shark presence between
July 2002 to December 2018, the total number of observations
with morphometric data (small juveniles with size information,
larger blue sharks with size and sex and high-position quality for
electronic tags) was 265,595 (see SM,Table SM.1). The small
juvenile class was not split by sex as they tend to mix in
temperateandsubarcticwaters(Nakano, 1994). The size
classes considered were 1) small juveniles of both sexes
(hereafter SJ) with fork length (FL) below 125 cm (n = 60,904
observations, 23%), 2) large juvenile females (hereafter LJF) with
FL from 125 to 180 cm (n = 54,611, 21%), 3) large juvenile males
(hereafter LJM) with FL from 125 to 190 cm (n = 90,792, 34%), 4)
adult females (hereafter AF) with FL above 180 cm (n = 29,773,
11%) and adult males (hereafter AM) with FL above 190 cm (n =
29,515, 11%). Any presence information that was collected after
the model calibration process or for which the environmental
information was missing (e.g., missing satellite-derived
chlorophyll-a) was used for validation only of the habitat
model (see Model Performance section). After filtering (one
high-precision location per day), the electronic tagging data
accounted for 14,559 fishery-independent observations (906 for
SJ, 4,160 for LJF, 2,355 for LJM, 2,350 for AF, and 4,788 for AM),
representing 5.5% of the analyzed data. The overall data
distribution by size and sex classes were: 22.9% for SJ, 20.6%
for LJF, 34.2% for LJM, 11.2% for AF and 11.1% for AM.
Figure 1 shows the spatial and seasonal distributions of the
analyzed data (see SM for density maps by size and sex classes).
Although the South-East Atlantic, the North and East of the
Indian Ocean, and the Northwest Pacific are not well-
represented in our dataset, the data spans a wide range of
latitudes and productive ecosystems in the other areas, and
likely accounts for most of the environmental variability of
blue shark habitat.
From Ecological Traits to Modeling
Foraging Habitat
The selection of environmental variables for modeling was
guided by their relevance to reflect the main ecological traits of
the blue shark. These choices were therefore based on past
habitat analysis and expert knowledge as described in the
following three sections. This species is known to have a wide
distribution from equatorial to temperate latitudes (Vandeperre
et al., 2014a;Coelho et al., 2018;Maxwell et al., 2019), inhabiting
contrasting environments in terms of productivity. Similar to
other large marine predators, the blue shark is attracted by
mesoscale features such as fronts or eddies (Queiroz et al.,
2012;Vandeperre et al., 2014b;Scales et al., 2018;Braun et al.,
2019) in relatively plankton-rich surface waters (epipelagic and
mesotrophic areas: defined hereafter as the depth layer from 0 to
1.5-times the euphotic depth with surface chlorophyll content in
the range of 0.1-4.0 mg.m
-3
). However, blue shark also makes use
of relatively poor surface waters (oligotrophic, with surface
chlorophyll content below about 0.1 mg.m
-3
) and feeds in the
upper mesopelagic layer (depth layer from 1.5 to 4.5-times the
euphotic depth; e.g., Vedor et al., 2021a), with cephalopods as
dominant prey (Cordova-Zavaleta et al., 2018;Konan et al.,
2018). To reflect these contrasting environments, we retained
surface chlorophyll-a fronts and mesopelagic micronekton as
proxies for foraging behavior. Additionally, blue shark diving
profiles suggest that behavioral thermoregulation has an
important effect on hunting tactics (Campana et al., 2011;
Braun et al., 2019). The blue shark diving profiles often show
higher contrasts in temperature levels between the mesopelagic
and the surface layer in tropical areas than when swimming
within the intermediate layer in temperate latitudes. The primary
function of these dives may be to follow the diurnal vertical
migration of prey, with deeper dives during the day or during a
full moon at night (Vedor et al., 2021a), but can also potentially
be related to orientation (Campana et al., 2011;Musyl et al.,
2011;Elliott, 2020). Similarly, blue sharks were shown to remain
for longer durations at depth in the warmer anticyclonic eddies
than in the cooler cyclonic eddies (Braun et al., 2019) or in the
warmer Gulf Stream than in contiguous waters of the Labrador
Current (Campana et al., 2011). Both of these behaviors would
release the species from thermal constraints while foraging. Sea
surface height anomaly (SSHa) is mainly influenced by seasonal
changes in temperature and the geostrophic currents that
characterize eddies and gyres that are known to shape the
vertical and horizontal distribution of the full pelagic food web
(Polovina et al., 2001;Tew Kai and Marsac, 2010;Godø et al.,
2012), including the blue shark (Vandeperre et al., 2014b). Thus,
water temperature within the mesotrophic or oligotrophic
environments (sea surface temperature, SST,andthe
temperature 100 m below the mixed layer depth, T
mld+100
), as
well as sea surface height anomaly in regards to the mesoscale
activity, would be expected to play a key role in the global
distribution of blue sharks and were selected as highly
discriminant variables in the present habitat modeling. Besides
the abiotic environmental covariates retained for habitat
modeling (SST,T
mld+100
and SSHa), global fields of biotic
(chlorophyll-a fronts, mesopelagic micronekton) were
included, as they could substantially help define the foraging
habitat of the species. All environmental variables were extracted
as mean values over a 25 km radius centered on the calibration
presence data as this distance is relevant to most location
uncertainty of blue shark observer and electronic tagging data,
and is at the scale or below of oceanic features of importance
(e.g., fronts or eddies) (see also Figure SM.6 for data
integration). We particularly accounted for the ecological
traits of the species and literature knowledge that highlighted
some under-sampling of extreme environments in our
presence dataset.
The Biotic Environmental Variables
The main proxies used for blue shark foraging in our model are
the chlorophyll-a-derived productivity fronts in the mesotrophic
Druon et al. Global-Scale Blue Shark Habitat
Frontiers in Marine Science | www.frontiersin.org April 2022 | Volume 9 | Article 8284124
areas and the mesopelagic micronekton in oligotrophic areas.
Incorporating both of these is justified by i) the increased
occupancy of blue shark in surface layers (upper 100 m) driven
by highly productive (and cold) waters, and in deeper layers in
oligotrophic (and warm) environments (Vedor et al., 2021a;
Fujinami et al., 2021), ii) the efficient detection of mesoscale
productivity features by satellite observation in relatively
productive areas, and iii) the prediction of potential prey
sources at depth (mesopelagic micronekton) in relatively poor-
production environments where satellite observation is
unsuitable. The lack of detection by satellite ocean color
sensors of maximum chlorophyll-a in the subsurface
oligotrophic environments due to light attenuation
(exponential decrease of light with increasing depth)
FIGURE 1 | Distribution of both calibration and validation presence data (FL < 125 cm and FL > 125 cm with sex information, upper map, n = 265,595, number of
presence data per grid cell of 0.25°), and monthly distribution by size and sex class (‘All’corresponds to all presence data contemporary to environmental data
independently of size and sex availability). These data are mainly derived from fisheries observer (94.5%) programs, but include a small amount of data from
electronic tags (one position per day of higher accuracy, 5.5%). See in Supplementary Material the Figures SM.1-5 on the distributions by size and sex classes,
and seven video animations of habitat in regions of high-density electronic tagging data.
Druon et al. Global-Scale Blue Shark Habitat
Frontiers in Marine Science | www.frontiersin.org April 2022 | Volume 9 | Article 8284125
underestimates productivity in these oligotrophic environments.
Therefore, we considered the best performance of both biotic
variables between the high dynamics of productivity fronts, of
importance for large pelagic predators and sensed by satellite
observation, in mesotrophic waters and the deeper and lower
dynamics of the simulated mesopelagic micronekton in
oligotrophic environments.
In mesotrophic areas, the daily detection of productive
oceanic features (chlorophyll-a fronts) from ocean color
satellite sensors (currently MODIS-Aqua) is a good generic
proxy for food availability to fish populations (Druon et al.,
2021). When productivity fronts are active long enough (from
weeks to months) to allow the development of mesozooplankton
populations (Druon et al., 2019), they were shown to attract epi-
and mesopelagic fish and top predators (Olson et al., 1994;
Polovina et al., 2001;Briscoe et al., 2017;Druon et al., 2017;
Baudena et al., 2021). After an initial development phase of 3-4
weeks, the mesozooplankton biomass reached substantial levels
in the persistent and not necessarily stationary chlorophyll-a
fronts, i.e. where chlorophyll-a gradient is high (Druon et al.,
2019). This biomass aggregation may represent concomitant
foraging hotspots for small pelagic fish, and result in active
aggregation of highly mobile predators (e.g., bluefintuna
Thunnus thynnus in Druon et al., 2016;fin whale Balaenoptera
physalus in Panigada et al., 2017, basking shark Cetorhinus
maximus in Miller et al., 2015). Similarly, blue sharks have
been shown to be attracted by frontal oceanic features (Bigelow
et al., 1999). Daily chlorophyll-a (CHL,mg.m
−3
)datawere
gathered from the MODIS-Aqua ocean color sensor (years
2002–2018; 1/24° resolution) using the Ocean Color Index
(OCI) algorithm (Hu et al., 2012) and extracted from the
NASA portal (https://oceancolor.gsfc.nasa.gov/l3/; reprocessed
January 2018). Small and large chlorophyll-a fronts were derived
from and refer to different levels of chlorophyll-a gradient values
(see the SM for details on the chlorophyll-a gradient calculation).
Based on the link described above between productivity fronts
and biomass of low and high trophic levels (e.g., Olson et al.,
1994;Polovina et al., 2001;Druon et al., 2019;Druon et al., 2021;
Baudena et al., 2021), high chlorophyll-a gradient levels are
assumed to correspond, when persistent, to productivity fronts
with an important capacity to sustain well-developed food chains
and foraging opportunities for predators. The five-day
chlorophyll-a gradient values were extracted in a 25 km radius
centered on the (day and location of) blue shark presence data
for each size and sex class, and a minimum of 33% coverage was
set to accept the mean gradient value (5-day and 25 km-radius
mean value). The histogram of extracted log-transformed
gradient values was used to derive a dependent linear
predictor, which is the main component of the daily foraging
habitat in the mesotrophic environment.
In oligotrophic environments, the estimate of mesopelagic
micronekton (‘micronekton upper mesopelagic & micronekton
migrant upper mesopelagic’, in wet weight g.m
-2
) was extracted
from the global ocean low and mid trophic levels biomass
content hindcast model, available through the EU-Copernicus
Marine Environment Monitoring Service (https://marine.
copernicus.eu/access-data). These variables are simulated by
the SEAPODYM-LMTL (SEAPODYM for Low and Mid-
Trophic Level organisms) model (Lehodey et al., 2015, see SI
for details). Weekly data were aggregated by month and linearly
interpolated from the original grid at 1/12° resolution to the
habitat grid at 1/24° resolution.
For foraging proxies in the habitat model, we thus used the
mesopelagic micronekton in oligotrophic areas (CHL < CHL
min
)
and productivity fronts in mesotrophic areas (CHL > CHL
min
),
noting that CHL
min
was a relatively low chlorophyll-a value used
as a threshold between both foraging proxies. The specific value
of CHL
min
for each blue shark class was identified using cluster
analysis (the k-means clustering technique; see SM for
full description).
The Abiotic Environmental Variables
The abiotic variables included in the habitat model were sea
surface temperature (SST), sea surface height anomaly (SSHa),
and the temperature 100 m below the mixed layer depth (T
mld
+100
). The variables that were not included in the habitat model
were sea surface salinity, sea surface current intensity, surface
oxygen content, and the depth of the mixed layer since they
presented large variabilities (see SM) and were not identified as
primary discriminant variables by the expert knowledge. A range
of favorable conditions for each abiotic variable was defined and
subsequently used to exclude unsuitable habitats. The 0.3
th
and
99.7
th
percentile values were selected for the favorable range of
SST and SSHa for each size and sex class. This range was large
enough to encompass some extreme environmental conditions
that were poorly represented in the presence data, and included
areas with contrasting SSTs yet high seasonal blue shark presence
(e.g., in the Northwest Atlantic; see Results). More generally, we
considered the expert knowledge (e.g., known presence in local
extreme conditions) in the choice of these threshold values since
the dataset likely over-represents some environments and under-
represents others. The majority of blue shark observations came
from fishery-dependent data, which are known to be spatially
and temporally biased (Saul et al., 2020). We also excluded an
intermediate range of SST for the blue shark female classes (LJF
and AF) using the cluster analysis, noting a substantially low
presence in this range (90
th
and 10
th
percentile values of two
different clusters, see Results). Unlike the other abiotic
exclusions, this intermediate SST levels within the favorable
range for large females appeared to be actively avoided rather
than corresponding to physical intolerance. Finally, our analyses
showed that the temperature 100 m below the mixed layer depth
(T
mld+100
) was a relevant variable for identifying the upper limit
of the mesopelagic layer. Micronekton was extracted from the
mesopelagic layer, representing depths of 138 ± 32 m. This upper
depth of the mesopelagic layer corresponds to a vertical position
where large blue sharks spend a large proportion of the night,
while deeper dives during the day are alternatively done with
shallower dives in response to thermoregulation needs
(Campana et al., 2011;Braun et al., 2019). Consequently, T
mld
+100
was considered to represent an averaged-dive temperature,
i.e. identifying the appropriate mean temperature representing a
dive at the approximate interface between the epipelagic and
upper mesopelagic layers. We selected the same minimum
Druon et al. Global-Scale Blue Shark Habitat
Frontiers in Marine Science | www.frontiersin.org April 2022 | Volume 9 | Article 8284126
temperature value for T
mld+100
as for SST since the minimum
levels of SST and T
mld+100
were considered to be the surface and
mean-dive extreme temperature tolerance for each size and sex
class. The fields of temperature, mixed layer depth, and sea
surface height anomalies were extracted from the EU-
Copernicus Marine Environment Monitoring Service global
model (https://marine.copernicus.eu/access-data). All abiotic
data were first linearly interpolated from the original grid at
1/12° resolution to the habitat grid at 1/24° resolution, and then
linearly interpolated from monthly to daily values for estimating
the daily habitat (see also Figure SM.6 for data integration).
The Blue Shark Environmental Envelope
The environmental envelope of a species is defined as the set of
environments within which it is believed that are necessary to
maintain viable populations (Walker and Cocks, 1991). All biotic
and abiotic variables were integrated (mean value) over a 25 km
radius centered on each presence data for the identification of the
environmental envelope. This radius was selected as
representative of the geolocation precision of most of the
presence data (below about 50 km). Blue sharks of any size can
easily cover this distance within a day, as the lower and upper
bounds of the 95% confidence interval for the mean movement
rate of large blue sharks was 34 ± 9 km.day
-1
and 52 ± 18
km.day
-1
(1.42 ± 0.38 km.h
-1
and 2.15 ± 0.73 km.h
-1
,Kai and
Fujinami, 2020). The habitat model for blue sharks evaluates
positive relationships with foraging proxies (productivity fronts
or mesopelagic micronekton) after unsuitable abiotic
environments are excluded (based on SST,T
mld+100
and SSHa).
The resulting modeling envelope has two main components
depending on the level of surface chlorophyll-a: the
oligotrophic (CHL < CHL
min
)andmesotrophic(CHL
min
<
CHL < CHL
max
) foraging habitats that use mesopelagic
micronekton and productivity fronts (chlorophyll-a horizontal
gradients), respectively (Figure 2). Both components were
associated with the abiotic variables, i.e., temperature and
SSHa, where a value of 1 was set for favorable levels and a
value of 0 otherwise, therefore excluding unfavorable levels from
the habitat. As described above, a minimum temperature value in
the upper mesopelagic layer (T
mld+100
) was used to exclude
waters in oligotrophic environments that were too cold for
diving blue sharks, while a suitable range of SST was used to
exclude unsuitably warm or cold waters from the habitat in
mesotrophic waters. When a specific size class avoided certain
sea surface temperatures (large females, LJF and AF), the
associated daily foraging habitat in this intermediate SST range
was set to 0.
The same calibration method that was used to estimate the
quality of foraging habitat (i.e., prediction of foraging
opportunities) was applied to mesopelagic micronekton and
productivity fronts (MMnekton and gradCHL, respectively).
The habitat model for both these foraging proxies has two
parameters besides the distinct CHL range on which they
apply (CHL < CHL
min
for mesopelagic micronekton and
CHL
min
< CHL < CHL
max
for productivity fronts): a minimum
and intermediate value of mesopelagic micronekton
(MMnekton) and a horizontal gradient of chlorophyll-a
(gradCHL,Table 1). The minimum and intermediate threshold
values for MMnekton define the slope of daily habitat quality in
the oligotrophic environments (and similarly for gradCHL in the
mesotrophic environments). The intermediate threshold value
(the upper part of the habitat slope) is the minimum value of
MMnekton or gradCHL that corresponds to a habitat index of 1.
The suitable CHL range and the minimum threshold of foraging
proxies (MMnekton
min
and gradCHL
min
) were identified using
cluster analysis (see SM for details on the clustering) and
FIGURE 2 | Scheme of the blue shark envelope modeling linking the environmental variables in oligotrophic (CHL < CHL
min
) and mesotrophic (CHL
min
< CHL <
CHL
max
) environments with mesopelagic micronekton and productivity fronts (chlorophyll-a horizontal gradients) as foraging proxies, respectively. The abiotic
variables (SST, T
mld+100
and SSHa) were used to exclude unsuitable environments. CHL, surface chlorophyll-a content; SST, sea surface temperature; T
mld+100
,
Temperature 100 m below the mixed layer depth; SSHa, sea surface height anomaly.
Druon et al. Global-Scale Blue Shark Habitat
Frontiers in Marine Science | www.frontiersin.org April 2022 | Volume 9 | Article 8284127
represent the boundary values of suitable foraging environments
for each size and sex class. The intermediate value of each proxy
(MMnekton
int
and gradCHL
int
), which define minimum levels of
MMnekton and gradCHL corresponding to the maximum daily
habitat of 1, was identified using the values of MMnekton
min
and
gradCHL
min
and the preferred range of MMnekton and
gradCHL. This preferred range of each foraging proxy was
defined by the maximum slope of the cumulative distribution
that corresponds to presence data in the respective CHL range
(CHL < CHL
min
for MMnekton and CHL
min
< CHL < CHL
max
for gradCHL). A break in the daily habitat index between 0 and
0.3 was inserted to reflect sub-optimal foraging opportunities
(i.e., no effective foraging). However, such areas are still
important for a highly mobile predator to detect prey gradients
and actively move towards higher densities. Overall, there is no
direct correspondence between the daily foraging habitat
function for both foraging proxies (MMnekton and gradCHL),
even though each proxy quantitatively reflects the level of
foraging opportunities (see Discussion).
Habitat Index Equations by Size and Sex
Classes
Increasing levels within each foraging proxy, from small to large
productivity fronts or from low to high mesopelagic
micronekton levels were standardized from 0 to 1 in the daily
productive habitat indices. Daily foraging habitat was then
filtered by the various abiotic limitations, scored as 1 (inside
the favorable range) or 0 (outside). Thus, daily foraging habitat
was defined in each grid cell as satisfying the following equations
for suitable environmental conditions:
Feeding HabitatDay,Cell
=
0if CHL < CHLmin and MMnekton < MMnektonmin
ðÞor
CHLmin < CHL < CHLmax and gradCHL < gradCHLmin
ðÞor
CHL > CHLmax
0to 1if ∗CHL < CHLminMMnektonmin < MMnekton < MMnektonint
ðÞor
∗∗ CHLmin < CHL < CHLmax and gradCHLmin < gradCHL < gradCHLint
ðÞ
1if CHL < CHLmin and MMnekton > MMnektonint
ðÞor
CHLmin < CHL < CHLmax and gradCHL > gradCHLint
ðÞ
8
>
>
>
>
>
>
>
>
>
>
>
>
>
<
>
>
>
>
>
>
>
>
>
>
>
>
>
:
*linear function from 0 to 1 as follows:
Foraging HabitatDay,Cell =1+ MMnekton −MMnektonint
MMnektonint −MMnektonmin
**linear function from 0 to 1 as follows:
ForagingHabitatDay,Cell =1+ ln gradCHLðÞ−ln gradCHLint
ðÞ
ln gradCHLint
ðÞ−ln gradCHLmin
ðÞ
where ln is the natural logarithm, CHL
min
and CHL
max
define the
suitable range of chlorophyll-a, and MMnekton
min
and
MMnekton
int
are the minima and intermediate thresholds of
the mesopelagic micronekton concentration. The latter two
thresholds bound the increasing linear function for each
suitable habitat index between 0 and 1 (the same reasoning
applies to the gradCHL linear function).
The areas that met the daily biotic and abiotic requirements of
the habitat model were then integrated over time to create
monthly foraging habitat suitability maps. The time
composites are expressed in frequency of suitable habitat
occurrence (%) computed as a mean of 0-to-1 daily values
quantitatively associated with either the respective foraging
proxy. The multi-annual composites showing seasonal patterns
were computed from the monthly means to set an equal weight
between months. This is particularly relevant since lower habitat
coverage occurs during winter due to the higher cloud presence
and the subsequent greater number of missing estimates of
satellite-derived chlorophyll-a.
Model Performance
Performance of the habitat model was evaluated using an
independent set of presence data (hereafter, validation data).
The validation data are composed of either data that were
gathered after the model calibration, or consisted of
observations for which the coverage of satellite-derived
chlorophyll-a was not meeting the selection criterion for the
initial calibration. The validation criterion of minimum habitat
coverage was more stringent on the habitat coverage closer to the
observation (see the SI for criteria details), while the calibration
criterion focused on a mean habitat coverage at the scale of
TABLE 1 | Calibration parameters for the biotic variables of the blue shark foraging habitat model.
Biotic parameter values for
blue shark foraging habitat
Mesopelagic micronekton CHL (mg.m
-3
)Horizontal CHL gradient
MMnekton (g.m
-2
wet weight) gradCHL (mg.m
-3
.km
-1
)
min. *int. ** min.* max. * min. * int. **
Small juveniles (SJ) 0.89 3.91 0.12 0.74 0.00055 0.0048
Large juvenile females (LJF) 0.91 1.89 0.13 0.55 0.00054 0.0029
Large juvenile males (LJM) 0.91 1.86 0.09 0.48 0.00040 0.0052
Adult females (AF) 0.93 1.67 0.13 0.47 0.00047 0.0044
Adult males (AM) 0.98 2.70 0.125 0.55 0.00064 0.0075
(See Materials and Methods for the selection of the variables based on past habitat analysis and expert knowledge). CHL, sea surface chlorophyll-a content; gradCHL, horizontal gradient
of CHL; SJ, small juvenile blue sharks; LJF, large juvenile females; LJM, large juvenile males; AF, adult females; AM, adult males.
*Values identified using the cluster analysis (15
th
and 85
th
percentile values of the low/high-level cluster(s) [mesotrophic cluster(s) for CHL and gradCHL, and oligotrophic cluster for
MMnekton], see text for details and SM for the clustering),
**Values identified using the cumulative distribution function (gradCHL or MMnekton, see Figure 3).
Druon et al. Global-Scale Blue Shark Habitat
Frontiers in Marine Science | www.frontiersin.org April 2022 | Volume 9 | Article 8284128
variation of the environment (about 25 km). To assess model
performance, we defined the minimum level of habitat quality
that allowed for effective foraging at 30%, as lower levels
correspond to sub-optimal foraging opportunities (i.e., no
effective foraging and daily habitat set to 0 when the daily
function of foraging proxies led to values below 0.3). Recall
that monthly habitat quality was the mean frequency of daily 0-
to-1 values associated with the foraging proxies. From the 5-day
integrated habitat computed around the validation data, we
assessed model performance three ways. First, we calculated
the rate of positive occurrences within the core habitat from
thepresencedata.Second,wecomputedthedistanceof
validation presence data to the closest habitat boundary
because the rate of true positives does not account for presence
data that should often occur in the vicinity of the core habitat.
Predators spend a substantial time searching for their prey, and
should frequently occur near potential foraging areas. Third, we
computed the relative surface of the core habitat on a monthly
basis for each hemisphere (see SM) to evaluate the degree to
which the model discriminates suitable habitats.
RESULTS
The Foraging Niche: Mesotrophic
Versus Oligotrophic
The specific niche for foraging by class has an oligotrophic (CHL
<CHL
min
)andmesotrophic(CHL
min
<CHL<CHL
max
)
component, using mesopelagic micronekton and productivity
fronts, respectively, as proxies for suitable foraging habitat. Using
SJ blue sharks as an example, their daily habitat values are
defined by the MMnekton envelope(orangelinesegment,
Figure 3, upper left panel) by its slope, i.e., mostly by the
maximum cumulative distribution of the corresponding
presence data (dashed and solid blue lines, same panel). These
MMnekton values, scaled between 0 and 1, represent the range of
foraging opportunities that exist between the lowest and highest
micronekton levels in oligotrophic waters. Because the relative
presence in the oligotrophic areas varies between blue shark
classes, the daily habitat values defined by the MMnekton
envelope differ with respect to life stage (Figure 3, left panels).
The same approach is applied to productivity fronts in
mesotrophic environments (Figure 3, right panels) except that
the corresponding CHL levels are between CHL
min
and CHL
max
(dashed and solid green lines). The range of favorable CHL values
(CHL
min
and CHL
max
) and minimum level of each foraging
proxy (MMnekton
min
and gradCHL
min
) are identified for each
class using the cluster analysis (15
th
and 85
th
percentile values,
Table 1). As an example, the cluster analysis for SJ shows that the
suitable CHL range in mesotrophic areas is between about 0.12
and 0.74 mg.m
–3
(CHL
min
and CHL
max
). The range represents
the 15
th
and 85
th
percentiles of the intermediate and highest
CHL-cluster, the red and blue clusters respectively (Figure SM.7,
upper left panel). Note the blue cluster is always the largest in
size. The lower CHL-cluster defining oligotrophic habitats (CHL
< CHL
min
, green cluster) shows that the minimum gradCHL
value (gradCHL
min
; the 15
th
percentile of the CHL-intermediate
cluster) is 5.5.10
−4
mg.m
−3
.km
−1
, and the minimum value for
MMnekton (MMnekton
min
; the 15
th
percentile of the CHL.-lower
cluster) is 0.89 g.m
-2
wet weight (Figure 3 and Table 1). This
minimum gradCHL valueandthemaximumslopeofthe
cumulative distribution of SJ blue shark presence define the
daily habitat index for productivity fronts. Subsequently, they
also define the lowest gradCHL value for which the daily habitat
index reaches the maximum value of 1 (the intermediate
gradCHL value, gradCHL
int
) at 4.68.10
−3
mg.m
−3
.km
−1
(i.e., ln
(gradCHL) of −5.4 in Figure 3). The same approach is used for
mesopelagic micronekton although the daily habitat function is
defined for CHL < CHL
min
,andMMnekton is not log-
transformed. The resulting value of MMnekton
int
for SJ blue
shark, i.e., the minimum value for which the daily habitat reaches
the value of 1, is 3.91 g.m
-2
wet weight. The biotic calibration
parameters for the other size and sex classes (LJF, LJM, AF, and
AM) are presented in Figure 3 and Figures SM.7-11 (see the
summary of values in Table 1). While the CHL
min
and
MMnekton
min
values that notably delimit meso- and
oligotrophic waters are similar between classes (0.09-0.13
mg.m
-3
and 0.89-0.98 g.m
-2
wet weight), the major differences
that define the habitats arise from the intermediate values (i.e.,
the slopes), which directly derive from the frequency of presence
(0.0029-0.0075 mg.m
-3
.km
-1
and 1.67-3.91 g.m
-2
wet weight).
The distribution of blue shark presence relative to both foraging
proxies and CHL levels, i.e., MMnekton for CHL < CHL
min
(blue
line in Figure 3, left panels) and gradCHL for CHL
min
< CHL <
CHL
max
(green line in Figure 3, right panels) shows that adults
(AF and AM) are mostly located in oligotrophic environments,
while small juveniles (SJ) are more prevalent in mesotrophic
environments, and large juveniles (LJF and LJM) have a balanced
presence in both. This suggests that blue sharks transition from
mesotrophic to oligotrophic environments through their
lifespan, and tend to be found in higher and lower latitudes,
respectively. However, all stages are still present in both
environments, confirming the widespread distribution of blue
shark populations.
The Abiotic Niche
Using the abiotic variables to exclude unfavorable environments
meant a large percentile range (0.3
th
and 99.7
th
percentile values,
Table 2) for abiotic variables was required to encompass the
extreme conditions of the Northwest Atlantic, which were
under-sampled in our dataset and not sufficiently highlighted
in the cluster analysis with two or three clusters only. The
proximity of the cold Labrador Current to the warm Gulf
Stream south of Nova Scotia (South-East Canada) during
winter caused an extreme situation where the temperature of
surface waters is colder (and fresher) than below the mixed layer.
These SST levels are too cold for blue sharks (below about 11°C,
Figure SM.16) yet are suitable at depth (T
mld+100
above about
12°C). Because historic data (Campana et al., 2006) and some of
our electronic tagging data demonstrate the substantial presence
of blue sharks in winter in this area, we accounted for this
peculiarity in the model by selecting the 0.3
th
and 99.7
th
percentile values of the global distribution by size and sex class
for the range of SST and SSHa. This range was wide enough to
Druon et al. Global-Scale Blue Shark Habitat
Frontiers in Marine Science | www.frontiersin.org April 2022 | Volume 9 | Article 8284129
FIGURE 3 | Frequency of blue shark presence in each foraging proxy (colored histograms, relative units) and identification of the respective daily linear function of
foraging habitat (orange line segment). The distribution of mesopelagic micronekton (oligotrophic foraging, the colored histogram on left panels; MMnekton; g.m
-2
wet weight) and chlorophyll-a gradient (mesotrophic foraging, the colored histogram on the right panels; gradCHL; log-transformed, mg.m
-3
.km
-1
) at the location of
blue shark presence data by size and sex classes (SJ, LJF, LJM, AF, and AM) are compared to the global ocean distribution (light grey histogram). The distribution of
MMnekton and gradCHL at the location of blue shark presence and for CHL < CHL
min
and CHL
min
< CHL < CHL
max
, respectively, are delineated by the blue and
green lines. The maximum slope of the cumulative distribution of blue shark presence for each foraging proxy (blue and green dashed line for MMnekton and
gradCHL, respectively) defines the slope of the daily habitat linear function (orange line; see also Table 1; also using MMnekton
min
and gradCHL
min
).
Druon et al. Global-Scale Blue Shark Habitat
Frontiers in Marine Science | www.frontiersin.org April 2022 | Volume 9 | Article 82841210
suitably capture blue shark presence in that area (see regional
validation in Figures SM.17-18). The parameterization of abiotic
variables (Table 2) allows for tolerance of lower temperature
levels, which was exhibited by the SJ and LJF classes (12.3°C and
11.6°C, respectively). Compared to the other classes, AM had the
highest threshold to minimum temperature (14.8°C), indicating
the lowest tolerance to cold temperatures. The maximum
temperature tolerance is less variable among classes than the
minimum, although adults display slightly higher levels than the
other classes (c.f. 28.9-29.3°C with 28.3-28.6°C, Table 2). Similar
to temperature, the preferred SSHa ranges of SJ and LJF classes
are lower than the LJM and adult classes, indicating preferential
use of highly productive and temperate areas than less
productive and tropical areas. This can be seen by the
differences in their frequency of presence in mesotrophic and
oligotrophic environments (Figure 3). An unexpected pattern
indicated by the environmental-presence data relationship was
the apparent avoidance of large juveniles and adult females in an
intermediate range of SST (Figure 4). Our results highlight that
the presence of large females is three to four-fold lower than large
males in the range of 21.7-24°C and 22.1-23.4°C for LJF and AF,
respectively (5% presence for LJF compared to 17% for LJM, and
4% presence for AF compared to 13% for AM). These
intermediate ranges of SST for LJF and AF were used to set the
daily habitat to 0 (Table 2). To highlight the impact of this
apparent avoidance by large females, we show the 21.7°C and
24°C isotherms on the mean seasonal habitat map for large males
(next section).
Mean Seasonal Foraging Habitat by Sex
and Size Class
We used the favorable foraging conditions of blue sharks for each
class as defined by the model calibration to extrapolate daily
global blue shark habitat. The mean seasonal foraging habitat by
size and sex class for the period 2003-2018 was computed using
monthly mean levels (Figures 5–9, see also Figure SM.6 for data
integration and the seven habitat animations with the overlay of
electronic tagging and observer data as separate SM). The
frequency of suitable habitat occurrence (%) differed primarily
among seasons and the five blue shark classes. We plotted the
chlorophyll-a isocontour of CHL
min
on each map to distinguish
the distribution of foraging habitat arising from the mesopelagic
micronekton (oligotrophic) and productivity front
(mesotrophic) proxies (CHL < CHL
min
and CHL
min
< CHL <
CHL
max
, respectively). Oligotrophic environments tend to
contain higher levels of suitable foraging habitat as compared
to mesotrophic areas. This notably reflects the more stable
estimated levels of mesopelagic micronekton compared to the
scattered frequency of productivity fronts occurring in
mesotrophic areas. Favorable conditions for oligotrophic
foraging occur for all classes in the central part of the large
subtropical gyres, although not in the most oligotrophic areas
(see, e.g., the habitat differences between the subtropical
northeast and poorer southeast Pacific, Figures 5-9). Instead,
mesotrophic foraging mostly arises in temperate and equatorial
latitudes, both of which display a seasonal contraction during the
cold months alternately in both hemispheres, resulting in a
latitudinal oscillation between the seasons in the equatorial area.
Although blue shark foraging habitat is generally widespread
from about 50°S to 50°N, major unsuitable areas for foraging
occur in the central part of the main tropical gyres, such as in the
West and South-East Pacific or in the South-West of the North
Atlantic Oceans, due to particularly low productivity and
extremely high SSTs (Figures SM.13-15). In the Indian Ocean,
SST is too high in the northern basin (see in Figure SM.14 the
SST isocontour of 28.7°C enhancing the mean upper SST
limitation between all classes, 28.7°C ±0.38) and productivity is
too low in the latitudes from 15°S to 25°S (Figure SM.13). Given
their more restricted temperature tolerances, a larger extent of
the basin is unsuitable for males than for females, mostly due to
higher SSTs than the tolerated levels (Figure SM.14). The SJ and
AM classes show the most contrast in suitable environmental
conditions and predicted foraging habitats, particularly in the
northern hemisphere where seasonal variability is higher. For
example, SJ has a higher latitudinal range in the North Pacific
and Atlantic Oceans, yet a greater longitudinal range in the
Mediterranean Sea since the western basin is further North. The
other main difference between size and sex classes arises from the
apparent avoidance of females in an intermediate range of SST
levels. The SST avoidance isocontours of LJF shown on the mean
seasonal maps of males (Figures 7,9) highlight that habitat
overlapbetweenfemalesandmalesislackingmostlyfrom
January to June in the southern hemisphere, from April to
December in the North Atlantic, and from July to March in
the North Pacific. The overall observer and electronic tagging
data generally agree well with the predicted lack of habitat
overlap (see also Figures SM.22-24 comparing the males and
females’habitat with blue shark catches in (Coelho et al., 2018),
except in specific areas and seasons (e.g., North-east Pacificin
April-June, south-west Indian Ocean in October-December).
Global blue shark habitat declined during 2003-2018, as
indicated by changes in the frequency of suitable habitat
occurrence of about -0.13%.yr
-1
,(Figure SM.26). Declines
were unevenly distributed, with substantial negative trends (in
excess of -1.5%.yr
-1
of favorable habitat occurrence) mostly in the
northern and central Indian Ocean for the larger classes, and in
the tropical northern Pacific Ocean. Habitat expansion into
higher latitudes (above 30°) in both hemispheres was indicated
by positive trends (in excess of -1.5%.yr
-1
in the North-west
Atlantic, South Pacific and south Indian oceans). Overall, the
higher losses of habitat over time for blue sharks appeared to
occur in the equatorial and tropical Indian Ocean and in the
tropical North Pacific, specifically for the juveniles. The major
gains in habitat were observed in the Northwest Atlantic and the
Mediterranean Sea for the larger size classes, and in the
temperate South Pacific and Indian Oceans.
Distance of Presence Data to Closest
Habitat and Seasonal Habitat Size
We used fairly strict minimum criteria for habitat coverage (see
Materials and Methods and SM) to evaluate model performance
from the independent validation data, representing 47% (125,825
Druon et al. Global-Scale Blue Shark Habitat
Frontiers in Marine Science | www.frontiersin.org April 2022 | Volume 9 | Article 82841211
TABLE 2 | Calibration parameters for the abiotic variables of the blue shark foraging habitat model.
Abiotic parameter values for blue shark foraging habitat SST (°C) SST
intermediate avoidance
(°C) SSHa (m) T
mld+100
(°C)
Min. = Minimum min. *max. *min. ** max. ** min. *max. *min. ***
Max. = Maximum
Small juveniles (SJ) 12.3 28.3 N/A N/A -0.69 0.82 12.3
Large juvenile females (LJF) 11.6 28.5 21.7 24.0 -0.62 0.83 11.6
Large juvenile males (LJM) 13.1 28.6 N/A N/A -0.54 0.92 13.1
Adult females (AF) 13.3 29.3 22.1 23.4 -0.55 0.94 13.3
Adult males (AM) 14.8 28.9 N/A N/A -0.53 1.00 14.8
See Materials and Methods for the selection of the variables based on past habitat analysis and expert knowledge. SST, sea surface temperature; SSHa, sea surface height anomaly; T
mld
+100
, the temperature at 100 m below the mixed layer depth; SJ, small juvenile blue sharks; LJF, large juvenile females; LJM, large juvenile males; AF, adult females; AM, adult males.
*Values identified using the global distribution by size and sex class (0.3
th
and 99.7
th
percentile values, Figure 4),
**Values identified using the clusters linked to SST avoidance by females (90
th
and 10
th
percentile values, Figure 4 and SM),
***Minimum values of T
mld+100
were set identical to SST
min
(see Methods for details).
FIGURE 4 | Global distribution of sea surface temperature (SST) by size and sex class of blue shark. A continuous SST niche for males and small juveniles (left
panels) and a fragmented SST niche for large juvenile and adult females (right panels) is evident, with three to four-fold lower presence of females than males in the
SST range of 21.7-24°C and 22.1-23.4°C, respectively.
Druon et al. Global-Scale Blue Shark Habitat
Frontiers in Marine Science | www.frontiersin.org April 2022 | Volume 9 | Article 82841212
FIGURE 5 | Mean seasonal distribution of blue shark foraging habitat for the small juveniles (SJ, 2003-2018, in frequency of suitable habitat occurrence, %). The
chlorophyll-a isocontour of 0.12 mg.m
-3
(CHL
min
) separates the mean area of oligotrophic foraging (below this value using mesopelagic micronekton as foraging
proxy) and mesotrophic foraging (above this value using productivity fronts). Presence data (calibration and validation) are represented as pink dots for observer data
and colored line transects for electronic tagging data (start and end of months are shown by a black star).
Druon et al. Global-Scale Blue Shark Habitat
Frontiers in Marine Science | www.frontiersin.org April 2022 | Volume 9 | Article 82841213
FIGURE 6 | Mean seasonal distribution of blue shark foraging habitat for the large juvenile females (LJF, 2003-2018, in frequency of suitable habitat occurrence, %).
The chlorophyll-a isocontour of 0.13 mg.m
-3
(CHL
min
) separates the mean area of oligotrophic foraging (below this value using mesopelagic micronekton as foraging
proxy) and mesotrophic foraging (above this value using productivity fronts). Presence data (calibration and validation) are represented as pink dots for observer data
and colored line transects for electronic tagging data (start and end of months are shown by a black star).
Druon et al. Global-Scale Blue Shark Habitat
Frontiers in Marine Science | www.frontiersin.org April 2022 | Volume 9 | Article 82841214
FIGURE 7 | Mean seasonal distribution of blue shark foraging habitat for the large juvenile males (LJM, 2003-2018, in frequency of suitable habitat occurrence, %).
The chlorophyll-a isocontour of 0.09 mg.m
-3
(CHL
min
) separates the mean area of oligotrophic foraging (below this value using mesopelagic micronekton as foraging
proxy) and mesotrophic foraging (above this value using productivity fronts). Presence data (calibration and validation) are represented as pink dots for observer data
and colored line transects for electronic tagging data (start and end of months are shown by a black star). The SST isocontours of LJF avoidance (21.7°C and 24°C),
slightly larger than for AF, allow evaluating the potential lack of habitat overlap with LJM.
Druon et al. Global-Scale Blue Shark Habitat
Frontiers in Marine Science | www.frontiersin.org April 2022 | Volume 9 | Article 82841215
FIGURE 8 | Mean seasonal distribution of blue shark foraging habitat for the adult females (AF, 2003-2018, in frequency of suitable habitat occurrence, %). The
chlorophyll-a isocontour of 0.13 mg.m
-3
(CHL
min
) separates the mean area of oligotrophic foraging (below this value using mesopelagic micronekton as foraging
proxy) and mesotrophic foraging (above this value using productivity fronts). Presence data (calibration and validation) are represented as pink dots for observer data
and colored line transects for electronic tagging data (start and end of months are shown by a black star).
Druon et al. Global-Scale Blue Shark Habitat
Frontiers in Marine Science | www.frontiersin.org April 2022 | Volume 9 | Article 82841216
FIGURE 9 | Mean seasonal distribution of blue shark foraging habitat for the adult males (AM, 2003-2018, in frequency of suitable habitat occurrence, %). The
chlorophyll-a isocontour of 0.125 mg.m
-3
(CHL
min
) separates the mean area of oligotrophic foraging (below this value using mesopelagic micronekton as foraging
proxy) and mesotrophic foraging (above this value using productivity fronts). Presence data (calibration and validation) are represented as pink dots for observer data
and colored line transects for electronic tagging data (start and end of months are shown by a black star). The SST isocontours of LJF avoidance (21.7°C and 24°C),
slightly larger than for AF, allow evaluating the potential lack of habitat overlap with AM.
Druon et al. Global-Scale Blue Shark Habitat
Frontiers in Marine Science | www.frontiersin.org April 2022 | Volume 9 | Article 82841217
FIGURE 10 | Model performance estimated in distances (km) of validation presence data to closest core habitat boundary (above 30% favorable) for each size and
sex class (SJ, LJF, LJM, AF, and AM) and corresponding cumulative distribution showing distance values for each decile bin of presence data (negative are
observations within favorable habitat). For the SJ class for instance, 74% of validation presence data were within the core habitat (above 30%) and the distance of
the 80
th
percentile of presence data to the closest core habitat boundary was 2 km (n = 25,563).
Druon et al. Global-Scale Blue Shark Habitat
Frontiers in Marine Science | www.frontiersin.org April 2022 | Volume 9 | Article 82841218
observations) of the available information on blue shark presence
(proportion by class ranges from 42 to 52%, Table SM.1).
Globally, the validation data had a similar distribution as the
calibration data, although with slightly higher representation in
the North-East and South-East Pacific from additional longline
observer data and in the North Atlantic from electronic tagging
data. The model performance was estimated through i) the
fraction of validation data within the highest quality habitat
(where the minimum level for effective foraging was 30%,
hereafter, core habitat), integrated over five days centered on
the day of the observation, ii) the distance of 80% of the
validation data to the closest core habitat boundary (how close
to core habitat are most of the predator presence data) and iii)
monthly changes in the relative surface of the core habitat to
evaluate how discriminant the prediction is. The fraction of
validation presence data within the core habitat reflects the
potential for effective foraging, and was 74% for SJ (n =
25,563), 68% for LJM (n = 44,298), 58% for LJF (n = 26,366),
47% for AM (n = 14,004) and 61% for AF (n = 15,594). A
predator searching for prey should remain in the vicinity of the
core habitat, and the 80
th
percentile of presence data to the
closest core habitat boundary displayed the same ranking of
performances by class: with 2 km for SJ, 3 km for LJM, 21 km for
LJF, 29 km for AM and 10 km for AF (Figure 10).
All classes present a large proportion of observations (86-99%)
closer than 50 km, which matches the level of geolocation
uncertainty of the presence data (Figure 10). This range is 79-
97% between classes for observations closer than 25 km to the core
habitat. Considering the large distribution of the species compared
to the relatively low surface area of the predicted core habitat (about
18-22 ± 2% and 23-29 ± 2% in the north and south hemispheres,
respectively, with larger and lower extents for SJ and AM,
respectively, Figure SM.12), we conclude that the habitat model
is fairly discriminant and accurately predicts habitat suitability
among classes. Finally, we present in the SM as separate files
seven regional habitat animations with the overlay of observer and
electronic tagging data for a qualitative appreciation of the model
results and a visualization of the seasonal movements of blue
sharks by size and sex classes.
DISCUSSION
Consistency of Model Results With Known
Traits and Habitats
Quantitative validation associated with a relatively restricted
surface area of the core habitat in the global ocean showed that
the model can discriminate habitats by sex and size class for blue
sharks. Furthermore, our results are consistent with previously
published catch data in the contrasting SST conditions of the
Northwest Atlantic (Figures SM.17-18,Campana et al., 2006)
and in the central-eastern Atlantic, where a maximum of catch
per unit of effort (CPUE, kg/set) corresponds to the spatial
dominance of different habitat classes (Figures SM.19-20). At-
sea observer data suggests large juvenile females predominate in
relatively cold waters in the southeast Indian Ocean, which is
consistent with our selected lower SST
min
for this class than for
males and adults (Figure SM.24). The model is qualitatively
consistent with observations from other areas where little or no
calibration data were used, especially in the Pacific Ocean. In the
northwest Pacific, Nakano (1994) found a maximum CPUE of
blue shark (in number for all classes) in the area 25-45°N, 140°E-
160°W, with lower levels in the equatorial area (also in Strasburg,
1958). This pattern is well captured by the model (see
interannual mean habitats in Figures SM.25). High CPUEs in
number were observed to occur up to 55°N in the northeast
Pacific in July (Nakano and Nagasawa, 1996), also in agreement
with our results (Figures 5-9). Strasburg (1958) denoted a
maximum CPUE of newborn blue sharks in the area 30-40°N
in the Pacific in summer, which agrees with the highest habitat
quality of the small juveniles for that season and represents the
most persistent habitat in that region year-round (Figure 5).
This observation of nurseries in the North Pacific is consistent
with the presence of late-stage pregnant blue sharks in spring
(30-40°N in the Northwest Pacific, Fujinami et al., 2021) and
with the favorable habitat of adult females in the same season
(Figure 8). In the Southwest Pacific, the mean prey biomass off
eastern Australia is about nine-fold higher for blue sharks
(n = 147, mean FL = 137 cm) south of 32°S compared to north
(Young et al., 2010), which agrees well with our results for
juveniles (Figures 5-7). In the tropical Southeast Pacific, our
validation observer data matches well with the presence of large
females (Figures 6,8). Differences among sex and size classes are
consistent with the observed size composition of blue shark
catches in the Australasian region, with mostly juveniles south of
35°S and adults north of 35°S (West et al., 2004;Neubauer et al.,
2021). The blue shark is the most landed shark species in the
Peruvian shark fisheries, representing 42% of total landings
(Gonzalez-Pestana et al., 2016), which indirectly corroborates
our findings of suitable habitat for small juveniles and large
females (Figure SM.25). In the Atlantic and Indian oceans, other
catch data display consistent distributions with our results.
Mostly adult blue sharks (195-320 cm total length) are caught
by the artisanal driftnet fishery in the Guinea current (off Ivory
Coast, Central-eastern Atlantic; (Konan et al., 2018), which
correlates with the medium habitat quality of adults and the
lack of favorable habitat of smaller specimens from our model
(Figure SM.25). Notably, sex ratio and fork length distributions
in the South Atlantic and the Indian Ocean lack female
representation, which is consistent with the avoidance of
intermediate SST by large females in the modeling (Figures
SM.22-23). These intermediate SST levels also occur in the
vicinity of the Azores (Figures 6,8, and habitat animations in
the SM), and large females are absent in catch data from July to
November (Vandeperre et al., 2014b;Coelho et al., 2018).
The ecological benefit of sex segregation from females avoiding
SST within their physiologically-tolerated range requires further
investigation. It is possible that these environments correspond to
suboptimal temperatures for pregnant blue sharks between the
cooler parturition grounds and warmer subtropical waters, aiding
fertilization and embryonic growth at the early development stage
(Hazin et al., 1994;Fujinami et al., 2021). These intermediate SST
Druon et al. Global-Scale Blue Shark Habitat
Frontiers in Marine Science | www.frontiersin.org April 2022 | Volume 9 | Article 82841219
levels likely correspond to poor nutritional conditions for pups
compared to temperate waters (Nakano, 1994;Fujinami et al.,
2021). These temperature levels may also correspond to the
optimal temperature range for males’gonad development and
maturation (‘thermal-niche fecundity’,Wearmouth and Sims,
2008). Furthermore, large males are indiscriminate feeders
(McCord and Campana, 2003), and could likely prey upon
newly-born progeny if it would not be in the females’
evolutionary interest to stay away from males around the time
of pupping. Finally, because of the aggressiveness of mating males
and given the potential for injury (Sims, 2005), females may
venture into these areas only when mating is strictly necessary.
The twice-thicker skin of females (Pratt, 1979) could also result in
higher tolerance to low temperatures (Vandeperre et al., 2014a),
given that the LJF class had the lowest SST
min
value (Table 2)
compared to males. While temperature is fundamental to the
habitat predictions, a direct comparison of minimum levels of
averaged dive temperature (SST
min
and T
mld+100
) with absolute
physiological thresholds in literature is difficult due to the high
gradient of temperature experienced during the dives. Braun et al.
(2019) reported a mean dive temperature of adult males in the
relatively cold cyclonic eddies of about 15°C from tagging
experiments in the Northwest Atlantic, which is consistent with
our findings (14.8°C). Furthermore, blue sharks swiftly end their
deep dives when muscle cools to 15°C, with a range of water
temperature met during the dives spanning from 7 to 26°C (Carey
et al., 1990). While finding the same minimum muscle
temperature, Watanabe, Nakamura, and Chiang (Watanabe
et al., 2021) concluded that the regular deep-diving behavior of
this species can be parsimoniously explained by their motivation
for maintaining body temperature within a narrow range
while foraging in the stratified water columns. Overall, blue
shark classes have variable tolerance to minimum temperature
(LJF < SJ < [LJM, AF] < AM), which is valuable habitat
information for identifying bycatch mitigation measures
targeting the most vulnerable classes.
Approach Limitations
The accuracy of the blue shark niche and habitat prediction is
limited by geolocation uncertainties of the observations, the
intrinsic limitations of the environmental variables, and the
necessary model simplification of a complex species ecology.
The geolocation uncertainties of observer data and of the light/
SST-derived electronic tagging, all may have influenced the
precise identification of the ecological niche. Given the
maximum length of longline sets (100km), it was not possible
to determine a single location with a precision greater than 50
km. Similarly, the precision of light-based geolocation estimates
in equatorial areas, in regions with low SST gradients, or during
equinoxes in tropical areas can be low, even following estimation
using Hidden Markov Models (Braun et al., 2015;Braun et al.,
2018). We allowed for some positional inaccuracy by integrating
each environmental variable within a 25 km radius (centered on
the observations). Physiologically, blue sharks are likely able to
explore such an area on a daily basis (Kai and Fujinami, 2020).
Overall, both datasets (observer and electronic tagging) are
complementary for describing the habitat niche since the
fishery-dependent data are numerous and tends to be
concentrated while the independent data are less frequent and
more spread out. The optical remote sensing data from which
CHL is derived is limited by cloud coverage, which frequently
precludes data collection in relatively high latitudes (about 45-
70°N) and in the equatorial belt. The effect of data gaps on
habitat prediction was partially compensated for by using
monthly habitat distributions to compute seasonal mean
composites, therefore giving equal weight to each month. The
other foraging proxy, the mesopelagic micronekton, has
limitations inherent to the model used for its prediction. First,
the parameters of the mesopelagic micronekton model were
calibrated based on limited data, including observations
derived from the literature (Lehodey et al., 2010), and trawling
and acoustic survey data (GUID CMEMS, 2021). Although the
optimization was successfully tested in this model, the biomass
estimation from single-frequency acoustic data remains an issue
for rigorous parameter estimation (Lehodey et al., 2015).
Additionally, the empirical definition of the upper mesopelagic
layer, which was based on the euphotic depth, needs to be further
verified once complementary data become available.
Another limitation of the habitat model relates to the
comparability of both foraging proxies. Large scale analysis of
habitat quality is difficult because blue sharks modify their
strategy over their lifespan, from more frequent foraging in
relatively cold and mesotrophic waters (likely in the epipelagic
layer) as a small juvenile, to less frequent and deeper foraging in
relatively warm surface waters (likely in the mesopelagic layer) as
an adult. Thus, habitat values do not represent absolute levels of
foraging capacity in that they are not comparable in oligotrophic
(micronekton, CHL < CHL
min
) or mesotrophic conditions (CHL
gradient, CHL
min
<CHL<CHL
max
). In other words,
oligotrophic daily habitat values may reach the maximum
value of 1 even if the foraging capacity is functionally much
lower than in mesotrophic environments. Small juveniles and
adults are both adapted to foraging in relatively productive/cold
and poor/warm niches, respectively, to meet their energetic
needs. This means that the habitat suitability represents both
the foraging capacity within each proxy and the energetic needs
of individuals. These characteristics of balanced energy intake
and expenditure should be kept in mind when interpreting
the habitat results. It should be further noted that the scaling
of both foraging proxies was done independently from one
another using the respective cumulative distribution functions
of presence.
Finally, our decision to only use foraging proxies as attracting
factors within the model means that biological processes such as
reproduction are not explicitly accounted for. Because female
parturition occurs in suitable environments for small juveniles
(temperate waters, Nakano, 1994;Fujinami et al., 2021), we
would expect a substantial overlap due to reproduction to be
accounted for in our analyses. However, mating behavior may
remain unaccounted for in the modeling if disconnected from
foraging behavior. This might particularly be the case for AM for
which the reproduction behavior may explain the lowest model
Druon et al. Global-Scale Blue Shark Habitat
Frontiers in Marine Science | www.frontiersin.org April 2022 | Volume 9 | Article 82841220
performance among the five size and sex classes. Overall, the
predicted habitat describes the suitable areas for foraging, and
any organism allocates the energy obtained from it to the
competing functions of growth, reproduction and longevity/
survival (investment in maintenance, storage and repair,
Litchman et al., 2013). These authors underline that organisms
shall optimize their behavior and energy allocation pattern to
maximize the fitness of the individual in a particular
environment, which may apply to the blue shark foraging
strategies in their lifespan from relatively cold and rich
environments to warmer and poorer surface waters.
Perspectives for Research, Management,
and Policy
There is an increasing need for appropriate management of blue
shark stocks (e.g., in the Atlantic by ICCAT, European Union,
2020;European Union, 2021) due to greater fishing pressure in
recent decades. While the species appears relatively resilient
in some areas the catch of blue shark has dramatically declined
in others. In the Atlantic, total catch has been relatively
consistent over time (ICCAT, 2020) even though blue shark
represents about 60% of the Spanish and Portuguese longline
bycatch (Porsmoguer et al., 2015). Conversely, abundance in the
Mediterranean Sea declined by 96.5% over 56 years (Ferretti
et al., 2008). Furthermore, the level of uncertainty in the data
inputs and structural model assumptions for blue shark stocks
are often high enough to prevent reaching a consensus on
specific management recommendations. Mitigation of shark
bycatch mostly consists of measures related to fishing gears,
specifying hook or lure or bait types, type of branch lines, and/or
fishing depth (e.g., Sacchi, 2021) or prohibiting retention
onboard, transhipment, landing or storing for certain sharks
and fleets (e.g. IATTC Resolution C-19-05 for shark species).
However, there is increasing interest from RFMOs to implement
spatio-temporal management strategies (e.g., IOTC-2021-17
th
Working Party on Ecosystems and Bycatch), especially under a
dynamic management approach (e.g., Hazen et al., 2018;Lopez
et al., 2019) that may highly reduce bycatch (Pons et al., 2022),
which would require a good understanding of spatial population
structure. Historically, the lack of long time-series of data specific
to sex and size has limited the consideration of population
structure (e.g. size and sex segregation) in stock assessment
models (Mucientes et al., 2009). This means that any risk
associated with differential fishing mortality rates on specific
components of the stock cannot be accounted for. The data
compiled for our study shows that morphological data are more
readily available in recent years (53% of observations in our
dataset, 2003-2018), which suggests that sex- and spatially-
structured stock assessments could potentially be developed,
elucidating the need for spatio-temporal measures to protect
particularly vulnerable size and sex classes from fishing.
In general, blue sharks are relatively tolerant to the hardships
of being hooked on a longline or entangled in a net for several
hours but exhibit variable at-vessel mortality levels (3%, Kotas
et al., 2000; 5%, Megalofonou et al., 2005; 4.9-31%, Hutchinson
et al., 2021 and review herein) and higher post-release rates (15%
with 95% CI, 8.5–25.1%, Musyl et al., 2011; 17% with 95% CI, 11-
26%, Musyl and Gilman, 2019; 38%, Hutchinson et al., 2021)
depending on the gear and regional practice. Coelho et al. (2012)
also reported a relationship between mortality and size, with the
smaller specimens having higher mortality rates than the larger
sharks. Provided that best practices for handling and release are
followed (Musyl and Gilman, 2019;Hutchinson et al., 2021), it is
likely that releasing live blue sharks caught by longline fisheries is
a complementary viable management tool to protect biomass in
blue shark populations (as has been suggested in relation to the
apparent recovery of the Southwest Pacific stock, Neubauer et al.,
2021). It is likely mortality could be further reduced if spatio-
temporal information on core habitats of the main life-history
stages could be used to minimize catches of the most vulnerable
classes. Vulnerability is likely greatest for small juveniles given
their susceptibility to capture-induced mortality as well as adult
females because they are pivotal to population productivity. As
with any large elasmobranch, reproduction dynamics of blue
shark are characterized by slow growth, late age of maturity and a
relatively low reproductive rate (Myers and Worm, 2005).
Spatio-temporal mitigation measures can also be related to
depth. Depending on the specificlonglinefishery (e.g., the
Canadian and Japanese longline fleets deploy gear within the
top 15 m of the water column and at about 200 m, respectively),
monthly mean habitat maps of the most vulnerable classes could
highlight when the blue shark likely surface or exhibit
mesopelagic foraging, using the CHL
min
level as a boundary
between both core habitats. Furthermore, it is noteworthy that
young-of-the-year blue sharks spend most of their time at depths
shallower than 40 m (Nosal et al., 2019) or that hooking lines in
warm surface waters (e.g., > 28°C) may increase post-release
mortality due to thermoregulation needs (Musyl and Gilman,
2018). Finally, temporal or spatial variation in stock structure,
such as that provided by class-based habitat predictions, can
improve specification of the selectivity pattern or growth models
used in stock assessment, which can substantially reduce
uncertainty in assessment results (Carvalho et al., 2015;
Neubauer et al., 2021). There would also be the potential to
use the extent and temporal trends of suitable habitats (Figures
SM.25-26) to partially account for environmental variation
affecting indices of abundance, to improve the standardization
of catch per unit effort (Maunder et al., 2006). If strong inter-
annual changes in habitats are detected, it would be meaningful
to account for these habitat modifications in the stock
projections used to determine catch advice, noting that
substantial habitat trends (Figure SM.26) may serve as
reasonable estimates for the near future. Our finding that
suitable habitat extends poleward during the warm months
and contracts during the cold months, as supported by satellite
tracking, implies that blue shark populations may seasonally
concentrate in the northern or southern basin ocean, alternately
in both hemispheres, exhibiting higher vulnerability to fishing
during periods when waters are the coldest (generally
November-May and May-September in the north and south
hemispheres, respectively, except the LJF class, see Figure
SM.12). The decadal positive trend of habitat in high latitudes
Druon et al. Global-Scale Blue Shark Habitat
Frontiers in Marine Science | www.frontiersin.org April 2022 | Volume 9 | Article 82841221
is likely due to the warming of surface waters, except for the
negative trend in the subpolar North Atlantic where cooling was
observed (Hu and Fedorov, 2020). A general poleward
movement of the population and catches is likely to occur in
the next decades. Future developments will focus on products for
bycatch mitigation combining the core habitat of the classes to
protect in priority (e.g. SJ and AF) with weights defined by
management objectives. Such products, based on monthly
habitat in the recent years, may serve as avoidance maps for
fisheries, distinguishing with a color code the likely surface or
deeper presence.
Understanding the ecological niche and core habitat of large
pelagic species is key to promoting conservation and progressing
in fishery management. We have shown how to incorporate
environmental information available at suitable resolution
together with detailed fisheries and fishery-independent data to
meet this need at a global scale. Blue sharks, and other shark
species even more vulnerable to fishing, should benefit from this
approach, as it could be used to reduce uncertainty in stock
assessment and inform different conservation and management
options to ultimately promote long-term sustainability.
DATA AVAILABILITY STATEMENT
The blue shark habitat data are available on the JRC Data
Catalogue (https://data.jrc.ec.europa.eu/dataset?q=fish%
20habitat) (or upon request) and habitat maps are available
online (https://fishreg.jrc.ec.europa.eu/web/fish-habitat/
habitatmapping). Chlorophyll-a data are available online
(https://oceancolor.gsfc.nasa.gov/cgi/l3). The abiotic
environmental and mesopelagic micronekton data are available
online (https://marine.copernicus.eu/access-data).
AUTHOR CONTRIBUTIONS
J-ND headed the collaboration and performed the computations.
SC, FV, FH, HB, RC, NQ, FS, FA, DD, MM, JL, BB, PA, HD, PS,
BF, AZ, PB, IS, FG, DS, JN, PC, AL, GD, MD, ER, and AJ
analyzed the results, and wrote sections of the manuscript. SC,
FV, FH, HB, RC, NQ, FS, FA, DD, MM, JL, BB, PA, HD, PS, BF,
PB, IS, FG, DS, JN, PC, AL, GD, MZ, MD, ER, AJ, ML, MF, and
PT provided data. All authors contributed to manuscript
revision, read, and approved the submitted version.
ACKNOWLEDGMENTS
The authors would like to particularly thank the following
organizations and associated scientists for the data contributions
on blue shark presence: Bedford Institute of Oceanography,
Fisheries & Oceans Canada, Tagging of PacificPelagics-
Packard and Moore Foundations (TOPP), National Oceanic and
Atmospheric Administration (NOAA, Pacific Islands Regional
Office, Honolulu, HI, and Southwest Fisheries Science Center,
Monterey, CA), Inter-American Tropical Tuna Commission
(IATTC), Instituto do Mar (IMAR), University of Azores,
COSTA project (costaproject.org), Universidade Federal Rural de
Pernambuco, Institut de Recherche pour le Développement (IRD),
Pelagic Research Group Hawaii, Portuguese Institute for the Ocean
and Atmosphere (IPMA), International Commission for the
Conservation of Atlantic Tunas (ICCAT), AZTI, MEDLEM,
Hellenic Centre for Marine Research (HCMR), MedBlueSGen
project, CSIRO Marine and Atmospheric Research (Hobart),
Ministry for Primary Industries –Fisheries New Zealand/
National Institute of Water and Atmospheric Research (NIWA),
Marine Biological Association of the UK, University of
Southampton, CIBIO/In-BIO at the Universidade do Porto,
Universidad Autónoma de Nuevo León/INAPESCA/PNAAPD/
FIDEMAR, Indian Ocean Tuna Commission (IOTC), Spanish
Institute of Oceanography (IEO), Organization of Associated
Producers of Large Tuna Freezers (OPAGAC), University of
Bari, Association Ailerons, Association des Requins et des
Hommes. We also particularly thank the environmental data
providers for the quality and availability of their ocean color
products: NASA Ocean Biology (OB.DAAC), Greenbelt, MD,
USA, and the EU-Copernicus Marine Environment Monitoring
Service. The authors additionally thank Jose Blasco Munoz (EC-
JRC) for the videos' editing. The authors are thankful to the
reviewers for helping in editing the paper and providing valuable
comments. This publication is dedicated to FH, he is and will be,
missed for his extensive scientific knowledge, his contributions to
improving shark management and conservation, and his
exceptional human values and enthusiasm.
SUPPLEMENTARY MATERIAL
The Supplementary Material for this article can be found online
at: https://www.frontiersin.org/articles/10.3389/fmars.2022.
828412/full#supplementary-material
Supplementary Video 1 | LJF_NE_Pacific - Habitat animation with overlay of
observations for large juvenile females of blue shark in the NE Pacific.
Supplementary Video 2 | LJM_N_Atlantic - Habitat animation with overlay of
observations for large juvenile males of blue shark in the N Atlantic.
Supplementary Video 3 | AF_NE_Pacific - Habitat animation with overlay of
observations for adult males of blue shark in the NE Pacific.
Supplementary Video 4 | AF_Indian_NW_Pacific - Habitat animation with
overlay of observations for adult females of blue shark in the Indian and NW Pacific.
Supplementary Video 5 | AM_NE_Pacific - Habitat animation with overlay of
observations for adult males of blue shark in the NE Pacific.
Supplementary Video 6 | SJ_N_Atlantic - Habitat animation with overlay of
observations for small juveniles of blue shark in the N Atlantic.
Supplementary Video 7 | LJF_N_AtlanticHabitat animation with overlay of
observations for large juvenile females of blue shark in the N Atlantic.
Supplementary Data Sheet 1 | Supplementary Information.
Druon et al. Global-Scale Blue Shark Habitat
Frontiers in Marine Science | www.frontiersin.org April 2022 | Volume 9 | Article 82841222
REFERENCES
Adams, G. D., Flores, D., Galindo Flores, O., Aarestrup, K., and Svendsen, J. C.
(2016). Spatial Ecology of Blue Shark and Shortfin Mako in Southern Peru:
Local Abundance, Habitat Preferences and Implications for Conservation.
Endanger. Species Res. 31, 19–32. doi: 10.3354/esr00744
Aires-da-Silva,A.M.,andGallucci,V.F.(2007).DemographicandRiskAnalyses
Applied to Management and Conservation of theBlueShark(PrionaceGlauca)inthe
North Atlantic Ocean. Marine Freshw. Res. 58 (6), 570–580. doi: 10.1071/MF06156
Baudena, A., Ser-Giacomi, E., D’Onofrio, D., Capet, X., Cotté, C., Cherel, Y., et al.
(2021). Fine-Scale Structures as Spots of Increased Fish Concentration in the
Open Ocean. Sci. Rep. 11 (1), 1–13. doi: 10.1038/s41598-021-94368-1
Bigelow, K. A., Boggs, C. H., and He, X. I. (1999). Environmental Effects on
Swordfish and Blue Shark Catch Rates in the US North Pacific Longline
Fishery. Fish Oceanogr. 8 (3), 178–198. doi: 10.1046/j.1365-2419.1999.00105.x
Braun, C. D., Galuardi, B., and Thorrold, S. R. (2018). HMMoce: An R Package
for Improved Geolocation of Archival-Tagged Fishes Using a Hidden
Markov Method. Methods Ecol. Evol. 9 (5), 1212–1220. doi: 10.1111/2041-
210X.12959
Braun, C. D., Gaube, P., Sinclair-Taylor, T. H., Skomal, G. B., and Thorrold, S. R.
(2019). Mesoscale Eddies Release Pelagic Sharks From Thermal Constraints to
Foraging in the Ocean Twilight Zone. Proc. Natl. Acad. Sci. 116 (35), 17187–
17192. doi: 10.1073/pnas.1903067116
Braun, C. D., Kaplan, M. B., Horodysky, A. Z., and Llopiz, J. K. (2015). Satellite
Telemetry Reveals Physical Processes Driving Billfish Behavior. Anim.
Biotelem. 3 (1), 1–16. doi: 10.1186/s40317-014-0020-9
Briscoe, D. K., Hobday, A. J., Carlisle, A., Scales, K., Paige Eveson, J., Arrizabalaga,
H., et al. (2017). Ecological Bridges and Barriers in Pelagic Ecosystems. Deep
Sea Res. Part II: Top Stud. Oceanogr. 140, 182–192. doi: 10.1016/j.dsr2.
2016.11.004
Calich, H. J., and Campana, S. E. (2015). Mating Scars Reveal Mate Size in
Immature Female Blue Shark Prionace Glauca. J. Fish Biol. 86 (6), 1845–1851.
doi: 10.1111/jfb.12671
Campana, S. E. (2016). Transboundary Movements, Unmonitored Fishing
Mortality, and Ineffective International Fisheries Management Pose Risks for
Pelagic Sharks in the Northwest Atlantic. Can. J. Fish Aquat. Sci. 73 (10), 1599–
1607. doi: 10.1139/cjfas-2015-0502
Campana, S. E., Dorey, A., Fowler, M., Joyce, W., Wang, Z., Wright, D., et al.
(2011). Migration Pathways, Behavioural Thermoregulation and
Overwintering Grounds of Blue Sharks in the Northwest Atlantic. PloS One
6 (2), e16854. doi: 10.1371/journal.pone.0016854
Campana, S. E., Joyce, W., Fowler, M., and Showell, M. (2016). Discards, Hooking,
and Post-Release Mortality of Porbeagle (Lamna Nasus), ShortfinMako
(Isurus Oxyrinchus), and Blue Shark (Prionace Glauca) in the Canadian
Pelagic Longline Fishery. ICES J. Marine Sci. 73 (2), 520–528. doi: 10.1093/
icesjms/fsv234
Campana, S. E., Marks, L., Joyce, W., and Kohler, N. E. (2006). Effects of
Recreational and Commercial Fishing on Blue Sharks (Prionace Glauca) in
Atlantic Canada, With Inferences on the North Atlantic Population. Can. J.
Fish Aquat. Sci. 63 (3), 670–682. doi: 10.1139/f05-251
Carey, F. G., Scharold, J. V., and Kalmijn, A. J. (1990). Movements of Blue Sharks
(Prionace Glauca) in Depth and Course. Marine Biol. 106 (3), 329–342.
doi: 10.1007/BF01344309
Carpentieri, P., Nastasi, A., Sessa, M., and Srour eds, A. (2021). Incidental Catch of
Vulnerable Species in Mediterranean and Black Sea Fisheries –A Review. Gen.
Fish Comm. Mediterr. Stud. Rev. 101, 320. doi: 10.4060/cb5405en
Carvalho, F., Ahrens, R., Murie, D., Bigelow, K., Aires-Da-Silva, A., Maunder, M.
N., et al. (2015). Using Pop-Up Satellite Archival Tags to Inform Selectivity in
Fisheries Stock Assessment Models: A Case Study for the Blue Shark in the
South Atlantic Ocean. ICES J. Marine Sci. 72 (6), 1715–1730. doi: 10.1093/
icesjms/fsv026
Carvalho, F. C., Murie, D. J., HV Hazin, F., Hazin, H. G., Leite-Mourato, B., and
Burgess, G. H. (2011). Spatial Predictions of Blue Shark (Prionace Glauca)
Catch Rate and Catch Probability of Juveniles in the Southwest Atlantic. ICES J.
Marine Sci. 68 (5), 890–900. doi: 10.1093/icesjms/fsr047
Castro, J. A., and Mejuto, J. (1995). Reproductive Parameters of Blue Shark,
Prionace Glauca, and Other Sharks in the Gulf of Guinea. Marine Freshw. Res.
46 (6), 967–973. doi: 10.1071/MF9950967
Castro, J. I., Woodley, C. M., and Brudek, R. L. (1999). “A Preliminary Evaluation
of the Status of Shark Species,”in FAO Fisheries Techn. Pap, vol. 380. (Rome:
FAO).
Clarke, S. C., McAllister, M. K., Milner-Gulland, E. J., Kirkwood, G. P.,
Michielsens, C. G. J., Agnew, D. J., et al. (2006). Global Estimates of Shark
Catches Using Trade Records From Commercial Markets. Ecol. Lett. 9 (10),
1115–1126. doi: 10.1111/j.1461-0248.2006.00968.x
Coelho, R., Fernandez-Carvalho, J., Lino, P. G., and Santos, M. N. (2012). An
Overview of the Hooking Mortality of Elasmobranchs Caught in a Swordfish
Pelagic Longline Fishery in the Atlantic Ocean. Aquat. Living Resour. 25 (4),
311–319. doi: 10.1051/alr/2012030
Coelho, R., Mejuto, J., Domingo, A., Yokawa, K., Liu, K.-M., Cortés, E., et al.
(2018). Distribution Patterns and Population Structure of the Blue Shark
(Prionace Glauca) in the Atlantic and Indian Oceans. Fish. Fish 19 (1), 90–106.
doi: 10.1111/faf.12238
Cordova-Zavaleta, F., Mendo, J., Briones-Hernandez, S. A., Acuna-Perales, N.,
Gonzalez-Pestana, A., Alfaro-Shigueto, J., et al. (2018). Food Habits of the Blue
Shark, Prionace Glauca (Linnaeus 1758), in Waters Off Northern Peru. Fish.
Bull. 116 (3–4), 310–324. doi: 10.7755/FB.116.3-4.9
Cruz, M. M., Elenara Szynwelski, B., and RO de Freitas, T. (2021). Biodiversity on
Sale: The Shark Meat Market Threatens Elasmobranchs in Brazil. Aquat.
Conserv: Marine Freshw. Ecosyst.,1–14. doi: 10.1002/aqc.3710
Desbiens,A.A.,Roff,G.,Robbins,W.D.,Taylor,B.M.,Castro-Sanguino,C.,
Dempsey, A., et al. (2021). Revisiting the Paradigm of Shark-Driven Trophic
Cascades in Coral Reef Ecosystems. Ecology 102 (4), e03303. doi: 10.1002/ecy.3303
Dolgov, A. V., Drevetnyak, K. V., and Gusev, E. V. (2005). The Status of Skate
Stocks in the Barents Sea. J. Northwest Atlantic Fishery Sci. 35, 249–260. doi:
10.2960/J.v35.m522
Druon, J.-N., Chassot, E., Murua, H., and Lopez, J. (2017). Skipjack Tuna
Availability for Purse Seine Fisheries Is Driven by Suitable Feeding Habitat
Dynamics in the Atlantic and Indian Oceans. Front. Marine Sci. 4.
doi: 10.3389/fmars.2017.00315
Druon, J.-N., Fromentin, J.-M., Hanke, A. R., Arrizabalaga, H., Damalas, D.,
Tičina, V., et al. (2016). Habitat Suitability of the Atlantic Bluefin Tuna by Size
Class: An Ecological Niche Approach. Prog. Oceanogr. 142, 30–46.
doi: 10.1016/j.pocean.2016.01.002
Druon, J.-N., Gascuel, D., Gibin, M., Zanzi, A., Fromentin, J.-M., Colloca, F., et al.
(2021). Mesoscale Productivity Fronts and Local Fishing Opportunities in the
European Seas. Fish Fish 22, 1227–1247. doi: 10.1111/faf.12585
Druon, J.-N., Hélaouët, P., Beaugrand, G., Fromentin, J.-M., Palialexis, A., and
Hoepffner, N. (2019). Satellite-Based Indicator of Zooplankton Distribution for
Global Monitoring. Sci. Rep. 9 (1), 1–13. doi: 10.1038/s41598-019-41212-2
Dulvy, N. K., Pacoureau, N., Rigby, C. L., Pollom, R. A., Jabado, R. W., Ebert, D. A.,
et al. (2021). Overfishing Drives Over One-Third of All Sharks and Rays
Toward a Global Extinction Crisis. Curr. Biol. 31 (21), 4773–4787.
doi: 10.1016/j.cub.2021.08.062
Elliott, R. (2020). “Spatio-Temporal Patterns in Movement, Behaviour and Habitat
Use by a Pelagic Predator, the Blue Shark (Prionace Glauca), Assessed Using
Satellite Tags.”University of Auckland, New Zealand: PhD Thesis,
ResearchSpace@ Auckland.
European Union (2020). “COUNCIL REGULATION (EU) 2020/123 of 27 January
2020 Fixing for 2020 the Fishing Opportunities for Certain Fish Stocks and
Groups of Fish Stocks, Applicable in Union Waters and, for Union Fishing
Vessels, in Certain Non-Union Waters”.Official Journal of the European Union.
European Union (2021). “COUNCIL REGULATION (EU) 2021/92 of 28 January
2021 Fixing for 2021 the Fishing Opportunities for Certain Fish Stocks and
Groups of Fish Stocks, Applicable in Union Waters and, for Union Fishing
Vessels, in Certain Non-Union Waters.”
Ferretti, F., Myers, R. A., Serena, F., and Lotze, H. K. (2008). Loss of Large
Predatory Sharks From the Mediterranean Sea. Conserv. Biol. 22 (4), 952–964.
doi: 10.1111/j.1523-1739.2008.00938.x
Ferretti, F., Worm, B., Britten, G. L., Heithaus, M. R., and Lotze, H. K. (2010).
Patterns and Ecosystem Consequences of Shark Declines in the Ocean. Ecol.
Lett. 13 (8), 1055–1071. doi: 10.1111/j.1461-0248.2010.01489.x
Frid, A., Baker, G., and Dill, L. (2008). Do Shark Declines Create Fear-Released
Systems? Oikos 117 (2), 191–201. doi: 10.1111/j.2007.0030-1299.16134.x
Fujinami, Y., Shiozaki, K., Hiraoka, Y., Semba, Y., Ohshimo, S., and Kai, M. (2021).
Seasonal Migrations of Pregnant Blue Sharks Prionace Glauca in the
Druon et al. Global-Scale Blue Shark Habitat
Frontiers in Marine Science | www.frontiersin.org April 2022 | Volume 9 | Article 82841223
Northwestern Pacific. Marine Ecol. Prog. Ser. 658, 163–179. doi: 10.3354/
meps13557
Godø, O. R., Samuelsen, A., Macaulay, G. J., Patel, R., Sætre Hjøllo, S., Horne, J.,
et al. (2012). Mesoscale Eddies Are Oases for Higher Trophic Marine Life. PloS
One 7 (1), e30161. doi: 10.1371/journal.pone.0030161
Gonzalez-Pestana, A., Kouri, C., and Velez-Zuazo, X. (2016). Shark Fisheries in
the Southeast Pacific: A 61-Year Analysis From Peru. F1000Research 3, 164.
doi: 10.12688/f1000research.4412.2
GUID CMEMS. (2021). “GLOBAL_MULTIYEAR_BGC_001_033, Quality
Information Document.”Available at: https://catalogue.marine.copernicus.eu/
documents/QUID/CMEMS-GLO-QUID-001-033.pdf.
Hazen, E. L., Scales, K. L., Maxwell, S. M., Briscoe, D. K., Welch, H., Bograd, S. J.,
et al. (2018). A Dynamic Ocean Management Tool to Reduce Bycatch and
Support Sustainable Fisheries. Sci. Adv. 4 (5), eaar3001. doi: 10.1126/
sciadv.aar3001
Hazin, F., Kihara, K., Otsuka, K., Boeckman, C. E., and Leal, E. C. (1994).
Reproduction of the Blue Shark Prionace Glauca in the South-Western
Equatorial Atlantic Ocean. Fish Sci. 60 (5), 487–915. doi: 10.2331/fishsci.60.487
Holden, M. J. (1974). Problems in the Rational Exploitation of Elasmobranch
Populations and Some Suggested Solutions. Sea Fish Res. 117–37.
Hu, S., and Fedorov, A. V. (2020). Indian Ocean Warming as a Driver of the North
Atlantic Warming Hole. Nat. Commun. 11 (1), 1–11. doi: 10.1038/s41467-020-
18522-5
Hu, C., Lee, Z., and Franz, B. (2012). Chlorophyll a Algorithms for Oligotrophic
Oceans: A Novel Approach Based on Three-Band Reflectance Difference.
J. Geophys. Res.: Oceans 117 (C1), 1–25. doi: 10.1029/2011JC007395
Hutchinson, M., Siders, Z., Stahl, J., and Bigelow, K. (2021). Quantitative Estimates
of Post-Release Survival Rates of Sharks Captured in Pacific Tuna Longline
Fisheries Reveal Handling and Discard Practices That Improve Survivorship.
PIFSC Data Rep., DR-21-001. doi: 10.25923/0m3c-2577
ICCAT (2015) “Report of the 2015 ICCAT Blue Shark Stock Assessment Session.”
ICCAT Madrid. Available at: https://www.iccat.int/Documents/SCRS/DetRep/
BSH_SA_ENG.PDF.
ICCAT (2020) “Report for Biennial Period 2018-19 PART II, Madrid, Spain.”
Collective Volume of Scientific Papers ICCAT 2 (SCRS). Available at: https://
www.iccat.int/Documents/BienRep/REP_EN_18-19_II-2.pdf.
Kai, M., and Fujinami, Y. (2020). Estimation of Mean Movement Rates for Blue
Sharks in the Northwestern Pacific Ocean. Anim. Biotelem. 8 (1), 1–8.
doi: 10.1186/s40317-020-00223-x
Konan, K. J., Yves-Narcisse Kouamé, K., Issa Ouattara, N., and Koné, T. (2018).
Feeding Habits of the Blue Shark Prionace Glauca (Linnaeu) Off the Coastal
Waters of Ivory Coast (West Africa). J. Biodivers. Environ. Sci. (JBES) 12 (3),
192–200.
Kotas, J. E., dos Santos, S., Guedes de Azevedo, V., Meneses de Lima, J. H., Dias
Neto, J., and Fernández Lin, C. (2000). Observations on Shark By-Catch in the
Monofilament Longline Fishery Off Southern Brazil and the National Ban on
Finning. IBAMA–REVIZEE Res. 8.
Lehodey, P., Conchon, A., Senina, I., Domokos, R., Calmettes, B., Jouanno, J., et al.
(2015). Optimization of a Micronekton Model With Acoustic Data. ICES J.
Marine Sci. 72 (5), 1399–1412. doi: 10.1093/icesjms/fsu233
Lehodey, P., Murtugudde, R., and Senina, I. (2010). Bridging the Gap From Ocean
Models to Population Dynamics of Large Marine Predators: A Model of Mid-
Trophic Functional Groups. Prog. Oceanogr. 84 (1), 69–84. doi: 10.1016/
j.pocean.2009.09.008
Litchman, E., Ohman, M. D., and Kiørboe, T. (2013). Trait-Based Approaches to
Zooplankton Communities. J. Plankton Res. 35 (3), 473–484. doi: 10.1093/
plankt/fbt019
Lopez, J., Lennert-Cody, C., Maunder, M., Xu, H., Brodie, S., Jacox, M., et al.
(2019). “Developing Alternative Conservation Measures for Bigeye Tuna in the
Eastern Pacific Ocean: A Dynamic Ocean Management Approach,”in
American Fisheries Society & The Wildlife Society 2019 Joint Annual
Conference. AFS. Available at: https://www.iattc.org/Meetings/Meetings2019/
SAC-10/INF/_English/SAC-10-INF-D_Bigeye%20tuna%20Dynamic%
20Ocean%20Management.pdf.
Mancusi, C., Baino, R., Fortuna, C., Gil De Sola, L., Morey, G., Nejmeddine Bradai,
M., et al. (2020) “MEDLEM Database, a Data Collection on Large
Elasmobranchs in the Mediterranean and Black Seas.”Available at: 10.12681/
mms.21148.
Maunder, M. N., Sibert, J. R., Fonteneau, A., Hampton, J., Kleiber, P., and Harley,
S. J. (2006). Interpreting Catch Per Unit Effort Data to Assess the Status of
Individual Stocks and Communities. ICES J. Marine Sci. 63 (8), 1373–1385.
doi: 10.1016/j.icesjms.2006.05.008
Maxwell, S. M., Scales, K. L., Bograd, S. J., Briscoe, D. K., Dewar, H., Hazen, E. L.,
et al. (2019). Seasonal Spatial Segregation in Blue Sharks (Prionace Glauca)
by Sex and Size Class in the Northeast Pacific Ocean. Divers. Distrib. 25 (8),
1304–1317. doi: 10.1111/ddi.12941
McCord, M. E., and Campana, S. E. (2003). A Quantitative Assessment of the Diet
of the Blue Shark (Prionace Glauca) Off Nova Scotia, Canada. J. Northwest
Atlantic Fishery Sci. 32, 57–63. doi: 10.2960/J.v32.a4
Megalofonou, P., Damalas, D., and De Metrio, G. (2009). Biological
Characteristics of Blue Shark, Prionace Glauca, in the Mediterranean Sea. J.
Marine Biol. Assoc. United Kingdom 89 (6), 1233–1242. doi: 10.1017/
S0025315409000216
Megalofonou, P., Damalas, D., Yannopoulos, C., De Metrio, G., Deflorio, M., de la
Serna, J. M., et al. (2000). By Catches and Discards of Sharks in the Large
Pelagic Fisheries in the Mediterranean Sea. Final Rep. Project 97/50.
Megalofonou, P., Yannopoulos, C., Damalas, D., De Metrio, G., de la Serna, J. M.,
and Macias, D. (2005). Incidental Catch and Estimated Discards of Pelagic
Sharks From the Swordfish and Tuna Fisheries in the Mediterranean Sea. Fish.
Bull. 103 (4), 620–634.
Miller, P. I., Scales, K. L., Ingram, S. N., Southall, E. J., and Sims, D. W. (2015).
Basking Sharks and Oceanographic Fronts: Quantifying Associations in the
North-East Atlantic. Funct. Ecol. 29 (8), 1099–1109. doi: 10.1111/1365-
2435.12423
Mucientes, G. R., Queiroz, N., Sousa, L. L., Tarroso, P., and Sims, D. W. (2009).
Sexual Segregation of Pelagic Sharks and the Potential Threat From Fisheries.
Biol. Lett. 5 (2), 156–159. doi: 10.1098/rsbl.2008.0761
Musick, J. A., Burgess, G., Cailliet, G., Camhi, M., and Fordham, S. (2000).
Management of Sharks and Their Relatives (Elasmobranchii). Fisheries 25 (3),
9–13. doi: 10.1577/1548-8446(2000)025<0009:MOSATR>2.0.CO;2
Musyl, M. K., Brill, R. W., Curran, D. S., Fragoso, N. M., McNaughton, L. M.,
Nielsen, A., et al. (2011). Postrelease Survival, Vertical and Horizontal
Movements, and Thermal Habitats of Five Species of Pelagic Sharks in the
Central Pacific Ocean. Fish. Bull. 109 (4), 341–368.
Musyl, M. K., and Gilman, E. L. (2018). Post-Release Fishing Mortality of Blue
(Prionace Glauca) and Silky Shark (Carcharhinus Falciformes) From a
Palauan-Based Commercial Longline Fishery. Rev. Fish Biol. Fish 28 (3),
567–586. doi: 10.1007/s11160-018-9517-2
Musyl, M. K., and Gilman, E. L. (2019). Meta-Analysis of Post-Release Fishing
Mortality in Apex Predatory Pelagic Sharks and White Marlin. Fish Fish 20 (3),
466–500. doi: 10.1111/faf.12358
Myers, R. A., and Worm, B. (2005). Extinction, Survival or Recovery of Large
Predatory Fishes. Philos.Trans.R.Soc.B:Biol.Sci.360 (1453), 13–20.
doi: 10.1098/rstb.2004.1573
Nakano, H. (1994). Age, Reproduction and Migration of Blue Shark [Prionace] in
the North Pacific Ocean. Bulletin-National Res. Institute Far Seas Fish (Japan)
31, 141–256.
Nakano, H., and Nagasawa, K. (1996). Distribution of Pelagic Elasmobranchs
Caught by Salmon Research Gillnets in the North Pacific. Fish Sci. 62 (6), 860–
865. doi: 10.2331/fishsci.62.860
Neubauer, P., Large, K., and Brouwer, S. (2021). Stock Assessment of Southwest
Pacific Blue Shark,”WCPFC-SC17-2021/SA-WP-03. Report to the WCPFC
Scientific Committee. Seventeenth Regular Session, 66.
Nosal, A. P., Cartamil, D. P., Wegner, N. C., Lam, C. H., and Hastings, P.
A. (2019). Movement Ecology of Young-Of-the-Year Blue Sharks
Prionace Glauca and Shortfin Makos Isurus Oxyrinchus Within a Putative
Binational Nursery Area. Marine Ecol. Prog. Ser. 623, 99–115. doi: 10.3354/
meps13021
Okes, N., and Sant, G. (2019). An Overview of Major Shark Traders, Catchers and
Species. TRAFFIC Cambridge UK.
Olson, D. B., Hitchcock, G. L., Mariano, A. J., Ashjian, C. J., Peng, G., Nero, R. W.,
et al. (1994). Life on the Edge: Marine Life and Fronts. Oceanography 7 (2), 52–
60. doi: 10.5670/oceanog.1994.03
Pacoureau, N., Rigby, C. L., Kyne, P. M., Sherley, R. B., Winker, H., Carlson, J. K.,
et al. (2021). Half a Century of Global Decline in Oceanic Sharks and Rays.
Nature 589 (7843), 567–571. doi: 10.1038/s41586-020-03173-9
Druon et al. Global-Scale Blue Shark Habitat
Frontiers in Marine Science | www.frontiersin.org April 2022 | Volume 9 | Article 82841224
Panigada, S., Donovan, G. P., Druon, J.-N., Lauriano, G., Pierantonio, N., Pirotta,
E., et al. (2017). Satellite Tagging of Mediterranean Fin Whales: Working
Towards the Identification of Critical Habitats and the Focussing of Mitigation
Measures. Sci. Rep. 7 (1), 1–12. doi: 10.1038/s41598-017-03560-9
Polovina, J. J., Howell, E., Kobayashi, D. R., and Seki, M. P. (2001). The Transition
Zone Chlorophyll Front, a Dynamic Global Feature Defining Migration and
Forage Habitat for Marine Resources. Prog. Oceanogr. 49 (1–4), 469–483.
doi: 10.1016/S0079-6611(01)00036-2
Pons,M.,Watson,J.T.,Ovando,D.,Andraka,S.,Brodie,S.,Domingo,A.,etal.(2022).
Trade-Offs Between Bycatch and Target Catches in Static Versus Dynamic Fishery
Closures. Proc. Natl. Acad. Sci. 119, (4). doi: 10.1073/pnas.2114508119
Porcher, I. F., Darvell, B. W., and Ziegler, I. (2021). Shark Conservation–Analysis
and Synthesis. Preprints doi: 10.20944/preprints202102.0145.v4
Porsmoguer, S. B., Bănaru, D., Boudouresque, C. F., Dekeyser, I., and Almarcha, C.
(2015). Hooks Equipped With Magnets Can Increase Catches of Blue Shark
(Prionace Glauca) by Longline Fishery. Fish Res. 172, 345–351. doi: 10.1016/
j.fishres.2015.07.016
Pratt, HAROLD L. (1979). Reproduction in the Blue Shark, Prionace Glauca.
Fishery Bull. 77 (2), 445–470.
Queiroz, N., Humphries, N. E., Couto, A., Vedor, M., Da Costa, I., MM
Sequeira, A., et al. (2019). Global Spatial Risk Assessment of Sharks Under
the Footprint of Fisheries. Nature 572 (7770), 461–466. doi: 10.1038/s41586-
019-1444-4
Queiroz, N., Humphries, N. E., Noble, L. R., Santos, A. M., and Sims, D. W. (2012).
Spatial Dynamics and Expanded Vertical Niche of Blue Sharksk in
Oceanographic Fronts Reveal Habitat Targets for Conservation. PloS One 7
(2), e32374. doi: 10.1371/journal.pone.0032374
Sacchi, J. (2021). Overview of Mitigation Measures to Reduce the Incidental Catch
of Vulnerable Species in Fisheries. Gen. Fish Comm Mediterranean Stud. Rev.
100, 1–124. doi: 10.4060/cb5049en
Saul, S., Brooks, E. N., and Die, D. (2020). How Fisher Behavior Can Bias Stock
Assessment: Insights From an Agent-Based Modeling Approach. Can. J. Fish
Aquat. Sci. 77 (11), 1794–1809. doi: 10.1139/cjfas-2019-0025
Scales, K. L., Hazen, E. L., Jacox, M. G., Castruccio, F., Maxwell, S. M., Lewison, R.
L., et al. (2018). Fisheries Bycatch Risk to Marine Megafauna Is Intensified in
Lagrangian Coherent Structures. Proc. Natl. Acad. Sci. 115 (28), 7362–7675.
doi: 10.31230/osf.io/nmvwz
Serena, F., and Silvestri, R. (2018). Preliminary Observations on Juvenile Shark
Catches as By-Catch of the Italian Fisheries With Particular Attention to the
Tuscany Coasts.”in. Proc. 7 Congress Codice Armonico 18b/Scientific Section.
Rosignano Marittimo (Li), 158–169.
Simpfendorfer, C. A., and Dulvy, N. K. (2017). Bright Spots of Sustainable Shark
Fishing. Curr. Biol. 27 (3), R97–R98. doi: 10.1016/j.cub.2016.12.017
Sims, D. W. (2005). Differences in Habitat Selection and Reproductive Strategies of
Male and Female Sharks. Sexual Segregation Vertebrates, 127–147.
Stevens, J. D. (1999). Sharks. Subsequent Edition (March 1, 1999) (New York:
Checkmark Books), p. 240.
Stevens, J. D., Bonfil, R., Dulvy, N. K., and Walker, P. A. (2000). The Effects of
Fishing on Sharks, Rays, and Chimaeras (Chondrichthyans), and the
Implications for Marine Ecosystems. ICES J. Marine Sci. 57 (3), 476–494.
doi: 10.1006/jmsc.2000.0724
Strasburg, D. W. (1958). Distribution, Abundance, and Habits of Pelagic Sharks in
the Central Pacific Ocean. Fisheries 1, 2S.
Tanaka, K. R., Van Houtan, K. S., Mailander, E., Dias, B. S., Galginaitis, C.,
O’Sullivan, J., et al.(2021). North Pacific Warming Shifts the Juvenile Range of a
Marine Apex Predator. Sci. Rep. 11 (1), 1–9. doi: 10.1038/s41598-021-82424-9
Tew Kai, E., and Marsac, F. (2010). Influence of Mesoscale Eddies on Spatial
Structuring of Top Predators’Communities in the Mozambique Channel.
Prog. Oceanogr 86 (1–2), 214–223. doi: 10.1016/j.pocean.2010.04.010
Vandeperre, F., Aires-da-Silva, A., Fontes, J., Santos, M., Serrão Santos, R., and
Afonso, P. (2014a). Movements of Blue Sharks (Prionace Glauca) Across Their
Life History. PloS One 9 (8), e103538. doi: 10.1371/journal.pone.0103538
Vandeperre, F., Aires-da-Silva, A., Santos, M., Ferreira, R., Bolten, A. B., Serrao
Santos, R., et al. (2014b). Demography and Ecology of Blue Shark (Prionace
Glauca) in the Central North Atlantic. Fish Res. 153, 89–102. doi: 10.1016/
j.fishres.2014.01.006
Vandeperre, F., Aires-da-Silva, A., Lennert-Cody, C., Serrão Santos, R., and
Afonso, P. (2016). Essential Pelagic Habitat of Juvenile Blue Shark (Prionace
Glauca) Inferred From Telemetry Data. Limnol Oceanogr 61 (5), 1605–1625.
doi: 10.1002/lno.10321
Vedor, M., Mucientes, G., Hernández-Chan, S., Rosa, R., Humphries, N., Sims, D.
W., et al. (2021a). “Oceanic Diel Vertical Movement Patterns of Blue Sharks
Vary With Water Temperature and Productivity to Change Vulnerability to
Fishing. Front. Marine Sci. 891. doi: 10.3389/fmars.2021.688076
Vedor, M., Queiroz, N., Mucientes, G., Couto, A., da Costa, I., Dos Santos, A.,
et al. (2021b). “Climate-Driven Deoxygenation Elevates Fishing Vulnerability
for the Ocean’s Widest Ranging Shark. Elife 10, e62508. doi: 10.7554/
eLife.62508
Walker, P. A., and Cocks, K. D. (1991). HABITAT: A Procedure for Modelling a
Disjoint Environmental Envelope for a Plant or Animal Species. Global Ecol.
Biogeogr Lett. 1, 108–118. doi: 10.2307/2997706
Watanabe, Y., Nakamura, I., and Chiang, W.-C. (2021). Behavioural
Thermoregulation Linked to Foraging in Blue Sharks. Marine Biol. 168 (11),
161. doi: 10.1007/s00227-021-03971-3
Wearmouth, V. J., and Sims, D. W. (2008). Sexual Segregation in Marine Fish,
Reptiles, Birds and Mammals: Behaviour Patterns, Mechanisms and
Conservation Implications. Adv. Marine Biol. 54, 107–170. doi: 10.1016/
S0065-2881(08)00002-3
West, G., Stevens, J., and Basson, M. (2004). “Assessment of Blue Shark Population
Status in the Western South Pacific(Hobart, Tasmania, Australia: AFMA
Project R01/1157.”CSIRO Marine Research).
Young, J. W., Lansdell, M. J., Campbell, R. A., Cooper, S. P., Juanes, F., and Guest,
M. A. (2010). Feeding Ecology and Niche Segregation in Oceanic Top
Predators Off Eastern Australia. Marine Biol. 157 (11), 2347–2368.
doi: 10.1007/s00227-010-1500-y
Conflict of Interest: MM is employed by Pelagic Research Group LLC.
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.
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.
Copyright © 2022 Druon, Campana, Vandeperre, Hazin, Bowlby, Coelho, Queiroz,
Serena, Abascal, Damalas, Musyl, Lopez, Block, Afonso, Dewar, Sabarros, Finucci,
Zanzi, Bach, Senina, Garibaldi, Sims, Navarro, Cermeño, Leone, Diez, Zapiain,
Deflorio, Romanov, Jung, Lapinski, Francis, Hazin and Travassos. 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.
Druon et al. Global-Scale Blue Shark Habitat
Frontiers in Marine Science | www.frontiersin.org April 2022 | Volume 9 | Article 82841225