Copy igh © 2021 by he au ho (s). Published he e unde license by he Resilience Alliance.
Rubio, I., J. Hileman, and E. Ojea. 2021. Social connec i i y and adap i e capaci y s a egies in la ge-scale ishe ies. Ecology and
Socie y 26(2):42. h ps://doi.o g/10.5751/ES-12395-260242
Resea ch
Social connec i i y and adap i e capaci y s a egies in la ge-scale ishe ies
I a xe Rubio 1,2, Jacob Hileman 3 and Elena Ojea 2
ABSTRACT. La ge-scale ishe ies a e impo an social-ecological sys ems ha a e inc easingly being h ea ened by global clima e
change. Adap i e capaci y is key o mo ing ishe ies on o clima e esilien pa hways, howe e , implemen ing policies o imp o e
adap i e capaci y is challenging gi en he many di e se s akeholde s in ol ed in ishe ies. P e ious esea ch sugges s social ne wo ks
a e in eg al o adap i e capaci y because social connec i i y can enable, o cons ain, knowledge and in o ma ion sha ing. We examine
he ne wo k o communica ion among s akeholde s in he Basque opical una eeze pu se seine ishe y in he eas e n A lan ic Ocean.
We use clus e analysis, desc ip i e s a is ics, and exponen ial andom g aph models o assess whe he di e en ypes o ac o s, occupying
di e en ne wo k posi ions, alue simila adap i e capaci y s a egies. The esul s indica e ha many ac o ypes a e equen ly connec ed
wi hin he ishe y. P e e ences o adap i e capaci y s a egies a y wi hin and ac oss ac o ypes, and he p e e ences o highly cen al
ac o s a e gene ally mo e homogeneous and na owly ocused. All ac o s ag ee on he impo ance o he social o ganiza ion domain
om adap i e capaci y, while ishing indus y ep esen a i es end o ha e he mos holis ic pe spec i e on adap i e capaci y o e all.
We discuss he implica ions o hese indings as hey ela e o policies o suppo ing adap i e capaci y and clima e esilien ishe ies.
Key Wo ds: adap a i e capaci y; global clima e change; go e nance; social ne wo k analysis; opical unas
INTRODUCTION
La ge-scale ishe ies a e inc easingly being h ea ened by a ange
o an h opogenic p essu es, including pollu ion, o e ishing,
habi a deg ada ion, and clima e change (Halpe n e al. 2019).
Clima e change, in pa icula , has he po en ial o ans o m
ma ine ecosys ems a a global scale, and is al eady impac ing he
demog aphy, dis ibu ion, and phenology o ma ine species
(Ga cía Molinos e al. 2016, Poloczanska e al. 2016, F ee e al.
2019). I is widely ecognized ha hese impac s will esul in
ishe ies p oduc ion being al e ed and ca ch po en ial
edis ibu ed egionally, and ha his has signi ican
socioeconomic implica ions o hose wi h ishe ies-dependen
li elihoods (Lam e al. 2016, F ee e al. 2019). Tha hese
dis ibu ion shi s a e ansbounda y, ha hey occu ac oss
human-de ined poli ical and adminis a i e bounda ies, u he
sugges s con lic o e new sha ed ishe ies may a ise (e.g., Spijke s
and Boons a 2017), which poses an addi ional bu den o he
sus ainable go e nance o ishe ies (Pinsky e al. 2018). Indeed,
many o he expec ed impac s o clima e change a e al eady being
el in bo h he ecological and social dimensions o ishe ies (IPCC
2019).
Fishe ies a e complex social-ecological sys ems (SES)
cha ac e ized by in e dependencies among nume ous ecological,
social, economic, and ins i u ional componen s ac oss mul iple
scales (Be kes 2011, Pa elow 2018). Human beha io al
esponses, in his case ishing p ac ices, eac o ecological change
and mul iple social and economic cons ain s, and also d i e
ecological change (Pe y e al. 2011). An eme ging iew o
ishe ies managemen emphasizes he need o in eg a e he holis ic
pe spec i e o SES app oaches o imp o e solu ions o cu en
ishing esou ce dilemmas. These include sus ainable
managemen o ishe ies and p omo ing esilien ishe ies
(Pa elow and Boda 2015, Cenek and F anklin 2017, Ma shall e
al. 2018, Holsman e al. 2019), which a e especially impo an in
he con ex o apid clima ic change (Ogie e al. 2020). Howe e ,
accoun ing o such in e connec ions is a di icul ask in
empi ical esea ch.
The use o ne wo k analysis as a means o unde s anding
in e dependencies in complex SES has eme ged as i s own sub-
ield o go e nance esea ch o e he pas wo decades (Janssen
e al. 2006, Bodin and Tengö 2012, Bodin 2017, Bodin e al. 2019,
Ba nes e al. 2020), including a subs an ial numbe o empi ical
s udies o ishe ies (e.g., Fulle e al. 2017, Alexande e al. 2018,
Yle yinen e al. 2018). Social ne wo ks ha e played a c ucial ole
in cases whe e di e se public and p i a e ac o s ha e come
oge he o e ec i ely add ess esou ce go e nance challenges
(Olsson e al. 2008, Cinne e al. 2012, Lleona e al. 2014, Ri e a
e al. 2014, Pi a e al. 2016). Mo eo e , social ne wo ks can help
enable collec i e ac ion and knowledge sha ing a ound adap i e
capaci y (Adge 2003, Pelling e al. 2008), which is key o he
e ec i e go e nance and long- e m sus ainabili y o ishe ies in
he ace o global clima e change (Lindeg en and B ande 2018).
Adap i e capaci y is de ined as “ he abili y o sys ems,
ins i u ions, humans and o he o ganisms o adjus o po en ial
damage, o ake ad an age o oppo uni ies, o o espond o
consequences” (IPCC 2018:542). We d aw on esea ch om
Cinne e al. (2018) o ame adap i e capaci y in he social
dimension o ishe ies go e nance, which is composed o i e
domains: social o ganiza ion, asse s, lexibili y, lea ning, and
agency. This amewo k is mo e comp ehensi e han con en ional
app oaches, whe e in es men s o imp o e indi iduals’ capaci ies
o adap o changes end o ocus on a na ow iew o adap i e
capaci y, o example, building inancial asse s (Ba nes e al.
2020). Thus, unde s anding which domains a e mos alued by
ac o s in a ishe y can aid he de elopmen o speci ic adap a ion
s a egies and policies o suppo ing sus ainable and esilien
ishe ies.
1Basque Cen e o Clima e Change BC3, 2Fu u e Oceans Lab, CIM-Uni e sidade de Vigo, Spain, 3S ockholm Resilience Cen e, S ockholm
Uni e si y, S ockholm, Sweden
Ecology and Socie y 26(2): 42
h ps://www.ecologyandsocie y.o g/ ol26/iss2/a 42/
In his pape , we examine he social ne wo k o communica ion
among s akeholde s in a la ge-scale in e na ional ishe y: he
Basque opical una ishe y. The objec i e is o in es iga e
whe he social connec i i y is linked o ac o s’ p e e ences o
adap i e capaci y s a egies, wi h he goal o main aining he
ishe y in he u u e (i.e., o be sus ainable) and in he ace o
global clima e change (i.e., o be esilien ). Adap a ion ac ions a e
mani es a ions o adap i e capaci y (Smi and Wandel 2006), and
ou aim he e is o iden i y in which adap i e capaci y domains
ac ions a e likely o ha e he g ea es in luence, and which ac o s
migh acili a e hem. To his end, we use desc ip i e s a is ics and
exponen ial andom g aph models (ERGMs) o explo e wo
b oad esea ch ques ions: (1) Which adap i e capaci y s a egies
a e alued by which ac o s? (2) Do he adap i e capaci y s a egies
alued by ac o s a y acco ding o hei posi ions in he ne wo k?
To add ess hese esea ch ques ions, we i s ame he Basque
opical una ishe y as a complex SES and iden i y i s co e
componen s. Examining he composi ion o he SES is necessa y
o co ec ly iden i y all ele an s akeholde s in he go e nance
ne wo k. Nex , we desc ibe he in-dep h, high-le el expe
in e iews we used o alida e he cha ac e iza ion o he ishe y
as a SES, measu e social connec i i y among ac o s, and elici
ac o s’ p e e ences o di e en adap i e capaci y s a egies. We
employ ne wo k analysis o s udy social connec i i y, i.e.,
communica ion, in he ishe y. Ou empi ical indings
demons a e he link be ween adap i e capaci y p e e ences and
he s uc u e o social connec i i y, and sugges se e al a ge ed
adap a ion s a egies. Mo eo e , he app oach helps iden i y
cen al ac o s wi h pa icula iews on adap i e capaci y. We hen
discuss he indings as hey ela e o adap i e capaci y o
sus ainable and clima e esilien ishe ies, and highligh policy
implica ions and conside a ions o u u e esea ch.
THEORY
La ge-scale ishe ies as complex social-ecological sys ems
Social-ecological sys ems heo y desc ibes how human socie ies
and he en i onmen a e in e connec ed and co-e ol e ac oss
spa ial and empo al scales, o ming complex adap i e sys ems
(P eise e al. 2018). In his pape , we d aw on he wo k o Os om
and McGinnis (Os om 2009, McGinnis and Os om 2014) o
es ablish he heo e ical aming o s udying he Basque opical
una SES. McGinnis and Os om (2014) p o ide a amewo k o
s udying SES ha has been b oadly used in esou ce go e nance
se ings, including wa e , o es s, and ishe ies, among o he s, as
a ool o diagnose sus ainabili y (Pa elow 2018). In he con ex
o ishe ies, he SES amewo k has mos ly been applied o he
s udy o small-scale ishe ies (e.g., Leslie e al. 2015, London e
al. 2017). A p esen , ela i ely ew empi ical s udies ha e applied
he amewo k o la ge-scale ishe ies (see Pa elow 2018). Gi en
he economic impo ance o la ge-scale ishe ies a ound he
wo ld, coupled wi h he impac s global clima e change is al eady
ha ing on ishe ies, expanding he scope o empi ical SES s udies
o include la ge-scale ishe ies is a key con ibu ion o he
li e a u e.
Eps ein e al. (2014) in es iga ed how he SES amewo k could
be scaled up om small-scale o la ge-scale ishe ies, using he
A lan ic Blue in una ishe y as a case s udy and ocusing on he
In e na ional Commission o he Conse a ion o A lan ic
Tunas (ICCAT) go e nance egime. They highligh ed he
omission o esou ce use s om he ICCAT as a majo issue, and
discussed how i de i ed om bo h a lack o in o ma ion and he
in e na ional na u e o he go e nance sys em. Acco ding o he
au ho s, ule making occu ed a he in e na ional le el wi h li le
di ec inpu om esou ce use s, which dis ega ds he p inciple
o use pa icipa ion and collec i e-choice ules ha SES heo y
has iden i ied as necessa y o e ec i e and sus ainable
go e nance (Os om 1990, Cox e al. 2010). In iew o hese
indings, o add ess he esou ce use s limi a ion we de ine a la ge-
scale ishe y as an SES using a bo om-up app oach; we include
local-le el esou ce use s om an in e na ional ishe y and hei
in e ac ions wi h ac o s a highe le els, i.e., na ional, egional,
and in e na ional o ganiza ions. The esou ce use s he e a e a
he base o he SES de ini ion in he con ex o he ICCAT
go e nance egime. As a no e on e minology, we use “ac o s”
and “o ganiza ions” in e changeably.
F aming adap i e capaci y in la ge-scale ishe ies
We use he coas al communi ies adap i e capaci y amewo k
de eloped by Cinne e al. (2018), as well as ecommenda ions
om o he ecen s udies (Whi ney e al. 2017, F eduah e al.
2018), in ou o mula ion o adap i e capaci y in la ge-scale
ishe ies. As p e iously men ioned, his amewo k di ides
adap i e capaci y o clima e change in o i e domains: asse s,
lexibili y, social o ganiza ion, lea ning, and agency. Asse s
include all he p i a e and public esou ces ac o s can access in
imes o need, e.g., sa ings, echnology, and heal h ca e. Flexibili y
e e s o he ins i u ional o indi idual capaci y o swi ching
be ween adap a ion s a egies o deal wi h change, e.g.,
di e si ying income sou ces. Social o ganiza ion desc ibes how
ac o s a e in e connec ed, and whe he such connec ions enable
o inhibi coope a ion, collec i e ac ion, and knowledge sha ing,
e.g., building connec ions be ween he ishing indus y and
scien i ic o ganiza ions o c ea e access o in o ma ion ha
acili a es adap a ion. Lea ning e e s o ac o s’ capaci ies o
gene a e and p ocess in o ma ion, and assess po en ial esponse
s a egies, e.g., being in o med abou he s a us o ish s ocks.
Agency is he abili y o ac o s o choose om mul iple s a egies
when esponding o en i onmen al change, and o mobilize he
o he adap i e capaci y domains.
These i e b oad domains o m he concep ual ounda ion o he
dimensions o adap i e capaci y. Thus, in o de o ope a ionalize
he amewo k and apply i o ou case s udy o he Basque
opical una ishe y, we de ine a se o speci ic adap i e capaci y
s a egies o each domain o adap i e capaci y, d awing on
examples om he li e a u e (Whi ney e al. 2017, F eduah e al.
2018; Table 1).
Social connec i i y and adap i e capaci y p e e ences
A social ne wo k is an in e connec ed sys em comp ising a se o
ac o s (nodes) such as o ganiza ions, indi iduals, ins i u ions, o
communi ies, and he o mal and/o in o mal ela ionships (links)
ha exis among hem (Sco and Ca ing on 2014). In s udies o
social ne wo ks, he ocus is on how he s uc u e o social
ela ions among ac o s may in luence he beha io o indi iduals,
g oups o ac o s, o he sys em as a whole (Bodin and C ona 2009,
Sco and Ca ing on 2014). In he con ex o na u al esou ce
go e nance, pa e ns o collabo a ion among ac o s can shed ligh
on pa icipa ion, g oup lea ning, esou ce dis ibu ion, and o he
Ecology and Socie y 26(2): 42
h ps://www.ecologyandsocie y.o g/ ol26/iss2/a 42/
social p ocesses (Lubell 2013, Mancilla-Ga cía and Bodin 2019),
as well as biophysical and socio-poli ical ou comes (de Lange e
al. 2019). S udies o esou ce go e nance ne wo ks o en
speci ically seek o explain why ce ain ac o s do, o do no , adop
ce ain managemen p ac ices o policy belie s, accoun ing o
bo h endogenous, i.e., ne wo k s uc u e, and exogenous
a iables, i.e., a ibu es o he ac o s o hei ela ionships.
Table 1. Adap i e capaci y domains by Cinne e al. (2018) and
he speci ic adap i e capaci y s a egies examined in ou s udy.
Domains o
adap i e capaci y
om Cinne e al.
(2018)
Adap i e capaci y s a egies used in his s udy†
Asse s (a) Speci ic in as uc u e (e.g., po s, essels)
(b) C edi o sa ings
(c) Public unding
O ganiza ion (d) Well connec ed wi h o he ac o s
(e) Pa icipa ion in in e na ional/na ional/local
o ums
( ) Quali y o go e nance and leade ship
Flexibili y (g) Income di e si y
(h) Le el o dependence on na u al esou ces
Lea ning (i) Knowing ha he en i onmen is con inuously
changing including he species and ecosys ems
(j) In e gene a ional lea ning capaci y
(k) Resou ce s a us assessmen and eedback
mechanisms
Agency (l) Abili y o eac o changes (e.g., dis ibu ion,
abundance, clima ic isks)
†Based on Whi ney e al. (2017), Cinne e al. (2018), and F eduah e al.
(2018).
The speci ic ocus o his pape is on he ela ionship be ween
communica ion (links) among o ganiza ional s akeholde s
(nodes) in a la ge-scale ishe y, and s akeholde s’ p e e ences o
adap i e capaci y s a egies in he con ex o clima e change. To
he ex en ha social connec i i y is associa ed wi h he adap i e
capaci y domains ha s akeholde s p e e , i is possible o de elop
a ge ed s a egies o suppo ing adap i e capaci y in ishe ies
ha a e in o med by social connec i i y. Highly cen al ac o s a e
well posi ioned o sha e in o ma ion widely h oughou hei
ne wo ks (Gibbs 2008, Bodin and C ona 2009, de Lange e al.
2019). I hese ac o s ha e a mo e comp ehensi e iew o adap i e
capaci y domains, ha hey a e able o sp ead his in o ma ion
eadily h ough hei ne wo k may be an ad an ageous ea u e o
he go e nance sys em. Howe e , i cen al ac o s ha e na owe
iews on adap i e capaci y domains, hen his may limi o he
ac o s’ iews and po en ially educe adap i e capaci y in he
ishe y. Also, i is no uncommon ha simila ac o s sha e simila
iews, a phenomenon known as “homophily” in social ne wo ks
(McPhe son e al. 2001), and he e o e communica ion ac oss
g oups may be impo an o suppo ing a holis ic, indus y-wide
app oach o adap i e capaci y.
CASE STUDY
Backg ound
Tuna species a e among he mos aluable ma ine species in he
wo ld, accoun ing o 20% o he alue o all ma ine cap u e
ishe ies, and o e 8% o all globally aded sea ood (FAO 2018).
T opical unas, i.e., Skipjack una (Ka suwonus pelamis),
Yellow in una (Thunnus albaca es), and Bigeye una (Thunnus
obesus), ep esen 75% o una and una-like global ca ches (FAO
2018). These ma ine esou ces a e a ge ed by la ge-scale pu se
seine s among o he gea s. T opical una eeze pu se seine s
owned by Spanish companies ish a ound 10% o he global ca ch
o opical unas (Ugalde and Samano 2019), and he majo i y
o hese essels a e Basque, i.e., con olled by Basque capi al
(Ugalde 2014). In o al, 66 Spanish and Basque una eeze
essels (including con enience lags) ope a e in he Paci ic,
A lan ic, and Indian Oceans (Ugalde and Samano 2019). Among
hem, 18 Basque una eeze pu se seine s ope a ed in he eas e n
A lan ic Ocean in 2019. Ou case s udy ocuses speci ically on his
ishe y and builds on p e ious esea ch o clima e change
dis ibu ion impac s. A he ecological le el, una habi a
dis ibu ion limi s a e al eady shi ing polewa d as a esponse o
clima e change (E auskin-Ex amiana e al. 2019). A he ishe y
le el, pu se seine s ha e also shi ed in he las 30 yea s as a
combina ion o clima ic, echnological, and managemen o ces
(Rubio e al. 2020).
Rega ding he go e nance egime, he ICCAT is he Regional
Fishe ies Managemen O ganiza ion (RFMO) esponsible o he
managemen and conse a ion o una and una-like species in
he A lan ic Ocean. Decisions a e aken by con ac ing pa ies
(each coun y ge s one o e, excep o coun ies wi hin he
Eu opean Union [EU] who a e ep esen ed unde a single o e o
he EU). In his con ex , since 2005 he Bigeye una s ock has a
o al allowable ca ch (TAC) alloca ed o coun ies, al hough i is
o e ished and o e ishing is ac i ely occu ing (ICCAT “ ed
s a us”). The eas e n Yellow in una s ock has had a gene al TAC
since 2012, which has no been alloca ed o ishing coun ies, and
ye is conside ed o be doing well (ICCAT “g een s a us”). Finally,
he eas e n Skipjack una s ock has no TAC and is also doing
ela i ely well (ICCAT, “g een s a us”; ICCAT 2016, 2019).
SES cha ac e iza ion o he case s udy
We cha ac e ize he Basque opical una ishe y as a complex
SES comp ising he ou co e sub-sys ems desc ibed by Os om
(2009): esou ce sys em, esou ce uni s, go e nance sys em, and
use s (co esponding o he b oade “ac o s” subsys em as
de ined in McGinnis and Os om 2014). All o he cha ac e is ics
o he social, economic, and poli ical se ings, ela ed ecosys ems,
and hei in e ac ions complemen hese ou sub-sys ems
(McGinnis and Os om 2014). Use s ep esen he indus ial
ex ac i e ac i i y o opical unas by Basque eeze pu se seine
essels in he eas e n A lan ic Ocean. We concep ually illus a e
he ishe y as a SES (Fig. 1), and highligh he impac s o clima e
change on he SES as d i e s, and adap i e capaci y and
sus ainabili y as ou comes. Many di e en o ganiza ions a e
p esen in he Basque opical una SES, and we dis inguish he
ollowing ou g oups o ac o s: he ishing indus y, go e nmen
bodies and he RFMO, esea ch bodies, and non-go e nmen al
o ganiza ions (NGOs) and o he s, e.g., non-p o i o ganiza ions.
The ishing indus y includes he p ima y use s engaged in he
ex ac ion o opical unas in he A lan ic Ocean, which a e ou
Basque companies ep esen ed in wo ishe ies associa ions.
Go e nmen bodies ac i ely pa icipa e in he decision-making
p ocess abou alloca ion o ishing possibili ies and esou ce
managemen a he in e na ional le el as pa o he RFMO, i.e.,
he ICCAT. We iden i y ou go e nmen g oups: (A) he egional
Ecology and Socie y 26(2): 42
h ps://www.ecologyandsocie y.o g/ ol26/iss2/a 42/
Fig. 1. Concep ual ep esen a ion o he Basque opical una ishe y as a complex social-
ecological sys em (SES) unde clima e change. Adding up numbe s wi hin boxes, 23 o ganiza ions
a e accoun ed o , which co espond o he ne wo k nodes in Figu e 4 (20 o hem a e speci ic
o ganiza ions om which 13 we e in e iewed). The o iginal igu e used du ing in e iews is Figu e
A1.1. RFMO = Regional Fishe ies Managemen O ganiza ion.
go e nmen o he Eu opean Union; (B) o he egional
go e nmen s ha a e lag coun ies o Basque essels (non-EU);
(C) o he egional go e nmen s wi h public o p i a e ag eemen s
o he ishe y; and (D) na ional and local go e nmen s. G oups
B and C a e non-exclusi e; one egional go e nmen can be pa
o bo h g oups depending on i s ela ionship wi h he esou ce
use s. Al hough he na ional go e nmen egula es he ishe y a
he coun y le el, he local go e nmen plays a mo e speci ic ole
in he ishe y, e.g., p ocessing eques s o cons uc ing new
essels. Resea ch bodies in o m decision making in he ishe y
wi h he bes a ailable science and de elop esea ch p ojec s wi h
di e en ac o s ( wo esea ch bodies a e ela ed o he Basque
una SES). NGOs and o he o ganiza ions ypically ocus on
conse a ion o esou ces and he long- e m sus ainabili y o
ishe ies. In o al, he e a e eigh NGOs and o he o ganiza ions
ha pa icipa e in he go e nance sys em o he Basque opical
una ishe y. Las , i is impo an o no e he e he “high-le el”
na u e o his s udy. We a e no looking a e e y indi idual ac o ,
e.g., essels, c ewmembe s, e c., bu a he a o ganiza ions and
ins i u ions in o de o cap u e coo dina ion a a highe le el o
decision making.
DATA AND METHODS
The Basque opical una ishe y ep esen s a la ge-scale ishe y
wi h impo an economic alue and p essing sus ainabili y
challenges, and p o ides a compelling con ex o empi ically
add essing ou esea ch ques ions on social connec i i y and
p e e ences o adap i e capaci y s a egies. We begin by using he
SES amewo k o desc ibe he ishe y and iden i y he ac o s
in ol ed, who we hen in i ed o in-dep h in e iews o collec
in o ma ion on hei adap i e capaci y p e e ences and
communica ions wi h o he s akeholde s. We used clus e analysis
o g oup ac o s wi h simila p e e ences, and used social ne wo k
analysis o in es iga e he s uc u e o communica ion among
ac o s and how hey ela e o adap i e capaci y p e e ences. All
analyses we e pe o med using he R En i onmen o S a is ical
Compu ing (R Co e Team 2018), and he sc ip s wi h hei
wo k low a e a ailable he e: h ps://gi hub.com/i ubio/
opi una_SNA_Adap Capaci y.
S akeholde in e iews
We designed a ques ionnai e (a ailable in Appendix 1) o ca ying
ou in-dep h, semi-s uc u ed in e iews wi h he ele an
s akeholde s in he Basque opical una SES. One high-le el
ep esen a i e, such as he di ec o , om each o ganiza ion was
in i ed o pa icipa e (Table 2). O ganiza ional s akeholde s we e
iden i ied h ough he au ho s’ p o essional ne wo ks and
knowledge o he empi ical se ing, and pa icipan s we e also
gi en he chance o nomina e o he s akeholde s du ing he
in e iews, d awing on hei own expe ience and knowledge
Ecology and Socie y 26(2): 42
h ps://www.ecologyandsocie y.o g/ ol26/iss2/a 42/
(B yman 2016). A e disclosing he gene al objec i e o he s udy
o in i ed pa icipan s, wi hou p o iding in o ma ion ha could
al e hei esponses, and in o ming hem abou con iden iali y
and hei igh s du ing he in e iew p ocess, hey we e asked o
hei olun a y pa icipa ion in he in e iews. We we e able o
conduc hi een in-dep h in e iews, co esponding o 65% o
he wen y speci ic o ganiza ions iden i ied o be di ec ly ela ed
o he Basque ishe y (nomina ed in Figu e 1). We did no
in e iew na ional go e nmen s a he in e na ional scale because
o lack o access o ep esen a i es o hese go e nmen s, who
co espond o “O he egional go e nmen s” in Figu e 1.
Table 2. Ac o ype, posi ion, numbe (n = 13), and Os om’s
social-ecological sys ems (SES) amewo k ca ego y co esponding
o s akeholde s in e iewed. RFMO = Regional Fishe ies
Managemen O ganiza ion.
Ac o ype Posi ion o pe son(s)
in e iewed (numbe )
Os om’s SES amewo k
ca ego y
Go e nmen o
RFMO
Di ision di ec o s (2) Go e nance sys em ac o s
Resea ch Coo dina o o esea ch a ea (1)
P incipal esea che (1)
NGO, o he s Di ision ice p esiden (1)
Fishe ies esponsible (2)
Fishing indus y Managing di ec o s (4)
Depu y di ec o (1)
A ea di ec o (1)
Use s (o ac o s in
McGinnis and Os om
2014)
The in e iews we e ca ied ou o e Janua y o Ap il 2019,
including 11 in-pe son in e iews and wo ideo calls. In e iews
las ed one hou and 20 minu es, on a e age. All in e iewees we e
male and had been wo king in he ishing sec o o an a e age o
18 yea s. Fo y-six pe cen o in e iewees we e be ween 35 and
44 yea s old, 31% be ween 45 and 54, and 23% be ween 55 and
64. Despi e ha ing a ge ed ep esen a i es o op o ganiza ions,
mos o he esponden s conside ed hei o ganiza ion o ha e
li le in luence in decision-making p ocesses. As much as 54%
conside ed hei o ganiza ion o ha e e y li le in luence, 38%
mode a e in luence, and only 8% a lo o in luence.
Du ing he in e iew, s akeholde s we e asked abou h ee main
opics: (1) hei unde s anding o he Basque opical una SES;
(2) communica ion exchanges be ween hei o ganiza ion and
o he o ganiza ions; and (3) he adap a ion ac ions and adap i e
capaci y s a egies needed o main ain he ishing ac i i y in he
long e m (see ques ions in Appendix 1).
Fo he s akeholde s’ unde s anding o he SES (1), in e iewees
we e asked abou he componen s o he case s udy using he
diag am o he concep ual model o he Basque una SES (i.e.,
Fig. A1.1). This p ocess allowed o he iden i ica ion o any
missing in o ma ion abou he SES.
Fo communica ion exchanges (2), in e iewees we e p esen ed
wi h a lis o he o ganiza ions iden i ied in he ishe y and we e
asked abou hei equency o communica ion wi h each o hem,
in e ms o discussing ishe y ma e s (Table A1.1).
Communica ion equency op ions we e no o en (1–4 imes/
yea ), some imes (5–11 imes/yea ), o en (1–4 imes/mon h,
mon hly), and a lo (mo e han once a week). This in o ma ion
allows o he measu emen o social connec i i y in he ishe y,
in he o m o a communica ion ne wo k among o ganiza ions.
As pa o he in e iew p ocess, some s akeholde s iden i ied
addi ional o ganiza ions in he ishe y wi h whom hey
communica e, and hese in e ac ions we e also eco ded.
In e ac ions wi h non-in e iewed o ganiza ions (5 go e nmen
o RFMO, and 5 NGOs and o he s) we e also collec ed, bu we
we e unable o assess in e ac ions be ween non-in e iewed
o ganiza ions. In his way, we could map he o ganiza ions, o
g oups o o ganiza ions, iden i ied in he SES.
Fo adap i e capaci y p e e ences (3), in e iewees we e i s
asked o indica e wha hey belie ed we e he mos impo an
adap a ion ac ions o he long- e m sus ainabili y o he ishe y
(A1 Ques ion 5 and SM Table 2 in Rubio, Hobday, and Ojea,
unpublished manusc ip ). A e wa d, hey we e p esen ed wi h he
lis o 12 adap i e capaci y s a egies (see Table 1 and Table A1.2),
and asked o alue hei impo ance using a 5-ca ego y Like
scale, including 1 (no impo an ), 2 (sligh ly), 3 (mode a ely), 4
( e y), and 5 (ex emely impo an ). Responden s epea ed his
p ocess h ee imes, anking he impo ance o adap i e capaci y
s a egies o hei h ee mos impo an adap a ion ac ions. The
speci ic ques ion was, “How impo an would he ollowing
s a egies be on a scale om 1 o 5 o achie e he mos impo an
adap a ion ac ions o seek a sus ainable ishe y?” In his way,
pe cep ions o he impo ance o adap i e capaci y domains a e
linked o eal adap a ion ac ions in he ishe y. This esul ed in
each esponden ha ing h ee answe s pe adap i e capaci y
s a egy, which we e a e aged o ob ain a single alue. Resul s
we e classi ied acco ding o a 4-ca ego y scale: alues be ween 1
and 2 we e less impo an ; alues be ween 2 and 3 we e sligh ly
impo an ; alues be ween 3 and 4 we e impo an ; and alues
be ween 4 and 5 we e e y impo an . Finally, he impo ance o
adap i e capaci y s a egies was also a e aged by adap i e
capaci y domain.
Da a analysis
In o de o iden i y g oups o ac o s ag eeing on he impo ance
o speci ic adap i e capaci y s a egies o he sus ainabili y o he
Basque opical una ishe y, we pe o med a hie a chical clus e
analysis o he s akeholde s’ esponses. This me hod o analysis
is sui able o small sample sizes, i.e., ypically less han 250
obse a ions (Kau man and Rousseeuw 2005). Clus e s we e
de e mined h ough Wa d’s linkage me hod (Wa d 1963) wi h
Euclidean dis ances, implying a minimum inc ease in he o al
wi hin-clus e a iance (Mu agh and Legend e 2014). We hen
plo ed a hea map o he esul s and calcula ed a inal impo ance
sco e o each adap i e capaci y g oup esul ing om he
hie a chical clus e analysis by a e aging he o iginal impo ance
alues o adap i e capaci y s a egies pe domain o Cinne e al.
(2018).
A e clus e ing o ganiza ions wi h he same p e e ences o
adap i e capaci y s a egies and domains, we analyzed he social
ne wo k o communica ion be ween he o ganiza ions. The
ne wo k is composed o nodes, which ep esen o ganiza ions o
g oups o o ganiza ions (n = 23, Fig. 1), and links, which indica e
communica ion be ween o ganiza ions. We speci ically elec ed o
include only communica ion occu ing mo e han once pe
mon h, in o de o ocus on he mos ac i e o ganiza ions in he
ishe y. We cons uc ed he ne wo k as a se o di ec ed
communica ion ies be ween o ganiza ions, i.e., Ac o A
communica es equen ly wi h Ac o B, and/o Ac o B
communica es equen ly wi h Ac o A. We op ed o include only
Ecology and Socie y 26(2): 42
h ps://www.ecologyandsocie y.o g/ ol26/iss2/a 42/
incoming ies (in-deg ee), in o de o accoun o he ac ha no
e e y o ganiza ion ag eed o be in e iewed du ing he da a
collec ion p ocess, and he e o e no e e y o ganiza ion
con ibu ed ies (ou -deg ee). In his sense, all ac o s had he same
chance o being nomina ed as ecei e s o communica ion.
A e wa d, we plo ed he ne wo k, including bo h adap i e
capaci y g oups and o ganiza ion ypes in he isualiza ion.
Nex , we calcula ed desc ip i e s a is ics o he di ec ed ne wo k
using he “ig aph” package (Csa di and Nepusz 2006). A he
ne wo k le el, densi y measu es how many ies be ween ac o s
exis ela i e o he maximum numbe o possible ies be ween
ac o s. The a e age in-deg ee cap u es he mean numbe o imes
each ac o was named by o he ac o s when asked o iden i y who
hey equen ly communica e wi h. The a e age pa h leng h is he
mean numbe o ies sepa a ing any wo ac o s, and ep esen s
he eachabili y o ac o s wi hin he sys em. A he node le el, in-
deg ee cen ali y is he numbe o incoming ies an ac o possesses,
and be weenness cen ali y ep esen s he ex en o which an ac o
indi ec ly acili a es connec i i y wi h o he ac o s in he ne wo k
(Bodin and C ona 2009). Ac o s wi h high in-deg ee cen ali y
sco es can be hough o as con e sa ional hubs because o he
ac o s communica e equen ly wi h hem. We also measu ed
be weenness cen ali y o ies in he ne wo k, which cap u es he
ex en o which a ie makes o he connec ions possible. In-deg ee
cen ali y and be weenness cen ali y o bo h nodes and ies we e
no malized be ween 0 and 1 o ease o in e p e a ion.
Las , we pe o med an ERGM analysis using he “s a ne ”
package in R (Handcock e al. 2019; see goodness-o - i
diagnos ics in Appendix 2). In an ERGM, he ne wo k is ea ed
as he dependen a iable, and ne wo k o ma ion p ocesses (e.g.,
cen aliza ion and homophily), ope a ionalized as mic o-le el
s uc u al pa ame e s, a e he p edic o a iables (Le y and
Lubell 2018). ERGMs use Mon e Ca lo Ma ko Chain maximum
likelihood simula ion o es ima e he ex en o which a se o
de ined s uc u al pa ame e s a e o e - o unde - ep esen ed in
he obse ed ne wo k (C anme and Desma ais 2011). The esul s
o an ERGM a e a se o coe icien s o he speci ied ne wo k
pa ame e s included in he model, which ep esen he change in
he log-odds o ie o ma ion, and indica e he di ec ion and
magni ude o each pa ame e (Hileman and Lubell 2018).
ERGMs a e cons uc ed by selec ing model e ms ha a e
in o med by bo h heo y and knowledge o he empi ical se ing
(Alexande e al. 2015), and he e o e we include he ollowing
e ms ha add ess he ela ionship be ween social connec i i y
and adap i e capaci y p e e ences: geome ically weigh ed
edgewise sha ed pa ne s (gwesp) cap u es closu e; geome ically
weigh ed in-deg ee dis ibu ion (gwideg) cap u es cen aliza ion
(no e: he e m ac ually measu es an i-cen aliza ion, and
he e o e a nega i e coe icien deno es a endency o ac o s o
o m ies wi h high-deg ee ac o s); and homophily (nodema ch)
cap u es whe he ac o s o he same ype and belonging o he
same adap i e capaci y g oup end o o m ies wi h o he ac o s
o he same ype and g oup. We also include he ollowing con ol
pa ame e s: baseline a es o ie o ma ion (node ac o ) o each
ac o ype and adap i e capaci y g oup, and whe he o no an
ac o was in e iewed du ing he da a collec ion p ocess; ne wo k
densi y (edges); and ac o s who a e no connec ed o he ne wo k
(isola es).
RESULTS
Ac o p e e ences o adap i e capaci y s a egies and domains
Hie a chical clus e analysis esul ed in ou dis inc adap i e
capaci y g oups (Fig. 2), each o which ep esen s di e en se s
o p e e ences o adap i e capaci y s a egies. We obse e how
he same ac o ypes do no necessa ily belong o he same
adap i e capaci y g oup, i.e., sha e he same iews on adap i e
capaci y s a egies. G oup 1 and G oup 2 a e he smalles , and
include only wo ac o s each; G oup 3 is he la ges wi h six
o ganiza ions, and all ou o ganiza ion ypes a e ep esen ed;
and G oup 4 con ains h ee ac o s, all belonging o he ishing
indus y. The impo ance gi en o each adap i e capaci y s a egy
by each o ganiza ion, which was used o iden i y he g oups, is
also displayed in Figu e 2. Pa icipa ion in in e na ional/na ional/
local o ums (e) is he only adap i e capaci y s a egy conside ed
o be impo an o e y impo an by all o ganiza ions.
Con e sely, all o ganiza ions anked c edi /sa ings (b) and
income di e si y (g) as less o sligh ly impo an .
Fig. 2. P e e ences o adap i e capaci y s a egies by ac o
ype. The hea map shows hie a chical clus e ing o he 13
o ganiza ions in he Basque una social-ecological sys ems ha
we e in e iewed ( ows), acco ding o he alue hey assigned o
he di e en adap i e capaci y s a egies (columns). Impo ance
alues ank om 1 (da k o ange, less impo an ) o 5 (da k
blue, e y impo an ). The ou adap i e capaci y g oups
esul ing om he clus e analysis a e ep esen ed by numbe s
on he le side o he dendog am. See Table 1 o he
co espondence be ween le e s (columns) and ull names o
adap i e capaci y s a egies.
Figu e 3 shows he a e age alues o he esul ing adap i e
capaci y g oups pe domain om Cinne e al. (2018). G oup 1
is cha ac e ized by s ong p e e ences o o ganiza ion; G oup 2
is cha ac e ized by i s lack o s ong p e e ences o o agains
any o he domains; G oup 3 is cha ac e ized by s ong
p e e ences o o ganiza ion and agency; and G oup 4 is
cha ac e ized by i s p e e ence o all adap i e capaci y domains,
as each domain was conside ed ei he impo an o e y
Ecology and Socie y 26(2): 42
h ps://www.ecologyandsocie y.o g/ ol26/iss2/a 42/
impo an . All g oups hough o ganiza ion is impo an o e y
impo an , bu di e ed on he impo ance gi en o o he domains.
G oups 2, 3, and 4 all conside ed lea ning o be an impo an
adap i e capaci y domain, while G oup 1 conside ed lea ning o
be only sligh ly impo an . G oup 4 conside ed bo h asse s and
lexibili y o be impo an adap i e capaci y domains, while he
h ee emaining g oups all conside ed hese wo domains o be o
lesse impo ance (G oup 1), sligh ly impo an (G oup 2), o a
mix o bo h (G oup 3). Al hough clus e ed sepa a ely, hese
g oups did no necessa ily ha e opposing iews. To ecap, G oup
1 ga e impo ance o o ganiza ion and agency; G oup 2 o
o ganiza ion and lea ning; G oup 3 o o ganiza ion, agency, and
lea ning; and G oup 4 o all adap i e capaci y domains. Excep
o G oup 4, he es had na owe iews on adap i e capaci y.
Fig. 3. P e e ences o adap i e capaci y domains pe adap i e
capaci y g oup. Final impo ance sco es a e calcula ed o each
adap i e capaci y g oup (subplo s) om he clus e analysis by
a e aging o iginal impo ance alues (y-axis) gi en o adap i e
capaci y s a egies pe domain (x-axis).
Social connec i i y and adap i e capaci y p e e ences
Figu e 4 p esen s he communica ion ne wo k among
o ganiza ions in he Basque una SES. In-deg ee cen ali y and
edge be weenness cen ali y a e g aphically ep esen ed by node
size and ie colo , espec i ely. O ganiza ions in he cen e o he
igu e a e hose wi h mo e incoming ies, and include wo ac o s
om he ishing indus y, wo go e nmen en i ies, and one
esea ch o ganiza ion. Ties wi h he highes edge be weenness link
ac o s wi h impo an oles as connec o s; he ne wo k may
ac u e i hese ac o s we e o s op communica ing equen ly.
Ne wo k-le el desc ip i e s a is ics indica e all o ganiza ions in
he ne wo k a e easily eachable (a e age pa h leng h = 1.99), and
on a e age possess mul iple incoming ies om o he
o ganiza ions (a e age in-deg ee = 3.22). E en gi en i s small size,
he ne wo k is s ill ela i ely dense (densi y = 0.15).
Figu e 4 also shows nodes (ac o s) colo ed by adap i e capaci y
g oups. The mos cen al ac o s ep esen ed almos all adap i e
capaci y g oups (G oups 1, 3, and 4). Ac o s belonging o adap i e
capaci y G oup 1 ended o ha e he highes in-deg ee cen ali y
and be weenness cen ali y in he ne wo k (Table 3). They we e
ollowed by G oups 3 and 4, which had simila in-deg ee alues,
bu G oup 3 had a highe be weenness cen ali y. G oup 2
possessed he lowes be weenness cen ali y o all adap i e
capaci y g oups in he ne wo k (Table 3). When looking a
o ganiza ion ypes as g oups, he esea ch g oup had he highes
in-deg ee cen ali y and be weenness cen ali y alues, ollowed
by he ishing indus y (Table 3). In spi e o mode a e in-deg ee
cen ali y, go e nmen en i ies possessed he lowes be weenness
cen ali y sco e a e NGOs and o he s. NGOs and o he s also
had he lowes in-deg ee.
Table 3. Mean deg ee cen ali y and be weenness cen ali y alues
by adap i e capaci y g oup and o ganiza ion ype. Regional
Fishe ies Managemen O ganiza ion.
In-deg ee
cen ali y (± SD)
Be weenness
cen ali y (± SD)
Adap i e capaci y g oup
G oup 1 0.85 (± 0.21) 0.80 (± 0.29)
G oup 2 0.20 (± 0.14) 0.01 (± 0.00)
G oup 3 0.42 (± 0.15) 0.23 (± 0.26)
G oup 4 0.43 (± 0.32) 0.15 (± 0.23)
O ganiza ion ype
Fishing indus y 0.43 (± 0.25) 0.25 (± 0.40)
Go e nmen o RFMO 0.34 (± 0.23) 0.08 (± 0.17)
NGOs, o he s 0.12 (± 0.13) 0.09 (± 0.23)
Resea ch 0.70 (± 0.42) 0.32 (± 0.39)
Table 4 displays he esul s o he ERGM. In he model, we see a
posi i e and signi ican coe icien o closu e in he
communica ion ne wo k, i.e., he mo e pa ne s wo ac o s sha e
in common, he mo e likely hey hemsel es a e o sha e a ie.
Cen aliza ion also had a posi i e and signi ican coe icien in
he model, which means he e was no a la ge a ia ion in he
numbe o ies each ac o possessed in he ne wo k. The model
indica ed homophily was also a d i e o ie o ma ion in he
ne wo k, speci ically among ac o s o he same o ganiza ion ype,
e.g., ishing indus y ac o s ended o communica e mo e wi h
o he ac o s in he indus y. Howe e , homophily as a unc ion
o adap i e capaci y g oup was no signi ican in he model.
Table 4. Exponen ial andom g aph model o he ne wo k o
equen ly connec ed o ganiza ions in he Basque una social-
ecological sys em. Coe icien s o each e m included in he
model a e shown wi h espec i e s anda d e o s in b acke s.
Ne wo k
Ne wo k
p ocesses
Closu e (gwesp)
Cen aliza ion (gwideg)
0.906** (0.279)
2.954* (1.322)
Homophily Adap i e capaci y (nodema ch)
Ac o ype (nodema ch)
0.284 (0.473)
1.176*** (0.330)
Con ol
pa ame e s
Baseline a es (node ac o )
Adap i e capaci y (node ac o )
AC G oup 2 -1.400** (0.514)
AC G oup 3 -1.072* (0.434)
AC G oup 4 -1.248** (0.453)
Ac o ype (node ac o )
Go e nmen 0.716* (0.350)
NGOs/o he s 0.140 (0.296)
Scien is s 0.202 (0.343)
In e iewed? (node ac o ) 2.873*** (0.739)
Densi y (edges)
Isola es (isola es)
-7.329*** (1.091)
2.114 (1.552)
*p < 0.1, **p < 0.05, ***p < 0.01
Ecology and Socie y 26(2): 42
h ps://www.ecologyandsocie y.o g/ ol26/iss2/a 42/
Fig. 4. Ne wo k showing highly connec ed (communica ing mo e han once a mon h) o ganiza ions
in he Basque una social-ecological sys em (n = 23). Node shapes ep esen he ac o ype, node
colo s ep esen adap i e capaci y g oups, node sizes ep esen deg ee cen ali y, and ie boldness
ep esen s edge be weenness alues. A label is only shown o go e nmen o Regional Fishe ies
Managemen O ganiza ions (RFMO) abo e he co esponding node. Communica ion was no
eco ded be ween “NA” o ganiza ions as hey we e no in e iewed.
DISCUSSION
The impo ance ha di e en s akeholde s in he Basque opical
una ishe y gi e o clima e adap i e capaci y domains is he co e
ocus o his s udy, along wi h assessing why his is he case. Ou
esul s e lec he in e na ional na u e o he go e nance egime
whe e una ishe ies a e embedded (Allen 2010) because he
o ganiza ion domain om Cinne e al. (2018) is he only domain
ha all ac o s uni e sally ag ee is impo an o sus aining he
ishe y in he ace o global clima e change. Mo e speci ically, all
s akeholde s ag ee ha pa icipa ion in in e na ional, na ional,
and/o local o ums is an impo an adap i e capaci y s a egy.
In addi ion o pa icipa ion in o ums, being well connec ed wi h
o he ac o s and ha ing good go e nance and leade ship (in he
social o ganiza ion domain), and ha ing esou ce s a us
assessmen and eedback mechanisms (in he lea ning domain)
a e highly alued by almos all ac o s. This is impo an in he
se ing o he ICCAT go e nance egime, whe e decision making
abou managemen measu es is in o med by s ock assessmen s,
and limi s o he ishing ac i i y o opical unas is de i ed om
s ock s a us (ICCAT 2016, 2019). These limi s can be mo e o less
es ic i e as a esul o poli ical in e es s a he in e na ional le el
(Webs e 2009). O he adap i e capaci y s a egies a e gene ally
seen as less impo an , e.g. ha ing c edi , sa ings, o income
di e si y. This is also likely ela ed o he in e na ional na u e o
he ishe y, whe e c edi al eady exis s, bu inclusi eness o local
s akeholde s in he decision-making p ocesses is almos non-
exis en because ule making occu s among egional go e nmen s
(Eps ein e al. 2014, Pen z e al. 2018).
The esul s o he social ne wo k analysis shed ligh on he
ela ionship be ween s uc u es o communica ion among
s akeholde s in he ishe y and hei p e e ed adap i e capaci y
s a egies. We ound he ishe y was cha ac e ized by s ong
communica ion among di e en ac o ypes, which gene ally can
help s eng hen adap i e capaci y (Whi ney e al. 2017). Linking
he ne wo k s uc u e wi h he go e nance egime, Pe e sson e
al. (2019) ound ha non-s a e ac o s, such as ishing indus y
ep esen a i es o NGOs, a e in e es ed in go e nance o ishe ies
and egula ly a end RFMO mee ings globally. These mee ings
a e a good oppo uni y o s eng hening ne wo ks, as
ep esen a i es om many di e en o ganiza ions a e able o
in e ac wi h each o he . Non-s a e ac o s, including scien i ic
o ganiza ions and ep esen a i es o he ishing indus y, we e he
mos equen ly communica ed wi hin he ne wo k. In addi ion,
we ound ha cen al o ganiza ions ended o sha e simila
p e e ences o adap i e capaci y s a egies (McPhe son e al.
2001, Shephe d 2017), ega dless o he ype o ac o , excep o
one ishing indus y ep esen a i e who had a holis ic iew o
adap i e capaci y (adap i e capaci y G oup 4).
Digging deepe in o ne wo k s uc u es and hei ela ionship o
adap i e capaci y, highly cen al ac o s a e well-posi ioned o
sha e in o ma ion widely h ough he ne wo k (Gibbs 2008, Bodin
and C ona 2009, de Lange e al. 2019), and can ca alyze
widesp ead adop ion o new beha io s h ough hei social
in luence (Valen e and Pumpuang 2007). S a egies o os e
adap i e capaci y in he ishe y could be coo dina ed h ough he
ac o s ha occupy cen al posi ions and ep esen
communica ion hubs. In ou s udy, e en hough he e is no a
la ge a ia ion in he deg ee o ac o s in he ne wo k, he e is
ne e heless a co e g oup o mo e cen al ac o s. Howe e , he
majo i y o hese cen al ac o s ha e na owe adap i e capaci y
Ecology and Socie y 26(2): 42
h ps://www.ecologyandsocie y.o g/ ol26/iss2/a 42/
p e e ences, ocusing p ima ily on o ganiza ion and agency, and
o a lesse ex en lea ning. This may po en ially educe adap i e
capaci y in he ishe y i i simila ly in luences o he ac o s’
p e e ences (Bodin 2017). Tha said, we may no necessa ily
expec all di e en ypes o ac o s o conside all adap i e capaci y
domains as highly impo an , and i may e en be na u al o
di e en ypes o ac o s o “specialize” in pa icula adap i e
capaci y domains, o pa icula sui es o p omo e adap i e
capaci y. In e -g oup collabo a ion, as indica ed by he ERGM
esul s, may signal he e is a heal hy mix o s a egies o imp o e
o e all adap i e capaci y in he ishe y.
This aises he ques ion o whe he all domains o adap i e
capaci y need o be p esen o ishe ies o be esilien o clima e
change. O , in o he wo ds, whe he he e is subs i u abili y among
he elemen s o adap i e capaci y. Yohe and Tol (2002) p oposed
he weakes link hypo hesis, which en ails ha adap i e capaci y
may be es ic ed by he weakes o i s unde lying de e minan s.
A e pe o ming a li e a u e e iew, Cinne e al. (2018) sugges
he e is limi ed subs i u abili y be ween domains o adap i e
capaci y wi h espec o shocks and long- e m change. Howe e ,
he au ho s also highligh ha he weakes link idea has no ye
been es ed in ishe ies. We do no di ec ly add ess he
subs i u abili y hypo hesis, bu ou esul s poin in he di ec ion
o ce ain domains being pe cei ed as mo e impo an han
o he s, i.e., social o ganiza ion. This sugges s ha some adap i e
capaci y op ions a e conside ed seconda y (o a e seen as less
impo an ), which could esul in a less esilien ishe y. Ano he
possibili y is ha hese adap i e capaci y op ions a e conside ed
seconda y because hey migh al eady be co e ed wi hin he
ishe y.
Mo eo e , simila ac o ypes did no necessa ily sha e he same
iews on he adap i e capaci y s a egies hey conside ed mos
impo an o sus aining he ishe y, which was he case among
less han hal o in e iewed o ganiza ions. Two NGOs and o he s
( om G oup 3), wo go e nmen s ( om G oup 3), and hal o
he ishing indus y (G oup 4) had he same iews on adap i e
capaci y wi hin hei own ac o ype g oup (o sec o ). Howe e ,
we ound ha ac o s do no end o o m ies wi h ac o s om
he same adap i e capaci y g oup, bu hey do end o o m ies
wi h ac o s om he same sec o . The endency o closu e wi hin
he ne wo k is p ima ily being d i en by hese wi hin-sec o ies.
Resea ch o ganiza ions and he ishing indus y se e as b idging
ac o s and p o ide mo e in e -g oup connec i i y wi h NGOs and
o he ac o s, o example. I is use ul o iden i y such b idging
ac o s because hey can help build suppo ac oss g oups when
a emp ing o add ess en i onmen al p oblems ha equi e
ans o ma ional changes in managemen and pe cep ions
(Wes ley e al. 2013, Olsson e al. 2014).
Ano he conce n a ising om his esea ch ela es o he
consequences o p io i izing adap i e capaci y ac ions om a
policy-making pe spec i e. Policies ha p omo e adap i e
capaci y a e di icul o implemen when s akeholde s ha e
a ying p e e ences o adap i e capaci y s a egies. In he case
o he Basque opical una ishe y, i would likely be easy o
implemen policies ha a ge he o ganiza ional domain o
adap i e capaci y, gi en he e was b oad ag eemen among
s akeholde s conce ning he impo ance o his pa icula
domain. Howe e , i he in en ion is o p omo e esilien ishe ies,
an e o should be made o os e a mo e holis ic pe spec i e o
adap i e capaci y. Gi en mo e holis ic pe spec i es on adap i e
capaci y a e ep esen ed by a ew ac o s om he ishing sec o ,
cen al ac o s could be in ol ed in e o s o mobilize and
encou age mo e in eg a ed pe spec i es h ough s eng hening
hei communica ion wi h hese ac o s om he ishe ies sec o ,
hus imp o ing he esilience o he go e nance sys em o he
SES as a whole.
CONCLUSIONS
In his pape , we adop a ne wo k-based app oach o s udy he
ela ionship be ween p e e ences o adap i e capaci y s a egies
and social connec i i y, in he o m o communica ion, among
s akeholde s in ol ed in an in e na ional la ge-scale ishe y. We
speci ically in es iga e he Basque opical una ishe y in he eas
A lan ic Ocean, which is pa o he ICCAT go e nance egime.
Ou esul s allow o he iden i ica ion o adap a ion s a egies
ha a e highly suppo ed by all ac o s, which in his case
cons i u es s a egies a ge ing he o ganiza ional domain o
adap i e capaci y. Mo e speci ically, p omo ing pa icipa ion in
in e na ional, na ional, and local o ums would be well suppo ed
by all ac o s. When linking adap i e capaci y p e e ences and
social connec i i y, he p e e ences o cen al ac o s a e simila ,
because he majo i y ocus on ela i ely ew adap i e capaci y
domains, i.e., social o ganiza ion, lea ning, and agency. This
could be de imen al o p omo ing a ully esilien ishe y,
especially gi en ac o s om he ishing sec o a e he only ones
who conside all adap i e capaci y domains o be impo an .
Thus, boos ing connec i i y wi h pa icula ac o s who ha e
comp ehensi e iews on adap i e capaci y could help sp ead hese
iews and p omo e a esilien ishe y. Finally, he communica ion
ne wo k does connec many di e en ac o ypes in he ishe y,
which is an encou aging sign and could be used om a policy-
making pe spec i e o p omo e adap i e capaci y h ough cen al
ac o s ac ing as “communica ion hubs.” Fu u e esea ch can help
con ex ualize his s udy by examining egional go e nmen s
independen ly, and u he in es iga ing adap i e capaci y
pe cep ions including iews o he RFMO.
Responses o his a icle can be ead online a :
h ps://www.ecologyandsocie y.o g/issues/ esponses.
php/12395
Acknowledgmen s:
This esea ch was suppo ed by he p ojec CLOCK, unde he
Eu opean Ho izon 2020 P og am, ERC S a ing G an Ag eemen
nº679812 unded by he Eu opean Resea ch Council. I is also
suppo ed by he Basque Go e nmen h ough he BERC 2018-2021
p og am and by he Spanish Minis y o Economy and
Compe i i eness MINECO h ough he BC3 Ma ía de Maez u
excellence acc edi a ion MDM-2017-0714. We hank, wi hou
implica ing, Ö. Bodin, I. Fe ze , and J.Rocha o scien i ic ad ice,
and all he pa icipan s o he s udy who made i possible.
Ques ion 6. How impo an would he ollowing op ions (Table A1.2) be on a scale om
1 (no impo an ) o 5 ( e y impo an ) o achie e he ac ions chosen in he
p e ious ques ion? Answe by ac ion (A1, A2 and A3).
Table A1.2 Lis o adap i e capaci y s a egies.
A1
A2
A3
Speci ic in as uc u e (po s, essels…)
C edi o sa ings
Public unding
To be well connec ed wi h o he playe s
To pa icipa e in in e na ional/na ional/local o ums
Good quali y o go e nance and leade ship
Income di e si y
Le el o dependence on na u al esou ces
Knowing ha he en i onmen is con inuously changing including he species
and ecosys ems
In e gene a ional lea ning capaci y
Resou ce s a us assessmen and eedback mechanisms
Abili y o eac o he changes (e.g. dis ibu ion, abundance, clima ic isks)
Appendix 2. ERGM GOODNESS-OF-FIT DIAGNOSTICS
