E alua ing ish o aging beha iou on non-indigenous Aspa agopsis
axi o mis using a emo e ideo o aging sys em
Saha Chebaane
a,b,c,*
, Aschwin Hilleb and Engelen
d
, Miguel Pessanha Pais
b,e
, Rod igo Sil a
a
,
F ancesca Gizzi
a
, Raül T iay-Po ella
a,
, Ma a Flo ido
g
, Jo˜
ao Gama Mon ei o
a,h
a
MARE - Ma ine and En i onmen al Sciences Cen e / ARNET - Aqua ic Resea ch Ne wo k, Regional Agency o he De elopmen o Resea ch, Technology and
Inno a ion (ARDITI), Funchal, Po ugal
b
Depa amen o de Biologia Animal, Faculdade de Ciˆ
encias, Uni e sidade de Lisboa, Po ugal
c
Biological and En i onmen al Sciences and Enginee ing (BESE), King Abdullah Uni e si y o Science and Technology (KAUST), Thuwal, Saudi A abia
d
CCMa , Uni e sidade do Alga e, Campus de Gambelas, 8100-139, Fa o, Po ugal
e
MARE - Ma ine and En i onmen al Sciences Cen e / ARNET - Aqua ic Resea ch Ne wo k, Faculdade de Ciˆ
encias, Uni e sidade de Lisboa, Po ugal
G upo en Biodi e sidad y Conse aci´
on, IU-ECOAQUA, Uni e sidad de Las Palmas de G an Cana ia, Las Palmas, Spain
g
Labo a o io de Biología Ma ina, Depa amen o de Zoología, Facul ad de Biología de la Uni e sidad de Se illa, A . de la Reina Me cedes, 41012, Se illa, Spain
h
Facul y o Li e Sciences, Uni e si y o Madei a, 9000, Funchal, Po ugal
ARTICLE INFO
Keywo ds:
Fish-mac oalgae dynamics
Biological in asion
RVFS
Feeding p e e ence
ABSTRACT
The p oli e a ion o pes and in asi e ma ine mac oalgae h ea ens coas al ecosys ems, wi h bio ic in e ac ions,
including di ec e ec s such as g azing and indi ec e ec s such as he ophic cascades, whe e one species
indi ec ly a ec s ano he h ough i s in e ac ions wi h a hi d species, play a c i ical ole in de e mining he
esis ance o local communi ies o hese in asions. This s udy examines he o aging beha iou and p e e ence o
na i e ish communi ies owa d na i e (Halop e is scopa ia, Sa gassum ulga e) and non-indigenous (Aspa agopsis
axi o mis) mac oalgae using he Remo e Video Fo aging Sys em (RVFS). Fi y- ou weedpops we e deployed
ac oss h ee loca ions o p esen hese mac oalgae, while associa ed epi aunal assemblages we e also collec ed.
Video analysis e ealed ha ou common ish species displayed p e e ence owa ds na i e mac oalgae, possibly
due o by he p esence o zooben hos a he han he bi o y. This obse a ion sugges s ha hese ish species
iden i ied he mac oalgae as a habi a ha ha bou ed hei p e e ed ood i ems. In con as , A. axi o mis was
consis en ly a oided, sugges ing limi ed in eg a ion in o he local ood web. Si e-speci ic a ia ions in ish-
mac oalgae in e ac ions and epi aunal di e si y highligh ed he complexi y o hese dynamics. This s udy con-
ibu es o unde s anding o he ecological implica ions o in asi e mac oalgae and suppo s he use o RVFS as a
ool o assessing local bio ic esis ance agains non-indigenous species in coas al ecosys ems globally.
1. In oduc ion
Oceanic islands a e gene ally enowned o hei dis inc i e ecosys-
ems and gene ic di e si y, cha ac e ised by an a ay o unique species
and subs an ial endemism (Gillespie, 2007; ´
A ila e al., 2018). The
Maca onesian islands, in pa icula , se e as ese oi s o biodi e si y,
hos ing a di e se assemblage o ma ine li e, including a signi ican
numbe o mac oalgae species (F ei as e al., 2019). These islands ac as
o sho e e uges o a di e se a ay o ma ine o ganisms, la gely due o
he s able clima ic condi ions ha pe sis ed h ough he Pleis ocene
glacia ions, which os e ed unique gene ic lineages in hese isola ed
habi a s (C owley, 1981; P laumann e al., 2003; Hayes e al., 2005;
Xa ie e al., 2010).
Despi e hei ecological impo ance, oceanic islands ace signi ican
h ea s om non-indigenous species (NIS) in oduc ions (Micael e al.,
2014; Cas o e al., 2022). The ulne abili y o hese islands o ma ine
NIS is closely linked o he isola ion o hei shallow-wa e ecosys ems,
which limi s na u al colonisa ion and eco e y om dis u bances
(Pa ish, 1989; Hachich e al., 2015). Al hough some ma ine pop-
ula ions on oceanic islands main ain occasional connec ions o coas al
a eas, his connec i i y is highly a iable and o en species-speci ic,
in luenced by local oceanog aphic ba ie s such as deep ocean
* Co esponding au ho . MARE - Ma ine and En i onmen al Sciences Cen e / ARNET - Aqua ic Resea ch Ne wo k, Regional Agency o he De elopmen o
Resea ch, Technology and Inno a ion (ARDITI), Funchal, Po ugal.
E-mail add ess: [email p o ec ed] (S. Chebaane).
Con en s lis s a ailable a ScienceDi ec
Ma ine En i onmen al Resea ch
jou nal homepage: www.else ie .com/loca e/ma en e
h ps://doi.o g/10.1016/j.ma en es.2024.106766
Recei ed 1 July 2024; Recei ed in e ised o m 17 Sep embe 2024; Accep ed 23 Sep embe 2024
Ma ine En i onmen al Resea ch 202 (2024) 106766
A ailable online 24 Sep embe 2024
0141-1136/© 2024 The Au ho s. Published by Else ie L d. This is an open access a icle unde he CC BY-NC license ( h p://c ea i ecommons.o g/licenses/by-
nc/4.0/ ).
enches, s ong cu en s, and empe a u e g adien s (Rocha e al., 2007;
Hogan e al., 2012). These ma ine ecosys ems o en ha e lowe na i e
species ichness, smalle popula ions, and simpli ied ophic s uc u es,
which educe compe i ion and bio ic esis ance o in ade s (Vi ousek,
1990; Micael e al., 2014). Fu he mo e, demog aphic isola ion and
esou ce limi a ions, such as es ic ed ood and space, heigh en he
suscep ibili y o hese ecosys ems o NIS impac s (Micael e al., 2014).
Consequen ly, oceanic islands ypically exhibi lowe unc ional di-
e si y and simple ecological ne wo ks han mainland coun e pa s,
lea ing hem ulne able o in asions (Pea son, 2009).
In he Maca onesian a chipelagos, mac oalgae a e he mos domi-
nan NIS, wi h 31 iden i ied species (Cas o e al., 2022). The in o-
duc ion o in asi e mac oalgae can dis up he ecological balance and
gene ic in eg i y o hese ma ine communi ies, leading o signi ican
biodi e si y losses and al e ed ecosys em unc ions (Ka sane akis e al.,
2014; Gebu zi and McCa hy, 2018).
Mac oalgae a e essen ial in he ma ine ood web, as p ima y ood
sou ce bu also by p o iding habi a o di e se o ganisms ha a e p ey
o nume ous highe -le el consume s, including non-pho osyn he ic
bac e ia, p o is s, in e eb a es, and ish (Po apo a e al., 2005). In
addi ion, he e is also conside a ion o he cascading e ec along he
ood web: mac oalgae se e as a p ima y nu i ional sou ce o he bi -
o ous in e eb a es and ishes (Lim e al., 2016; Chen e al., 2021),
which in u n, indi ec ly suppo highe -le el consume s, such as
in e i o ous and pisci o ous ishes ( an Lie e al., 2018; Wenge e al.,
2018). By consuming mac oalgae-associa ed epi auna and being p ey o
meso- and apex p eda o s, in e i o ous ishes play a c ucial ole in
acili a ing he ans e o ene gy o he nex ophic le el. This p ocess
highligh s he i al ophic links ha connec p ima y p oduce s o
uppe -le el consume s, o ming he “bed ock” o ma ine biodi e si y
and ecosys em unc ioning. These ophic in e ac ions span ac oss all
ophic le els wi hin ma ine ood webs, c ea ing a bi-di ec ional dy-
namic wi h bo h op-down and bo om-up e ec s (Chen e al., 2021).
The in oduc ion o NIS can lead o biological in asions ha signi -
ican ly dis up he local ecosys ems and p omo e phase shi s (Lesse and
Sla e y, 2011; Edelis e al., 2013), loss o ood web complexi y (By nes
e al., 2007) and decline o na i e axa o lineages (Micael e al., 2014;
Thomsen e al., 2016; Gebu zi and McCa hy, 2018). In pa icula , NIS
mac oalgae ha e he po en ial o ou compe e o exclude na i e mac o-
algae and o he sessile o ganisms (Ges oso e al., 2012; Palomo e al.,
2016), as well as o ha e di ec and indi ec impac s on die a y habi s,
ood a ailabili y, ophic in e ac ions and, ul ima ely, on he local ood
webs (Ka sane akis e al., 2014; Thomsen e al., 2016). When explo ing
hese p ocesses in ol ing NIS mac oalgae, i is impo an o ecognise
he ole o epi auna in he mac oalgae-consume s in e ac ions, since
hei ole as a habi a p o ide s can in luence ophic in e ac ions dy-
namics by a ec ing he s eng h o p eda o -p ey ela ionships (i.e.,
enhancing p eda o o aging e iciency on epi auna o educing p ey
ulne abili y o p eda ion due o habi a use ai s) (Klecka and Boukal,
2013). In his con ex , he dynamics o consume – ood esou ce in-
e ac ions assume inc eased signi icance, as mac oalgae consume s and
epi auna consume s can exe ei he acili a i e o cons aining in-
luences on he p oli e a ion o NIS: consume p e e ence o na i e
mac oalgae and associa ed epi auna can indi ec ly acili a e NIS p oli -
e a ion by educing he p essu e on NIS, whe eas p e e ence o NIS
mac oalgae o epi auna may con ibu e o bio ic esis ance o in asions
(Thomsen e al., 2016; Chebaane e al., 2024).
The po en ial in luence o NIS mac oalgae on local ophic in-
e ac ions unde lines he impo ance o asce aining whe he local
consume s mani es selec i e, ejec ion, o gene alis eeding beha iou s
ha ei he p omo e o hinde he dominance o NIS mac oalgae. These
eeding beha iou s cons i u e a pi o al mechanism and undamen al
ac o s in luencing he esis ance, suscep ibili y and esilience o local
communi ies o biological in asions (San ama ía e al., 2021; Chebaane
e al., 2024). This s udy assesses whe he local ish display p e e ences
when p esen ed wi h a selec ion o na i e and NIS mac oalgae and hei
associa ed epi auna. Unde s anding hese p e e ences is c ucial o
de e mining whe he consume eeding beha iou s can con ibu e o
bio ic esis ance agains NIS. Addi ionally, his esea ch examines he
ole o mac oalgae bo h as a p ima y ood sou ce and as a habi a p o-
ide o speci ic epi auna assemblages, and how hese ac o s may in-
luence he die o local ish, which we e ca ego ised as he bi o ous,
in e i o ous o omni o ous. Ou hypo heses we e ha ish would
display a p e e ence o na i e mac oalgae and associa ed epi auna,
po en ially acili a ing he p oli e a ion o NIS mac oalgae, and ha
hese p e e ences would a y acco ding o he eeding ca ego ies o ish
(he bi o ous, in e i o ous, and omni o ous). The s udy’s indings aim
o shed ligh on he po en ial ole o consume p e e ences in media ing
he impac s o NIS mac oalgae on local ecosys ems.
2. Ma e ials and me hods
2.1. Expe imen al design and s udy si e
In his pilo s udy, one non-indigenous mac oalgal species (Aspa -
agopsis axi o mis) (Cas o e al., 2022) and wo indigenous mac oalgal
species (Halop e is scopa ia and Sa gassum ulga e) we e selec ed o
conduc he he bi o y p e e ence expe imen . The selec ion o hese
mac oalgae was d i en by hei ecological signi icance, as hey ep e-
sen he dominan communi ies in he s udied a ea, which is Madei a
Island in he NE A lan ic (Fig. 1). The en i e expe imen al p ocedu e was
conduc ed in si u in an unde wa e en i onmen , employing scuba di -
ing echniques. The wa e empe a u e was on a e age 18 ◦C (±1◦), and
he expe imen was pe o med a a dep h ange o 7–8 m. Th ee di e en
loca ions in he sou he n egion o Madei a Island we e chosen o
eplica ion o he expe imen .
2.1.1. Remo e ideo o aging sys em design
To ensu e a uni o m p esen a ion o mac oalgae o ish ac oss all
s udy loca ions, s anda d expe imen al uni s (Fig. 2) we e cons uc ed,
inspi ed by he He bi o y Assay p o ocol known as “Weedpops,” which
had been de eloped by The Ma ine Global Ea h Obse a o y (Ma ine-
GEO) and adap ed om he esea ch o Hay (1981). Addi ionally, ou
s udy used he app oach o he Remo e Video Fo aging Sys em (RVFS)
(Chebaane e al., 2022, 2024). While he RVFS expe imen adi ionally
employed ouling communi ies, ou adap a ion u ilises he concep ually
simila weedpops as he expe imen al uni s, designa ing hem as he
‘s udied bai ’ in his in es iga ion. Despi e he a iances in he speci ic
bai employed, we explici ly employ he RVFS p o ocol in ou s udy o
he assessmen o o aging beha iou .
In his s udy, each expe imen al uni , e e ed o as a “weedpop
Expe imen al uni ,” consis ed o h ee lines (1 m o ound plai opes
made om polyp opylene) called o as weedpop. To achie e he desi ed
e ical o ien a ion o each weedpop in he wa e , one end was equipped
wi h 1-kg D- ing weigh s commonly used in scuba di ing, while he
opposi e end ea u ed a plas ic sponge loa e . Each line had ou ags
placed equidis an ly, wi h an indi idual om each species o mac o-
algae andomly posi ioned adjacen o h ee ags, while he ou h space
was le emp y o se e as a con ol (as illus a ed in Fig. 2).
The mac oalgae specimens we e collec ed on he e y day o he
expe imen , wi hin he same di e, om he exac loca ion whe e he
expe imen a ion ook place. A each si e, a eam o h ee scuba di e s
deployed a o al o six weedpop expe imen al uni s, wi h h ee uni s
deployed each day o e wo consecu i e days. Each weedpop expe i-
men al uni was equipped wi h a ipod, suppo ing a ideo came a
(PARALENZ Vaqui a Unde wa e 4K Came a) se up o eco d ideos o
he weedpop expe imen al uni . To ensu e he came a’s s abili y, e e y
ipod was ancho ed wi h 2 kg o D- ing di e weigh s, a choice in lu-
enced by bo h i s ease o anspo a ion o he wa e and i s p o en e -
icacy in main aining came a s abili y.
On he i s day o he expe imen , a eam o h ee di e s deployed
h ee weedpop expe imen al uni s, including he weedpops and ipods
S. Chebaane e al.
Ma ine En i onmen al Resea ch 202 (2024) 106766
2
equipped wi h came as (Fig. 2, s ep 1). Following his, mac oalgae we e
collec ed and a ached o he weedpops (Fig. 2, s eps 2 and 3). The
posi ion o each mac oalga on e e y weedpop was documen ed by one
di e using a p e-p epa ed shee (Fig. 2, s ep 4). A e wa d, he came as
we e ac i a ed o ideo eco ding, and he eam exi ed he si e (Fig. 2,
s ep 5). Two hou s la e , wo di e s e u ned o he si e o u n o he
came as (Fig. 2, s ep 6). The p esence o absence o mac oalgae on he
weedpops was hen eco ded by one di e on he same p e-p epa ed
shee (Fig. 2, s ep 7). The came as and ipods we e subsequen ly
emo ed while he mac oalgae emained a ixed o he weedpops
(Fig. 2, s ep 8).
A e a 24-h in e al, a eam o h ee di e s e isi ed he si e. The
p esence o absence o mac oalgae was eco ded ini ially (Fig. 2, s ep 9),
ollowed by he de achmen o hese mac oalgae om he weedpops
(Fig. 2, s ep 10). A new se o mac oalgae was andomly posi ioned on
each weedpop, and he p ocedu es om he p e ious day we e epea ed
on he nex day a he same ime (Fig. 2, s ep 1–10). This p ocess
spanned 2 days pe loca ion, equi alen o 48 h o expe imen a ion.
P e ious s udies (Willis e al., 2006; Bi e al., 2012) ha e shown
ha ac o s such as ime o day and subs a e ype signi ican ly in luence
ish beha iou and abundance. These s udies ha e also epo ed minimal
day- o-day a iabili y wi hin ou con ex . The e o e, we conduc ed he
expe imen a he same ime each day o con ol o a iabili y associ-
a ed wi h ime o day and exclude ‘day’ as a ac o in ou analysis. This
app oach esul ed in six eplica es pe block a he han he ini ial
planned h ee. By s anda dising he iming o he expe imen , we
minimised he impac o ime-o -day a iabili y, enabling a ocused
analysis o mac oalgae emo al, ish p e e ence, and abundance ac oss
he h ee dis inc loca ions.
2.1.2. Epi aunal sampling design
In pa allel wi h RFVS deploymen and ials, samples o Halop e is
scopa ia, Sa gassum ulga e, and Aspa agopsis axi o mis we e collec ed
om each o he h ee s udy loca ions: Funchal, Ga ajau, and Quin a do
Lo de. Th ee eplica es o each mac oalga pe loca ion we e collec ed,
each sepa a ed by a ew me e s (Fig. 2, s ep 11). Mac oalgae indi iduals
a a dep h o 7–8 m we e ca e ully de ached om he subs a e, enclosed
in a plas ic zip bag o minimize o ganism loss, and p omp ly anspo ed
o he labo a o y on he same day.
In he labo a o y, each sample was insed h ough a 0.2 mm mesh
sie e wi h esh wa e o cap u e all mobile mac o auna. Subsequen ly,
each mac oalga was examined unde a s e eomic oscope o de ach any
emaining epi auna species ha we e s ill a ached. Each sample om
each mac oalga was placed in a con aine and p ese ed wi h 70%
e hanol. O ganisms we e iden i ied o he o de le el, as his le el o
iden i ica ion su ices o communi y dis inc ion (Timms e al., 2013;
O e o-Fe e e al., 2019; Gue a-Ga cía e al., 2021). The quan i ica ion
was done using a dissec ing mic oscope.
The mac oalgae we e placed on il e pape o d ain he wa e .
Subsequen ly, each sample o mac oalga was weighed using a digi al
scale wi h millig am p ecision. Epi auna abundance was assessed by
s anda dising he coun o a numbe o indi iduals pe 10 g o we
mac oalgal weigh .
2.2. Da a analysis
The analysis was conduc ed using P ime 7, while da a isualisa ion
was ca ied ou using bo h P ime 7 and R ( e sion 4.2.1), u ilising he
“ggplo 2” (Wickham, 2016) and “ne wo kD3”) (Allai e e al., 2017)
packages.
2.2.1. Mac oalgae emo al ac oss space
To examine a ia ions in mac oalgae emo al ac oss di e en loca-
ions, p esence-absence da a we e ga he ed om he weedpops, and a
uni a ia e mixed-e ec s PERMANOVA analysis was conduc ed,
employing Euclidean dis ance ma ices among he samples (Ande son
and Robinson, 2003). The ac o ‘Loca ion’ was ea ed as a ixed ac o
wi h h ee le els, he ac o ‘Block’ was a andom ac o nes ed wi hin
‘Loca ion’ wi h six eplica es pe loca ion, and he ac o ‘Mac oalgae’
was a ixed ac o wi h h ee le els. This analysis was i s applied o he
ini ial 2-h in e al, which se ed as a baseline, and subsequen ly o he
ull 24-h pe iod. S a is ical signi icance was assessed h ough 9999
Fig. 1. Map showing he loca ion o he s udy a ea in he sou he n egion o Madei a Island, NE A lan ic. The map highligh s he h ee si es whe e he bi o y
p e e ence expe imen s we e conduc ed.
S. Chebaane e al.
Ma ine En i onmen al Resea ch 202 (2024) 106766
3
Fig. 2. Expe imen al se up o he Remo e Video Fo aging Sys em (RVFS) ea u ing weedpops as he expe imen al uni s. Each expe imen al uni comp ises h ee
polyp opylene opes equipped wi h D- ing weigh s and loa e s ( e e ed o as weedpop), accommoda ing mac oalgae specimens. U ilising ideo came a on a ipod
o eco d he ish in e ac ion wi h he expe imen al uni s o 2 h, he p o ocol includes he ollowing s eps (1) deploymen o he weedpop uni s wi h ipods and
came as, (2) collec ion and (3) a achmen o mac oalgae o he weedpops; (4) documen a ion o posi ion a T0; (5) ini ia ion o ideo eco ding; (6) s opping he
ideo eco ding; (7) sco ing o mac oalgae p esence o absence a e 2 h (2h), (9) sco ing and (10) de achmen o mac oalgae a e 24-h in e al (24h). (11) collec ion
o mac oalgae o epi aunal sampling. The en i e p ocess o als a 48-h expe imen pe iod.
S. Chebaane e al.
Ma ine En i onmen al Resea ch 202 (2024) 106766
4
pe mu a ions a a con idence le el o
α
=0.05. I he coun o unique
pe mu a ions was insu icien o s a is ical in e ences a a signi icance
le el o 0.05, Mon e Ca lo P- alues we e used (Ande son and Robinson,
2003; Ande son, 2008).
Simila i y ma ices we e calcula ed using he B ay-Cu is index when
compa ing quan i a i e species composi ion and Euclidean dis ances in
all o he cases. This s anda disa ion was used o s eamline he
desc ip ion o simila i y measu es ac oss he di e en analyses.
2.2.2. Fish abundance and composi ion
A sole obse e employing a scan sampling echnique documen ed
ish species abundance, composi ion, and beha iou s. The obse e
conduc ed scans second by second wi hin in e als o app oxima ely 4
min and 30 s due o he au oma ic spli ing o ideo by he came a.
Howe e , all eco ded seconds we e included in he analysis o ensu e
comp ehensi e da a co e age. A o al o 24 h, 38 min, and 51 s o ideo
oo age we e eco ded ac oss he h ee s udy loca ions, wi h app oxi-
ma ely 1 h and 30 min o ideo con en pe weedpop expe imen al uni .
Fish abundance was quan i ied using he MaxN analysis me hod (Cappo,
2010; Whi ma sh e al., 2017), which es ima es he maximum numbe o
indi iduals om a single species obse ed wi hin an indi idual ideo
ame. The de i ed MaxN da a we e hen used o calcula e he
B ay-Cu is simila i y ma ix, ollowed by he applica ion o a Mul i-
a ia e One-way analysis o co a iance PERMANCOVA (Ande son,
2001), wi h ‘Loca ion’ as a ixed ac o . The o al du a ion in seconds pe
weedpop expe imen al uni was included in he design as a co a ia e.
Subsequen ly, a Simila i y Pe cen ages P ocedu e (SIMPER) analysis
was conduc ed o illus a e he con ibu ions o ish species o he
obse ed simila i ies in species composi ion ac oss he a ious loca ions.
Addi ionally, ish composi ion was assessed wi hin each loca ion using a
p esence-absence me hodology. These da a we e used o calcula e a
Jacca d ma ix. Once again, a PERMANCOVA analysis was applied,
using he same design as he p e ious analysis, o in es iga e po en ial
di e ences in species composi ion ac oss he s udy loca ions. T ophic
le els and die s o each ish species we e de e mined based on he da a
p o ided by FishBase (F oese and Pauly, 2000).
2.2.3. Fish beha iou
To u he explo e ish beha iou dynamics, he analysis ocused on
h ee ca ego ies: p esence, in e es and bi ing. P esence was quan i ied
as he amoun o ime (in seconds) a ish was isible on he ideo. In-
e es was de ined as he du a ion (in seconds) du ing which a ish swam
owa d o paused in on o speci ic species o mac oalgae, and bi ing
was de ined as he du a ion (in seconds) du ing which a ish ed on he
mac oalgae (Chebaane e al., 2022). The p esence o ish pe second
ac oss di e en loca ions was assessed using a uni a ia e PERMANOVA
analysis. This analysis employed a nes ed design, wi h he ‘Loca ion’
ac o being ixed, and he ‘Weedpop expe imen al uni ’ nes ed wi hin
each loca ion as a andom ac o . The o al du a ion in seconds pe
weedpop expe imen al uni was included in he design as a co a ia e. To
assess he p e e ence o ish o speci ic mac oalga species in e ms o
in e es and bi ing, a uni a ia e PERMANCOVA analysis was employed.
This analysis ollowed a comple e andomised block design, wi h he
‘Loca ion’ ac o ixed, he ‘block’ ac o nes ed andomly wi hin
‘Loca ion’, and he ‘mac oalgae’ ac o ixed. The p esence, in e es , and
bi ing beha iou s o each ish species we e hen isualised in a Sankey
diag am. These diag ams we e c ea ed using he “ne wo kD3” package
in R (Allai e e al., 2017).
2.2.4. Epi auna associa ed wi h he mac oalgae
The epi aunal abundance da a om each sample we e ini ially
compu ed by di iding he epi auna coun by he o iginal weigh o he
mac oalgae sample, s anda dising i o 10 g o compa abili y. Subse-
quen ly, no mali y was assessed using he Shapi o-Wilk es , and he
homogenei y o a iances was examined using Le ene’s es in IBM SPSS
S a is ics 27, wi h esul s indica ing ha no mal da a dis ibu ion and
homogenei y in a iances we e achie ed. A wo-way PERMANOVA es
was conduc ed a e calcula ing he B ay-Cu is simila i y ma ix o
epi auna abundance o in es iga e di e ences in epi auna communi y
composi ion among he s udied mac oalgae species wi hin each loca-
ion. The ac o ‘loca ion’ was ixed wi h h ee le els (Funchal, Ga ajau,
and Quin a do Lo de), and he ac o ‘mac oalgae’ was ixed wi h h ee
le els (Halop e is scopa ia and Sa gassum ulga e and Aspa agopsis axi-
o mis). Da a we e in en ionally le un ans o med o a oid magni ying
a e epi aunal species; o example, The mo e complex mo phology o
H. scopa ia compa ed o A. axi o mis and S. ulga e esul s in mo e
a ailable space, which could lead o a highe pe cen age o a e species.
P incipal Coo dina es Analysis (PCO) was used o isualise he epi aunal
composi ion by axonomic class o each mac oalgae species a each
loca ion. The same PERMANOVA design was used o es di e ences in
he numbe o epi aunal axa, o al epi aunal abundance, and he
Shannon-Wiene di e si y index (H
′
) be ween he s udied mac oalgae
species in each loca ion a e calcula ing he Euclidean dis ance simi-
la i y ma ix.
3. Resul s
3.1. Mac oalgae emo al ac oss space
Si e-speci ic mac oalgae emo al was obse ed, ega dless o he
du a ion o exposu e o g aze . Du ing he ini ial 2-h pe iod, mac oalgae
we e no iceably emo ed only om Funchal. Speci ically, bo h na i e
and b own mac oalgae, Halop e is scopa ia and Sa gassum ulga e, we e
emo ed om he Funchal weedpop. Howe e , s a is ical signi icance
was no achie ed due o he ela i ely low p opo ion o hese emo als
(Table 1; Fig. 3).
Following a 24-h in e al, he selec i i y o H. scopa ia and S. ulga e
in Funchal became s a is ically signi ican as a g ea e p opo ion o
hese mac oalgae we e emo ed. As o he non-indigenous ed mac o-
algae Aspa agopsis axi o mis, emo al was obse ed in a single weedpop
loca ed in Quin a do Lo d. Fu he mo e, in Ga ajau, na i e b own
mac oalgae H. scopa ia we e also emo ed, bu his was limi ed o a
small p opo ion, esul ing in s a is ical non-signi icance (Table 1;
Fig. 3).
3.2. Fish abundance and composi ion
A o al o 109 ish, ep esen ing 13 di e en species, we e obse ed
wi hin he 18 weedpop expe imen al uni s ac oss a ious loca ions
(Fig. 4). These loca ions displayed signi ican di e ences in bo h ish
composi ions and abundance (Table S1). Due o a came a mal unc ion a
he Ga ajau loca ion, he eco ded oo age o one o he weedpops
Table 1
Resul s o he uni a ia e PERMANOVA o he p opo ion o mac oalgae
emo al ac oss loca ions du ing he ini ial 2 h and a e 24 h. Abb e ia ions
used: EXP - Expe imen al uni ; F - Funchal; Q - Quin a do Lo de; G - Ga ajau; A -
Aspa agopsis axi o mis; H - Halop e is scopa ia; S - Sa gassum ulga e. Bold on
indica es s a is ical signi icance a he
α
=0.05 le el. D - deg ees o eedom; MS
- mean squa e sum; P (MC) - p- alues o he pe mu a ion using he Mon e-Ca lo
es . The pe mu a ion es was conduc ed wi h 9999 pe mu a ions.
Sou ce D 2-h 24-h
MS Pseudo-
F
P
(MC)
MS Pseudo-
F
P
(MC)
Loca ion 2 0.67 5.00 0.02 3.91 15.29 0.001
Mac oalgae 2 0.17 1.67 0.20 1.19 7.11 0.001
EXP (Loca ion) 15 0.13 1.33 0.24 0.26 1.53 0.15
Loca ion x
Mac oalgae
4 0.17 1.67 0.18 1.32 7.94 0.001
Residuals 30 0.10   0.17  
Pai wise
compa isons
F∕= (Q =G) Fo F: A ∕= (S=H)
Q and G: A =S =H
S. Chebaane e al.
Ma ine En i onmen al Resea ch 202 (2024) 106766
5

blocks was sho e compa ed o he o he samples. To accoun o his
di e ence, ideo du a ion was included as a co a ia e in all subsequen
analyses.
The s a is ical analysis ca ied ou h ough PERMANOVA e ealed
di e ences in MaxN among he loca ions (Table S1), indica ing a ia-
ions in he maximum numbe o obse ed ish. Howe e , PERMDISP
analysis did no show any signi ican di e ences in dispe sion ac oss he
loca ions (F2,15 =2.5809; p =0.1), sugges ing a ela i ely consis en
dis ibu ion pa e n (Table S1).
Rega ding species composi ion, he PERMANOVA analysis also
demons a ed signi ican di e ences ac oss he loca ions (Table S1),
indica ing a ia ions in he species p esence. Consis en wi h he MaxN
analysis, PERMDISP analysis did no e eal signi ican di e ences in
dispe sion ac oss loca ions (F2,15 =2.984; p =0.1), implying ela i ely
uni o m species dis ibu ion pa e ns (Table S1).
In Funchal, 45 indi iduals om 7 di e en ish species we e
obse ed, wi h an a e age obse a ion a e o 5.5 indi iduals pe hou .
No ably, Ch omis limba a and Spa isoma c e ense we e he mos abundan
species in his a ea (Fig. 4; Table S2).
In Ga ajau, 20 indi idual ish, ep esen ing 6 di e en species, we e
eco ded, wi h an a e age obse a ion a e o 2.8 indi iduals pe hou .
The dominan species obse ed in his loca ion we e Thalassoma pa o
and Sa pa salpa (Fig. 4; Table S2).
Las ly, in Quin a do Lo de, 44 indi idual ish om 8 di e en species
we e documen ed, wi h an a e age obse a ion a e o 5 indi iduals pe
hou . The dominan species in his a ea we e Can higas e capis a a and
Sphoe oides ma mo a us (Fig. 4; Table S2).
Mos ish species obse ed in he h ee loca ions we e classi ied as
ca ni o ous, wi h a ophic le el ange o 3.0–4.0, p edominan ly
eeding on zooben hos. Among he ish iden i ied, Sa pa salpa was he
sole he bi o ous species, wi h a ophic le el 2.0. Kyphosus sec a ix has
a ophic le el ange o 2.0–2.19, and Spa isoma c e ense had a ophic
le el o 2.6, indica ing ha hey consume mac oalgae and small
in e eb a es.
Fig. 3. P opo ion o mac oalgae emo ed om weedpops pe loca ion (Funchal, Ga ajau, and Quin a do Lo de) du ing he ini ial 2 h (T2) and a e 24 h (T24). The
x-axis ep esen s he ime in e als (T0: ini ial se up, T2: a e 2 h, and T24: a e 24 h), while he y-axis shows he p opo ion o mac oalgae missing. Each poin
ep esen s he mean p opo ion missing o each mac oalga species (Aspa agopsis axi o mis, Halop e is scopa ia, and Sa gassum ulga e), wi h e o ba s indica ing he
s anda d e o .
Fig. 4. Shade plo illus a es he a e age maximum numbe o indi idual ish pe hou o each species, ac oss ideo se ups (n =6 pe loca ion), a a ious loca ions.
S. Chebaane e al.
Ma ine En i onmen al Resea ch 202 (2024) 106766
6
3.3. Fish beha iou
Rega dless o speci ic ish species, he empo al p esence o ish
displayed signi ican simila i y ac oss he h ee s udied loca ions
(Table S3; Fig. 5). Ch omis limba a was he p edominan species in e ms
o empo al p esence in Funchal, while in Ga ajau, Sa pa salpa and
Spa isoma c e ense we e he mos equen ly obse ed species. In Quin a
do Lo de, he h ee mos equen ish species we e Can higas e cap-
is a a, Sphoe oides ma mo a us, and Thalassoma pa o (Table S4).
Signi ican di e ences in ish p e e ences we e obse ed when
iden i ying mac oalgae species ha a ac ed pa icula in e es .
Sa gassum ulga e and Halop e is scopa ia we e p e e ed o e Aspa -
agopsis axi o mis ac oss all loca ions (Table S5). Th ee ish species,
Ch omis limba a, Thalassoma pa o, and Spa isoma c e ense, consis en ly
in e ac ed wi h all h ee p esen ed mac oalgae species. Only hese h ee
ish species showed in e es o he non-indigenous ed mac oalgae
A. axi o mis. Addi ionally, Can higas e capis a a showed a dis inc
p e e ence o S. ulga e along wi h he h ee men ioned ish species. Fo
H. scopa ia, all he p e iously men ioned ish, along wi h Can higas e
capis a a, Similipa ma lu ida, and Balis es cap iscus, showed in e es in
his mac oalga species.
Rega ding bi ing beha iou , only Spa isoma c e ense, Can higas e
capis a a, Thalassoma pa o, and Similipa ma lu ida we e obse ed bi ing
(Fig. 5). Among he h ee mac oalgae species p esen ed, bo h H. scopa ia
and S. ulga e we e bi en, wi h H. scopa ia being he mos equen ly
bi en (Fig. 5; Table S5). On he o he hand, A. axi o mis, was no bi en
by any ish.
A Quin a do Lo de and Funchal, ish a ge ed only H. scopa ia.
Speci ically, in Funchal, Thalassoma pa o showed p e e ed o
H. scopa ia, while in Quin a do Lo de, Spa isoma c e ense displayed a
simila p e e ence. In Ga ajau, all ou species we e obse ed bi ing
H. scopa ia, and bo h Spa isoma c e ense and Can higas e capis a a we e
seen o bi ing no only H. scopa ia bu also S. ulga e.
In he con ex o he die o he in e ac ing ish species wi h he
p o ided mac oalgae, i was obse ed ha all ish p ima ily ed on
zooben hos, excep o S. c e ense, which displayed a mixed die a y
p e e ence o bo h zooben hos and mac oalgae.
3.4. Epi auna associa ed wi h he mac oalgae
A o al o 3621 epi auna indi iduals we e iden i ied o he o de
le el, encompassing 5 phyla, 13 classes, and 34 o de s. The o al
epi aunal abundance was p ima ily composed o A h opods, mainly
Amphipoda (90%), ollowed by Ha pac icoida (3%) and Decapoda
(2.5%). Molluscs accoun ed o 10% o he o al abundance, p edomi-
nan ly ep esen ed by Ca di ida (Bi al ia) a 36% and Caenogas opoda
Fig. 5. Sankey diag am illus a ing ish in e ac ions wi h mac oalgae ac oss he h ee s udy loca ions (Ga ajau, Funchal, and Quin a do Lo de). The diag am shows
low o ish beha iou s: P esence, In e es , and Bi ing; owa ds speci ic mac oalgae species (Aspa agopsis axi o mis, Sa gassum ulga e, and Halop e is scopa ia). The
nodes ep esen loca ions, mac oalgae, and ish species. Links be ween he nodes indica e how ish species in e ac wi h mac oalgae, wi h he wid h o each link
co esponding o he du a ion (in seconds) o each beha iou pe indi idual ish. The colou s o he links and nodes ep esen he speci ic ish species in ol ed, while
icons deno e ood i em p e e ences: zooben hos, zooplank on, and plan s. The igh side o he diag am highligh s he ish species obse ed exhibi ing bi ing
beha iou , which we e he only species eco ded pe o ming his ac ion du ing he s udy.
S. Chebaane e al.
Ma ine En i onmen al Resea ch 202 (2024) 106766
7
(Gas opoda) a 36%. Annelida cons i u ed 5.4% o he o al, wi h
Phyllodocida (57%) and Te ebellida (34%). Echinode ms accoun ed o
0.5% p ima ily Ophiu oidea (68%). Cnida ia was minimally p esen ,
wi h Ac inia ia ep esen ing 0.05% o he o al abundance, obse ed in a
single sample o Sa gassum ulga e loca ed in Funchal.
Epi auna associa ed wi h he h ee mac oalgae a each s udied
loca ion showed a signi ican di e ence among he mac oalgae species,
ega dless o loca ion (Table S6). S. ulga e had he highes ela i e
abundance o epi auna, ollowed by Halop e is scopa ia and Aspa agopsis
axi o mis. PERMANOVA esul s based on he epi aunal abundance
indica ed signi ican di e ences among mac oalgae and loca ions, wi h
an in e ac ion be ween hose ac o s. Subsequen pai wise compa isons
e ealed ha he epi aunal communi ies associa ed wi h each mac oalga
species a ied signi ican ly ac oss all loca ions (Table S6). In he
Fig. 6. P incipal Coo dina es Analysis (PCO) o epi auna axonomic composi ion and abundance pe 10 g o mac oalgae species (Aspa agopsis axi o mis, Sa gassum
ulga e, and Halop e is scopa ia) ac oss h ee s udy loca ions: Funchal, Ga ajau, and Quin a do Lo de (QL). Each loca ion includes h ee eplica e samples pe
mac oalgae species, indica ed by shapes, and mac oalgae species a e dis inguished by colou . Panels (a), (c), and (e): PCO o dina ions show Epi aunal communi y
s uc u e based on B ay-Cu is dis ance ma ices, wi h he axes ep esen ing he main a ia ion among samples: (a) A h opoda, (c) Mollusca, and (e) Annelida.
Bubble sizes e lec he ela i e abundance o each axonomic g oup, as indica ed in he legend. Panels (b), (d), and ( ): Ba cha s display he pe cen age con ibu ion
o each o de o he o al epi aunal abundance o each axonomic g oup pe mac oalgae species wi hin each loca ion: (b) A h opoda, (d) Molluscs, and ( ) Annelids.
The o de s wi hin each phylum a e depic ed wi h speci ic colou s as lis ed in he legend.
S. Chebaane e al.
Ma ine En i onmen al Resea ch 202 (2024) 106766
8
P incipal Coo dina es Analysis (PCO) o dina ion based on epi aunal
abundances pe mac oalga showed dis inc g oupings among samples
ep esen ing di e en mac oalgae species (Fig. 6). Howe e , samples o
H. scopa ia in Ga ajau me ged wi h hose om all loca ions o S. ulga e.
No dis inc g oupings eme ged among he sampling si es, sugges ing
ha he signi ican di e ences in epi aunal composi ion we e con ingen
upon speci ic algal species, hough a si e-speci ic e ec was seen due o
simila i ies be ween H. scopa ia in Ga ajau and S. ulga e (Table S6;
Fig. 6).
Rega ding o de ichness and di e si y, signi ican di e ences we e
obse ed among he mac oalgae species, wi h he si e-speci ic e ec
no ed (Table S6; Fig. 7). In Funchal, H. scopa ia hos ed he highes
numbe o o de s, ollowed by S. ulga e, while A. axi o mis hos ed he
ewes . Fo di e si y, H. scopa ia displayed g ea e di e si y han bo h
S. ulga e and A. axi o mis, which displayed equal le els.
In Ga ajau, H. scopa ia simila ly hos ed he highes numbe o o de s,
while A. axi o mis and S. ulga e displayed equi alen o de ichness.
Conce ning di e si y, A. axi o mis and H. scopa ia demons a ed he
highes le els, in con as o S. ulga e.
Fo Quin a do Lo de, A. axi o mis and H. scopa ia hos ed a highe
numbe o o de s compa ed o S. ulga e. In e ms o di e si y,
H. scopa ia showed g ea e di e si y compa ed o A. axi o mis and
S. ulga e, wi h S. ulga e displaying he lowes di e si y.
Wi h espec o axonomic composi ion, a h opods we e he mos
abundan phylum ac oss all h ee s udied loca ions and mac oalgae,
ollowed by molluscs and annelids. S. ulga e displayed he highes
abundance o a h opods in all loca ions when compa ed o A. axi o mis
and H. scopa ia (Table S7: Fig. 6). Howe e , in Funchal and Quin a do
Lo de, he e was no signi ican di e ence in he o al abundance o a -
h opods be ween A. axi o mis and H. scopa ia. In Ga ajau, A. axi o mis
hos ed he lowes o al abundance o a h opods (Table S7: Fig. 6).
In e es ingly, S. ulga e hos ed he lowes numbe o a h opods o -
de s ac oss all loca ions (Funchal (5), Ga ajau (4), and Quin a do Lo de
(5)), while H. scopa ia displayed he highes numbe o a h opod o de s
(Funchal (8), Ga ajau (6), and Quin a do Lo de (7)). Only in Quin a do
Lo de we e he numbe s o a h opod o de s equal be ween A. axi o mis
and H. scopa ia (7) (Fig. 6).
Rega ding molluscs, which cons i u e he phylum wi h he la ges
numbe o o de s (13), he highes o al abundance o molluscs ac oss all
loca ions was obse ed in H. scopa ia, ollowed by S. ulga e and
A. axi o mis. No ably, A. axi o mis hos ed he lowes o al abundance o
molluscs in Funchal and Ga ajau. In Quin a do Lo de, bo h A. axi o mis
and S. ulga e displayed an equal o al abundance o molluscs (Table S7;
Fig. 6).
H. scopa ia also displayed he highes numbe o mollusc o de s in
bo h Funchal (10) and Ga ajau (5). In hese loca ions, S. ulga e anked
second, hos ing 5 o de s o molluscs in Ga ajau and 8 in Funchal.
Con e sely, A. axi o mis had he lowes numbe o mollusc o de s in
hese loca ions, wi h 3 in Funchal and 1 in Ga ajau. Howe e , in Quin a
do Lo de, A. axi o mis hos ed he highes numbe o mollusc o de s (7),
su passing S. ulga e (5) and H. scopa ia (5).
Rega ding he annelids, he o al abundance o annelids was highes
in H. scopa ia, ollowed by S. ulga e, while A. axi o mis hos ed he
lowes abundance o annelids in Funchal. Howe e , in Quin a do Lo de
and Ga ajau, S. ulga e exhibi ed an equal o al abundance o annelids as
H. scopa ia and A. axi o mis, wi h H. scopa ia su passing A. axi o mis in
o al annelid abundance (Table S7; Fig. 6).
H. scopa ia, bo h in Funchal and Ga ajau, hos ed he highes numbe
o annelid o de s, wi h 5 in Funchal and 3 in Ga ajau, in con as o
A. axi o mis (2 in Funchal and 1 in Ga ajau) and S. ulga e (3 in Funchal
and 2 in Ga ajau). In Quin a do Lo de, all he mac oalgae displayed an
equal numbe o annelid o de s, which amoun ed o 2 (Fig. 6).
Fig. 7. (a) To al epi aunal abundance (densi y) pe 10 g o mac oalgae species ac oss all loca ions. (b) Numbe o o de s ( axonomic ichness) iden i ied pe
mac oalgae species wi hin each loca ion. (c) Shannon-Wiene index (H
′
) pe mac oalgae species wi hin each loca ion (di e si y). Le e s (a, b, c) abo e ba s indica e
s a is ically signi ican di e ences be ween mac oalgae species a he
α
=0.05 le el, based on pai wise compa ison o PERMANOVAs (de ailed esul s in Table S6).
Loca ions: F: Funchal; G: Ga ajau; Q: Quin a do Lo de.
S. Chebaane e al.
Ma ine En i onmen al Resea ch 202 (2024) 106766
9