109
E alua ing b own ou as a po en ial biological con ol agen o
signal c ay ish
Ma ilde Ma os1, Amílca Teixei a2, An ónio B. Noguei a1, Janeide Padilha1, Ronaldo Sousa1
1 CBMA – Cen e o Molecula and En i onmen al Biology/ARNET-Aqua ic Resea ch Ne wo k/ IB-S, Ins i u e o Science and Inno a ion o Bio-Sus ainabili y,
Depa men o Biology, Uni e si y o Minho, Campus Gual a , 4710-057 B aga, Po ugal
2 CIMO, LA SusTEC, Ins i u o Poli écnico de B agança, Campus de San a Apolónia, 5300-253 B agança, Po ugal
Co esponding au ho : Ronaldo Sousa ( [email p o ec ed])
Copy igh : © Ma ilde Ma os e al.
This is an open access a icle dis ibu ed unde
e ms o he C ea i e Commons A ibu ion
License (A ibu ion 4.0 In e na ional – CC BY 4.0).
Resea ch A icle
Abs ac
This s udy e alua es he po en ial o b own ou (Salmo u a) as a biological con ol o a ecen ly
es ablished signal c ay ish (Paci as acus leniusculus) popula ion in a p o ec ed a ea (Bacei o Ri e ,
Mon esinho Na u al Pa k, Po ugal). Fi e sampling si es we e moni o ed h oughou 2023. Re-
sul s indica ed ha b own ou we e able o p eda e on signal c ay ish bu did so in equen ly as
only 12.24% o he sampled ish showed signs o signal c ay ish in hei s omachs. The numbe
o signal c ay ish in he s omach con en s o b own ou was also low (only 2.13% o all p ey
i ems), bu accoun ed o 17.70% o he o al biomass o all e ie ed p ey i ems. P eda ion was
highe in he wa me mon hs and was size-dependen , wi h la ge ish mo e able o p eda e on his
non-na i e c ay ish species. These indings highligh ha , al hough b own ou can p ey on signal
c ay ish, hei e ec i eness as a biological con ol agen is limi ed due o hei lowe abundance
and p eda ion a es. Howe e , his si ua ion may change in he u u e since b own ou may in-
c easingly consume signal c ay ish as hey become mo e amilia wi h his no el p ey. The e o e,
i is impo an o conse e key na i e popula ions, and e en ein o ce hei abundance, o allow
communi ies o de elop e ec i e esis ance o non-na i e species. In he mean ime, a mul i ace ed
managemen app oach ha inco po a es addi ional con ol s a egies, such as mechanical emo al,
is ecommended o educe he abundance and biomass o signal c ay ish, po en ially mi iga ing
hei impac and helping o main ain ecosys em balance in his p o ec ed a ea.
Key wo ds: Biocon ol, eshwa e ecosys ems, Mon esinho Na u al Pa k, non-na i e species,
Paci as acus leniusculus, Salmo u a
In oduc ion
Non-na i e species a e a majo h ea o biodi e si y, pa icula ly in eshwa e
ecosys ems (S aye 2010; Galla do e al. 2016). Among hese, he signal c ay-
ish (Paci as acus leniusculus) s ands ou as one o he mos ecologically dis up i e
non-na i e species in Eu ope (K ieg e al. 2020). Na i e o No h Ame ica, his
species was in oduced p ima ily o aquacul u e and comme cial easons (Holdich
e al. 2014), bu has apidly expanded i s ange, in e ac ing wi h na i e species o
esou ces and habi a use and al e ing key ecological p ocesses (Galib e al. 2021;
Al es e al. 2025). I s p esence leads o cascading e ec s on eshwa e ecosys-
ems, including changes in ophic in e ac ions, educ ions in na i e biodi e si y,
Academic edi o : Ped o Anas ácio
Recei ed:
4 Ma ch 2025
Accep ed:
14 May 2025
Published:
7 Oc obe 2025
Ci a ion: Ma os M, Teixei a A,
Noguei a AB, Padilha J, Sousa R
(2025) E alua ing b own ou as a
po en ial biological con ol agen
o signal c ay ish. In: Anas ácio P,
Ribei o F, Chainho P (Eds) In asions
in Aqua ic Sys ems. NeoBio a 102:
109–123. h ps://doi.o g/10.3897/
neobio a.102.152018
NeoBio a 102: 109–123 (2025)
DOI: 10.3897/neobio a.102.152018
Ad ancing esea ch on alien species and biological in asions
A pee - e iewed open-access jou nal
NeoBio a
110
NeoBio a 102: 109–123 (2025), DOI: 10.3897/neobio a.102.152018
Ma ilde Ma os e al.: B own ou as a biological con ol
and habi a deg ada ion (Ca alho e al. 2022a, 2025). The i s documen ed
occu ence o signal c ay ish in Po ugal da es back o 1997 in he Maçãs Ri e
(Be na do e al. 2011), and since hen, his non-na i e species has been inc easing
in abundance and expanding i s dis ibu ion in he no heas pa o he coun y
(Anas ácio e al. 2019; Mei a e al. 2019; Ca alho e al. 2025).
T adi ional con ol measu es applied o signal c ay ish, such as mechanical
emo al and chemical ea men s, o en ail o p o ide long- e m solu ions
and can ha e unin ended ecological consequences (Moo house e al. 2014;
Peay e al. 2019). A p omising al e na i e is he use o na i e p eda o s such as
eels (Anguilla anguilla) o egula e non-na i e c ay ish popula ions, le e aging
na u al ecological in e ac ions o es o e balance wi hin a ec ed ecosys ems
(Aquiloni e al. 2010; Musseau e al. 2015). Gi en he ecological impac o sig-
nal c ay ish and he challenges o adi ional con ol me hods in moun ainous
oligo ophic i e s, in es iga ing he ole o po en ial na i e p eda o s, such as
b own ou (Salmo u a), is c ucial o unde s anding whe he his ish can
con ibu e as a biological con ol o his non-na i e species. The b own ou , a
widely dis ibu ed species in Eu opean eshwa e ecosys ems, is known o i s
oppo unis ic eeding beha io , including c ay ish species (B idcu and Gille
1995). As a isual p eda o , b own ou p ima ily ely on sigh o cap u e
p ey, wi h eeding ac i i y peaking a dawn and dusk (Kleme sen e al. 2003).
This beha io al pa e n aligns wi h he desc ibed noc u nal ac i i y o signal
c ay ish (Sb agaglia and B ei haup 2022), making his non-na i e species po-
en ially ulne able o p eda ion du ing c epuscula hou s. Addi ionally, la ge
b own ou ha e been obse ed o inco po a e p og essi ely la ge p ey in o
hei die , sugges ing ha adul indi iduals could exe signi ican p eda ion
p essu e on c ay ish popula ions (Kleme sen e al. 2003).
Al hough p eda ion by b own ou may con ibu e o he con ol o signal
c ay ish, i s e ec i eness depends on a ious ecological ac o s, including p ey
a ailabili y, en i onmen al condi ions, and he adap abili y o bo h species (Ca ls-
son e al. 2009). Mo eo e , he e may be a lag pe iod be o e na i e p eda o s ecog-
nize non-na i e c ay ish as a iable ood sou ce, po en ially allowing hei popula-
ions o expand be o e signi ican p eda ion p essu e is exe ed (Ca oll 2007; Cox
2013). In ac , se e al s udies epo inc eased consump ion (and e en egula ion)
on non-na i e popula ions wi h inc easing in asion ime (Siemann e al. 2006;
Ca lsson and S aye 2009; Ca lsson e al. 2011; San ama ía e al. 2022), o h-
e s indica e no such change (Ca pen e and Cappuccino 2005; Pin o and Bye s
2015). Gi en his backg ound, and because p e en ion (signal c ay ish is al eady
in he sys em) and e adica ion (signal c ay ish is al eady widesp ead in he s udied
p o ec ed a ea) a e no longe iable solu ions, unde s anding he po en ial ole o
b own ou as a na u al p eda o is essen ial o assessing he iabili y o biological
con ol as a sus ainable managemen s a egy (e.g. be e conse e o e en e-s ock
b own ou popula ions in in aded a eas). The e o e, his s udy aimed o: i) assess
he deg ee o p eda ion o b own ou on he signal c ay ish; ii) assess possible spa-
ial and empo al di e ences in p eda ion a es; and iii) e alua e i his p eda ion is
size-dependen . We hypo hesise ha : i) b own ou will p eda e on signal c ay ish,
wi h p eda ion a es being highe in si es whe e c ay ish a e mo e abundan ; ii) p e-
da ion a es will inc ease in wa me mon hs due o g ea e ac i i y o bo h species;
and iii) la ge ou will consume mo e c ay ish.
111
NeoBio a 102: 109–123 (2025), DOI: 10.3897/neobio a.102.152018
Ma ilde Ma os e al.: B own ou as a biological con ol
Ma e ial and me hods
S udy a ea
Ou s udy a ea comp ised i e sampling si es along he Bacei o Ri e (Fig. 1), wi h-
in he Mon esinho Na u al Pa k (NE Po ugal), a p o ec ed a ea c ea ed in 1979
(Cas o e al. 2010). This p o ec ed a ea has a high biodi e si y, hos ing, o example,
abou 80% o Po ugal’s mammal species. The Bacei o Ri e , wi h a o al leng h
o 60 km, o igina es in Spain and belongs o he Dou o basin (Sousa e al. 2019).
The a ea has e y low human dis u bance and ha bo s a ich aqua ic biodi e si y
wi h high conse a ion s a us (e.g. Py enean desman Galemys py enaicus, classi ied
as Endange ed by he IUCN Red lis ). Rega ding he ish communi y, only na i e
species a e p esen in he s udied a ea and comp ised he b own ou (Salmo u -
a) classi ied as Leas Conce n by he IUCN Red lis , he No he n s aigh -mou h
nase (Pseudochond os oma du iense) classi ied as Nea Th ea ened, he Ibe ian ba bel
(Lucioba bus bocagei) classi ied as Nea Th ea ened, and he No he n Ibe ian chub
(Squalius ca oli e ii) classi ied as Nea Th ea ened (Oli ei a e al. 2025). The i e
sampling si es ha e a simila wid h (a ound 6 m) and he al i ude a ies be ween 594
m in he mos downs eam si e (B1) and 835 m in he mos ups eam si e (B5). This
i e is an ideal a ea o s udying p eda o -p ey in e ac ions gi en he lack o o he
(besides biological in asions) signi ican human dis u bances, such as pollu ion and
he p esence o dams (Sousa e al. 2019, 2020). The signal c ay ish was i s de ec ed
in he Bacei o Ri e in 2013 (Sousa e al. 2015).
En i onmen al cha ac e iza ion
The ollowing en i onmen al pa ame e s we e measu ed using a HACH HQ2200
mul i-pa ame e p obe a e e y sampling si e and h oughou he s udy pe iod:
pH, o al dissol ed solids (mg/L), conduc i i y (μS/cm), dissol ed oxygen (mg/L),
and empe a u e (°C). This en i onmen al cha ac e iza ion was consis en ly pe -
o med in all i e sampling si es ac oss nine di e en ime pe iods ( om Ap il o
Decembe ) h oughou he en i e yea o 2023.
B own ou and signal c ay ish da a collec ion
B own ou eed, p e e ably, in he ea ly hou s o he day, so hei collec ion was
ca ied always in he mo ning o coincide wi h hei na u al eeding beha io .
Fish we e cap u ed using elec o ishing (Hans G assl™ ELT60II-GI; 300–600 V,
DC, 2200W) o a pe iod o 30-minu es a each sampling si e along a s anda d
leng h o 150 me e s o he i e channel, comp ising an a ea o abou 1000 m2
in each si e. To gua an ee si e independence be ween loca ions, and o educe
he possibili y o S. u a mo ing be ween sampling si es, loca ions we e sepa-
a ed by a leas 1 km. This allows o si e independence, since esiden b own
ou S. u a shows high si e ideli y, no mally s aying wi hin a ew hund ed
me e s o hei p e e ed spo s o much o he yea (O idio e al. 2002; Höjesjö
e al. 2015). The ela i e abundance o b own ou pe si e was exp essed as he
o al numbe o indi iduals pe ca ch pe uni o e o (ind. CPUE). We p e e
o use he e m abundance a he han densi y, as he sampling p ocedu e may
in oduce some bias and is unlikely o cap u e all he ou wi hin he 1000 m2
112
NeoBio a 102: 109–123 (2025), DOI: 10.3897/neobio a.102.152018
Ma ilde Ma os e al.: B own ou as a biological con ol
a ea su eyed. Ne e heless, since he sampling was consis en ly conduc ed by
he same expe ienced ope a o s, we a e con iden ha compa isons be ween
sampling si es and ac oss di e en imes o he yea emain alid.
A o al o 8 o 10 aps we e used pe si e o cap u e signal c ay ish, and le un-
de wa e o 24 hou s. These aps we e speci ically used o decapod c us aceans,
educing he chances o possible byca ch o o he species, namely ish. T aps we e
all cylind ical (43 cm d, 22 cm h; 1.5 cm mesh) and we e bai ed wi h dead ma ine
ish (T achu us achu us). These aps we e s a egically placed in a eas such as
pools, i les, a eas nea he banks, and he cen al pa o he i e channel. The
ela i e abundance o signal c ay ish pe si e was exp essed as he o al numbe o
indi iduals pe ca ch pe uni o e o (ind. CPUE).
This bio ic cha ac e iza ion was consis en ly pe o med a all i e sampling si es
ac oss nine di e en ime pe iods ( om Ap il o Decembe ) h oughou he en i e
yea o 2023. I should be no ed ha om Janua y o Ma ch, and due o he high p e-
cipi a ion and high i e low, i was impossible o pe o m elec o ishing and o place
he aps o cha ac e ize he b own ou and signal c ay ish popula ions, espec i ely.
B own ou die
A o al o 825 s omach con en s o b own ou we e analyzed o de e mine whe h-
e he signal c ay ish was p esen in hei die . This was done h ough a simple,
non-le hal echnique, which in ol ed squi ing wa e in o he s omach o induce
egu gi a ion ( ollowing Sánchez-He nández e al. 2010). Once he samples we e
collec ed, ou we e ca e ully e u ned o he i e o minimize s ess and mo al-
i y. Indi iduals p esen in he s omach con en we e iden i ied using Tache e al.
(2010) and we e coun ed. Then, indi iduals iden i ied o each axon we e d ied a
60 °C o 48 h o eco d hei biomass (d y weigh ).
Figu e 1. Map o he su eyed a ea showing he loca ion o he 5 sampling si es in Bacei o Ri e .
Map was p oduced using QGIS so wa e (QGIS De elopmen Team, 2022).
113
NeoBio a 102: 109–123 (2025), DOI: 10.3897/neobio a.102.152018
Ma ilde Ma os e al.: B own ou as a biological con ol
Da a analysis
A P incipal Componen Analysis (PCA) was conduc ed o assess how abio ic ac o s,
including empe a u e, oxygen, conduc i i y, o al dissol ed solids (TDS), and pH,
a y ac oss he i e sampling si es o e ime. The a iables we e no malized and ana-
lyzed enabling he ca ego iza ion o si es based on hei en i onmen al cha ac e is ics.
To assess he spa ial and empo al changes in b own ou and signal c ay ish
ela i e abundances, a wo-way ANOVA was pe o med o assess he e ec s o
Si e and Julian Day (JD) and hei in e ac ion. Pos hoc pai wise compa isons
we e conduc ed using Tukey’s Hones Signi ican Di e ence (HSD) es based on
es ima ed ma ginal means (EMMs) o Si e, JD, and hei in e ac ion o explo e
di e ences be ween g oups.
To assess whe he abio ic and bio ic ac o s in luenced c ay ish p eda ion by
b own ou , we used a Gene alized Linea Model (GLM) wi h a binomial dis-
ibu ion and logi link unc ion. The esponse a iable was c ay ish p esence in
b own ou s omach con en s (0 = absen , 1 = p esen ), and p edic o a iables
included b own ou leng h (cm), wa e empe a u e (°C), o al dissol ed solids
(TDS, mg/L), pH, c ay ish ela i e abundance (ind. CPUE/24h), sampling si e
(ca ego ical), and JD (ca ego ical). We es ed o mul icollinea i y among p e-
dic o a iables using he Va iance In la ion Fac o (VIF), conside ing alues > 5
as indica i e o mul icollinea i y issues. Due o a s ong nega i e co ela ion be-
ween empe a u e and dissol ed oxygen ( = -0.89), we excluded dissol ed oxygen
om he inal model o a oid edundancy. Simila ly, o al dissol ed solids (TDS)
showed a e y high co ela ion wi h conduc i i y ( = 0.99) and so conduc i i y
was also emo ed o p e en collinea i y issues. Model diagnos ics we e pe o med
by examining esidual plo s o ensu e he app op ia eness o he GLM. All analyses
we e conduc ed in R using he glm() unc ion om he base s a s package.
All s a is ical analyses we e ca ied ou using he R S udio so wa e (R Co e
Team, 2021).
Resul s
En i onmen al cha ac e iza ion
The esul s o he abio ic cha ac e iza ion pe si e and o e ime can be seen in
Suppl. ma e ial 1: able S1. The PCA ca ego ized he sampling si es in o wo majo
g oups (Fig. 2). The i s g oup, which includes si es B1, B2, and B3, was cha ac-
e ized by highe conduc i i y and TDS alues, while he second g oup, wi h si es
B4 and B5, showed lowe alues o hese a iables. PC1 explains 68.59% o he
o al a iance, wi h he p ima y con ibu ions coming om conduc i i y and TDS
(nega i e side). PC2 accoun ed o 18.68% o he o al a iance and was s ongly
in luenced by empe a u e. Dissol ed oxygen and pH had minimal in luence.
B own ou and signal c ay ish popula ion dynamics
B own ou abundance a ied ac oss space and ime. Si es B4 and B5 showed a
sha p inc ease, peaking a ound JD 226, ollowed by a decline (Fig. 3). In con-
as , si es B1, B2 and B3 exhibi ed mo e g adual inc eases wi h gene ally lowe
abundance (Fig. 3). ANOVA indica ed a ma ginal e ec o sampling si es on ou
114
NeoBio a 102: 109–123 (2025), DOI: 10.3897/neobio a.102.152018
Ma ilde Ma os e al.: B own ou as a biological con ol
abundance (p = 0.0566), wi h pos -hoc compa isons e ealing ha si e B5 had sig-
ni ican ly highe abundance han si e B1 (p = 0.046), while di e ences among o h-
e si es we e no s a is ically signi ican . ANOVA con i med a signi ican e ec o
JD on b own ou abundance (F(8, 45)=4.86, p < 0.001), wi h Tukey’s es e ealing
highe abundance on JD 208 (p = 0.037), 226 (p = 0.002), and 243 (p = 0.014)
compa ed o JD 107. Addi ionally, JD 226 had a signi ican ly highe abundance
han JD 166 (p = 0.018) bu a lowe abundance han JD 347 (p = 0.012), ein-
o cing he end o declining b own ou abundance owa d he la e s ages o he
s udy pe iod. B own ou leng h a ied be ween 4.0 and 28.7 cm.
Figu e 2. P incipal Componen s Analysis (PCA) showing he a angemen o he i e sampling si es based on he abio ic ac o s measu ed
h oughou he yea . PC1 explains 68.59% o all a iance and PC2 18.68%.
Figu e 3. Abundance o b own ou o e ime and sampling si es.
115
NeoBio a 102: 109–123 (2025), DOI: 10.3897/neobio a.102.152018
Ma ilde Ma os e al.: B own ou as a biological con ol
Signal c ay ish abundance also showed signi ican spa ial (F(5, 362)=22.99,
p < 0.001) and empo al (F(1, 366)=79.3, p < 0.001) di e ences (Fig. 4). Tukey’s es
indica ed signi ican ly highe c ay ish abundances in si es B2 and B3 compa ed o
si e B5 (p < 0.001), si e B1 compa ed o si e B2 (p = 0.0038), B3 (p = 0.0014) and
B5 (p = 0.0052). Addi ionally, B2 (p = 0.0032) and B3 (p < 0.001) had signi i-
can ly highe c ay ish abundance han B1. Tempo al analysis e ealed an inc ease
in abundance a e JD 140, peaking a ound JD 226, ollowed by a sha p decline
(Fig. 4), wi h a signi ican ly lowe c ay ish abundance on JD 107 compa ed o JD
188, 208, 226, 243, and 279 (all p < 0.05).
B own ou as a biological con ol o signal c ay ish
The p esence o signal c ay ish in he s omach con en s o b own ou was low.
Among he sampled b own ou , 12.24% showed e idence o signal c ay ish
in hei s omachs, wi h c ay ish ep esen ing only 2.13% o he o al p ey
i ems sampled (Suppl. ma e ial 1: able S2 o de ailed in o ma ion). Howe e ,
signal c ay ish accoun ed o 17.70% o he o al biomass o he p ey ound in
he s omach con en .
The GLM esul s (Table 1) indica ed ha b own ou leng h (p < 0.001), wa e
empe a u e (p < 0.001), and JD day (p = 0.002) we e signi ican p edic o s o
signal c ay ish p eda ion. La ge ou had a highe p obabili y o consuming c ay-
ish. Wa e empe a u e was posi i ely associa ed wi h signal c ay ish consump-
ion. Con e sely, JD showed a nega i e ela ionship wi h signal c ay ish p eda ion,
indica ing a decline in consump ion o e ime. Spa ial a ia ion also played a ole,
as si es B4 (p = 0.007) and B5 (p = 0.001) showed signi ican ly lowe p eda ion
a es (Table 1). Signal c ay ish abundance, howe e , was no a signi ican p edic o
o p eda ion (p = 0.394) (Table 1). Model diagnos ics indica ed no se e e mul i-
collinea i y (all VIF < 5), and esidual analysis sugges ed an adequa e model i
(Suppl. ma e ial 1: ig. S1).
Table 1. Summa y o he gene alized linea model (GLM) wi h a binomial dis ibu ion and logi link
unc ion. The esponse a iable is he p esence/absence o signal c ay ish in b own ou s omach con en s
(0 = absen , 1 = p esen ). The p edic o a iables include ou leng h (cm), wa e empe a u e (°C), o al
dissol ed solids (TDS, mg/L), pH, c ay ish abundance (ind. CPUE), sampling si e ( ac o wi h B1 as
e e ence le el), and Julian day. The able p esen s model es ima es, s anda d e o s, z- alues, and p- alues.
Signi ican e ec s (p < 0.05) a e ma ked wi h * and highly signi ican e ec s (p < 0.001) wi h ***.
Va iable Es ima e S d. E o z alue p- alue
(In e cep ) 27.547567 24.844294 1.109 0.26751
T ou Leng h (cm) 0.308655 0.032958 9.365 < 2e-16 ***
Wa e Tempe a u e (°C) 0.318505 0.075412 4.224 2.4e-05 ***
TDS -0.024971 0.018133 -1.377 0.16848
pH -4.78158 3.434486 -1.392 0.16385
C ay ish Abundance 0.01555 0.018281 0.851 0.39499
Si e B2 -0.777836 0.461591 -1.685 0.09197
Si e B3 0.075695 0.472005 0.16 0.87259
Si e B4 -1.935907 0.724084 -2.674 0.0075 *
Si e B5 -4.323871 1.351895 -3.198 0.00138 *
Julian Day -0.013684 0.004513 -3.032 0.00243 *
116
NeoBio a 102: 109–123 (2025), DOI: 10.3897/neobio a.102.152018
Ma ilde Ma os e al.: B own ou as a biological con ol
Discussion
The success o non-na i e species is equen ly associa ed wi h he lowe bio ic esis-
ance (lack o pa asi es, diseases, compe i o s and p eda o s) o he na i e commu-
ni ies (Colau i e al. 2004). Despi e he p esence o signal c ay ish in b own ou
s omach con en s, hei occu ence was in equen , hough hey con ibu ed signi i-
can ly o o al p ey biomass consumed by b own ou . This si ua ion, in addi ion
o he low densi y o b own ou (i we con e ou ela i e abundance da a in o
densi y alues based on he 1,000 m2 sampled a ea a each si e and on each occasion)
in compa ison wi h o he Eu opean i e s (Eklö e al. 1999), e en in he Ibe ian
Peninsula (Nicola e al. 2008), sugges s ha his p eda o is unlikely o se e as an
e ec i e and sole biological con ol o signal c ay ish in he Bacei o Ri e .
En i onmen al cha ac e iza ion
The esul s o he en i onmen al cha ac e iza ion showed clea spa ial and em-
po al di e ences. Downs eam si es (B1, B2, and B3) loca ed a a lowe al i ude
had highe conduc i i y and TDS compa ed o ups eam si es (B4 and B5), wi h
empe a u e playing a signi ican ole in seasonal a ia ion. Dissol ed oxygen and
pH emained ela i ely s able ac oss si es and ime o he yea . These en i onmen-
al pa e ns a e pa icula ly ele an because conduc i i y and TDS, which we e
highe in downs eam si es, may in luence habi a p e e ences o b own ou and
signal c ay ish. Likewise, empe a u e a ia ion ac oss si es aligns wi h obse ed
seasonal ends in bo h b own ou and c ay ish abundance. Anyway, i should be
no ed ha all si es p esen e y low human dis u bance, being he Bacei o Ri e an
oligo ophic sys em cha ac e ized by clea wa e s and low nu ien concen a ions.
The e o e, he spa ial changes de ec ed o he analyzed en i onmen al pa ame e s
e lec he g adien ha ypically occu s along he longi udinal p o ile o wa e -
cou ses, in he absence o human dis u bances, acco ding o he i e con inuum
concep (Vanno e e al. 1980).
Figu e 4. Abundance o signal c ay ish o e ime and sampling si es.
117
NeoBio a 102: 109–123 (2025), DOI: 10.3897/neobio a.102.152018
Ma ilde Ma os e al.: B own ou as a biological con ol
B own ou and signal c ay ish popula ion dynamics
The peak in b own ou abundance was in he mid- o-la e s ages o he s udy pe-
iod and aligns wi h seasonal pa e ns o salmonid popula ions ha a e in luenced
by en i onmen al ac o s such as empe a u e and ood a ailabili y (Lobón-Ce -
iá 2009; Blanch ield e al. 2023). This obse a ion is in line wi h he wo k o
Ellio and Ellio (2010), who emphasized he impo ance o seasonal empe a-
u e changes in egula ing b own ou popula ion dynamics. Howe e , i should
be no ed ha in hese mid- o-la e s ages o he s udy pe iod, he i e low is also
lowe , which may inc ease he e iciency o elec o ishing, con ibu ing o highe
abundance alues. The spa ial a ia ion in b own ou abundance, pa icula ly he
lowe o e all abundance in si es B1, B2, and B3, could be a ibu ed o di e enc-
es in habi a quali y, ood a ailabili y, he p esence o a highe abundance o he
non-na i e signal c ay ish, and he highe ishing p essu e (Oli ei a e al. 2025).
The empo al a ia ions in signal c ay ish abundance sugges a s ong empe a-
u e dependence, wi h c ay ish cap u es inc easing du ing wa me pe iods and de-
clining du ing colde mon hs due o educed ac i i y. As ec o he mic o ganisms,
hei me abolism and beha io a e empe a u e-dependen (Rod íguez Valido e
al. 2021). C ay ish abundance inc eased a e JD 140, peaking a ound day 226,
be o e declining, e lec ing ypical seasonal ac i i y pa e ns (Hudina e al. 2017).
The obse ed inc ease in abundance du ing wa me pe iods (JD 166–226) likely
co esponds o ele a ed me abolic a es and heigh ened ac i i y o signal c ay ish
in esponse o highe empe a u es and so inc easing he chances o being cap-
u ed (Rika dsen e al. 2006; Sousa e al. 2013). The subsequen decline in he
au umn and ea ly win e is likely due o dec eased ac i i y and bu owing beha io
in colde mon hs, p o iding he mal s abili y (Paye e and McGaw 2003; Hudina
e al. 2017). Signal c ay ish abundance also a ied ac oss si es, sugges ing ha local
ac o s such as ood a ailabili y, p eda ion p essu e, habi a condi ions o in e spe-
ci ic in e ac ions in luence dis ibu ion (Ya a and Magoulick 2018). Di e ences
may also ela e o in asion g adien s, whe e in asion on s exhibi lowe abun-
dances compa ed o co e a eas (Hudina e al. 2012; Al es e al. 2025; Ca alho
e al. 2025). I should be no ed ha he Bacei o Ri e si e B1 was he i s o be
colonized by his non-na i e species and he sp ead was in he ups eam di ec ion,
being he cu en in asion on loca ed a si e B5 ( i s eco ds in his si e a he
end o he summe 2022; Ronaldo Sousa pe sonal obse a ion).
B own ou as a biological con ol o signal c ay ish
Only 12% o he b own ou analysed p esen signal c ay ish in hei s omach
con en s. The numbe o consumed c ay ish in compa ison o all i ems iden i-
ied in he s omach con en s was e y low (2.13%), bu hei con ibu ion o he
biomass o consumed p ey was mo e conside able (17.70%). The low p e alence
and consump ion a es o signal c ay ish may be a ibu ed o se e al ecological
and beha io al ac o s. C ay ish possess ha d exoskele ons and de ensi e beha io s,
making hem mo e di icul o cap u e and consume han so e -bodied p ey (e.g.,
aqua ic insec s and small ish). Fu he mo e, p e ious s udies ha e shown ha ish
p eda o s o en exhibi an ini ial lag in ecognizing no el p ey (Ca oll 2007; Cox
2013), which may explain why b own ou do no ye e ec i ely inco po a e c ay-
ish in o hei die . Despi e being an oppo unis ic eede , whose hun ing ac i i y
