P e en i e Ve e ina y Medicine 221 (2023) 106062
A ailable online 2 No embe 2023
0167-5877/© 2023 The Au ho (s). Published by Else ie B.V. This is an open access a icle unde he CC BY license (h p://c ea i ecommons.o g/licenses/by/4.0/).
The economic impac o dec eased mo ali y and inc eased g ow h
associa ed wi h p e en ing, eplacing o imp o ing cu en me hods o
delousing a med A lan ic salmon in No way
Cecilie S iland Walde
a
,
*
, B i Bang Jensen
a
,
b
, Ma i S o moen
c
, F ank Asche
d
,
e
, Bå d Misund
,
Jos ein Mulde Pe e sen
g
a
No wegian Ve e ina y Ins i u e, Elizabe h S ephansens ei 1, Ås 1433, No way
b
Na ional Ins i u e o Aqua ic Resou ces, Technical Uni e si y o Denma k, Denma k
c
Facul y o Ve e ina y Medicine, No wegian Uni e si y o Li e Sciences, Elizabe h S ephansens ei 1, Ås 1433, No way
d
School o Fo es , Fishe ies and Geoma ics Sciences and Global Food Sys ems Ins i u e, Uni e si y o Flo ida, Gaines ille, FL, USA
e
Depa men o Sa e y, Economics and Planning, Uni e si y o S a ange , S a ange , No way
Depa men o Economics and Finance, Uni e si y o S a ange Business School, S a ange , No way
g
PHARMAQ AS, Indus i egen 50, O e halla N-7863, No way
ARTICLE INFO
Keywo ds:
Salmon lice ea men s
Biological cos s
Mo ali y
G ow h
P o i abili y
S ochas ic pa ial budge analysis
A lan ic salmon a ming
No way
ABSTRACT
Impac s o salmon lice is a majo conce n o a sus ainable p oduc ion o a med A lan ic salmon in No way. Mos
ea men me hods o emo al o salmon lice ha e associa ed inc eased mo ali y and dec eased g ow h in a
pe iod a e delousing, which a ec s he p o i abili y o he a me , and causes poo wel a e and sus ainabili y. In
addi ion, he a iance in mo ali y and g ow h, especially a e non-medicinal ea men me hods, is high, which
makes i ha d o a a me o decide which con ol measu e o apply o keep lice le els below he legal limi . In
his s udy, we ha e applied a s ochas ic pa ial budge app oach o assess he economic impac o educing
mo ali y and inc easing g ow h o a med A lan ic salmon by p e en ing, eplacing and imp o ing cu en
delousing me hods in No way. We ha e simula ed a p oduc ion cycle o wo di e en smol -g oups o ind he
ou come (ha es ed biomass, a e age end weigh o he salmon, numbe o dead ish and eed consump ion) o
p oduc ion cycles wi hou o wi h wo, h ee o ou delousing ea men s in he on-g owing phase a sea. The
esul s sugges ha accoun ing o he biological losses associa ed wi h lice ea men s is impo an when making
choices o delousing s a egies. The biological cos s o inc eased mo ali y and dec eased g ow h associa ed wi h
especially non-medicinal ea men s a e expec ed o be high, bu a ies subs an ially. The e o e, he economic
bene i o p e en ing o imp o ing can also be high. The calcula ions imply ha salmon p oduce s could in es a
conside able amoun in measu es o p e en ion o imp o emen o he mal ea men s be o e b eak-e en. Fo
example could a a me use on a e age 535,313
€
/cage/ 1-yea ling p oduc ion in measu e o p e en ou
he mal ea men s be o e i is no longe economical bene icial. Depending on he pe o mance o he ou
he mal ea men s a a me could use om 319,196–737,934
€
/cage/ 1-yea ling p oduc ion on measu es o
imp o emen . Replacing one he mal ea men wi h ano he immedia e ea men me hod has a mino eco-
nomic bene i . The esul s u he shows ha sales alue and eed consump ion cons i u es he la ges sha e o
he change in p o i be ween di e en ea men egimes. The esul s om his s udy also show ha no aking
in o accoun he isk o mo ali y and educed g ow h associa ed wi h he di e en ea men me hods o
delousing, could lead o unde es ima ing he bene i o imp o ing, p e en ing and eplacing ea men s.
1. In oduc ion
The e a e conside able conce ns ega ding he sus ainabili y o
No wegian p oduc ion o A lan ic salmon, limi ing g ow h in a highly
p o i able indus y (Osmundsen e al., 2022; Sik eland e al., 2022).
Impac s o salmon lice is one o he main conce ns and is now he
* Co espondence o: No wegian Ve e ina y Ins i u e, Tho møhlensga e 53C, 3. Floo , Be gen 5006, No way.
E-mail add ess: [email p o ec ed] (C.S. Walde).
Con en s lis s a ailable a ScienceDi ec
P e en i e Ve e ina y Medicine
jou nal homepage: www.else ie .com/loca e/p e e med
h ps://doi.o g/10.1016/j.p e e med.2023.106062
Recei ed 24 Feb ua y 2023; Recei ed in e ised o m 27 Oc obe 2023; Accep ed 27 Oc obe 2023
P e en i e Ve e ina y Medicine 221 (2023) 106062
2
de ining elemen o he egula o y sys em in e ms o whe he p oduc-
ion will inc ease, emain cons an o decline in each o 13 p oduc ion
a eas (Osmundsen e al., 2020). Managing he salmon lice challenge lead
o signi ican cos s o he a me due o los biomass as a consequence o
educed g ow h and inc eased mo ali y associa ed wi h ea men
(Walde e al., 2021; Walde e al., 2022), possible nega i e e ec o
ea men on he ile quali y causing down-g ading o eclaims, di ec
ea men cos s such as equipmen , labou and ene gy, as well as cos s o
con ol and p e en ion (Abolo ia e al., 2017; Cos ello, 2009; I e sen
e al., 2020; I e sen e al., 2017).
P oduc ion o salmon akes 2 ½−3 yea s om ha ching o eggs un il
ha es . The ish spend he i s 10–16 mon hs o hei li e in a land-
based eshwa e acili y. Subsequen ly, g oups o 150,000–200,000
salmon a e ans e ed and s ocked in open ne -cages a seawa e si es
o an on-g owing pe iod o he emaining 14–22 mon hs. Fish ans-
e ed o sea in he all he same yea hey ha ch a e commonly e e ed
o as 0-yea ling, while ish ans e ed he sp ing he yea a e hey
ha ched a e e e ed o as 1-yea ling. Since p oduc ion a sea occu in
open ne cages, he salmon a e exposed o seasonal en i onmen al
changes such as luc ua ing wa e empe a u es, ligh and salini y, in
addi ion o di e en pa hogens in he wa e column, including salmon
lice.
Following i s comme cial b eak h ough in he ea ly 1970 s, salmon
a ming has de eloped apidly om an owne -ope a ed small-scale
sec o o a la ge-scale in ensi e p oduc ion p ocess ope a ed by mul i-
na ional companies (Asche e al., 2022). This is pa ly due o a num-
be o inno a ions ha has educed p oduc ion cos and imp o ed
compe i i eness (A ewe ki e al., 2023; Asche, 2008) and inc eased he
scale o each si e (Asche e al., 2013), as well as dynamic egula o y
sys em ha has acili a ed his g ow h (He soug, 2021).
As in o he biological p oduc ion p ocesses, salmon aquacul u e is
a ec ed by di e en diseases. Many o he diseases a e de imen al o
he p oduc ion as hey educe g ow h a es and in wo s case induce
mo ali y, he eby educing heal h and wel a e, inc easing p oduc ion
cos and educing p o i abili y (I e sen e al., 2020). Howe e , p o-
duc ion losses also c ea e economic incen i es o p e en o ea he
diseases, and a apidly inc easing knowledge base has imp o ed he
indus y’s abili y o do so (A ewe ki e al., 2023).
As p oduc ion has inc eased, so has he salmon biomass along he
No wegian coas . In u n, his has inc eased he numbe o hos s o
di e en pa hogens, causing a challenge o he salmon indus y i sel as
well as an ex e nali y o wild salmon occupying he same wa e s (Dean
e al., 2021). The mos impo an challenge is he ec opa asi e salmon
lice, Lepeop hei us salmonis, which due o i s po en ial impac s on wild
salmon is he main ac o in he egula o y sys em in e ms o de e -
mining he indus y’s p oduc ion g ow h (Osmundsen e al., 2020). The
egula ions o lice in a med salmonids aim o p o ec bo h he a med
salmon and educe he spill-o e e ec o in es a ion o wild s ocks
(Fo sk i om lakselusbekjempelse, 2016; Jeong e al., 2023). In No way,
salmon p oduce s a e obliged by law o main ain lice le els below a legal
maximum limi (Fo sk i om lakselusbekjempelse, 2016). I he numbe
o lice inc eases beyond he salmon’s abili y o compensa e, i can cause
pa hology and e en ually dea h (Pike and Wadswo h, 1999).
Fo many yea s, he in es a ion p essu e o lice was kep unde
con ol by di e en medicinal eed o ba h ea men s. Howe e , du ing
2000–2012 he lice de eloped esis ance agains mos o hese ac i e
subs ances (Myh e Jensen e al., 2020). F om 2015, he domina ing
ea men p ac ice hus shi ed om medicinal ea men s o
non-medicinal ea men s in ol ing hea ed ba hs ( he mal ea men )
and lushing o b ushing he lice o he ish (mechanical ea men )
(O e on e al., 2018). While a a ie y o me hods exis o he man-
agemen o salmon lice, including he u iliza ion o cleane ish, he
implemen a ion o p e en a i e measu es such as semi-closed and sub-
me ged cages, and geog aphical managemen s a egies, he e is o en a
equi emen o one o se e al immedia e ea men s (Ba e e al.,
2020). Cu en ly, he con ol o salmon lice consis s o a mix u e o
se e al di e en p e en i e measu es and immedia e ea men s, mos ly
non-medicinal.
An impo an pa o he salmon lice challenge is high mo ali y due
o handling and subsequen ea men o he salmon, especially in he
las pe iod o he on-g owing phase a sea (Aunsmo e al., 2023; Bang
Jensen e al., 2020; Oli ei a e al., 2021; O e on e al., 2018; Pe sson
e al., 2022; Pincina o e al., 2021; T e e e al., 2023; Walde e al.,
2021). The mo ali y expe ience a e he non-medicinal ea men
me hods is shown o be many imes highe compa ed o he medicinal
ea men me hods (Walde e al., 2021).
Ano he impo an pa o he salmon lice challenge is he los
g ow h po en ial o he a med salmon due o a pe iod o eed wi h-
d awal p io o ea men and appe i e d op a e ea men (Walde
e al., 2022). I has been a gued ha he ocus on lice in he egula o y
sys em and ea men s agains salmon lice a e some o he easons o
he obse ed declining size a ha es in No wegian salmon aquacul u e
(Ba e e al., 2022; Oglend and Soini, 2020). In addi ion, delousing
ea men s may ha e nega i e e ec s on ile quali y due o inju ies and
wounds caused by he delousing ope a ion (Gisme ik e al., 2019;
Thompson e al., 2023).
The inc ease in mo ali y and dec ease in ish g ow h a ec s he
p o i abili y o he a me as i educes p oduc ion (Abolo ia e al.,
2017). Howe e , he e is a la ge a iabili y in he expe ienced mo ali y
a e di e en delousing ea men s, and he e ec on g ow h can a y
subs an ially be ween he di e en delousing ea men s (Walde e al.,
2021; Walde e al., 2022). The unce ain e ec on mo ali y and g ow h
om a delousing ea men can he e o e make i ha d o a a me o
decide which con ol measu e o apply o keep he le els o lice below
he legal maximum limi .
The objec i e o he p esen s udy is o desc ibe he impac on p o i s
o e a single p oduc ion cycle o salmon, om ei he 1) p e en ing
ea men s; 2) eplacing ea men s wi h o he ea men me hods; o 3)
imp o ing ea men s by including he biological losses o inc eased
mo ali y and dec eased g ow h associa ed wi h di e en delousing
ea men s.
2. Ma e ial and me hods
2.1. Bio-economic modelling
In he p esen s udy we apply a s ochas ic pa ial budge ing
app oach. Pa ial budge ing is a well-known economic ool o suppo
decision-making p ocesses in di e en a eas o animal p oduc ion
(Aunsmo e al., 2010; Pe e sen e al., 2016; Pe e sen e al., 2015;
Rush on, 2009). This ool quan i ies he economic consequences o a
speci ic change in he p oduc ion p ocedu e by compa ing he nega i e
and posi i e impac s o ind he economic ne bene i o he change. This
analysis does no desc ibe he p o i abili y o a p oduc ion cycle, bu
a he how p o i s a e a ec ed by choice o ea men me hod agains
salmon lice.
Fo he pa ial budge ing analysis, we applied a bio-economic model
ha consis ed o se e al scena ios, whe e a scena io was de ined as he
compa ison o a single p oduc ion cycle o ei he a 1 o 0-yea ling wi h
di e en delousing ea men egimes. The biological inpu a iables
we e based on dis ibu ions om wo da ase s, one desc ibing mo ali y
and he o he desc ibing p oduc ion and g ow h o salmon om s ocking
un il ha es . The biological ou pu a iables om he simula ed p o-
duc ion cycles and economic inpu a iables we e used o calcula e he
economic posi i e and nega i e impac s o changing a ea men egime.
The main ou pu o in e es was he economic ne bene i o a scena io,
exp essed as he change in p o i .
The bioeconomic model was buil in Excel (Mic oso Co po a ion)
wi h he Mon e Ca lo simula ion add-in ool @Risk (Palisade Co po a-
ion, NY, USA) which enables isk analysis by subs i u ing single poin
es ima es o unce ain inpu s wi h dis ibu ions sampled andomly by
se e al i e a ions pe simula ion.
C.S. Walde e al.
P e en i e Ve e ina y Medicine 221 (2023) 106062
3
2.1.1. Da a/ma e ial
No wegian salmon a ming companies eco d da a o hei p oduc-
ion, such as numbe o s ocked ish, a e age ish weigh , eeding ( ype
and amoun ), mo ali y, ea men s, en i onmen al da a e c. a he cage-
le el (NFD, 2008). The da ase in his s udy is based on daily da a om
h ee la ge No wegian A lan ic salmon a ming companies (companies
ope a ing mo e han 20 si es) and has p e iously been applied in Walde
e al. (2021) (da ase I) and Walde e al. (2022) (da ase II). The da ase
includes cage-le el his o ical p oduc ion da a ela ed o p oduc ion and
salmon lice ea men s.
The da ase applied in Walde e al. (2021) (da ase I) desc ibe es i-
ma ed dis ibu ions o change in mo ali y a e a e 4,644 delousing
ope a ions. This change was calcula ed by sub ac ing he mo ali y a e
se en days a e delousing wi h he mo ali y a e se en days be o e
delousing. Walde e al. (2021) desc ibes he equa ion o calcula ing he
change in mo ali y a e, and he backg ound o choosing a se en-day
in e al (b ie ly; choosing a se en day in e al be o e and a e he
ea men minimises he isk o in oducing e ec s o o he diseases o
ea men s, while s ill p o iding a su icien ly long pe iod o single-day
a ia ions o no in luence he esul s oo much). In he cu en s udy an
addi ional 165 ea men s we e excluded om he da ase desc ibed in
Walde e al. (2021) due o missing alues o change in mo ali y a e
se en days a e ea men . The inal da ase I consis ed o 4,479 ea -
men s o 1,756 ish-g oups om ou yea -classes, 2014–2017, and 158
si es. The es ima ed dis ibu ions o change in mo ali y a e
(Δmo a e) s a i ied on ea men me hod was used as s ochas ic inpu
a iable in he bio-economic model.
Walde e al. (2022) (da ase II) es ima ed he sho - e m e ec o
di e en delousing me hods on g ow h. The da ase applied in his s udy
con ained he same sou ce o da a as he one applied in Walde e al.
(2021) (da ase I), howe e only hose g oups o ish ha could be aced
om he ime o s ocking un il ha es we e included in he s udy. The
g ow h a e was exp essed as he he mal g ow h coe icien (TGC) (Cho,
1992). This was calcula ed by sub ac ing he se en-day mean o daily
TGC a e delousing by he i e-day mean o daily TGC be o e delousing.
Walde e al. (2022), desc ibes he calcula ion o change in g ow h a e
and he backg ound o he choice o ime in e al. In he cu en s udy,
an addi ional i e ish-g oups we e excluded due o p oduc ion leng h
sho e han 200 days. In addi ion, 21 ish-g oups only had ea men s
agains amoebic gill disease (AGD) du ing he p oduc ion cycle, and
hese we e also excluded. The inal da ase II consis ed o 609
ish-g oups, 302 1-yea lings and 307 0-yea lings. These came om ou
yea -classes, 2014–2017, and 94 di e en si es. They we e ea ed a
o al o 2,281 imes. The es ima ed dis ibu ions o change in a e age
daily g ow h a e (ΔTGC) s a i ied on ea men me hod was used as
s ochas ic inpu a iable in he bio-economic model in addi ion o o he
biological a iables desc ibing p oduc ion.
2.1.2. Delousing ea men s
The ea men me hods used we e he mal, mechanical, hyd ogen
pe oxide ba h, eshwa e ba h and medicinal ba h. Table 1 shows he
di e en ca ego ies applied and numbe o delousing ope a ions wi hin
each ca ego y. The medicinal ea men s we e combined in one ea -
men ca ego y, as he e we e oo ew ea men s o each compound o
include hem sepa a ely. Handling is assumed o be he main d i e o
mo ali y du ing medicinal delousing, ega dless o ac i e subs ance
used. Howe e , i was no possible o ca ego ise delousing as ea men
in cage o well-boa , as his in o ma ion was no consis en ly eco ded
(Walde e al., 2021; Walde e al., 2022).
2.1.3. Biological inpu pa ame e s
The biological inpu a iables in he model we e: numbe o s ocked
salmon, numbe o p oduc ion days, es ima ed a e age weigh a
s ocking, mon h o s ocking, a e age mon hly seawa e empe a u e a
3 m dep h, baseline mon hly mo ali y (%), baseline mon hly g ow h
a e (exp essed as TGC), days o eed wi hd awal p io o ea men ,
biological eed con e sion a io (bFCR), change in mo ali y a e
(Δmo a e) and change in g ow h a e (ΔTGC) (Table 2).
The inpu alues o change in mo ali y (Δmo a e) and g ow h
(ΔTGC) was made s ochas ic by ep esen ing hese inpu s as dis ibu-
ions (Table 2). This was done by applying he obse a ions o Δmo a e
(da ase I) and ΔTGC (da ase II) o each ea men me hod. Dis ibu-
ions o Δmo a e and ΔTGC we e i ed based on he obse a ions
using he unc ion “Fi ” in @Risk. The dis ibu ion i wi h he lowes
Akaike in o ma ion c i e ion alue was p e e ed. To a oid un eason-
able alues and hea y ails, each dis ibu ion was unca ed a he
minimum and maximum alues in he espec i e da ase s a e he dis-
ibu ions we e i ed. Co ela ion be ween he inpu a iables
Δmo a e and ΔTGC (n =2 281) was calcula ed using he “co ela e”
command in S a a (S a aCo p, 2017). The dis ibu ions we e sampled
andomly by 10,000 i e a ions pe simula ion.
The alue o he de e minis ic biological inpu a iables; numbe
s ocked, weigh a s ocking, mon h o s ocking, leng h o p oduc ion and
days o s a a ion we e chosen based on desc ip i e s a is ics o da ase
II. The de e minis ic inpu alue o each o hese a iables was chosen
based on he mos ep esen a i e measu e o he unde lying dis ibu ion,
which was ei he he app oxima e mean o he app oxima e median
alue. The a e age mon hly empe a u e was es ima ed by a e aging
he mon hly empe a u es o obse a ions in yea 2017 and 2018. The
1-yea lings had a mean s ocking weigh o 133 g and he 0-yea lings o
109 g. In he model, he s ocking weigh o bo h smol ypes was se o
100 g o ease o compa ison o he wo smol ypes. The baseline
mo ali y pe cen (mo
baseline
) and g ow h a e (TGC
baseline
) de ined
mo ali y and g ow h in mon hs wi hou ea men s. The mean g ow h
a e om s ocking un il i s ea men was close o no mally dis ib-
u ed, wi h a mean o 2.9. Howe e , he alue o 2.8 was chosen as he
baseline in he model o ensu e ha ha es ed weigh be ween bo h
smol ypes was wi hin he same weigh ca ego y. The biological eed
con e sion (bFCR) a io a ied acco ding o he inbound weigh o he
ish each mon h. All he mone a y alues we e eco ded as No wegian
k one (NOK), bu con e ed o Eu o (
€
), whe e 1
€
=10.104 NOK
(yea ly a e age 2022).
2.1.4. Biological ou pu pa ame e s
The biological ou pu pa ame e s om he di e en p oduc ion
Table 1
Ca ego iza ion (n =5) o he immedia e ea men ope a ions o a med A lan ic
salmon in h ee No wegian companies om 2014 o 2019.
Ca ego ies o ea men
ope a ions
Desc ip ion o ca ego y o delousing ope a ion
The mal Non-medicinal ea men using hea ed seawa e .
Includes all ea men s using:
a. Op ilice ®
b. The molice
c. Hea ed seawa e
Mechanical Non-medicinal ea men using b ushing o lushing.
Includes all ea men s using:
a. FLS A luse sys em
b. Hyd olice
c. SkaMik
d. Flushing o mechanical ea men
Hyd ogen pe oxide Hyd ogen pe oxide (H2O2) ba h in cage o well boa
agains salmon lice
F eshwa e ba h F eshwa e ba h in cage o well boa agains salmon lice
Medicinal ba h Medicinal ba h in cage o well boa using one o he
ollowing ac i e subs ances:
a. Azame iphos
b. Cype me h in
c. Del ame h in
d. Imidaclo id
e. O he
. Coho s ea ed wi h wo di e en combina ions o
ac i e subs ances a-e o hyd ogen pe oxide and one o
he ac i e subs ances a-e
C.S. Walde e al.
P e en i e Ve e ina y Medicine 221 (2023) 106062
4
cycles we e he ha es ed weigh (Eq. 1.1) and biomass (1.5), and
accumula ed amoun o eed used (Eq. 1.6) a he end o each p oduc ion
cycle.
The weigh gain in a non- ea men mon h was calcula ed by se ing
he TGC in Eq. 1.3 equal o TGC
baseline
he en i e pe iod o he mon h.
The weigh gain (wg
) du ing a ea men mon h was calcula ed in h ee
s eps:
1) Du ing eed wi hd awal (wg
s a
): A ea men is ini ia ed by a
pe iod o eed wi hd awal. Du ing his pe iod, he TGC in Eq. 1.3
would be equal o ze o, hus he weigh gain (wg
s a
) would also be
ze o.
2) Pos ea men (wg
pos ea
): The weigh gain se en days ( =7) a e
a ea men (wg
pos ea
) was calcula ed by subs i u ing TGC in Eq.
1.3 wi h TGC
pos ea
(Eq. 1.4).
3) Fo he emaining pa o he mon h (wg
em
): A e he pe iod o
eed wi hd awal and pos ea men pe iod ( =14), he weigh gain
Table 2
O e iew o he a iables used in pa ial budge model. The a iables a e equal
o bo h 0 and 1-yea ling unless o he wise is speci ied. D=de e minis ic,
S=s ochas ic. All p ices a e exp essed as 2022-NOK.
Va iable Value used in
model
Sou ce Type Dis ibu ion
( alue om
da a)
Numbe o
s ocked
150, 000 Da ase II D Mean
(158,909)
P oduc ion
days
488 Da ase II D Mean (483
days)
A e age
weigh a
s ocking
100 g Da ase II D Median (104
g)
Mon h o
s ocking 1-
yea ling/0-
yea ling
Ap il/Augus Da ase II D Mean (Ap il/
Augus )
Temp pe
mon h (⁰C)
Jan.=6.2,
Feb.=5.1,
Ma ch=4.2,
Ap il=5.2,
May=8.2,
June=9.9,
July=12.0,
Aug=14.7,
Sep =13,9,
Oc =12.3,
No =9.8,
Dec=7.5
Da ase II D Mean mon h
emp 2017
and 2018
Days o
inc eased
mo ali y/
dec eased
g ow h a e
ea men
7 Walde e al. (2021)
Walde e al. (2022)
D
Baseline
mon hly
mo ali y
0.2% Oli ei a e al. (2021) D
Baseline
mon hly
g ow h
2.8 Da ase II D Mean (2.9)
Days o eed
wi hd awal
p io
ea men
5 Da ase II D Mean (5.6)
Biological eed
con e sion
a io (bFCR)
Sk e ing’s
Rela i e
G ow h Index
Table
( eadimage.aspx
(sk e ingguidelines.
com)
D
Change in
mo ali y
a e
(Δmo a e)
se en days
a e
ea men
The mal 0.000766 Da ase I S Laplace
T unca ed
(−0.0169,
0.0347)
Mechanical 0.000646 Da ase I S Loglogis ic
T unca ed
(0.0075,
0.0706)
Hyd ogen
pe oxide
0.000727 Da ase I S Loglogis ic
T unca ed
(−0.0054,
0.0637)
F eshwa e 0.0000905 Da ase I S Loglogis ic
T unca ed
(−0.0032,
0.0265)
Medicinal 0.0000105 Da ase I S Laplace
T unca ed
(−0.0054,
0.0266)
Table 2 (con inued)
Va iable Value used in
model
Sou ce Type Dis ibu ion
( alue om
da a)
Change in
g ow h a e
(ΔTGC)
se en days
a e
ea men
The mal -0.95581 Da ase II S No mal
T unca ed
(−3.9846,
2.9028)
Mechanical -0.92332 Da ase II S Logis ic
T unca ed
(−3.5377,
1.8403)
Hyd ogen
pe oxide
-0.56656 Da ase II S Logis ic
T unca ed
(−3.4213,
1.8779)
F eshwa e -0.88385 Da ase II S Pe
T unca ed
(−2.6792,
1.1462)
Medicinal -0.36908 Da ase II S Logis ic
T unca ed
(−3.7493,
4.3831)
Feed p ices
(p
eed
)
14.60 NOK/
kg d y eed
In a ish.no
Ilaks.no
D
Handling dead
cos (p
mo
)
2.12 NOK/kg
ound weigh
Pe e sen e al.
(2015)
D
Ha es ing cos
(p
ha
)
4.05 NOK/kg
ound weigh
(Pe e sen e al.,
2015)
D
T ea men cos
(p
ea
)
NOK/kg li e
weigh
The mal 0.37 I e sen e al. (2017) D
Mechanical 0.26 I e sen e al. (2017) D
Hyd ogen
pe oxide
0.50 I e sen e al. (2017) D
F eshwa e 1.33 I e sen e al. (2017) D
Medicinal 0.37 I e sen e al. (2017) D
Sales p ice (p)
pe weigh
class
NOK/kg
head on
gu ed (HOG)
week 47,
2022
NASDAQ
NASDAQ Salmon
Index
(nasdaqomx ade .
com)
3–4 76.53 D
4–5 79.90 D
5–6 85.72 D
6–7 96.00 D
7–8 99.31 D
8–9 100.86 D
9 +101.99 D
C.S. Walde e al.
P e en i e Ve e ina y Medicine 221 (2023) 106062
5
in he emaining pa o he mon h (wg
em
) was calcula ed by se ing
he TGC in Eq. 1.3 equal o he TGC
baseline
and =numbe o
emaining days wi hin he ea men mon h. This implies an
assump ion o no compensa o y g ow h a e a ea men .
weigh ha es =weigh s ocking +∑
n=17
T=1(wg s a +wg pos ea
+wg em)(1.1)
weigh T=weigh T−1+wg s a +wg pos ea +wg em
(1.2)
wg =weigh −1−{weigh −1(1
3)+(TGC
1000 x empT∗ )3}(1.3)
TGCpos ea =TGCbaseline +ΔTGC (1.4)
T=mon h, =days, n =numbe o mon hs, wg=weigh gain, wi h
wi hd awal pe iod de aul =7 days, pos ea men pe iod de aul =7
days, emp =a e age mon hly empe a u e, pos ea =pos ea men ,
em = emaining (weigh gain emaining pe iod).
The mo ali y a e, Δmo a e, was ans o med o an inciden isk
(To e al., 2004) and he mon hly inbound numbe o ish (N
T
) was
calcula ed by sub ac ing he baseline mo ali y and he numbe o dead
se en days a e a ea men om he inbound numbe o ish he p e-
ious mon h (N
T
-1) (Eq. 1.5). In no- ea men mon h Δmo a e would
be equal o ze o. I was assumed ha he ea men s occu ed in he
beginning o he mon h, ini ia ed by he p e- ea men eed wi hd awal
pe iod o i e days. This gene a ed a small bias in numbe o dead, as
bo h he baseline mo ali y and he ea men mo ali y was calcula ed
om he inbound numbe a he s a o he mon h. Howe e , he
accumula ed e ec om ou ea men s was mino and consis en
ac oss all scena ios, and he bias hus ega ded o unimpo an o he
modelled ou come.
NT=NT−1− (mo baseline ×NT−1) − (NT−1×e−Δmo a e× )(1.5)
aB =∑
n=17
T=1
(NT×weigh T)(1.6)
aB=accumula ed biomass, n =numbe o mon hs
aFeed =∑
n=17
T=1(wgT
1000 ×bFCR)NT(1.7)
aFeed=accumula ed amoun o eed in kg, bFCR=biological eed
con e sion a io
aBDead =∑
n=17
T=1
[weigh T((mo baseline ×NT) + (NT×e−Δmo a e× )) ] (1.8)
aBDead=accumula ed biomass o dead, n =numbe o mon hs
aBT ea =∑
n
T=1
(weigh T×NT)(1.9)
aBT ea =accumula ed ea ed biomass, n =numbe o ea men s.
2.1.5. Economic inpu pa ame e s
The p ice componen s in he model we e: eed (p
eed
) handling o
dead ish, (p
mo
), ha es ing, (p
ha
), and ea men (p
ea
) (Table 2).
The eed p ices we e exp essed as NOK /kg d y eed. The handling o
dead ish and ha es ing p ices we e exp essed as NOK/ kg p oduced
ound weigh (Pe e sen e al., 2015). P ice o di e en ea men
me hods pe kg ea ed ish (p
ea
) we e ob ained om I e sen e al.
(2017). These es ima es included he cos o mo ali y pe kg ea ed
ish, and we e equal o he mal, mechanical and eshwa e ea men s,
0.17 NOK/kg ea ed. The cos o mo ali y was sub ac ed om hese
numbe s since ea men speci ic mo ali y was e ie ed om da ase I
(Table 2). All he p ices we e adjus ed o in la ion o 2022 NOK using
he mon hly consume p ice index epo ed by S a is ics No way. The
p ice inc ease om 2015 and 2017–2022 was 22.8% and 16.4%,
espec i ely.
Since he objec i e was o es ima e he economic bene i (ΔΠ) o a
change, only a iable cos s om delousing ope a ions we e included in
he model and i was assumed ha ixed cos s would no change be ween
di e en ea men me hods o p oduc ion cycles o same leng h.
Salmon p ices o di e en weigh classes we e p o ided om Nas-
daq (NASDAQ Salmon Index (nasdaqomx ade .com)). The Nasdaq
Salmon indices consis s o 11 indi idual p ice indices o weekly e-
po ed sales p ices o No wegian a med salmon o he highes quali y
classi ica ion called supe io quali y (Table 2). Nine o he indices a e
p ices o nine di e en weigh ca ego ies (1–2, 2–3, 3–4, 4–5, 5–6, 6–7,
7–8, 8–9 and 9 +kilog ams), while he o he wo a e weigh ed a e age
p ices ac oss all, o a selec ion o he mos sold (3–6 kilos), weigh
ca ego ies. In he model, he sales p ice (p) was made dependen on
ha es ed weigh , using he Nasdaq p ice ca ego ies o di e en weigh
ca ego ies. The spo sales p ices o week 47, 2022 was used as inpu in
he model (Table 2).
2.1.6. Economic ou pu pa ame e s
The cos o handling dead ish (M
cos
), ha es ing cos (H
cos
), eed
cos (F
cos
) and ea men cos (T
cos
) we e calcula ed by Eqs. 1.10–1.13.
The ha es ed weigh and biomass, and accumula ed biomass o dead
ish was con e ed o head on gu ed (HOG) o ound weigh as
app op ia e, by using a con e sion ac o o 1.067 o 1.2, espec i ely
(Di ec o a o Fishe ies, 2019). The main ou pu a iable o in e es , he
di e ence in p o i (ΔΠ) (Eq. 1.14) was calcula ed o each scena io.
Mcos =aBDead
1.07 ×pmo (1.10)
Hcos =aB
1.07 ×pha (1.11)
Fcos =aFeed ×p eed (1.12)
Tcos =aBT ea ×p ea (1.13)
ΔΠscen n ={(p×aB
1.2)− (Mcos +Hcos +Fcos +Tcos )}p od x
−{(p×aB
1.2)− (Mcos +Hcos +Fcos +Tcos )}p od y
(1.14)
Scen =scena io,n=scena io numbe ,p od =p oduc ion cycle.
We assumed he en i e ha es ed biomass o be classi ied as “supe-
io ” quali y, he ha es ed weigh o each ish wi hin he ish-g oup o
be he same as he a e age weigh o he ish-g oup, and he weigh o
he dead ish o be he same as he a e age weigh o he ish-g oup he
mon h o dea h. We u he assumed ha he (bFCR) was no a ec ed by
he ea men s.
2.1.7. Modelled p oduc ion cycle and baseline ea men egime
The numbe o ea men s du ing a p oduc ion cycle, when hey
occu ed, he ime elapsed be ween ea men s and he ype o ea -
men s was de ined as a ea men egime. The ea men egime was
selec ed by desc ip i e s a is ics o da ase II. In da ase II, he mos
common ype o ea men in he ecen yea classes (2016 and 2017)
was he mal ea men s, and he ish g oups we e on a e age ea ed
ou imes du ing he p oduc ion cycle. Fo bo h smol ypes, hal o all
ea men s we e pe o med 20–40 days apa , and numbe o days om
C.S. Walde e al.
P e en i e Ve e ina y Medicine 221 (2023) 106062
6
las ea men un il ha es was posi i ely skewed wi h a mode o 31
days. The occu ence o he ea men s, howe e , di e ed be ween he
smol ypes (Fig. 1). The empe a u e anged om he lowes empe a-
u e in Ma ch o 5.3⁰C o a peak in Augus o 14.4⁰C. The highes
occu ence o ea men s o he 1-yea ling is in Ap il (n =129) and o
he 0-yea ling in Sep embe (n =202). The selec ed baseline ea men
egime in he model was he e o e ou he mal ea men s ha occu ed
in he second yea in sea o bo h smol ypes (Fig. 1). The 1-yea ling was
ea ed in Ap il, May, June and July and ha es ed one mon h la e
(Fig. 1). The 0-yea ling was ea ed in Augus , Sep embe , Oc obe and
No embe and ha es ed one mon h la e (Fig. 1). The p oduc ion a sea
las ed o 488 days o bo h smol ypes, and bo h we e ea ed o he
i s ime 365 days a e s ocking. The weigh o he 1-yea ling anged
om 2.5 kg o 3.4 kg and o he 0-yea ling 2.3 −3.8 kg in he ea men
mon hs.
2.1.8. Scena ios
A p oduc ion cycle wi h he baseline ea men egime was
compa ed wi h a p oduc ion cycle whe e one o se e al o he he mal
ea men s we e ei he p e en ed, eplaced o imp o ed. In scena io 1
he baseline ea men egime was compa ed wi h a p oduc ion cycle
wi hou ea men s. In addi ion, we also compa ed ou mechanical
ea men s o a p oduc ion cycle wi hou ea men s (scena io 2). In
scena io 3 and 4, wo ou o ou he mal ea men s we e p e en ed,
ei he he i s and second o he hi d and ou h. In scena io 5 and 6
one ou o ou he mal ea men s was p e en ed, ei he he i s o he
ou h. In scena io 7–10 he i s he mal ea men was eplaced wi h
ei he a mechanical (scena io 7), hyd ogen pe oxide (scena io 8),
eshwa e (scena io 9) o medicinal ba h (scena io 10). In scena io
11–13 we looked a he a iance in p o i i ou he mal ea men s
we e wi hin he 5% wo s pe o ming ea men s wi h espec o mo -
ali y and g ow h o he salmon, compa ed o an expec ed pe o mance
(scena io 11), expec ed compa ed o 5% bes (scena io 12) and 5% wo s
compa ed o 5% bes (scena io 13). Al oge he , his c ea ed 13 di e en
scena ios lis ed in Table 3.
2.2. Sensi i i y analysis
To assess he e ec om a ying he alues o he de e minis ic inpu
a iables on he ou pu alue, sensi i i y analyses we e pe o med on
he ollowing inpu a iables: baseline g ow h a e, baseline mo ali y,
days o eed wi hd awal, and eed and sales p ices. The ange o he low
and high inpu alues o he a iables in he sensi i i y analyses was
based on he desc ip i e s a is ics o da ase II and li e a u e. Baseline
g ow h a e was assumed o be no mally dis ibu ed in he sensi i i y
analysis, wi h 5 h and 95 h pe cen iles se o 2.3 and 3.2 o TGC.
Mon hly mo ali y was assumed a uni o m dis ibu ion wi h a minimum
o 0.1 and a maximum 1.0%. Days o eed wi hd awal had a uni o m
dis ibu ion wi h minimum ou and maximum se en days o eed
wi hd awal. Feed p ices we e assumed a iangula dis ibu ion wi h
+/- 20% min/max, and sales p ices we e assumed a iangula dis i-
bu ion wi h +/- 40% min/max. The eg ession coe icien s o he inpu
a iables we e compa ed in a To nado cha by applying he unc ion o
sensi i i y analysis in @Risk.
In da ase I and II, he e we e <14 obse a ions o he mal ea -
men s in he yea 2014. We suspec ed ha hese migh ha e an e ec on
he expec ed ou pu , as he mo ali y a e hese ea men s was qui e
high. We he e o e compa ed he ou pu wi h and wi hou hese
Fig. 1. Shows he modelled p oduc ion cycle o he wo smol ypes wi h mon h o s ocking (S), ea men mon hs (nume a ed 1–4), and mon h o ha es (H), in
addi ion o desc ip ion o he seasonal change in empe a u e in Celsius deg ees (do ed line) and he o al numbe o ea men s pe mon h ca ego ised by smol
ypes (ba g aph) om da ase II. C ea ed wi h BioRende .com.
C.S. Walde e al.
P e en i e Ve e ina y Medicine 221 (2023) 106062
7
obse a ions.
3. Resul s
3.1. The economic bene i (ΔΠ) o changing delousing ea men egime
Fo all scena ios, excep scena io 9, he mos impo an d i e o he
expec ed economic bene i o a change in ea men egime, is he in-
c ease in e enue, ollowed by inc eased eed cos s in scena io 1–6
(Table 4). Fo scena io 9 ( eplacing a he mal ea men wi h a esh-
wa e ea men ), he inc ease in ea men cos is he mos impo an
d i e . The highe eed cos in scena io 1 and 2 is a esul o educed
mo ali y and highe g ow h a e, which esul in inc eased eed con-
sump ion and subsequen ly a la ge biomass a he ime o ha es .
The change in p o i o mos measu es a e sligh ly highe o he 0-
yea ling compa ed o he 1-yea ling (Table 4).
The economic bene i o changing he delousing ea men egime
a ies, especially ega ding he p e en ion o ou he mal ea men
(Fig. 2).
The measu es anged by posi i e economic bene i a e p e en ion,
imp o emen and eplacemen .
3.1.1. P e en ion
The model does no include he di ec cos s o a ious p e en i e
measu es because o he unce ain e ec on educ ion in numbe o
immedia e ea men s as well as he di ec cos s. The economic bene i
would he e o e indica e how much a a me could use on p e en i e
measu es pe cage be o e i is no longe economical bene icial. P e-
en ing o a oiding he mal and mechanical ea men s has a la ge ex-
pec ed economic posi i e bene i (Table 4, scena io 1–6). Fo ins ance,
he expec ed inc ease in p o i by p e en ing ou he mal ea men s is
535,313
€
/cage/p oduc ion cycle o a 1-yea ling and 692,346 o a 0-
yea ling (Table 4).
Fo 1-yea lings, p e en ing he las o he las wo ea men s (sce-
na io 6 and 4) ha e a g ea e e ec han p e en ing he i s o he i s
wo ea men s (scena io 5 and 3) (Table 4 and Fig. 2). Fo he 0-yea -
ling, his is opposi e; p e en ing he i s o he i s wo ea men s
ha e a g ea e e ec han p e en ing he las o he las wo ea men s
(Table 4).
3.1.2. Replacemen
Fo bo h smol ypes, he model shows ha eplacing he i s he -
mal ea men wi h ano he ea men measu e has a mino expec ed
economic posi i e bene i , compa ed o he o he measu es o p e en -
ing o imp o ing (Table 4, scena io 7–10). In ac , eplacing a he mal
ea men wi h a eshwa e ea men (scena io 9) has a modelled ex-
pec ed nega i e bene i (Table 4).
Table 4 shows, ha o he 1-yea ling eplacing he i s he mal
ea men wi h a mechanical ea men has a modelled g ea e posi i e
impac compa ed o eplacing i wi h a hyd ogen pe oxide ea men .
This is he opposi e o he 0-yea ling whe e eplacing he he mal
ea men wi h a hyd ogen pe oxide ea men has a g ea e posi i e
impac on p o i change compa ed o eplacing i wi h a mechanical
ea men .
3.1.3. Imp o emen
The di ec cos o imp o ing ea men s is no included in he model,
hus he economic bene i o imp o ing indica es wha could be spen
pe cage pe p oduc ion cycle be o e b eak e en. Imp o ing ea men s
ha e a high economic impac , especially i he a me is able o imp o e
he baseline ea men egime o ou he mal ea men s om being
Table 3
The di e en scena ios (n =13) o bo h 1 and 0-yea ling. Scena io 1, and 3–10
a e compa ed wi h he baseline scena io ha in ol es ou he mal ea men s.
In con as , Scena io 2 is compa ed wi h ou mechanical ea men s. The sce-
na ios a e ca ego ised based on h ee c i e ia: p e en ion, eplacemen o
imp o emen . Finally, he imp o emen scena ios (11–13) speci ically compa es
he 5% wo s , expec ed, and 5% bes ou comes de i ed om he Mon e Ca lo
simula ion.
Desc ip ion o scena io Scena io
numbe
P e en No ea men s (All ou he mal ea men s emo ed) 1
No ea men s (All ou mechanical ea men s
emo ed)
2
Fi s wo (2/4) he mal ea men s emo ed 3
Las wo (2/4) he mal ea men s emo ed 4
Fi s (1/4) he mal ea men emo ed 5
Las (1/4) he mal ea men emo ed 6
Replace Fi s (1/4) he mal ea men eplaced wi h one
mechanical ea men
7
Fi s (1/4) he mal ea men eplaced wi h one
hyd ogen pe oxide ba h
8
Fi s (1/4) he mal ea men eplaced wi h one
eshwa e ba h
9
Fi s (1/4) he mal ea men eplaced wi h one
medicinal ba h
10
Imp o e 4/4 he mal ea men s, 5% wo s compa ed o
expec ed
11
4/4 he mal ea men s, expec ed compa ed o 5%
bes
12
4/4 he mal ea men s, 5% wo s compa ed o 5%
bes
13
Table 4
Shows he expec ed o al change in cos s, e enue and p o i o he di e en scena ios in Eu o (
€
). Scena io 11–13 a e no included as hey e lec he a iance in
scena io 1. A nega i e sign indica es a dec ease in cos o e enue.
€
o al 1-yea ling
Scena io no 1 2 3 4 5 6 7 8 9 10
Cos s
Feed 152,708 150,369 60,445 94,835 26,151 49,003 574 2,005 - 137 5,117
T ea men - 53,552 - 43,588 - 22,427 - 29,272 - 10,702 - 8,828 - 2,165 6,725 37,154 2,171
Slaug he 38,244 37,073 14,741 22,541 5,966 12,068 265 538 - 175 1,946
Mo ali y - 1,839 - 1,542 - 749 - 849 - 354 - 549 - 63 - 15 74 - 339
Re enue 670,874 650,322 258,585 395,409 104,647 211,696 4,650 9,446 - 3,070 34,134
P o i (ΔΠ) 535,313 508,011 206,574 308,153 83,586 160,002 6,039 193 - 39,985 25,239
€
o al 0-yea ling
Scena io no 1 2 3 4 5 6 7 8 9 10
Cos s
Feed 211,795 209,332 121,728 83,367 62,284 35,665 819 5,360 526 10,008
T ea men - 57,460 - 46,790 - 20,271 - 33, 231 - 8, 073 - 17,682 - 1,991 6,440 34,570 2,335
Slaug he 51,066 49,804 28,217 21,170 14,365 9,641 334 1,388 - 15 3,222
Mo ali y - 1,959 - 1,636 - 663 - 1,160 - 254 - 620 - 61 - 5 75 - 322
Re enue 895,786 873,654 494,974 371,364 251,981 169,119 5,851 24,353 - 271 56,518
P o i (ΔΠ) 692,346 662,944 365,963 301,218 183,659 142,115 6,751 11,169 - 35,426 41,275
C.S. Walde e al.
P e en i e Ve e ina y Medicine 221 (2023) 106062
8
among he 5% wo s o he 5% bes pe o ming (Fig. 2 and supple-
men a y able). The modelled economic bene i o imp o ing ou
ea men s om he 5% wo s o expec ed, expec ed o 5% bes and om
wo s o bes pe o ming has an economic bene i o 319,196, 418,738
and 737,934
€
/cage o a 1-yea ling and 355,761, 476,703 and
832,464
€
/cage o a 0-yea ling (supplemen a y able, scena io 11–13).
3.2. Sensi i i y analysis
The sensi i i y analysis o he inpu a iables; baseline g ow h a e
(TGC
baseline)
, baseline mo ali y (mo
baseline),
mo ali y a e a e ea -
men (Δmo a e), g ow h a e a e ea men (ΔTGC), pe iod o eed
wi hd awal, sales p ice and eed p ice, shows ha o scena io 1–6 he
model is mos sensi i e o baseline g ow h a e and he leng h o he
pe iod o eed wi hd awal (scena io1, 2, 4, 5) ollowing mo ali y a e
a e ea men o sales p ices dependen on he scena io (Fig. 3). The
economic loss inc eases wi h inc easing baseline TGC, s a a ion pe iod
and sales p ices. Fo all scena ios, dec ease o inc ease in baseline
mo ali y and eed p ices a e o lesse impo ance. The mos impo an
ac o in scena io 6–10 is he ea men mo ali y, ollowed by he
g ow h a e a e ea men .
Remo ing he 14 he mal ea men s om 2014 had no e ec on he
simula ed ou pu , hey we e he e o e kep in he dis ibu ion.
4. Discussion
Ou esul s sugges ha accoun ing o he biological losses associ-
a ed wi h lice ea men s is impo an when making choices o delousing
s a egies. The biological cos s o inc eased mo ali y and dec eased
g ow h associa ed wi h especially non-medicinal ea men s a e
Fig. 2. Shows he a iance in p o i change o he 1-yea ling (A) and 0-yea ling (B) measu ed in en housand
€
/cage/p oduc ion cycle. The scena ios a e anged by
he median inc ease in p o i o he 1-yea ling. The ed e e ence line indica es no change in p o i by changing delousing egime. A posi i e change in p o i
indica es a posi i e economic impac o changing delousing egime.
C.S. Walde e al.
P e en i e Ve e ina y Medicine 221 (2023) 106062
9
expec ed o be high, bu a ies subs an ially. The e o e, he economic
bene i o p e en ing o imp o ing can also be high. Salmon p oduce s
could hus in es a conside able amoun in measu es o p e en ion o
imp o emen o he mal ea men s be o e b eak-e en. Replacing one
he mal ea men wi h ano he immedia e ea men me hod has a
mino e ec . The esul s u he shows ha changes in sales alue and
eed consump ion cons i u e he la ges sha e o he change in p o i
be ween he scena ios.
Feed and sales p ices a e he wo mos impo an d i e s o he
ou come o a change in ea men egime. Inc eased e enue is due o
dec eased mo ali y and inc eased g ow h, which means ha es ing
mo e and la ge salmon. The inc eased ha es weigh migh also shi
mo e salmon in o a highe weigh class, hus ex a p ice p emium will
gi e addi ional e enue. As eed cons i u es abou 50% o he p oduc ion
cos s (I e sen e al., 2020; Misund e al., 2017), mos o he inc eased
p oduc ion cos when p e en ing ea men s is due o inc eased o al
amoun o eed used in he p oduc ion cycle. This inc ease is explained
by a oiding ei he , o bo h, pe iods o p e- ea men eed wi hd awal
and dec eased appe i e a e ea men . In addi ion, mo ali y a e
ea men s is a oided, hus a la ge numbe o ish needs o be ed. The
cos o ha es ing and handling o dead ish a e mino compa ed o eed
and sales p ices, and do no a ec he p o i change o he same deg ee as
eed cos s and e enue.
O e all, he modelled economic bene i o he di e en scena ios is
highe o he 0-yea ling compa ed o he 1-yea ling (Table 4). This can
be explained by highe seawa e empe a u es in he ea men mon hs
o he 0-yea ling (9.4–14.7 ⁰C) compa ed o he 1-yea ling (5.2–12.0 ⁰C)
(Fig. 1). P e en ing ea men s hus leads o a la ge educ ion in g ow h
loss o he 0-yea ling compa ed o he 1-yea ling, which ul ima ely
leads o a di e ence in ha es ed biomass. The model acco dingly in-
dica es ha p e en ing ea men s in he mon hs wi h he highes
seawa e empe a u es, would be mo e bene icial han p e en ing
ea men s in mon hs wi h lowe empe a u es, because he po en ial
g ow h loss is educed. Howe e , an excep ion o his gene al obse a-
ion has been epo ed o he mal ea men s, in which sea wa e
empe a u es in he lowe anges a e associa ed wi h nega i e impac on
g ow h a es, possibly due o he la ge in e al be ween ea men - and
sea wa e empe a u e (Walde e al., 2022). The p essu e o lice is highe
in he mon hs wi h highe seawa e empe a u es, hus ea men s in
hese mon hs could be ha de o a oid. In addi ion, mechanical ea -
men s a low seawa e empe a u es a e associa ed wi h a highe isk o
win e -wounds (And ews e al., 2015; Somme se e al., 2022). The
model does no accoun o he possible in e ac ion be ween g ow h a e
and empe a u e, no inc eased isk o seconda y diseases.
No su p isingly, p e en ing all ou non-medicinal ea men s,
especially poo ly execu ed he mal ea men s, has he la ges modelled
economic bene i (Table 4, scena io 1–2). A ecen s udy demons a ed
ha combining di e en p e en i e measu es could educe he numbe
o delousing e en s by 25% du ing a p oduc ion cycle (Oldham e al.,
2023). In ou model, his would be equi alen o p e en ing one he mal
ea men . Table 4 shows ha i p e en ing he i s he mal ea men
(scena io 5) a a me could jus i y using 83,586 (1-yea ling) o 183,659
(0-yea ling)
€
/cage, as long as he p e en i e measu es do no a ec he
mo ali y no he g ow h o he salmon.
In I e sen e al. (2017) he cos o a he mal ea men is es ima ed o
be 0.054
€
pe kg salmon ea ed, including a mo ali y cos o 0.017
€
pe kg ea ed ish. Howe e , in hei es ima ions he economic alue o
a dead salmon is equal o he p oduc ion cos pe kg, whe eas in ou
es ima ions he cos o mo ali y also includes oppo uni y cos associ-
a ed wi h e enue loss, in addi ion o g ow h loss which would be
equi alen o 0.14–0.30
€
/kg. This shows ha no including he isk o
mo ali y and g ow h loss associa ed wi h delousing ea men s, in
addi ion o he al e na i e cos o los e enue could lead o unde -
es ima ing he cos o a he mal ea men .
In p ac ice, i is challenging o p e en all ea men s du ing a p o-
duc ion cycle, and he cu en con ol o delousing no mally consis s o a
mix u e o p e en i e and immedia e ea men measu es (Ba e e al.,
2020). Scena io 11–13 (supplemen a y able) highligh he impo ance
o imp o ing he quali y o he mal ea men s, showing ha he
g ea es po en ial e ec on p o i s comes om imp o ing om he 5%
wo s o he 5% bes pe o ming ea men s. Howe e , he di ec cos o
imp o ing ea men s is no included in he model because he e is li le
esea ch done iden i ying po en ial isk ac o s ela ed o inc eased
mo ali y and dec eased g ow h a e he mal and mechanical delous-
ing. As in he case o p e en ion, he model he e o e indica es wha
could be spen pe cage pe p oduc ion cycle on imp o ing he ea -
men s. The a ia ion in mo ali y and g ow h wi hin he di e en
ea men ca ego ies is la ge, especially wi hin he mal ea men s
(Walde e al., 2021; Walde e al., 2022). An impo an ac o explaining
his could be a me s apply he me hods in a ying designs. Rega ding
he he mal ea men s, he model shows ha i would be economical
jus i iable o use up o 740,000
€
/cage (1-yea ling) o 830,000
€
/cage
(0-yea ling) (supplemen a y able, scena io 13) o imp o e ou poo
pe o ming he mal ea men s o be among he bes pe o ming. Thus,
he e is a subs an ial economic incen i e o p io i ising esea ch on
iden i ying ac o s ela ed o issues such as handling p ocedu es, he
ea men ig, p io heal h s a us o he ish, and iming o he ea men ,
e c. o imp o e he mal ea men s by educing mo ali y and ensu ing
Fig. 3. To nado plo o sensi i i y analysis showing he anking o he eg ession coe icien o he selec ed inpu a iables compa ing a p oduc ion cycle wi h ou
he mal ea men s (baseline) o a p oduc ion cycle wi hou any ea men s (scena io 1).
C.S. Walde e al.
