Content uploaded by Tim Dempster
Author content
All content in this area was uploaded by Tim Dempster on Jun 26, 2018
Content may be subject to copyright.
AQUACULTURE ENVIRONMENT INTERACTIONS
Aquacult Environ Interact
Vol. 10: 279–289, 2018
https://doi.org/10.3354/aei00270 Published June 26
INTRODUCTION
Trophic subsidies occur via the flow of energy be -
tween ecosystems, as both matter and organisms
(Larsen et al. 2016). Subsidies influence food web and
ecosystem dynamics in recipient environments, with
the quantity and quality of the subsidy exerting a
strong effect on the overall impact (Marcarelli et al.
2011). While trophic subsidies occur naturally, in -
creasingly, anthropogenic subsidies are driving
changes to ecosystems and food webs by altering the
distribution, abundance, growth and reproduction of
consumers in recipient environments (Marczak et al.
2007, Oro et al. 2013). Given over half of the human
population lives within 60 km of the coast (UNEP
2016), it is unsurprising that anthropogenic subsidies
© The authors 2018. Open Access under Creative Commons by
Attribution Licence. Use, distribution and reproduction are un -
restricted. Authors and original publication must be credited.
Publisher: Inter-Research · www.int-res.com
*Corresponding author: camille.white@utas.edu.au
Aquaculture-derived trophic subsidy boosts
populations of an ecosystem engineer
C. A. White1,2, 5,*, R. J. Bannister3, S. A. Dworjanyn4, V. Husa3, P. D. Nichols2,
T. Dempster1
1School of BioSciences, University of Melbourne, Parkville, Victoria 3010, Australia
2Oceans and Atmosphere, Commonwealth Scientific and Industrial Research Organization, Castray Esplanade, Hobart,
Tasmania 7000, Australia
3Institute for Marine Research, PO Box 1870, 5817 Bergen, Norway
4National Marine Science Centre, Southern Cross University, Coffs Harbour, New South Wales 2450, Australia
5Present address: Institute of Marine and Antarctic Studies, University of Tasmania, Nubeena Crescent, Taroona,
Tasmania 7053, Australia
ABSTRACT: Environmental management of coastal aquaculture is focused on acute impacts of
organic and nitrogenous wastes close to farms. However, the energy-rich trophic subsidy that
aquaculture provides may create cascades with influences over broader spatial scales. In a fjord
region with intensive fish farming, we tested whether an ecosystem engineer, the white urchin
Gracilechinus acutus, was more abundant at aquaculture sites than control sites. Further, we
tested whether diets influenced by aquaculture waste altered reproductive outputs compared
with natural diets. Urchins formed barrens at aquaculture sites where they were 10 times more
abundant (38 urchins m−2) than at control sites (4 urchins m−2). Urchins were on average 15 mm
larger at control sites. In the laboratory, urchins fed aquafeed diets had 3 times larger gonad
indices than urchins fed a natural diet. However, their reproduction was compromised. Eggs from
females fed an aquafeed diet had 13% lower fertilisation success and 30% lower larval survival
rates at 10 d compared with females fed a natural diet. A reproductive output model showed that
enhanced numbers of 10 d old larvae produced by the dense aquaculture-associated aggregations
of G. acutus will supersede any detrimental effects on reproduction, with larval outputs from
aquaculture sites being on average 5 times greater than control sites. The results show that aqua-
culture waste can act as a trophic subsidy in fjord ecosystems, stimulating aggregations of urchins
and promoting the formation of urchin barrens. Where finfish aquaculture is concentrated, com-
bined effects on the wider environment may produce ecosystem-level consequences.
KEY WORDS: Aquaculture · Echinus acutus · Gracilechinus acutus · Larval survival · Norway ·
Population density · Reproductive output · Sea urchin · Trophic subsidy · Urchin barren
O
PEN
PEN
A
CCESS
CCESS
Aquacult Environ Interact 10: 279–289, 2018
280
are common in marine systems. Examples include the
bulk input of nitrogenous and organic wastes (Gor-
man et al. 2009), fisheries discards (Oro et al. 2013)
and, increasingly, waste products from finfish and
shellfish aquaculture (Fernandez-Jover et al. 2011a).
Aquaculture of carnivorous fish in coastal waters
releases dissolved nitrogen and organic carbon to re-
ceiving environments, particularly through waste
feed and faecal material (Carroll et al. 2003, Bannister
et al. 2014). Outputs from cage aquaculture can drive
community change in the immediate surrounds of a
farm, in both benthic (Keeley et al. 2012) and pelagic
(Riera et al. 2014) systems, often leading to a prolifer-
ation of opportunistic taxa (Macleod et al. 2004, Kutti
et al. 2007). Less well understood are the impacts
of aquaculture subsidies on a broader scale, where
waste may be delivered in quantities more readily as-
similated by the wider ecosystem, with potential con-
sequences for the marine food web (Bannister et al.
2016, Broch et al. 2017). As aquafeed inputs are high
in lipid, waste from aquaculture is an energy-rich re-
source in the marine environment, with wild fauna
benefiting energetically from consumption (Parrish
2009). Wild marine fauna, including fish (Fernandez-
Jover et al. 2011b) and mobile invertebrates (Olsen
et al. 2012, White et al. 2017) consume aquaculture
waste. Proxy fitness measures, such as somatic and
liver condition indices, are higher in farm-associated
wild fish than wild fish caught distant from farms
(Dempster et al. 2011). However, potential repercus-
sions to the fitness of individuals and dynamics of
populations that receive aquaculture-derived trophic
subsidies remain un explored, as are the mechanisms
through which subsidies could cascade through eco-
systems on a broader scale.
As well as bulk quantity, the quality of a trophic re -
source determines ecosystem-level outcomes (Mar -
carelli et al. 2011). Modern aquaculture feeds have
lipid compositions that are relatively alien in the mar-
ine environment, as they are rich in shorter chain
(C18) polyunsaturated fatty acids (PUFA) derived
from terrestrial vegetable oils and meals, and low in
omega-3 long-chain (≥C20) PUFA (n-3 LC-PUFA),
produced by marine phytoplankton (Turchini et al.
2009, Nichols et al. 2014). As wild marine fauna typi-
cally have diets high in n-3 LC PUFA (Twining et al.
2016), high consumption of aquaculture waste repre-
sents a substantial quantitative and qualitative bio-
chemical shift in dietary intake. A shift in nutritional
quality of diet, combined with other challenges asso-
ciated with near farm environments, such as heavy
metals, synthetic chemicals and persistent organic
pollutants (Burridge et al. 2010, Samuelsen et al.
2015), may supersede fitness benefits associated with
the bulk organic subsidy. In this manner, aquaculture
outputs could function as an ecological trap, whereby
individuals are attracted to the trophic subsidy, with
detrimental fitness and reproductive consequences
(Robertson & Hutto 2006, Hale & Swearer 2016).
Norway is the largest producer of farmed Atlantic
salmon Salmo salar globally, with an annual produc-
tion exceeding 1.3 million t and 990 licensed farms in
2015 (Directorate of Fisheries 2016). In total, the
industry releases approximately 60 000 t of carbon,
34000 t of nitrogen and 9750 t of phosphorus into
fjord and coastal ecosystems (Ta ranger et al. 2015),
where it is available as a trophic resource for wild
fauna. Farms in Norway attract wild fish with an esti-
mated 12 000 t of wild fish aggregating around farms
on any given day in summer (Dempster et al. 2009).
Benthic productivity can also increase in farm areas,
particularly in deep fjords where productivity is lim-
ited and the addition of organic waste contributes
significantly to food supply (Kutti et al. 2007, Olsen et
al. 2012).
Sea urchins are ecosystem engineers in many
coastal ecosystems, altering habitat structure and
function through grazing, with ramifications for the
entire food web (Graham 2004, Ling 2008). The white
sea urchin Gracilechinus acutus (formerly Echi nus
acutus) is an ecosystem driver in the Norwegian
fjords, largely due to high grazing pressure on kelp
vegetation, with barren formation observed in areas
with dense aggregations (Husa et al. 2014). G. acutus
is omnivorous and consumes mussel spat, epibenthic
invertebrates and detritus, and can capitalise on
aquaculture waste as a trophic resource (White et al.
2017). Whether consuming aquaculture waste im -
proves or reduces fitness of G. acutus is at present
unknown.
We investigated the effect of aquaculture on popu-
lation densities of wild G. acutus and tested the
physio logical and reproductive consequences of con-
suming an aquaculture-derived trophic subsidy. Pop-
ulation densities and reproductive outputs were then
combined to model the consequences of aquaculture
on populations of G. acutus. Current monitoring re -
gimes and management of aquaculture outputs are
generally focused on acute impacts directly associ-
ated with cage or lease zones. If the outcome of aqua-
culture subsidies is to drive broad-scale population
growth of a species able to act as an ecosystem engi-
neer, energy flow through coastal fjord ecosystems
may be altered, with this work having immediate re -
le vance in assessing ecosystem effects of salmon
aquaculture.
White et al.: Aquaculture waste subsidies boost urchin populations
MATERIALS AND METHODS
Urchin densities close to and distant from
aquaculture sites
In a region of western Norway with intensive
Atlantic salmon farming (Masfjorden; Fig. S1 in the
Supplement at www. int-res. com/ articles/ suppl/ q010
p279_ supp. pdf), we tested whether abundance and
size of the white urchin Gracilechinus acutus differed
between aquaculture and control sites. In August
2015, we assessed urchin abundance per m2 and col-
lected urchins to compare mean size at 4 salmon farm
sites and control sites 1.5−2.0 km away from the
nearest farm. Counts were done within 6 randomly
placed 1 m2 quadrants along a 50 m transect line at
5−10 m depth. At farm sites this was done in the sub-
tidal zone between shoreline and farm and as close
as possible to an active cage. The first 50 urchins
encountered per site were collected and measured
dorsoventrally at their widest point using callipers.
Effects of aquaculture waste feed in diets
on the reproductive outputs and physiological
responses of urchins
To determine whether the level of waste feed from
aquaculture contained in urchin diets affected spaw -
ning, subsequent development of larvae and their
survival, we fed urchins with manipulated diets and
followed the fate of the larvae. Urchins were col-
lected on SCUBA from Masfjorden, Hordaland, Nor-
way from a depth of 5−15 m in January 2015 with the
collection point >5 km from the nearest active farm.
Sixteen animals were randomly assigned to 1 of 15
aquaria (200 l) and supplied with flow-through sea-
water at ambient temperature and salinity (approxi-
mately 8.9°C and 34.8‰). Each of the aquaria were
given one of 3 diets for a period of 10 wk, with 5 repli-
cate aquaria per diet. Diet 1 contained a current com-
mercial Atlantic salmon Salmo salar feed (farm feed),
Diet 2 was a 1:1 combination of the commercial feed
and natural materials (composite feed) and Diet 3
contained only natural materials (natural feed). ‘Nat-
ural’ was defined as anything urchins were ob served
feeding on in the wild, or found within gut contents,
which were analysed from randomly collected wild
urchins prior to commencing the experiment. This
included macroalgae, including fucoid brown algae
Fucus vesiculosus and sugar kelp Sa charina lattis-
sima (70% w/w), mixed red algae (20% w/w),
encrusting flora and fauna from kelp (including
corallines, epiphytic red algae and bryozoans, 5%
w/w), mussel spat, gastropods and various crus-
taceans (5% w/w), all collected by diver from the
shallow subtidal zone.
Artificial diets for urchins were manufactured fol-
lowing the exact methodology outlined in White et
al. (2017). Samples of all 3 diets were retained and
stored at −80°C for subsequent analysis. Sea urchins
were held for 5 d without food prior to the start of the
feeding trial to standardize hunger. Diets were ana-
lysed for carbon, nitrogen and lipid to establish com-
parative energetics of each diet (Table 1). Animals
were fed once every 3 d during the experiment,
which ensured they were never food limited. All mor-
talities were recorded.
After 10 wk of feeding, urchins were induced to
spawn by injection of 2−3 ml of 1.0 M KCl. Eggs from
gravid females were collected in 500 ml beakers of
filtered seawater. Wild males were collected on the
morning of the spawning event and induced to
spawn, with sperm collected on petri dishes using
dry pipettes. Eggs were subsequently checked for
shape and integrity and sperm for motility. The eggs
of each female were fertilized by the sperm from
multiple (n = 5) males. Five lots of 200 eggs from each
female were placed in 100 ml sterile glass rearing
pots. The volume of sperm required to achieve a
sperm:egg ratio of 1000:1 was determined through
haemocytometer counts. The sperm was briefly acti-
vated in filtered seawater and added to containers
holding the eggs. Rearing containers were left for
10 min for fertilisation to occur, then flushed to re -
move excess sperm. Rearing containers were main-
tained at a temperature of 9.0°C (equivalent natural
fjord temperature) and flushed daily with filtered
seawater.
Each rearing container was scored for percent fer-
tilisation success (2 h post-fertilisation) and percent
survival to 10 d post-fertilisation by counting the
number of viable eggs and larvae in the rearing con-
281
Aquafeed Composite Natural
% carbon 48.0 21.8 9.1
% nitrogen 2.7 3.0 0.7
% total lipid 13.3 7.9 0.4
% cholesterol 0.07 0.02 0.009
n-3:n-6 0.96 0.95 1.70
Table 1. Carbon, nitrogen and total lipid values for experi-
mental dietary treatments. n-3:n-6 refers to the ratio of
omega-3 polyunsaturated fatty acids to omega-6 polyunsat-
urated fatty acids within dietary treatments. % cholesterol
is given as the % of total lipid45
Aquacult Environ Interact 10: 279–289, 2018
tainers. Photographs of fertilised eggs and larvae
post-metamorphosis were taken using an Olympus
SZX7 dissecting microscope, Olympus DP26 digital
camera and cellSens Entry v1.7 image capture soft-
ware. Diameters of 30 fertilized eggs were measured
from each rearing container using ImageJ (NIH). The
length and symmetry of 10 d old larvae were recor -
ded using the criteria of Sheppard-Brennand et al.
(2010).
At the conclusion of the experiment, gonad indices
were obtained for 5 randomly selected urchins from
each tank. Urchins were patted down to remove
excess external water and weighed. Gonads were
removed from the test and weighed. An index meas-
ure was obtained by dividing total weight by the
weight of the gonad. From 3 urchins per tank, 1
gonad was fixed in Bouin’s solution for histological
sectioning. Gonads were washed, dehydrated in
ethanol and soaked in a haemotoxylin, erythrosin
and saffron stain, then embedded in wax and cut into
3 µm sections before being set on slides for examina-
tion. The remaining 4 gonads were freeze dried, with
total lipid, n-3 LC-PUFA and n-6 PUFA content
measured using techniques described in White et al.
(2017).
Statistical analysis
We tested whether abundance and size of urchins
differed between aquaculture and control locations
using ANOVA with location (‘aquaculture’ or ‘con-
trol’) as a fixed factor. The effects of diet on gonad
index, lipid content and respiration were tested using
ANOVA with diet as a fixed factor. For analysis of lar-
val success parameters (fertilisation success, larval
survival, egg size, larval size and symmetry), a single
mean data point for each female derived from larvae
across all 5 rearing containers was determined. We
used a PERMANOVA (Anderson et al. 2008) to test
the overall effect of diet across the multiple parame-
ters of larval success (using PRIMER v7 and its com-
plementary software package PERMANOVA+ (v7))
Monte Carlo (MC) p-values of 0.05 were used to indi-
cate significant differences between treatments.
SIMPER analysis was subsequently used to assess
the contribution of each parameter to the dissimi -
larity between treatments. Following this, 1-way
ANOVA was used to test whether fertilisation suc-
cess, larval survival, egg size, larval size and symme-
try varied with diet. For larval success parameters,
both 2-way multivariate and univariate analysis was
performed first with tank as a factor nested within
diet. Where the effect of tank was highly non-signifi-
cant (p ≥0.2), the design was collapsed and effects
examined through 1-way analysis with diet as a fixed
factor (Quinn & Keough 2002). All data were checked
for assumptions of normality and homogeneity ac -
cording to Quinn & Keough (2002), and data were
square root transformed where appropriate. Where
diets differed significantly (p ≤0.05), Tukey-Kramer
post-hoc tests were conducted to detect differences
among means.
Reproductive output model
We modelled the reproductive outputs of farm-
associated and control urchin populations in Mas-
fjord by using our data on the abundance, size and
larval fitness after exposure to different diet types.
The number of 10 d old urchin larvae (N) produced
by 1 m2of habitat in farm and non-farm conditions
was calculated as:
N= (DU/S) × (E/ FS) × LS(1)
where DUrepresents the density of urchins per m2
observed from field data, Srepresents the male:
female ratio (assumed to be 0.5 for all scenarios) and
Erepresents the number of eggs released by each
female, which is size dependent. We assumed the
number of eggs released by urchins increased with
size and was directly proportional as for other inver-
tebrates (Levitan 1991). Fecundity estimates from the
experiment indicated that a ripe 5 cm female pro-
duced 5 million eggs, which was the value used to
scale size-dependent egg release. FSand LSrepre-
sent the fertilization and larval survival rates ob -
tained through the experiment. Nwas calculated
using experimental values obtained for both aqua-
culture feed and composite diets for the ‘farm’ condi-
tions, while values for the natural diet were used to
calculate Nfor ‘non-farm’ conditions.
RESULTS
Urchin densities close to and distant from
aquaculture sites
Abundance of urchins was 3 to 100 times greater at
aquaculture than control sites (F4, 44 = 31.6, p <
0.0001) (Table S1). All 4 aquaculture sites were char-
acterised by high densities of urchins (29−47 m–2)
and urchin barrens, compared with lower densities
(0.2−13.2 m–2) in control locations where only 1 of 4
282
White et al.: Aquaculture waste subsidies boost urchin populations
sites could be characterised as an urchin barren
(Fig. 1A). Mean test diameters were consistently
smaller (4−25 mm on average) at aquaculture than
control sites (F3, 390 = 33.2, p < 0.0001; Fig. 1B). There
was also a significant location effect for both abun-
dance (F4,44 = 3.2, p < 0.04) and test diameter (F3, 390 =
71.5, p < 0.0001) of urchins.
Effects of aquaculture waste feed in diets
on the reproductive outputs and physiological
responses of urchins
Mortality of adults throughout the 10 wk exposure
period was uniform across dietary treatments and did
not exceed 4 individuals from any tank. Females were
successfully induced to spawn from each treatment.
Diet altered larval survival and growth para meters
(F2,10 = 4.3, p(MC) = 0.03) (Tables S2–S4), with % sur-
vival 10 d post fertilisation, fertilisation success and
echinopluteal length identified by SIMPER as con-
tributing to over 70% of dissimilarity be tween diets.
Fertilisation success, egg diameter and echinopluteal
length all increased linearly from aquafeed to com-
posite to natural dietary treatments, while larval
asymmetry decreased (Fig. 2). Survival of larvae 10 d
post-fertilisation varied with dietary treatment (F2, 10
= 7.0, p = 0.01), with survival of larvae from the natu-
ral treatment over 30% higher than the aquafeed
283
Fig. 1. Results from the field survey of Gracilechinus acutus
(A) urchin abundance (m–2) and (B) average test size ± SE at
control and farm locations across 4 sites in Masfjord, Norway
Fig. 2. Mean (±SE) effects of dietary treatment on larval success and development of Gracilechinus acutus measured as (A) %
fertilisation success and % survival 10 d post fertilisation, (B) egg diameter (µm), (C) pluteal length, being the average length
of the pluteal arms (µm) and (D) % asymmetry obtained by the difference in length of pluteal arms. Superscript letters in (A)
denote significant treatment effects
Aquacult Environ Interact 10: 279–289, 2018
treatment and 24% higher than the composite treat-
ment (Fig. 2).
Urchins fed aquafeed diets had 1.5 and 4.0 times
larger gonads compared with the composite and
natural diets, respectively (F2,12 = 17, p < 0.0001;
Fig. 3A) (Table S5). When examined on a per unit
mass basis, total lipid content in gonads or eggs did
not differ with diet, with gonads ranging from 11.1 to
13.7 mg g−1 total lipid dry mass and eggs slightly
higher at 13.0−18.1 mg g−1 total lipid dry mass
(Fig. 3B). Likewise, the n-3 LC-PUFA:n-6 PUFA ratio
in gonads (F2,12 = 2.0, p = 0.2) and eggs (F2, 6 = 0.1, p =
0.9) was similar across diets, despite the n-3 LC-
PUFA:n-6 PUFA ratio being approximately double in
natural feed, compared with the aquafeed or com-
posite feeds (Fig. 3C, Table 1). Eggs were more
enriched in n-3 LC-PUFA compared with gonads for
all diets, varying between 1.65 and 1.75, which was
also the n-3 LC-PUFA:n-6 PUFA ratio of natural feed
(Fig. 3C, Table 1). Gonad histology indicated that
membrane-bound vesicles within the nutritive pha -
go cyte were emptier in urchins fed the aquafeed or
composite diet, but filled with varying granular con-
tents in urchins fed natural feed (Fig. 4).
Reproductive output model
Model results indicate that aggregations of adult
urchins at aquaculture sites will lead to a net increase
in the number of 10 d old larvae surviving in the
water column, compared with a natural scenario
(Fig. 5). Larval output from 1 m2of fjord in aqua -
culture locations subject to aquaculture feed or com-
posite diets were on average 5 times greater than
control locations with natural diets, although this var-
ied between locations (Fig. 5). Larval outputs under
the most ecologically relevant composite diet were
21% greater than for the aquaculture feed diet due to
greater survivorship of larvae.
DISCUSSION
We demonstrated that high densities of sea urchins
aggregate at aquaculture sites, while control sites
had far lower urchin densities. High densities of ur -
chins at aquaculture sites could form through attrac-
tion and aggregation of larval, juvenile or adult
urchins, or via reduced mortality of urchins at aqua-
culture sites compared with controls, or a combina-
tion of these processes. While we could not separate
mechanisms leading to aggregations of urchins,
aquaculture sites create suitable conditions for the
formation and persistence of urchin barrens, with
possible wider consequences for fjord ecosystems.
The occurrence of dense populations of urchins at
aquaculture sites places them directly at the source
of the greatest waste deposition, which they readily
consume (White et al. 2017). The trophic subsidy is
qualitatively different from natural feeds and creates
changes in the reproductive capabilities of urchins
that receive it. When abundance data and reproduc-
tive outputs data are combined, each aquaculture
site can be a population source for urchins, produc-
284
Fig. 3. Mean (±SE) effects of dietary treatment on Graci -
lechinus acutus lipid content measured as (A) standardized
gonad index, (B) total lipid (mg lipid g−1 dry mass) in gonads
and eggs, and (C) n-3 LC-PUFA:n-6 PUFA in gonads and
eggs. Superscript letters in (A) denote significant treatment
effects
White et al.: Aquaculture waste subsidies boost urchin populations
ing 5 times more competent 10 d old
larvae than control sites in the same
fjord. Overall, consuming an aqua -
culture waste subsidy affects both in -
dividuals and populations of ur chins,
with potential ecological consequen -
ces.
Effects on individuals
Gonad indices of Gracilechinus acu -
tus differed corresponding to the pro-
portion of aquafeed in the dietary
treatments. As gonads act as both a re-
productive and an energy storage or-
gan in sea urchins (Marsh & Watts
2001, Walker et al. 2001), it is un -
surprising that increases in dietary
carbohydrate, protein and lipid lead to
larger gonad indices. An energy-rich
subsidy in the wild could have impli-
cations for enhanced survival during
periods of food limitation, with urchins
able to reabsorb energy stores when
required to meet metabolic demands
(Kelly 2000). Given the importance of
maternal lipid reserves in facilitating
larval survival (Byrne et al. 2008, Car-
boni et al. 2012), offspring from the
aquafeed and composite dietary treat-
ments would have been expected to
have higher fertilisation and larval
success rates. Paradoxically, the oppo-
site oc curred, with the proportion of
aquafeed in maternal diet correspon-
ding to smaller eggs and decreased
fertilisation and larval survival at 10 d.
While urchins in the experiment were
never food limited, those fed lower-
energy diets may supply resources to
reproductive development in favour of
somatic growth (Kelly & Cook 2001,
Otero-Villanueva et al. 2004). Given
the lower lipid values of the natural
experimental diet, but similar lipid
concentrations in eggs compared with
other diets, combined with larger egg
size, there is some evidence this oc-
curred in females fed natural diets.
The biochemical composition of diet
may play an important role in the
increased success of larvae produced
285
Fig. 4. Histological examination of the gonads of Gracilechinus acutus using
haemotoxylin erythrosin saffron stain as (A) male urchins fed the aquafeed
treatment, (B) female urchins fed the aquafeed treatment, (C) male urchins
fed the composite treatment, (D) female urchins fed the composite treatment,
(E) male urchins fed the natural treatment and (F) female urchins fed the
natural treatment. NP: nutritive phagocyte; OO: oocyte; SP: spermatozoa.
Scale bar = 50 µm
Fig. 5. Model output showing the number of Gracilechinus acutus 10 d old
larvae (×106) present in the water column from a 1 m2section of fjord in both
aquaculture and control locations, under aquafeed (light grey), composite
(dark grey) and control (white) dietary conditions
Aquacult Environ Interact 10: 279–289, 2018
by females fed natural diets. Lipid composition of the
aquafeeds used were high in terrestrially derived C18
PUFA and low in marine-derived ≥C20 n-3 LC-PUFA,
which is the current global standard for grow-out
feeds in commercial finfish aquaculture (Turchini et
al. 2009, Nichols et al. 2014). Increased levels of ter -
res trially derivedoil in thediet, particularly in creased
n-6 PUFA, led to decreased sperm viability and fertil-
isation rates of guppies (Rahman et al. 2015) and ur -
chins (White et al. 2016), while high intake of marine-
derived n-3 LC-PUFA promotes growth and develop-
ment of urchin larvae (Carboni et al. 2012). There was
no difference in the n-3 LC-PUFA:n-6 PUFA ratio of
gonads or eggs among dietary treatments; however,
G. acutus can selectively accumulate particular fatty
acids from diet or biosynthesize essential fatty acids
from dietary substrates (White et al. 2017). This con-
siderably alters the fatty acid profile of body tissue
compared with diet and can help overcome limita-
tions in essential fatty acids (White et al. 2017). How-
ever, there is a limit to this capacity, and there are
unknown energetic costs (Laurel et al. 2010).
Performance of larvae may also be influenced by
other feed components. Although not measured in
this study, protein is important for successful larval
development and survival in sea urchins (Marsh &
Watts 2001) and interacts with lipid during uptake
and assimilation (Cook et al. 2007). As eggs pro-
duced by G. acutus females fed natural diet are
larger, yet contain similar concentrations of lipids to
the other diets, they may contain a higher proportion
of another metabolic component important to sur-
vival. In the wild, G. acutus forms dense aggrega-
tions on detrital macroalgae and a preference for
macroalgae is common in omnivorous urchins (e.g.
Echinometra chloroticus, Barker 2001; Psammechi-
nus miliaris, Kelly & Cook 2001). Therefore, specific
compounds that improve larval development may
occur in algal components of the natural diet, en -
hancing its overall nutritional value and therefore
larval success. Alternatively, another component of
the aquaculture feed itself may be detrimental to lar-
val success.
Population-level effects
Regardless of the physiological mechanism, female
G. acutus fed diets containing salmon aquaculture
feed produced larvae with lower survival rates at 10 d
than those fed a natural diet. Given the dense aggre-
gations of G. acutus at farms, by reducing re -
productive outcomes, aquaculture waste may act as
an ecological trap (Robertson & Hutto 2006, Hale &
Swearer 2016). The exact effect on populations will be
a balance between the physiological benefits that the
resource provides to individuals and populations, ver-
sus any detrimental effects on reproductive outputs.
Our reproductive model suggested that the numbers
of sea urchins at aquaculture sites negate any detri-
mental effects that consuming aquaculture feed have
on offspring. Barren areas are common where dense
aggregations of G. acutus are found, with urchins
able to maintain a lower limit of kelp vegetation (Husa
et al. 2014). Urchins were smaller at aquaculture com-
pared with control locations, with reduced body sizes
typical where densities of urchins are high (Levitan
1989, 1991). Subsequently, the energetic benefit of
the resource to the individual may be limited due to
high competition for the trophic subsidy. However,
the overall net effect on the population is positive, as
high population densities in broadcast spawners such
as G. acutus can en sure fertilisation success before
gametes become dilu ted (Quinn et al. 1993, Wahle &
Peckham 1999, Gascoigne et al. 2009). As such, in-
creased population densities of G. acutus around
aquaculture sites will increase larval abundances, de-
spite animals being smaller compared with more iso-
lated animals at control locations (Levitan 1991, Levi-
tan et al. 1992, Lundquist & Botsford 2011). As our
model could not account for this important Allee ef-
fect, relative fertilisation success at aquaculture sites
is likely under estimated.
Given the larval retention time for G. acutus is ap-
proximately 50 d (MacBride 1903, Gage et al. 1986),
larvae from aquaculture-associated aggregations will
disperse widely and settle beyond the aquaculture
zone where they were released. While this could in-
crease the spatial influence that aqua culture has on
urchin populations, it may also dilute its effect. How-
ever, in a typical fjordal circulation pattern, where a
thin low-salinity surface layer flows seaward, and
with a compensatory landward flow beneath this, a
large proportion of larvae are likely to be entrained
within the fjord. This phenomenon occurs for the sea
urchin Evechinus chloroticus in New Zealand fjord
systems (Lamare 1998, Wing et al. 2003), which pro-
duces similarly small, negatively buoyant plank-
totrophic larvae to G. acutus with similar development
times (Lamare 1998, Tyler & Young 1998). If entrain-
ment of larvae occurs in fjord systems with intensive
aquaculture, increases in larval production will be
concentrated, with effects compounded over time.
A more precise estimation of the impact of aqua-
culture on G. acutus populations across larger spatial
scales may be possible with further information on a
286
White et al.: Aquaculture waste subsidies boost urchin populations
number of key variables, such as the extent of suit-
able settling habitat within the fjord, and the spatial
and temporal persistence of aggregations around
aquaculture sites. The link between pelagic and ben-
thic systems and factors that may influence success-
ful metamorphosis and settling of G. acutus larvae
will also aid better estimates of the overall impact
that an aquaculture subsidy may have on a fjord-
wide scale. Moreover, as our survey depth only
extended to 10 m and G. acutus has a depth range
that potentially reaches 2000 m (Tyler & Young
1998), we were only able to capture a small propor-
tion of the total aggregation in our survey. Whether
the aquaculture subsidy actually acts as attractant,
causing G. acutus to migrate to shallower depths
around farms, requires further investigation.
The reproductive output model predicted an in -
crease in the number of larvae in the water column
each year due to aquaculture. When tracing assimila-
tion of farm waste, White et al. (2017) found urchins
consumed feed up to 350 m from farms. Using this
value as the demarcation between ‘aquaculture af -
fected’ and normal fjord, we can examine broader
effects of urchin aggregations around farms. Given
the total coastline of Masfjorden (70 km), this results
in 6% of coastline within the fjord affected by aqua-
culture. When coupled with outputs from the repro-
ductive model, this equates to an increase of approx-
imately 20% in the amount of urchin larvae in the
fjord system each year. Given that aquaculture com-
menced in the 1970s in Norway (Husa et al. 2014)
and has expanded with time, a long-term interaction
of this nature has broad implications for the creation
of urchin larvae population sources in fjord eco -
systems. Only observational data exist on urchin
populations prior to the commencement of aqua -
culture in the fjords (Jorde & Klavestad 1963), mak-
ing it difficult to fully assess the long-term impact of
aquaculture on urchins. However, the wider ecologi-
cal effects of urchin aggregations, such as barren for-
mation, suggest that enhanced production of urchin
larvae on a fjord-wide scale has the potential to drive
ecosystem-level change.
CONCLUSIONS AND IMPLICATIONS FOR
MANAGEMENT
Sea urchins occur naturally in Norwegian fjords,
and localised barren areas existed due to G. acutus
overgrazing prior to the introduction of large-scale
salmon aquaculture (Jorde & Klavestsd 1963, Husa et
al. 2014). However, our results show that these events
can be promoted by aquaculture subsidies. Aquacul-
ture waste is an energy-rich trophic subsidy and can
stimulate dense aggregations of urchins at farm sites.
Consumption of aquaculture waste alters biochemical
physiology and reduces larval success; however,
modelling indicates that adult density ef fects will su-
persede any detrimental effects on reproduction in
terms of net reproductive output. Like many urchin
species, G. acutus is an ecosystem engineer that can
drive ecosystem-level change via barren formation
through overgrazing on kelp. Risk-based assessment
on appropriate spatial and temporal scales is required
to fully understand the extent of interactions between
aquaculture-derived trophic sub sidies and popula-
tions of urchins in fjord eco systems.
The lack of regional baseline data for benthic com-
munities, on both soft and hard substrate, is not an
uncommon scenario in environments where aquacul-
ture has expanded from small-scale operations to
multi-farm operations with high regional densities.
An understanding of localised impacts may be ade-
quate where aquaculture operations are small. How-
ever, when aquaculture operates at high regional
densities, diffuse effects can become additive and
influence change on a much broader spatial scale,
where unfortunately, baseline data are often lacking.
As aquaculture continues to expand, it is critical to
capture a robust environmental baseline through
which broad-scale changes can be evaluated in the
future. Further exploration of the interaction be -
tween G. acutus and finfish aquaculture is warranted
to fully assess and subsequently mitigate any conse-
quences for broader ecosystem function in the Nor-
wegian fjords.
Acknowledgements. This work was supported by a Univer-
sity of Melbourne Overseas Research Experience Scholar-
ship (ORES) and the Norwegian Research Council (Project
no. 228871). The authors thank S. A. Olsen, K. A. Kvestad, B.
Haugland Taraldset, S. Woodcock, N. Keeley, B. Muir, F.
Oppedal, Ø and Ingrid Uglenes Fiksdal. Strand and T.
Strohmeier and the technical staff at the Institute of Marine
Research (IMR), Matre, for assistance.
LITERATURE CITED
Anderson MJ, Gorley RN, Clarke KR (2008) PERMANOVA+
for PRIMER: guide to software and statistical methods.
PRIMER-E, Plymouth
Bannister RJ, Valdemarsen T, Hansen PK, Holmer M, Ervik
A (2014) Changes in benthic sediment conditions under
an Atlantic salmon farm at a deep, well-flushed coastal
site. Aquacult Environ Interact 5: 29−47
Bannister RJ, Johnsen IA, Asplin L, Kutti T, Hansen PK
(2016) Near- and far-field dispersal modelling of organic
waste from Atlantic salmon aquaculture in fjord systems.
ICES J Mar Sci 73: 2408−2419
287
Aquacult Environ Interact 10: 279–289, 2018
Barker MF (2001) The ecology of Evechinus chloroticus. In:
Lawrence JM (ed) Edible sea urchins: biology and eco -
logy. Elsevier, Amsterdam, p 245−260
Broch OJ, Daae RL, Ellingsen IH, Nepstad R, Bendiksen EA,
Reed JL, Senneset G (2017) Spatiotemporal dispersal and
deposition of fish farm wastes: a model study from cen-
tral Norway. Front Mar Sci 4: 199
Burridge L, Weis JS, Cabello F, Pizarro J, Bostick K (2010)
Chemical use in salmon aquaculture: a review of current
practices and possible environmental effects. Aqua -
culture 306: 7−23
Byrne M, Sewell MA, Prowse TAA (2008) Nutritional eco logy
of sea urchin larvae: influence of endogenous and exoge-
nous nutrition on echinopluteal growth and phenotypic
plasticity in Tripneustes gratilla. Funct Ecol 22: 643−648
Carboni S, Vignier J, Chiantore M, Tocher DR, Migaud H
(2012) Effects of dietary microalgae on growth, survival
and fatty acid composition of sea urchin Paracentrotus
lividus throughout larval development. Aquaculture
324−325: 250−258
Carroll ML, Cochrane S, Fieler R, Velvin R, White P (2003)
Organic enrichment of sediments from salmon farming
in Norway: environmental factors, management prac-
tices, and monitoring techniques. Aquaculture 226:
165−180
Cook EJ, Kelly MS (2007) Effect of variation in the protein
value of the red macroalga Palmaria palmata on the
feeding, growth and gonad composition of the sea
urchins Psammechinus miliaris and Paracentrotus lividus
(Echinodermata). Aquaculture 270: 207−217
Cook EJ, Hughes AD, Orr H, Kelly MS, Black KD (2007)
Influence of dietary protein on essential fatty acids in the
gonadal tissue of the sea urchins Psammechinus miliaris
and Paracentrotus lividus (Echinodermata). Aquaculture
273: 586–594
Dempster T, Uglem I, Sanchez-Jerez P, Fernandez-Jover D,
Bayle-Sempere J, Nilsen R, Bjorn PA (2009) Coastal
salmon farms attract large and persistent aggregations of
wild fish: an ecosystem effect. Mar Ecol Prog Ser 385:
1−14
Dempster T, Sanchez-Jerez P, Fernandez-Jover D, Bayle-
Sempere J, Nilsen R, Bjorn PA, Uglem I (2011) Proxy
measures of fitness suggest coastal fish farms can act as
population sources and not ecological traps for wild
gadoid fish. PLOS ONE 6: e15646
Directorate of Fisheries (2016) Aquaculture statistics. www.
fiskeridir.no/English/Aquaculture/Statistics/Atlantic-
salmon-and-rainbow-trout (accessed September 2016)
Fernandez-Jover D, Martinez-Rubio L, Sanchez-Jerez P,
Bayle-Sempere JT and others (2011a) Waste feed from
coastal fish farms: a trophic subsidy with compositional
side-effects for wild gadoids. Estuar Coast Shelf Sci 91:
559−568
Fernandez-Jover D, Arechavala-Lopez P, Martinez-Rubio L,
Tocher DR and others (2011b) Monitoring the influence
of marine aquaculture on wild fish communities: benefits
and limitations of fatty acid profiles. Aquacult Environ
Interact 2: 39−47
Gage JD, Tyler PA, Nichols D (1986) Reproduction and
growth of Echinus acutus var. norvegicus Duben & Koren
and E. elegans Duben & Koren on the continental slope
off Scotland. J Exp Mar Biol Ecol 101:61– 83
Gascoigne J, Berec L, Gregory S, Courchamp F (2009) Dan-
gerously few liaisons: a review of mate-finding Allee
effects. Popul Ecol 51: 355−372
Gorman D, Russell BD, Connell SD (2009) Land-to-sea con-
nectivity: linking human-derived terrestrial subsidies to
subtidal habitat change on open rocky coasts. Ecol Appl
19: 1114−1126
Graham MH (2004) Effects of local deforestation on the
diversity and structure of Southern California giant kelp
forest food webs. Ecosystems 7: 341−357
Hale R, Swearer SE (2016) Ecological traps: current evi-
dence and future directions. Proc R Soc B 283:
20152647
Husa V, Steen H, Sjotun K (2014) Historical changes in
macroalgal communities in Hardangerfjord (Norway).
Mar Biol Res 10: 226−240
Jorde I, Klavestad N (1963) The natural history of the
Hardangerfjord. 4. The benthonic algal vegetation. Sar-
sia 9: 1−100
Keeley NB, Forrest BM, Crawford C, Macleod CK (2012)
Exploiting salmon farm benthic enrichment gradients to
evaluate the regional performance of biotic indices and
environmental indicators. Ecol Indic 23: 453−466
Kelly MS (2000) The reproductive cycle of the sea urchin
Psammechinus miliaris (Echinodermata: Echinoidea) in a
Scottish sea loch. J Mar Biol Assoc UK 80: 909−919
Kelly MS, Cook EJ (2001) The ecology of Psammechinus
miliaris. In: Lawrence JM (ed) Edible sea urchins: bio -
logy and ecology. Elsevier Science, Amsterdam,
p 217–224
Kutti T, Hansen PK, Ervik A, Hoisaeter T, Johannessen P
(2007) Effects of organic effluents from a salmon farm on
a fjord system. II. Temporal and spatial patterns in in -
fauna community composition. Aquaculture 262: 355−366
Lamare MD (1998) Origin and transport of larvae of the sea
urchin Evechinus chloroticus (Echinodermata: Echi-
noidea) in a New Zealand fiord. Mar Ecol Prog Ser 174:
107−121
Larsen S, Muehlbauer JD, Marti E (2016) Resource subsidies
between stream and terrestrial ecosystems under global
change. Glob Change Biol 22: 2489−2504
Laurel BJ, Copeman LA, Hurst TP, Parrish CC (2010) The
ecological significance of lipid/fatty acid synthesis in
developing eggs and newly hatched larvae of Pacific cod
(Gadus macrocephalus). Mar Biol 157: 1713−1724
Levitan DR (1989) Density dependent size regulation in
Diadema antillarum —effects on fecundity and survivor-
ship. Ecology 70: 1414−1424
Levitan DR (1991) Influence of body size and population
density on fertilization success and reproductive output
in a free-spawning invertebrate. Biol Bull 181: 261−268
Levitan DR, Sewell MA, Chia FS (1992) How distribution
and abundance influence fertilization success in the sea
urchin Stronglyocentrotus droebachiensis. Ecology 73:
248−254
Ling SD (2008) Range expansion of a habitat-modifying
species leads to loss of taxonomic diversity: a new and
impoverished reef state. Oecologia 156: 883−894
Lundquist CJ, Botsford LW (2011) Estimating larval produc-
tion of a broadcast spawner: the influence of density,
aggregation, and the fertilization Allee effect. Can J Fish
Aquat Sci 68: 30−42
MacBride EW (1903) The development of Echinus escu lentus,
together with some points in the development of E. mili -
aris and E. acutus. Philos Trans R Soc B 195: 285−327
Macleod CK, Crawford CM, Moltschaniwskyj NA (2004)
Assessment of long term change in sediment condition
after organic enrichment: defining recovery. Mar Pollut
Bull 49: 79−88
288
White et al.: Aquaculture waste subsidies boost urchin populations
Marcarelli AM, Baxter CV, Mineau MM, Hall RO (2011)
Quantity and quality: unifying food web and ecosystem
perspectives on the role of resource subsidies in fresh -
waters. Ecology 92: 1215−1225
Marczak LB, Thompson RM, Richardson JS (2007) Meta-
analysis: trophic level, habitat, and productivity shape
the food web effects of resource subsidies. Ecology 88:
140−148
Marsh AG, Watts SA (2001) Energy metabolism and gonad
development. In: Lawrence JM (ed) Edible sea urchins:
biology and ecology. Elsevier Science, Amsterdam,
p27–42
Nichols PD, Glencross B, Petrie JR, Singh SP (2014) Readily
available sources of long-chain omega-3 oils: Is farmed
Australian seafood a better source of the good oil than
wild-caught seafood? Nutrients 6: 1063−1079
Olsen SA, Ervik A, Grahl-Nielsen O (2012) Tracing fish farm
waste in the northern shrimp Pandalus borealis (Kroyer,
1838) using lipid biomarkers. Aquacult Environ Interact
2: 133−144
Oro D, Genovart M, Tavecchia G, Fowler MS, Martinez-
Abrain A (2013) Ecological and evolutionary implications
of food subsidies from humans. Ecol Lett 16: 1501−1514
Otero-Villanueva MDM, Kelly MS, Burnell G (2004) How
diet influences energy partitioning in the regular echi-
noid Psammechinus miliaris; constructing an energy
budget. J Exp Mar Biol Ecol 304: 159−181
Parrish CC (2009) Essential fatty acids in aquatic food webs.
Springer, New York, NY
Quinn GP, Keough MJ (2002) Experimental design and data
analysis for biologists. Cambridge University Press,
Cambridge
Quinn JF, Wing SR, Botsford LW (1993) Harvest refugia in
marine invertebrate fisheries —models and applications
to the red sea urchin, Stronglyocentrotus franciscanus.
Am Zool 33: 537−550
Rahman MM, Gasparini C, Turchini GM, Evans JP (2015)
Testing the interactive effects of carotenoids and poly -
unsaturated fatty acids on ejaculate traits in the guppy
Poecilia reticulata (Pisces: Poeciliidae). J Fish Biol 86:
1638−1643
Riera R, Sanchez-Jerez P, Rodriguez M, Monterroso O
(2014) Artificial marine habitats favour a single fish spe-
cies on a long-term scale: the dominance of Boops boops
around off-shore fish cages. Sci Mar 78: 505−510
Robertson BA, Hutto RL (2006) A framework for understand-
ing ecological traps and an evaluation of existing evi-
dence. Ecology 87: 1075−1085
Samuelsen OB, Lunestad BT, Hannisdal R, Bannister R and
others (2015) Distribution and persistence of the anti sea-
lice drug teflubenzuron in wild fauna and sediments
around a salmon farm, following a standard treatment.
Sci Total Environ 508: 115−121
Sheppard Brennand H, Soars N, Dworjanyn SA, Davis AR,
Byrne M (2010) Impact of ocean warming and ocean
acidification on larval development and calcification in
the sea urchin Tripneustes gratilla. PLOS ONE 5: e11372
Taranger GL, Karlsen O, Bannister RJ, Glover KA and others
(2015) Risk assessment of the environmental impact of
Norwegian Atlantic salmon farming. ICES J Mar Sci 72:
997−1021
Turchini GM, Torstensen BE, Ng WK (2009) Fish oil replace-
ment in finfish nutrition. Rev Aquacult 1: 10−57
Twining CW, Brenna JT, Hairston NG, Flecker AS (2016)
Highly unsaturated fatty acids in nature: what we know
and what we need to learn. Oikos 125: 749−760
Tyler P, Young CM (1998) Temperature and pressure toler-
ances in dispersal stages of the genus Echinus (Echino-
dermata: Echinoidea): prerequisites for deep-sea inva-
sion and speciation. Deep Sea Res II 45: 253−277
UNEP (United Nations Environment Programme) (2016)
World Ocean Assessment: Overview. GRID-Arendal,
Arendal
Wahle RA, Peckham SH (1999) Density-related reproductive
trade-offs in the green sea urchin, Strongylocentrotus
droebachiensis. Mar Biol 134: 127−137
Walker CW, Unuma T, McGinn NA, Harrington LM, Lesser
MP (2001) Reproduction of sea urchins. In: Lawrence JM
(ed) Edible sea urchins: biology and Ecology. Elsevier
Science, Amsterdam, p 5–26
White CA, Dworjanyn SA, Nichols PD, Mos B, Dempster T
(2016) Future aquafeeds may compromise reproductive
fitness in a marine invertebrate. Mar Environ Res 122:
67−75
White CA, Bannister RJ, Dworjanyn SA, Husa V, Nichols PD,
Kutti T, Dempster T (2017) Consumption of aquaculture
waste affects the fatty acid metabolism of a benthic
invertebrate. Sci Total Environ 586: 1170−1181
Wing SR, Gibbs MT, Lamare MD (2003) Reproductive
sources and sinks within a sea urchin, Evechinus
chloroticus, population of a New Zealand fjord. Mar Ecol
Prog Ser 248: 109−123
289
Editorial responsibility: Pablo Sánchez Jerez,
Alicante, Spain
Submitted: December 12, 2017; Accepted: May 2, 2018
Proofs received from author(s): June 12, 2018