Assimilation of inorganic nutrients from salmon (Salmo salar)farming by the macroalgae (Saccharina latissima)in an exposed coastal environment: implicationsfor integrated multi-trophic aquaculture

Xinxin Wang & Ole Jacob Broch & Silje Forbord &

Aleksander Handå & Jorunn Skjermo &Kjell Inge Reitan &

Olav Vadstein & Yngvar Olsen

Received: 19 July 2013 /Revised and accepted: 5 December 2013# Springer Science+Business Media Dordrecht 2013

Abstract This paper investigated the assimilation of dis-solved inorganic nitrogen (DIN) in Saccharina latissima inproximity to salmon cages in coastal waters. The bioassayswere performed on plants from three stations located in thevicinity of a salmon farm (Salmo salar) in exposed waters atTristein (63° 52′ N, 9° 37′ E) in Central Norway. The growth,the C and N content, and the nitrogen isotope ratios (δ15N) ofS. latissimawere monitored over 1 year. The DIN concentra-tions in seawater were higher at the salmon farm stations thanat the reference station during the winter, and the N/P ratio atthe salmon farm stations was higher from September toJanuary and in June. S. latissima at the salmon farm stationsgrew faster than at the reference station. The length ofS. latissima increased by 50 % when integrated with thesalmon farm compared to the reference station. The N contentof S. latissima was positively correlated to the DIN concen-tration in seawater (p<0.05), but the increased N supply fromsalmon did not result in N accumulation in S. latissima at thesalmon farm station because of the dilution by a higher growthrate. The δ15N in S. latissima was higher at the salmon farm

station from April to June and changed in the direction of theδ15N signature in urine. This indicated that N in S. latissima atthe salmon farm station partly originated from the salmon.One hectare of S. latissimamay absorb 0.8∼1.2 t N during onegrowth season. Large-scale cultivation of S. latissima shouldbe considered to mitigate the environmental effects of DINwastes from salmon farms.

Keywords Salmon farm . DIN . Saccharina latissima .

N content . δ15N . IMTA

Introduction

Salmon and trout production in Norway has increased from0.43million t in 1999 tomore than one million t in 2010 (FAO2012). The rapid development of salmonid aquaculture hascaused some concerns on the discharge of dissolved and solidwastes to the environment (Carroll et al. 2003; Wang et al.2012; Handå et al. 2012; Kutti et al. 2007a, b; Skogen et al.2009). About 57 % of feed N and 76 % of feed P were shownto be released into the environment from a Norwegian salmonfarm (Wang et al. 2013). Moreover, some 39 % of feed N and24 % of feed P were excreted as dissolved inorganic N and P(DIN and DIP), respectively (Wang et al. 2013). These nutri-ents may enrich the water column around fish cages(Karakassis et al. 2005; Nordvarg and Johansson 2002; Pittaet al. 2005; Sanderson et al. 2008) and may result in algalblooms and, in the worst case, in coastal eutrophication(Skogen et al. 2009; Cloern 2001; Dalsgaard and Krause-Jensen 2006). Macroalgae cultured in proximity to fish farmsmay utilise the excess inorganic nutrients, at the same time

X. Wang (*) :Y. OlsenTrondheim Biological Station, Department of Biology, NorwegianUniversity of Science and Technology, 7491 Trondheim, Norwaye-mail: xinxin.wang@bio.ntnu.no

O. J. Broch : S. Forbord :A. Handå : J. SkjermoSINTEF Fisheries and Aquaculture, P.O. Box 4762, Sluppen,7465 Trondheim, Norway

K. I. ReitanDepartment of Biology, Norwegian University of Science andTechnology, 7491 Trondheim, Norway

O. VadsteinDepartment of Biotechnology, Norwegian University of Science andTechnology, 7491 Trondheim, Norway

J Appl PhycolDOI 10.1007/s10811-013-0230-1

obtain increased biomass production (Buschmann et al. 2008;Sanderson et al. 2012; Zhou et al. 2006; Troell et al. 1997).

Macroalgae take up DIN in the water column and accumu-late it in their tissues, acting as long-term integrators of nutri-ents (Cohen and Fong 2005; Costanzo et al. 2001). Bioassaysof macroalgae have been used as a bioindicator of the nutrientavailability in surrounding waters (Dalsgaard and Krause-Jensen 2006). Furthermore, analysis of δ15N in macroalgaehas been applied to determine the dispersal of aquaculturederived inorganic nutrients from both land-based and offshorefarms (Garcia-Sanz et al. 2010). Animal wastes are enriched inthe heavy nitrogen isotope (15N) compared to natural nitrogensource. The δ15N of fish farm effluents has been reported to be8∼11.3‰ (Sanderson 2006; Vizzini and Mazzola 2004),while the δ15N of NO3

− in oceanic water was typically <3‰(Montoya et al. 2002). In an N-limited system, macroalgaemay take up DIN with small or no isotope fractionation, andcan reflect the δ15N of the N sources (Marshall et al. 2007;Högberg 1997). Therefore, taking up N from fish farms canresult in an increase in δ15N of macroalgae, which may enableus to trace the dissolved wastes from fish farms.

The objective of the present study was to investigate theseasonal assimilation of DIN by Saccharina latissima in prox-imity to salmon cages located in very exposed coastal waters.The sugar kelp S. latissima is one of the fast-growing specieswith a pronounced seasonal growth and storage of nutrients inEuropean waters (Broch and Slagstad 2012; Forbord et al.2012; Sjøtun 1993). In the present study, the growth, the C andN contents and the δ15N of S. latissimawere measured during1 year to study the seasonal assimilation of DIN byS. latissima in proximity to salmon cages in exposed coastalwaters.

Materials and methods

Macroalgae bioassays were performed at a salmon farm(Salmo salar) located in an exposed coastal area at Tristein(63° 52′ N, 9° 37′ E), north of the Trondheimsfjord, Norway(Fig. 1 from August 2010 to June 2011. The salmon farmconsisted of eight polar circle plastic cages, (circumference of157 m) with 15-m-deep net pens.

Temperature and salinity were measured with a CTD (SD204, Saiv Ltd, Norway) at each sampling day. Integrated watersamples (0∼8 m) were taken using a Ramberg water collector(a 2-m-length tube sampler with automatic opening and clos-ing valves, V=4.2 L) (Olsen et al. 2011). The water wasscreened through a 200-μm mesh net to remove larger zoo-plankton and debris, and was collected in 10-L light-protectedcontainers. Subsamples (1.5∼2 L) for the measurement ofChlorophyll a (Chl a) were filtered through pre-combusted(450 °C for 4 h) Whatman GF/C filters, and samples werestored at −20 °C until further analysis. Chl a was extracted

with methanol for 2 h at 4 °C prior to the measurement ofin vitro fluorescence with a Turner Design Fluorometer(Strickland and Parsons 1972). Inorganic nutrients weremeasured in the filtrate. NH4

+ was analysed in parallel ina fluorescence detector (DFL-10) auto-analyser accordingto Kerouel and Aminot (1997), and NO3

− + NO2− and

PO43− were analysed in parallel in a fluorescence detector

(DFL-10) auto-analyser according to Hansen and Koroleff(1999).

The release rate of DIN from the salmon farm during theexperimental period was estimated using a mass balancemodel (Wang et al. 2012).

Bioassay of S. latissima

Juvenile sporophytes of S. latissima were incubated in thelaboratory in Trondheim prior to the deployment at Tristein,as described by Forbord et al. (2012). S. latissimawere incu-bated at three stations with one station on the western side(Farm West), one on the eastern side (Farm East) and onereference station 4 km south of the farm. For each station, wepositioned a rope with two big buoys on each side with a ropeline attached to a 50-kg concrete cube weight on the seabed.The sporophytes were cultivated on 1-m-long ropes (n=3)attached at 2, 5 and 8 m depths on a vertical line.

The length (L) of the blade (n=10–20) of sporophytesat 5 m depth were measured each month. The averagegrowth rate of length increase (AGRL) was calculatedby the equation:

AGRL ¼ Lt−L0ð Þ=t ð1Þ

where L0 and Lt are the length at the start and end of eachperiod, respectively, and t is the time in days. The initial lengthof the sporophytes was around 0.5 cm.

We took three plants from each station each month. Themeristematic zone (growth area, new tissue) of the plants weresampled for C, N and N isotope ratios analysis. Macroalgaesamples were dried at 60 °C to a constant weight, ground witha mortar and pestle into a fine powder and homogenised.For each sample, around 1.5 mg of dry weight was transferredto tin capsules for analysis of total C and N content inan elemental combustion system (Costech AnalyticalTechnologies Inc., USA), with two replicates for eachsample.

Sampling of salmon feed and urine

Three salmon feed samples were taken from different monthsin 2009, and three salmon feed and two urine samples weretaken in 2012 for δ15N analysis. The urine samples were takenby stripping fish.

J Appl Phycol

Nitrogen isotope ratios

Samples (around 1.2 mg) were weighed and transferred into atin capsule. The isotopic ratios of N were measured on aEurovector EA3028 element analyser coupled to an IsotopeRatio Mass Spectrometer (IRMS) at the Institute for EnergyTechnology (IFE) at Kjeller, Norway. The isotopic ratios areexpressed as values per mil (‰) according to the followingequation:

δ15N ¼ Rsample=Rstandard

� �−1

� �� 103

The standard was N2 from air. The internal laboratorystandards were IFE trout.

Statistical analysis

All data sets were tested for normality using a Kolmogorov–Smirnov test, and for the homogeneity of variance using aLevene's test. The equality of means for samples taken ondifferent days from the same depth or taken on the same dayfrom different depths was tested by one-way ANOVA follow-ed by the Tukey HSD post-hoc comparison. The equality ofmeans for samples between sampling stations was tested by ttest. Data analyses were carried out using PASW Statistics 18for Windows. Relationships between N/C ratio of S. latissimaand DIN concentration in seawater and between growth rate in

length and N/C ratio of S. latissima at the Farm West andreference stations were examined by regression. The regres-sions were carried out using SigmaPlot 10.0. The significancelimits were set at 0.05. All data are given as mean ± SE.

Results

The water temperature (2, 5 and 8 m depth) was identical atthe three stations (p>0.05), except for a <2 °C variation(p<0.05) at the end of the study (June) (Fig. 2a). The watertemperature decreased fromAugust to February and increasedsteadily until the end of the experiment, and was in the rangeof 4.2∼13 °C, with an average of 6.6±0.26 °C, and withtypical winter and summer temperatures of 4∼5 and7∼13 °C, respectively. Also, the salinity was similar at thethree stations, with a maximum deviation of <0.5‰ (Fig. 2a).The salinity increased from 27.4‰ in August to 32.4‰ inOctober, whereupon it remained stable until April, followedby a small increase in June (34.2‰).

Generally, the Chl a concentrations were low at all stationsduring the autumn–winter period, started to increase inFebruary, and peaked in May. The Chl a concentrations werein the range of 0.06∼11.6 μg L−1, with an average of 2.2±0.58 μg L−1 (Fig. 2b). The Chl a concentrations were signif-icantly higher at the salmon farm stations than at the referencestation during the May–June period (p<0.05).

Fig. 1 Geographical location of the salmon farm and the experimental stations at the western (FarmWest) and eastern (Farm East) side of the farm and atthe reference station 4 km south of the farm from August 2010 to June 2011 at Tristein in Central Norway

J Appl Phycol

The DIN concentrations increased from September toMarch at the reference station and continued to increase inApril at the salmon farm stations. Thereafter, the concentra-tion decreased rapidly. The total range was 0 to 116 μg N L−1.The DIN concentrations at the fish farm stations were higherthan at the reference station during the November–Januaryperiod (Fig. 2c). The DIP concentrations were same at allstations except for higher values at the reference station inOctober (data not shown). The DIN/DIP ratio (Fig. 2d) in theseawater was highest during later autumn and winter and wasin the range of 1.5∼9.5 mg N mg P−1 at the salmon farmstations and 0.3∼7.4 mgNmg P−1 at the reference station. TheDIN/DIP ratios at the salmon farms stations were higher thanat the reference station during the September–January periodand the month of June.

Estimation of nutrients release rate from the salmon farm

Feed use and the salmon production varied among months,with the highest values in September (Fig. 3a). Total feed use

and salmon production from August to June were 4,960 and4,420 t, respectively. The feed conversion ratio (FCR, dry feedused per wet fish produced) ranged between 0.97 in May and1.7 in August, with an average of 1.16±0.07 (Fig. 3b). TheDIN release rate from the salmon farm varied among months,and peaked in August–October and January–February. Thelowest release rate was found in December, when the feedinput and the production were both at the minimum (Fig. 3c).The total DIN released from the salmon farmwas 115 t and thetotal DIP was 11 t. The release rate of DIP followed the sametrend as that of DIN (Fig. 3d). The DIN:DIP showed minorvariation with an average of 11.0±0.03 (SE) mg N mg P−1.

Growth of S. latissima

The juvenile sporophytes showed better growth at 5 m depththan at 2 and 8m depths (data not shown) and showed a strongseasonal variation in growth (Fig. 4a) at 5 m depth. The plantsreached a maximum length of 136±5.6 cm in June at the FarmWest station and 90±4.1 cm at the reference station. From

Month

Sep Oct Nov Des Jan Feb Mar Apr May Jun

DIN

:DIP

rat

io (

mg

N m

g P-

1)

0

2

4

6

8

10

Farm EastFarm WestReference

Month

Sep Oct Nov Des Jan Feb Mar Apr May Jun

DIN

(μg

N L

-1)

0

20

40

60

80

100

120

140

Farm EastFarm WestReference

Chl

a (μg

L-1

)

0

2

4

6

8

10

12

14

Farm WestFarm EastReferenceSalinity

Temperture

Tem

pert

ure

(o C)

0

2

4

6

8

10

12

14

Salin

ity (

ppm

)

24

26

28

30

32

34

36

Farm WestFarm EastReference

a b

c d

Fig. 2 aTemperature and salinity at the western (FarmWest) and eastern(Farm East) side of the farm and at the reference station 4 km south of thefarm from August 2010 to June 2011. b Chlorophyll a (Chl a) concen-tration at the Farm West and Farm East stations and at the reference

station from September 2010 to June 2011. cDissolved inorganic nitro-gen (DIN) concentration (μg L−1) and dN/P ratio of dissolved inorganicnutrients in seawater at the Farm West and Farm East stations and at thereference station from September 2010 to June 2011

J Appl Phycol

June and throughout the summer, epiphytes covered the spo-rophytes, which resulted in tissue losses and a decrease inlength. The plants at the salmon farm stations were longer thanthe plants at the reference station, and this difference wassignificant during the entire year (p<0.05). S. latissimashowed slow growth in length from August to March andrapid growth fromMarch to June (Fig. 4b). The growth rate ofS. latissima was at a maximum (0.99∼1.72 cm day−1) duringApril–May and negative during June–August. The growthrate at the salmon farm stations were higher than at thereference station, except for the periods of October–Januaryand February–March.

Carbon and nitrogen content of S. latissima

The C content (mg C g DW−1) of S. latissima showed amoderate seasonal variation during the experimental period(Fig. 5a). The C content at the Farm West and referencestations varied in the range of 213∼285 mg C g DW−1, with

an average value of 251±4.6 mg C g DW−1. The valuesshowed a significant difference between sampling days at bothstations (p<0.05), but no significant difference was foundbetween stations (p>0.05).

The N content (mg N g DW−1) of S. latissima was moredynamic than the C content, and there was a decrease from thestart of October to November at both stations, and thereafter asteady increase until April before it decreased again until theend of the experiment in June (Fig. 5b). The N content ofS. latissima varied in the range of 16∼38 mg N g DW−1, withan average value of 25±1.5 mg N g DW−1. Significant differ-ences were observed between sampling days at both stations,but no significant differences were found between stations(p>0.05), except for the higher N content at the referencestation in April (p<0.05).

The N/C ratio (Fig. 5c) of S. latissimawas in the range of65 to 156 mg N g C−1, with an average value of 102±6.5 mg N g C−1. The N/C ratio of S. latissima followed thesame pattern of variation as the N content per dry weight at

FCR

(dw

ww

-1)

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

(to

nne

mon

th-1

)

0

100

200

300

400

500

600

700a b

c d

Feed use Production

Month

Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun

DIN

rel

ease

rat

e (k

g N

mon

th-1

)

0

2000

4000

6000

8000

10000

12000

14000

16000

Month

Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun

DIP

rel

ease

rat

e (k

g P

mon

th-1

)

0

200

400

600

800

1000

1200

1400

Fig. 3 a Feed use (tonne month−1) and salmon production (tonnemonth−1), bFeed conversion ratio (FCR), cDIN release rate (kg month−1)and dDissolved inorganic phosphorus (DIP) release rate from the salmon

farm from August 2010 to June 2011 at Tristein in Central Norway. FCRwas expressed in terms of dry feed used per wet fish produced

J Appl Phycol

both stations. There were significant differences among sam-pling days (p<0.05), but no significant difference betweenstations (p>0.05). The N/P ratio of S. latissima varied duringthe experimental period and varied between 3.2 and12 mg N mg P−1 over the stations, with average valueof 7.3±0.60 mg N mg P−1 (Fig. 5d).

The N/C ratio of S. latissima was positively correlat-ed (r2=0.39, p<0.05) to the DIN concentration in sea-water at both stations (Fig. 6a). Moreover, the growth rate ofS. latissima increased significantly (r2=0.58, p<0.05) withN/C in S. latissima (Fig. 6b).

δ15N ratios of S. latissima, feed and urine

The average δ15N of salmon feed samples from 2009 to 2012were 9.46±0.22 and 4.51±0.24‰, respectively. The averageδ15N of urine from 2012 was 5.73±0.12‰ and was isotopi-cally heavier than feed with an enrichment of 1.2‰, thus theaverage δ15N of urine from 2009 was estimated to be around10.7±0.37‰.

The δ15N of S. latissima (Fig. 7) ranged from about −4.8 to3.3‰ with negative values in March and April. The highestδ15N was observed in June at the salmon farm station and inNovember at the reference station. The δ15N in S. latissimawas higher at the salmon farm station than at the referencestation from April to June and lower from October to March.No significant difference was found between the stations inNovember, March and May (p>0.05). No correlation wasfound between δ15N of S. latissima and the DIN concentra-tions at the stations (p<0.05).

Discussion

Effects of salmon farming on nutrients

The salmon farm released 115 t of DIN from August to June,translating to a DIN loading rate of 380 kg N day−1. The watermasses to which these nutrients drained into depend on hy-drodynamics and the depth of the site. The current speed of thesurface water in the fish farm was generally between 0 and20 cm s−1 (Handå et al. 2012). The size of the farm was 200×500 m2, with a depth of 15 m. If we assume that water entersthe farm in a plug-flow pattern with no dilution downstreamand an average current speed of 10 cm s−1, the total volumepassing the farm would be 4.5×107 m3 day−1. Consequently,the expected increase of DIN concentration as a result of thedischarge from the salmon would be 8.4 mg N m−3. Thisapproach certainly underestimated the real volume of thereceiving waters, because the nutrients were continuous-ly diluted by other water masses downstream. Thus, theabove expected increase of DIN concentration was anoverestimate and we can not expect a big change in theDIN concentrations.

The DIN concentration was 2∼31 μg N L−1 higher at thesalmon farm stations from September to January due to theDIN supply from salmon. The DIN concentration at the salm-on farm stations was lower from February to April. This maybecause the DIN released from salmon cages was taken up byphytoplankton (Dalsgaard and Krause-Jensen 2006). Thisagree with our findings that the Chl a concentrations at thesalmon farm stations was higher than at the reference stationfrom May to June (Fig. 2b). Another possible explanation isthat DIN released from sea cages were quickly diluted in alarge water volume and transported away from the cages fromFebruary to June (Troell et al. 1997). The N/P ratio at thesalmon farm stations was higher than at the reference stationfrom September to January and also in June. This was prob-ably because the N/P ratio of dissolved inorganic nutrientsreleased from salmon was higher than the natural supply, andthe N/P ratio at the salmon farm stations was a mixture of N/Pfrom salmon and the natural supply.

Aug-SepSep-Oct

Oct-DecDec-Jan

Jan-FebFeb-Mar

Mar-Apr

Apr-MayMay-Jun

Jun-Aug

Gro

wth

in le

ngth

(cm

d -

1 )

-1.0

-0.5

0.0

0.5

1.0

1.5

2.0

Farm West (5 m)Farm East (5 m)Reference (5 m)

Aug-10Sep Oct Dec Jan Feb Mar Apr May Jun Jul Aug-11

Len

gth

of la

min

a (c

m)

0

20

40

60

80

100

120

140

160

Farm West (5 m)Farm East (5 m)Reference (5 m)

Fig. 4 a Length (cm) and b the average growth rate of length increase(AGRL) of S. latissima at 5 m depth at the Farm West, Farm East andreference stations from August 2010 to August 2011. Mean±SE

J Appl Phycol

Seasonal growth of S. latissima

Our results showed a seasonal variation of length growth ratein S. latissima, with the maximum length growth rate seenwhen the DIN concentration in the seawater and N/C ratio inalgae tissues were high, after which the growth rate decreased,accompanied by a depleted DIN concentration. This suggeststhat S. latissimawas nutrient-limited in the present study.

Our results revealed an overall higher growth rate ofS. latissima at the salmon farm stations than at the referencestation, in agreement with previous findings that macroalgaecultured in close proximity to fish cages in open watersexhibited higher growth rates than at the control site (Reidet al. 2009; Troell et al. 1997, 1999; Sanderson et al. 2012;Zhou et al. 2006). The growth rate of S. latissima is reported tobe controlled by temperature, irradiance and nitrogen avail-ability (Fortes and Lüning 1980). Our results showed that thewater temperature and the salinity were same between sta-tions. There were also no differences in light between stations(Handå et al. 2013). Consequently, differences in growth and

N-metabolism of S. latissima found between stations cannotbe explained by temperature, salinity and light effects.Therefore, the higher growth rate of S. latissima at the salmonfarm stations from September to January was due to theincreased DIN from the salmon cages. However, S. latissimastill grew faster at the salmon farm stations during February–April when the DIN concentration was lower than referencestation. This may because macroalgae can assimilate DIN inthe water column and store it for later growth, acting as long-term integrators of nutrients (Cohen and Fong 2005; Costanzoet al. 2001). This makes it possible for macroalgae to benefitfrom nutrients released in pulses from the cages (Troell et al.1997). Thus, the higher growth rate of S. latissima at thesalmon farm stations during February-April could also bedue to the supply of DIN from the salmon cages.

The nitrogen content of S. latissima

Macroalgae have been reported to have higher N contents inclose proximity to the fish farm compared to control sites

C (

mg

g D

W-1

)

0

50

100

150

200

250

300

350a b

c d

Farm West

Reference

N (

mg

g D

W-1

)

0

10

20

30

40

50

Farm West

Reference

Month

Oct Nov Dec Jan Feb Mar Apr May Jun

N:C

rat

io (

mg

N g

C-1

)

0

20

40

60

80

100

120

140

160

180

Farm West

Reference

Month

Oct Nov Dec Jan Feb Mar Apr May Jun

N:P

rat

io (

mg

N m

g P-

1 )

0

5

10

15

20Farm West

Reference

Fig. 5 aCarbon content (mg g DW−1), bnitrogen content (mg g DW−1), cN/C ratio (mgN g C−1) and dN/P ratio (mgNmg P−1) of S. latissima from 2 to∼8 m depth at the Farm West and reference station from August 2010 to June 2011. Mean ± SE

J Appl Phycol

(Sanderson et al. 2012; Troell et al. 1999; Chopin et al. 1999,2007). In our study, the increased N-supply from the salmondid not result in N accumulation in S. latissima at the salmonfarm station due to the dilution by higher growth rate com-pared to the reference station, which is in agreement with(Fong et al. 2004). This suggests that S. latissima was N-limited during the production season. The N/C ratio ofS. latissima was positively correlated to the DIN, confirmingthat the N content in macroalgae was an indicator of biolog-ically available nutrients in the seawater (Sanderson et al.2012; Lin and Fong 2008). However, the N content ofS. latissima was not effective at tracing DIN released fromfish farms when the plants were N-limited. The growth rate ofS. latissima was positively correlated to the N/C ratio inS. latissima. This may suggest that macroalgae likeS. latissima respond in a similar way, as was found formicroalgae. The Droop model for nutrient-limited growth asa function of the internal concentration of the limiting nutrientis also suitable for macroalgae (Droop 1973).

The nitrogen isotope of S. latissima and urine

From October to March, S. latissima showed strong negativeδ15N values that were lower at the salmon farm station than atthe reference station. This may result from the fractionationand preferential uptake of 14N by the plants when the DINconcentration was high (Högberg 1997; Deutsch and Voss2006). Nitrogen fractionation has also been reported in terres-trial plants and phytoplankton (Högberg 1997; Marshall et al.2007). However, the concentration of NO3

− or NH4+ in the

previous studies was reported to be between 1.4 and 12 mM,which was much higher than the DIN concentration in ourstudy (Yoneyama et al. 1991; Högberg 1997).

From May to June, DIN was depleted and S. latissimatherefore took up both 14N and 15N, in accordance withHögberg (1997). The δ15N values of S. latissimawere higherat the salmon farm station than at the reference station fromApril to June. The δ15N of S. latissima changed in the direc-tion of the signature in urine (5.73∼10.7‰) (Fig. 7). Thisindicates that the origin of N in S. latissima was partly fromthe salmon farm. Our results thus confirmed that δ15N inmacroalgae reflected the δ15N of the sources of N underDIN depletion. Therefore, it is suitable only in caseswith nutrient depletion, because the δ15N in macroalgaeis a function of both the δ15N of the sources and thefractionation of N isotopes when N is excess (Marshallet al. 2007; Högberg 1997).

Implications for integration with salmon farm

Our results showed a successful integration of S. latissimawith salmon farming. The increased DIN supply from thesalmon farm resulted in better growth of S. latissima and the

DIN (mg l-1)

0 20 40 60 80 100 120 140

N:C

(m

g N

g C

-1)

0

20

40

60

80

100

120

140

160

180a

b

Farm WestReference

N:C (mg N g C-1)

0 20 40 60 80 100 120 140 160 180

Gro

wth

in le

ngth

(cm

day

-1)

-0.5

0.0

0.5

1.0

1.5

2.0

Farm WestReferences

Fig. 6 aRelationship between N/C ratio (mg N g C−1) of S. latissima andDIN concentration (μg L−1) in seawater (0∼8 m) and b relationshipbetween growth rate in length and N/C ratio of S. latissima at the FarmWest and reference stations

Oct Nov Mar Apr May Jun Salmon urine

δ15N

-6

-4

-2

0

2

4

6

8

10

12

Farm West ReferenceSalmon urine

Fig. 7 Nitrogen isotopic ratio (‰, δ 15N) of salmon urine and S. latissimafrom 2 to ∼8 m depth at the Farm West and reference stations fromOctober to November 2010 and March to June 2011. Mean ± SE

J Appl Phycol

length of S. latissima increased by 50 % when integrated withthe salmon farm compared to the reference station. The frondarea of S. latissima can be approximately calculated as 0.75×length×width (Broch et al. 2013). The frond area of individualplants at farm west station was 0.75×136×30=3,060 cm2.The frond area at reference station was 0.75×90×19=1,283 cm2. We assumed the weight per area of S. latissimawere same at the salmon farm and reference stations. Thus, thebiomass of individual plants at the reference station would be60 % lower than the plants at the salmon farm station after11 months of cultivation.

For the large-scale cultivation of S. latissima for one grow-ing season from August to June, a harvest of 220∼340 t wetweight ha−1 of S. latissima at the salmon farm station ispossible (Sanderson et al. 2012). If we assume a wet-to-dryweight ratio of 9:1, a hectare of S. latissima may absorb0.8∼1.2 t N. Large-scale cultivation of S. latissima integratedwith salmon farms should be considered to mitigate the envi-ronmental effects of DIN discharged from the salmon farm.

Acknowledgements This work was a part of the Research Council ofNorway project no. 199391/I10 (MACROBIOMASS). We are grateful tothe Norwegian Research Council and the China Scholarship Council forfinancial support. Thanks also to AquaCulture Engineering (ACE) forkindly providing research facilities and data on monthly feed use and fishproduction and to the Institute for Energy Technology (IFE) for stableisotope analysis.

References

Broch OJ, Slagstad D (2012) Modelling seasonal growth and composi-tion of the kelp Saccharina latissima. J Appl Phycol 24:759–776

Broch O, Ellingsen I, Forbord S, Wang X, Volent Z, Alver M, Handå A,Andresen K, Slagstad D, Reitan K, Olsen Y, Skjermo J (2013)Modelling the cultivation and bioremediation potential of the kelpSaccharina latissima in close proximity to an exposed salmon farmin Norway. Aquac Environ Interact 4:187–206

Buschmann A, Varela D, Hernandez-Gonzalez M, Huovinen P (2008)Opportunities and challenges for the development of an integratedseaweed-based aquaculture activity in Chile: determining the phys-iological capabilities of Macrocystis and Gracilaria as biofilters. JAppl Phycol 20:571–577

Carroll ML, Cochrane S, Fieler R, Velvin R, White P (2003) Organicenrichment of sediments from salmon farming in Norway: environ-mental factors, management practices, and monitoring techniques.Aquaculture 226:165–180

Chopin T, Yarish C, Wilkes R, Belyea E, Lu S, Mathieson A (1999)Developing Porphyra/salmon integrated aquaculture for bioremedi-ation and diversification of the aquaculture industry. J Appl Phycol11:463–472

Chopin T, Yarish C, Sharp G (2007) Beyond the monospecific approachto animal aquaculture—the light of integrated multi-trophic aqua-culture. In: Bert TM (ed) Ecological and genetic implications ofaquaculture activities. Springer, Dordrecht, pp 447–458

Cloern JE (2001) Our evolving conceptual model of the coastal eutrophi-cation problem. Mar Ecol Prog Ser 210:223–253

Cohen RA, Fong P (2005) Experimental evidence supports the use ofδ15N content of the opportunistic green macroalga Enteromorpha

intestinalis (Chlorophyta) to determine nitrogen sources to estuaries.J Phycol 41:287–293

Costanzo SD, O'DonohueMJ, DennisonWC, LoneraganNR, ThomasM(2001) A new approach for detecting and mapping sewage impacts.Mar Pollut Bull 42:149–156

Dalsgaard T, Krause-Jensen D (2006) Monitoring nutrient release fromfish farms with macroalgal and phytoplankton bioassays.Aquaculture 256:302–310

Deutsch B, VossM (2006) Anthropogenic nitrogen input traced bymeansof δ15N values in macroalgae: results from in-situ incubation exper-iments. Sci Total Environ 366:799–808

DroopMR (1973) Some thoughts on nutrient limitation in algae. J Phycol9:264–272

FAO (2012) World review of fisheries and aquaculture. The State ofWorld Fisheries and Aquaculture 2012. Food and AgriculturalOrganization of the United Nations, Rome

Fong P, Fong JJ, Fong CR (2004) Growth, nutrient storage, and release ofdissolved organic nitrogen by Enteromorpha intestinalis in responseto pulses of nitrogen and phosphorus. Aquat Bot 78:83–95

Forbord S, Skjermo J, Arff J, Handa A, Reitan KI, Bjerregaard R, LüningK (2012) Development of Saccharina latissima (Phaeophyceae)kelp hatcheries with year-round production of zoospores and juve-nile sporophytes on culture ropes for kelp aquaculture. J ApplPhycol 24:393–399

Fortes MD, Lüning K (1980) Growth rates of North Sea macroalgae inrelation to temperature, irradiance and photoperiod. HelgolMeeresunters 34:15–29

Garcia-Sanz T, Ruiz-Fernandez JM, Ruiz M, Garcia R, Gonzalez MN,Perez M (2010) An evaluation of a macroalgal bioassay tool forassessing the spatial extent of nutrient release from offshore fishfarms. Mar Environ Res 70:189–200

Handå A, Min H, Wang X, Broch OJ, Reitan KI, Reinertsen H, Olsen Y(2012) Incorporation of fish feed and growth of blue mussels(Mytilus edulis) in close proximity to salmon (Salmo salar) aqua-culture: Implications for integrated multi-trophic aquaculture inNorwegian coastal waters. Aquaculture 356–357:328–341

Handå A, Forbord S, Wang X, Broch OJ, Dahle S, Størseth TR, ReitanKI, Olsen Y, Skjermo J (2013) Seasonal- and depth-dependentgrowth of cultivated kelp (Saccharina latissima) in close proximityto salmon (Salmo salar) aquaculture in Norway. Aquaculture 414–415:191–201

Hansen HP, Koroleff F (1999) Determination of nutrients. In: GrasshoffK, Kremling K, Ehrhardt M (eds) Methods of seawater analysis, 3rdedn. Wiley, New York, pp. 159–228

Högberg P (1997) 15N natural abundance in soil–plant systems. NewPhytol 137:179–203

Karakassis I, Pitta P, KromMD (2005) Contribution of fish farming to thenutrient loading of the Mediterranean. Sci Mar 69:313–321

Kerouel R, Aminot A (1997) Fluorometric determination of ammonia insea and estuarine waters by direct segmented flow analysis. MarChem 57:265–275

Kutti T, Ervik A, Hansen PK (2007a) Effects of organic effluents from asalmon farm on a fjord system. I. Vertical export and dispersalprocesses. Aquaculture 262:367–381

Kutti T, Hansen PK, Ervik A, Hoisaeter T, Johannessen P (2007b) Effectsof organic effluents from a salmon farm on a fjord system. II.Temporal and spatial patterns in infauna community composition.Aquaculture 262:355–366

Lin DT, Fong P (2008) Macroalgal bioindicators (growth, tissue N, δ15N)detect nutrient enrichment from shrimp farm effluent enteringOpunohu Bay, Moorea, French Polynesia. Mar Pollut Bull 56:245–249

Marshall JD, Brooks JR, Lajtha K (2007) Sources of variation in thestable isotopic composition of plants. In:Michener R, Lajtha K (eds)Stable isotopes in ecology and environmental science, 2nd edn.Blackwell, London, pp 22–60

J Appl Phycol

Montoya JP, Carpenter EJ, Capone DG (2002) Nitrogen fixation andnitrogen isotope abundances in zooplankton of the oligotrophicNorth Atlantic. Limnol Oceanogr 47:1617–1628

Nordvarg L, Johansson T (2002) The effects of fish farm effluents on thewater quality in the Åland archipelago, Baltic Sea. Aquacult Eng 25:253–279

Olsen Y, Andersen T, Gismervik I, Vadstein O (2011) Marine heterotro-phic bacteria, protozoan and metazoan zooplankton may experienceprotein N or mineral P limitation in coastal waters. Mar Ecol ProgSer 436:81–100

Pitta P, Apostolaki ET, Giannoulaki M, Karakassis I (2005) Mesoscalechanges in the water column in response to fish farming zones inthree coastal areas in the Eastern Mediterranean Sea. Estuar CoastShelf Sci 65:501–512

Reid GK, Liutkus M, Robinson SMC, Chopin TR, Blair T, Lander T,Mullen J, Page F, Moccia RD (2009) A review of the biophysicalproperties of salmonid faeces: implications for aquaculture wastedispersal models and integrated multi-trophic aquaculture. AquacultRes 40:257–273

Sanderson JC (2006) Reducing the environmental impact of seacage fishfarming through cultivation of seaweed. The Open University, UK

Sanderson JC, Cromey CJ, Dring MJ, Kelly MS (2008) Distribution ofnutrients for seaweed cultivation around salmon cages at farm sitesin north-west Scotland. Aquaculture 278:60–68

Sanderson JC, Dring MJ, Davidson K, Kelly MS (2012) Culture, yieldand bioremediation potential of Palmaria palmata (Linnaeus)Weber & Mohr and Saccharina latissima (Linnaeus) C.E. Lane, C.Mayes, Druehl & G.W. Saunders adjacent to fish farm cages innorthwest Scotland. Aquaculture 354–355:128–135

Sjøtun K (1993) Seasonal lamina growth in two age groups of Laminariasaccharina (L.) Lamour in Western Norway. Bot Mar 36:433–441

Skogen MD, Eknes M, Asplin LC, Sandvik AD (2009) Modelling theenvironmental effects of fish farming in a Norwegian fjord.Aquaculture 298:70–75

Strickland JDH, Parsons TR (1972) A practical handbook of seawateranalysis. Bull Fish Res Board Canada 167:185–206

Troell M, Halling C, Nilsson A, Buschmann AH, Kautsky N, Kautsky L(1997) Integrated marine cultivation of Gracilaria chilensis(Gracilariales, Rhodophyta) and salmon cages for reduced environ-mental impact and increased economic output. Aquaculture 156:45–61

Troell M, Ronnback P, Halling C, Kautsky N, Buschmann A (1999)Ecological engineering in aquaculture: use of seaweeds for remov-ing nutrients from intensive mariculture. J Appl Phycol 11:89–97

Vizzini S, Mazzola A (2004) Stable isotope evidence for the environ-mental impact of a land-based fish farm in the westernMediterranean. Mar Pollut Bull 49:61–70

Wang X, Olsen LM, Reitan KI, Olsen Y (2012) Discharge of nutrientwastes from salmon farms: environmental effects, and potential forintegrated multi-trophic aquaculture. Aquac Environ Interact 2:267–283

Wang X, Andresen K, Handå A, Jensen B, Reitan K, Olsen Y (2013)Chemical composition and release rate of waste discharge from anAtlantic salmon farm with an evaluation of IMTA feasibility. AquacEnviron Interact 4:147–162

Yoneyama T, Omata T, Nakata S, Yazaki J (1991) Fractionation ofnitrogen isotopes during the uptake and assimilation of ammoniaby plants. Plant Cell Physiol 32:1211–1217

Zhou Y, Yang H, Hu H, Liu Y, Mao Y, Zhou H, Xu X, Zhang F (2006)Bioremediation potential of the macroalgaGracilaria lemaneiformis(Rhodophyta) integrated into fed fish culture in coastal waters ofnorth China. Aquaculture 252:264–276

J Appl Phycol