Atmospheric Freeze-Drying of Calanus finmarchicus and its effects on proteolytic and lipolytic activities

4th Nordic Drying Conference, Reykjavik, Iceland, June 17th to 19th 2009





Atmospheric Freeze-Drying of Calanus finmarchicus and its effects on proteolytic
and lipolytic activities




Michael Bantle1, Trygve M. Eikevik1 and Turid Rustad2

1. Norwegian University of Science and Technology, Department of Energy and Process
Engineering, NO-7491 Trondheim, Norway
2. Norwegian University of Science and Technology, Department of Biotechnology, NO-
7491 Trondheim, Norway
Email: Michael.Bantle@ntnu.no



Keywords: Zooplankton, Atmospheric Freeze Drying, Protease, Lipase, Fish Feed

ABSTRACT

Combining atmospheric freeze-drying (AFD) with a properly designed heat pump
system (HP) results in an energy-efficient drying process. This combination
makes HP-AFD competitive with vacuum freeze-drying (VFD). The advantages
of AFD compared to VFD are lower production costs and the ability to
manufacture a similar porous product. The zooplankton species Calanus
finmarchicus (CF) is attractive for use as aquaculture feed due to its high content
of valuable marine lipids and proteins. Atmospheric freeze-drying in a fluidized
bed has been evaluated as preservation method for CF. Three different trials of
drying techniques were conducted with CF. The drying temperatures were -5ºC, -
10ºC and a combination of -5ºC/+20ºC (initial -5ºC for 4 hours and final drying at
+20ºC). This paper evaluates the effects of drying conditions on proteolytic and
lipolytic activities, and describes particle size distribution and sorption isotherms.
Proteolytic activity was reduced for CF dried at a combined temperature of AFD-
5ºC/AD+20ºC, while no significant change occurred in CF that was only freeze-
dried. All dried CF showed a minor decrease in lipolytic activity. Drying rates for
AFD were significantly higher than for VFD.
ISBN 978-82-594-3406-7

---

4th Nordic Drying Conference, Reykjavik, Iceland, June 17th to 19th 2009

Introduction

With an annual production of more than 700 000 tons of farmed fish in 2006, Norway ranks as the ninth-
largest aquaculture nation in the world (SSB 2008). Salmon is the main species produced by Norwegian
aquaculture (SSB, 2008). In 2006 the aquaculture sector had a turnover of 2.7 billion USD, with double-
digit increases in annual production and turnover rates, especially for salmon and rainbow trout (SSB,
2008). The Royal Norwegian Society of Sciences and Letters (DKNVS) and the Norwegian Academy of
Technological Sciences (NTVA), (2006 and 1999) have described developing and exploiting biomarine
resources in the aquaculture sector as a potential value-added industry that can serve as an alternative to
the country’s limited natural oil and gas resources. The expectation is that by 2030, the added value from
the Norwegian marine industry will be on the order of the size of the oil industry. With its access to clean
and protected inshore/near shore waters, its sizable area for growing marine plant biomass and its healthy
climate for the production of high-grade seafood, Norway has a natural advantage in the marine industry
(DKNVS and NTVA, 2006). It is expected that the traditional capture fisheries industry will not grow
significantly in coming decades; instead, the real growth will be fish farming and aquaculture, including
the cultivation of new species such as shellfish and algae, biochemical/energy carriers and feed from
highly productive ocean areas. The exploitation of new raw materials for fish feed and feed formulation is
one main requirement for a sustainable aquaculture industry. DKNVS and NTVA (2006) identified feed
from highly productive ocean areas as a new market and predicted an added value of NOK 42.5 billion
(around 6 billion USD) by 2030. Their report states further that enormous unexploited resources can be
found that ought to be considered for high quality raw materials for fish feed, including small, hard-to-
catch or simply unattractive fish and zooplankton.
The zooplankton species Calanus finmarchicus (CF) is considered a possible new species for fish feed
(Anonymous, 2007). Their lipid composition includes a considerable amount of omega-3 polyunsaturated
fatty acids (PUFA) and rare phospholipids (Kattner, 1989). This makes CF interesting not only for
aquaculture feed, but also for biomedical purposes and human nutrition. Currently, CF is caught in
industrial batches using a fine mesh net similar to that used in trawling for krill. CF is then frozen and
sold without further processing, or is subjected to a separation process that isolates the valuable oil first.
CF is not currently used for the industrial production of fish feed, but several researchers are investigating
this possibility. Olsen et al. (2004) demonstrated that oil from CF can substitute for fish oil in the diet of
farmed Atlantic salmon, thus providing long chain omega-3 PUFA. Solgaar et al. (2007) showed high
proteolytic activity in CF and identified the high degree of post mortem degradation and a subsequent
leaching of highly valuable nutrients as possible problems for the use of CF as feed ingredient. Thus a
preservation method is needed to stabilize the raw material.
Drying and frozen storage are considered to be suitable preservation methods for a large number of food
and feed products. The base for the formulation of fish feed is mainly dry matter, such as fish meal.
Additives, for example fish oil, are mixed into the base with a binding agent, and this material is then
extruded as the fish feed pellet. This means that the initial high water content in CF must be reduced if it
is to be used as a base for pellets, and a drying process for CF will be needed in the process line. CF
should be dried until its water content is low enough to limit water activity. In general, water activity
between 0.2 and 0.6 results in reduced lipid oxidation, while water activities below 0.8 respectively 0.6
are recommended to ensure reduced enzymatic activity, and to limit mould and bacteria growth (Shahidi,
2007).
We examined a freeze-drying process for CF as a suitable drying process because CF is sensitive to
degradation. While vacuum freeze-drying (VFD) is considered costly (Mujumdar, 2007, chapter 11.5) an
atmospheric freeze-drying (AFD) process as described by Claussen et al. (2007a) is more economical and
gives product qualities similar to VFD (Donsi et al., 2001). Cost is an important consideration because
fish feed must be relatively low cost. Song (1990) developed an AFD process in a fluidized bed that
ISBN 978-82-594-3406-7

---

4th Nordic Drying Conference, Reykjavik, Iceland, June 17th to 19th 2009
involved a specific protocol called a temperature programme. In this procedure the product is first dried
with temperatures below its initial freezing point. As soon as the outer shell of the product is stable the
temperature of the drying air is increased above the freezing temperature of the product. This results in an
accelerated drying process, while the particle structure of the final product is the same as the structure of a
product that has been exclusively freeze-dried. The energy efficiency of AFD is significantly increased by
circulating and conditioning the drying air with a heat pump (Claussen et al., 2007b). Current efforts
include increasing the drying rate further by combining AFD with new technologies such as ultrasonic,
microwave, infrared radiation and pulse fluidization (Alves-Filho and Eikevik, 2008).
This paper investigates how CF can be dried in different AFD processes and the effects on the proteolytic
and lipolytic activity with regard to the use of CF as fish feed. The sorption isotherms at different
temperatures and the particle size distribution for the product were also measured. For the purpose of
comparison, the drying rate for CF in VFD was determined.
Materials and Methods

Calanus finmarchicus (CF):
Batches of CF were harvested between the islands of Hitra and Frøya (GPS: 63º30N, 9º55E) outside of
Trondheim Fjord, Norway at the end of April 2008. Industrial batches of CF can only be harvested with
current technologies during the spring bloom, when their concentration on the surface of the ocean is
high. For harvesting, a special surface trawl with a 500µm net was used. The catch consisted mostly of
CF at stages four and five, which is in
accordance with Tokle (2006). Batches of
CF were drained for 15 minutes, vacuumed
and frozen in a plate freezer (thickness
5cm). Storage on board was on dry ice
(around -80ºC), while on land the batches
were stored in a cryogenic freezer at -80ºC.
Twelve hours before AFD, the frozen CF
plates were placed in a freezer to be
temperature conditioned to the initial
temperature of the AFD. Therefore the CF
temperature and the drying temperature
were equal. The Calanus plates were
crushed into small particles inside the
freezer with a commercial meat grinder
two hours before drying.

Atmospheric Freeze-Drying (AFD):
The AFD system consisted of a cylindrical
drying chamber (diameter 20cm, height
35cm). Conditioned air was supplied from
the bottom through a 200µm meshed net

Figure 1: Heat pump assisted atmospheric freeze drying.
which was stretched over a supporting
structure. The drying chamber was closed
with a barred lid that was also spanned by a 200µm meshed net. The drying air was circulated and
conditioned with a heat pump (HP) system (Figure 1). The air humidity and temperature was decreased
by contact with cold surfaces of the HP evaporator. The dehumidified air was then reheated in the HP
condenser.

ISBN 978-82-594-3406-7

---

4th Nordic Drying Conference, Reykjavik, Iceland, June 17th to 19th 2009
An operating unit controlled temperatures in the HP system and the additional external condenser of the
drying system, which made it possible to adjust the humidity and temperature of the drying air while
recovering energy from the drying system. A ventilator controlled the volume flow of the drying air.
Therefore it was possible to have a fixed or stationary and fluidized particle bed in the drying chamber. A
fluidized bed requires a high volume flow of drying air but results in a fast drying process, while drying
in a fixed bed is slower but needs a reduced volume flow. Three different drying tests were performed in
the fluidized bed state. Calanus was dried at -10ºC and -5ºC only. A third test was conducted with an
initial drying temperature of -5 ºC for 4 hours and then increased to the final drying temperature of 20ºC
after 4 hours until the end of the drying test. This combined AFD-5ºC/AD+20ºC drying method was
recommended by Song (1990) and combines the advantages of the porous particle structure of the dried
product from AFD with the fast drying rate of conventional atmospheric drying (AD) above the freezing
point of the product. For all drying tests the temperature of the heat pump condenser was set to -25ºC.
This gave an average relative humidity in the drying air of 17.9% at -10ºC, 17.3% at -5ºC and 2.1% at
20ºC. The average approach velocity of the drying air was 1.9 m/s for all drying tests.

Vacuum Freeze-Drying (VFD):
The vacuum freeze-dryer model “Alpha 1-4 LD” from the company Martin Christ
Gefriertrocknungsanlagen, Osterode am Harz, Germany was used for comparative tests with an ice
condenser temperature of -60ºC and 0.01mbar pressure. Crushed CF was placed in a neck filter bottle and
connected to the vacuum chamber. Weight reduction was determined manually, by disconnecting and
weighing the bottle. The VFD apparatus was placed in the laboratory at ambient conditions (?20ºC), so
the heat for sublimation was conducted through the sample bottle.

Determination of proteolytic and lipolytic activity:
Samples of CF were homogenized for 20s in distilled water using an Ultra Turrax, stirred for 10min, and
centrifuged at 10400g for 20min at 4oC. The amount of water was double the amount of CF for fresh
material while it was 10 times the amount of CF for dried material. The supernatants were filtered
through glass wool and the volume of the extract determined. Protein concentrations in the extracts were
determined according to Lowry et al. (1951) with bovine serum albumin (Sigma A9647) as the standard.
General proteolytic activity was determined as described by Barret and Heath (1977) with minor
modifications. The incubation mixture consisted of 1.2mL of phosphate-citrate buffer (McIlvaine, 1921)
and 0.4mL of substrate (bovine hemoglobin Sigma H-2625, 1%). 2mL of 5% w/v TCA (trichloroacetic
acid, Merck) was also added to the zero sample. The samples were pre-incubated for 10 min in a water
bath at 30oC before 0.4mL suitably diluted enzyme extract was added. Incubation time was 1 hour. The
reaction was stopped by the addition of 2mL of 5% TCA. The samples were cooled for 30min and then
filtered before the amount of short peptides was determined according to Lowry et al. (1951). Activities
were expressed as microgram peptides (cut protein) per gram dry CF per minute. Activity was determined
at pH 7 and at 30ºC.
Lipolytic activity was determined by spectrofluorimetry according to the method described by Roberts
(1985) and later by Izquierdo and Henderson (1998) with minor modifications. The non-fluorescent
substrate 4-methyliumbelliferyl heptanoate was solubilized in a liposomal dispersion of soya lecithin.
This substrate was hydrolysed to heptanoic acid and the highly fluorescent compound, 4-methylium
belliferone through the action of lipases. Mixtures of 20?L substrate and 40?L of enzyme solutions,
suitably diluted in phosphate-citrate buffer (McIlvaine, 1921) at pH 7, were incubated for 15min in water
baths at 40ºC. Reactions were stopped by adding 3mL of cold 1mol/l Tris HCl, pH 7.5. For zero time
samples, extracts were incubated in water baths at 80ºC for 30 min, centrifuged at 420g for 10min, and
diluted and measured as the respective sample. Increase in emission at 450nm (excitation 365nm) was
measured using a Perkin Elmer 3000 spectrofluorometer. Activities were expressed as an increase in
fluorescence and given in arbitrary units (U) based on the mean of three measurements.

ISBN 978-82-594-3406-7

---

4th Nordic Drying Conference, Reykjavik, Iceland, June 17th to 19th 2009
Sorption Isotherm:
Sorption Isotherms for CF were determined with the water sorption analyser “CISORP” (C.I. Electronics
Ltd., Salisbury, UK). The relationship between the CF water content and water activity (aw) was recorded
when reaching equilibrium at constant temperature. The desorption characteristic of fresh CF was
measured first, followed by the adsorption characteristic (rehydration) of the same sample. The
temperatures were 10ºC and 30ºC. No commercial instrumentation is available to measure sorption
characteristics below 0ºC.

Particle Size Distribution (PSD):
The PSD of fresh CF was measured with a standard sieve analyser using frozen CF samples that were
ground inside the freeze storage room. The PSD of the dried CF was measured with the same analyser,
but at room temperature (?20ºC).
Results

Drying curves for the different AFD tests can be
seen in Figure 2. The drying rate was reduced
with temperature. During the first 3-5 hours the
bulk of CF was in a fixed bed state with channel
building rather than in a fluidized bed condition.
As the water content was reduced to around 65%,
the bed turned into a more fluidized state, which
increased mass exchange and the drying rate. At
water content of 20% to 25% the drying rate was
reduced again. This marked the beginning of the
second drying stage. After 24 hours the drying
process was stopped. The final moisture content

of the product dried at -10ºC and dried at -5ºC
Figure 2: Drying curve for Calanus finmarchicus at was 9.6% (w.b.). The final moisture content of
different temperatures.
the combined drying process (AFD-

5ºC/AD+20ºC) was 5.9% (w.b.). The VFD test
showed a much lower drying rate. During the first
24 hours, the moisture content was only reduced
to 58% and after 100 hours the final moisture
content was 2%. The weight reduction from VFD
test showed a linear inclination and it was not
possible to distinguish between different drying
stages.
The particle size of the CF used was around 1mm
(d50-distribution) after crushing the CF plates in
the grinder (Figure 3). After the drying process,
the purely freeze-dried particles were slightly
reduced in size. Size reduction was a result of
particle shrinkage and particle breakdown during
Figure 3: Particle size distribution of Calanus
finmarchicus before and after drying.
fluidization. Particles dried with the combined

drying process (AFD-5ºC/AD+20ºC) showed an
increased particle size distribution. CF particles
had a tendency to agglomerate during the first drying stage when they were in an unfrozen condition.
Under an optical microscope (60x) all particles showed a homogeneous structure and it was not possible
to detect complete single animals. The particle size for product produced using VFD was not measured.
ISBN 978-82-594-3406-7

---

4th Nordic Drying Conference, Reykjavik, Iceland, June 17th to 19th 2009


Figure 4: Proteolytic and lipolytic activity of Calanus finmarchicus before and after drying.

Figure 4 shows the measured lipolytic and proteolytic activity for fresh and dried CF as median with its
variance. Every bar in Figure 4 represents 16 tests (4 measurements performed on 4 samples). The
product from VFD was not analysed. All dried CF showed a small decrease in lipolytic activity compared
to untreated CF. Proteolytic activity only changed for the product from the combined drying (AFD-
5ºC/AD+20ºC). However the proteolytic activity of the product from -5ºC AFD varied over a wide range.
A water extract was made for determination of
proteolytic and lipolytic activity. The measured
activities thereby reflected the activities for
rehydrated CF.
Desorption and adsorption isotherms for fresh
Calanus at 30ºC and 10ºC are shown in Figure
5. Desorption of fresh CF was a drying process
which influenced the product structure
(shrinkage) and therefore the water activity for
the rehydrated CF (adsorption) was generally
lower. Below a moisture content of 25% to
30%, the sorption characteristics for CF did not
differ much between adsorption and
desorption. Above this moisture content,

desorption at 10ºC resulted in a lower water
Figure 5: Sorption isotherms for fresh Calanus
activity than desorption at 30ºC.
finmarchicus at 30ºC and 10ºC.



Discussion

Freshly caught CF is highly sensitive to degradation. If the raw material is not properly stored and
handled its quality plummets rapidly, so that after a few days or even hours the material is not longer
usable. The CF for this project was placed in the plate freezer within 1 hour after catching and stored until
further processing at -80ºC. This ensured that the material was of defined and high quality which cannot
be guaranteed with purchased material, where catching date, duration until freezing, freezing time and
storage conditions are not documented. The interpretation of biochemical analyses must be done with
regard of the complete process line. This includes catching, freezing and storage conditions. Because of
the rapid degradation in CF, a freeze-drying process was chosen as an appropriate dewatering method.
The expectation was that at higher drying temperatures the product would lose valuable components and
ISBN 978-82-594-3406-7

---

4th Nordic Drying Conference, Reykjavik, Iceland, June 17th to 19th 2009
the structural changes would be too high. For the same reason mechanical dewatering (e.g. in a press) was
excluded, because valuable lipids would be washed out in the liquid phase and structural changes due to
the high forces can be expected.
The particle size chosen for the drying process (d50 = 1mm) resulted in a high surface area for mass
transfer. As reported by Donsi et al. (2001), drying rates in AFD increase with decreasing particle size.
The drying rate achieved with the particle size we used was higher for the AFD than for the VFD.
Contrary to majority opinion (e.g. Mujumdar, 2007) it is possible for an AFD process to be faster than
VFD. It was not possible in AFD to properly fluidize the bulk of the CF that had high water content,
because of surface interactions. We suspect that so-called bound or unfrozen water (Wolfe et al., 2002)
near the particle surface froze the particles together as soon as the concentration gradient brought the
water to the surface. This effect will vanish when a dry front develops towards the particle centre, because
the dry part provides a distance to the next particle. The bulk of CF was therefore not in a fluidized state
at first, and the drying air streamed mostly on the bulk side or through channels past it. In this state some
drying air will remain in contact with the CF, but a laminar flow profile inside the bulk can be expected.
The mass transfer will therefore decrease and with it the drying rate. The drying rate will increase when
the particles are fluidized properly, because that means every particle can make good contact with the
drying air. This explains the accelerated drying towards the end of the first drying stage.
The proteolytic activity was measured for the rehydrated CF and showed the increase of short peptides in
the extract. Using CF as fish feed requires a product that is active when it comes in contact with water.
The proteolytic activity in freeze-dried CF was comparable to the activity in the fresh material. However,
the proteolytic activity in CF dried at -5ºC varied over a wide range. The combined dried CF (AFD-
5ºC/AD+20ºC) had a proteolytic activity that was reduced by around one-third. This could be related to
the higher drying temperature or the lower water content at the end of the drying. A minor decrease in
lipolytic activity occurred for all dried products, but no clear difference was visible for the different
drying processes. This indicates that lipase was more susceptible to denaturation during drying. The high
remaining enzymatic activity of CF from AFD could be a challenge for its use as standard fish feed, but
could be beneficial in starter feeds for aquaculture. Reduced proteolytic activity could be related to drying
conditions above the freeze range and should be investigated further.
All material was dried until its final water content was 10% or lower. This resulted in a water activity
clearly below 0.1 (Figure 5). In this range, enzymatic activity is very low or non-existent, so protease and
lipase will not occur and the product is preserved. However lipid oxidation is generally reduced at water
activities between 0.2 and 0.6 (Shahidi, 2007) and will be higher at a water activity below 0.1. Therefore
it can be expected that the dried CF has a high susceptibility towards lipid oxidation at this water content.
This will influence the quality (rancidity) of the dried product and could be avoided by drying CF to a
final water content of around 20% (w.b.). This would result in a water activity around 0.3 (Figure 5),
where lipid oxidation has its minimum.
The sorption isotherms show a hysteresis between desorption and adsorption. The explanation can be
found in the changed product structure during desorption. Fresh CF was dried slowly during analysis and
the equilibrium water activity after a certain time was noted at given temperatures. During this drying
process, CF shrank and started to degrade biologically. Adsorption in our measurement was a rehydrating
process and since the original structure of CF was diminished, less water than original could be adsorbed.
CF that had already been de- and rehydrated once showed no clear hysteresis. Figure 5 shows no
differences between desorption and adsorption at different temperatures for low moisture contents and
low water activities. Measuring a sorption isotherm takes several days, and as stated earlier, CF is very
sensitive to degradation. The reason why no clear difference was measured in this region could be the fact
that CF had already started degrading after the time needed for desorption to this low moisture content.
We were able to show with the drying tests that CF has an equilibrium at 20ºC, 5.9% moisture content
and 2.1% water activity and at -5ºC/-10ºC, 10% moisture content and ?17.5% water activity. Lower water
content and therefore reduced water activity can generally be reached at higher drying temperatures.
Storing CF in a dried condition requires low water activity to prevent quality losses due to enzymatic
ISBN 978-82-594-3406-7

---

4th Nordic Drying Conference, Reykjavik, Iceland, June 17th to 19th 2009
activity, oxidation and rancidity. Adding a drying period with temperatures above the freezing point at the
end of an AFD process reduces the water content and water activity of the product significantly. The
temperature step must be chosen with respect to the product quality. For CF a drying period with 20ºC
after freeze-drying reduced the proteolytic activity by approximately 30%. This has to be tolerated when
using CF as fish feed. The reduced water activity could also provide a longer storage period. Since CF
can only be caught seasonally, this might be an important advantage. With a combined process the long
term storage quality of a freeze-dried product can be improved while the drying time is reduced
significantly.
Conclusion

Fish feed is generally a low-cost product, and economic reasons make the use of AFD for CF a good
choice. We have shown that AFD did not affect proteolytic activity and only reduced the lipolytic activity
of CF to a minor extent. The quality of the freeze-dried and rehydrated product was almost equal to the
raw material and also shows the potential of AFD processes in general. Currently, CF can only be caught
in industrially significant amounts in the spring, because CF is only present in the surface layer of the sea
during their spring bloom. Additionally, CF has a higher lipid composition in the spring season. This
suggests the need for optimal long-term storage conditions, because CF is only available seasonally. The
quality of fresh and dried CF after different storage times is currently under investigation.
High energy consumption and low drying rates have until recently limited the use of AFD on an industrial
scale. Combining AFD with a heat pump made this drying process more competitive with VFD.
Depending on the product size, AFD can even achieve higher drying rates than VFD. Future research will
concentrate on improving the drying rate with so-called “hybrid” technologies that use ultrasonic, infrared
radiation or microwave in AFD (Alves-Filho and Eikevik, 2008). It is believed that AFD can be a cost-
effective alternative not only for VFD, but could substitute drying processes for temperature sensitive
products, which are currently dried in an unfrozen condition. Because of the low temperature, product
quality can be improved with AFD at drying rates that are similar to AD. Thus, AFD offers considerable
potential as a future drying technology.
Acknowledgments

This work was supported by the Research Council of Norway (project: 17264I-S40). Thanks to all the
members of the Calanus project for their support and to Christopher Rundel, Françoise Pomerleau and
Benoît de Sarrau for their assistance.
Notation

AD Atmospheric
Drying/Drier

AFD Atmospheric
Freeze-Drying/Drier

CF
Calanus finmarchicus (a zooplankton species)

HP Heat
Pump

PSD
Particle Size Distribution

VFD
Vacuum Freeze-Drying/Drier

w.b.
wet base

References

Alves-Filho, O., Eikevik, T.M. 2008. Hybrid heat pump drying technologies for porous material. 16th
International Drying Symposium (2008):1539-1545.
ISBN 978-82-594-3406-7

---

4th Nordic Drying Conference, Reykjavik, Iceland, June 17th to 19th 2009
Anonymous. 2007. Kan dyreplankton bli redningen? Nytt fra Havbruk 03/2007:8.
Barret, A.J., Heath, M.F. 1977. Lysosomes: a laboratory handbook. In: Dingle, J. T. Amsterdam: North-
Holland. p. 19-145.
Claussen, I.C., Andresen, T., Eikevik, T.M., Strømmen, I. 2007a. Atmospheric Freeze Drying - Modeling
and Simulation of a Tunnel Dryer. Drying Technology 25(12):1959 - 1965.
Claussen, I.C., Ustad, T.S., Strømmen, I., Walde, P.M. 2007b. Atmospheric Freeze Drying - A Review.
Drying Technology 25(6):947 - 957.
DKNVS (The Royal Norwegian Society of Science and Letters) and NTVA (Norwegian Academy of
Technological Sciences). 1999. Norges muligheter for verdiskaping innen havbruk. Trondeheim:
www.ntva.no/rapport/havbruk.pdf.
DKNVS (The Royal Norwegian Society of Science and Letters) and NTVA (Norwegian Academy of
Technological Sciences). 2006. Exploitation of Marine Living Resources – Global Opportunities for
Norwegian Expertise. Trondheim: http://www.ntva.no/seminarer/manus/havbruk-06.pdf.
Donsi, G., Ferrari, G., Di Matteo, P. 2001. Utilization of combined processes in freeze-drying of shrimps.
Food and Bioproducts Processing: Transactions of the Institution of Chemical Engineers, Part C
79(3):152-159.
Izquierdo, M.S., Henderson, R.J. 1998. The determination of lipase and phospholipase activities in gut
contents of turbot (Scophthalmus maximus) by fluorescence-based assays. Fish physiology and
biochemistry 19(2):153-162.
Kattner, G. 1989. Lipid composition of Calanus finmarchicus from the North Sea and the arctic. A
comparative study. Comparative Biochemistry and Physiology Part B: Biochemistry and
Molecular Biology 94(1):185-188.
Lowry, O.H., Rosebrough, N.J, Farr, A.L., Randall, R.J. 1951. Protein measurement with the Folin
phenol reagent. J. Biol. Chem. 193:265-275.
McIlvaine, T.C. 1921. A buffer solution for colorimetric comparison. J. Biol. Chem. 49:183-186.
Mujumdar, A.S. 2007. Handbook of industrial drying. Boca Raton, Fla.: Taylor & Francis.
Olsen, R.E., Henderson, R.J., Sountama, .J, Hemre, G.I., Ringø, E., Melle, W., Tocher, D.R. 2004.
Atlantic salmon, Salmo salar, utilizes wax ester-rich oil from Calanus finmarchicus effectively.
Aquaculture 240(1-4):433-449.
Roberts, I.M. 1985. Hydrolysis of 4-methylumbelliferyl butyrate: A convenient and sensitive fluorescent
assay for lipase activity. Lipids 20(4):243-247.
Shahidi F. 2007. Maximising the value of marine by-products. Boca Raton: CRC Press.
Solgaard, G., Standal, I.B., Draget, K.I. 2007. Proteolytic activity and protease classes in the zooplankton
species Calanus finmarchicus. Comparative Biochemistry and Physiology - B Biochemistry and
Molecular Biology 147(3):475-481.
Song, X. 1990. Low temperature, fluidized bed drying with temperature program. Trondheim: Norwegian
Institute of Technology, Division of Refrigeration Engineering. p. VI, 112 s.
SSB (Statistics of Norway). 2008. Aquaculture 2006. Oslo:
http://www.ssb.no/english/subjects/10/05/nos_fiskeoppdrett_en/nos_d401_en/nos_d401_en.pdf
Tokle, N.E. 2006. Are the ubiquitous marine copepods limited by food or predation? Experimental and
field-based studies with main focus on Calanus finmarchicus. Faculty of Natural Science and
Technology. Trondheim/Norway: Norwegian University of Science and Technology.
Wolfe, J., Bryant, G., Koster, K.L. 2002. What is 'unfreezable water', how unfreezable is it and how much
is there? Cryoletters 23(3):157-166.
ISBN 978-82-594-3406-7

---