Aubin Et Al. - 2009 - Assessment of The Environmental Impact of Carnivorous Finfish Production Systems Using Life Cycle Assessment - Jou
Aubin Et Al. - 2009 - Assessment of The Environmental Impact of Carnivorous Finfish Production Systems Using Life Cycle Assessment - Jou
Aubin Et Al. - 2009 - Assessment of The Environmental Impact of Carnivorous Finfish Production Systems Using Life Cycle Assessment - Jou
a r t i c l e i n f o a b s t r a c t
Article history: When evaluating the environmental impacts of finfish production systems, both regional impacts (e.g.,
Available online 21 September 2008 eutrophication) and global impacts (e.g., climate change) should be taken into account. The life cycle
assessment (LCA) method is well suited for this purpose. Three fish farms that represent contrasting
Keywords: intensive production systems were investigated using LCA: rainbow trout (Oncorhynchus mykiss) in
Life cycle assessment freshwater raceways in France, sea-bass (Dicentrarchus labrax) in sea cages in Greece, and turbot
Environmental impacts
(Scophtalmus maximus) in an inland re-circulating system close to the seashore in France. Two main
Production systems
characteristics differentiated the three farm systems: feed use and energy use. Emission of nitrogen and
Fish farming
phosphorus accounted for more than 90% of each farm’s potential eutrophication impact. In the trout and
sea-bass systems, feed production was the major contributor to potential climate change and acidifi-
cation impacts and net primary production use (NPPU). In these systems, the main source of variation
for environmental impacts was the feed conversion ratio. Results from this study indicate that the sea-
bass cage system was less efficient than the trout raceway system, with a higher level of potential
eutrophication (65% greater) and NPPU (15% greater). The turbot re-circulating system was a high
energy-consumer compared to the trout raceway system (four times higher) and the sea-bass cage
system (five times higher). Potential climate change and acidification impacts were largely influenced by
energy consumption in the turbot re-circulating system. In the turbot re-circulating system 86% of
energy use was due to on-site consumption, while in the sea-bass cage farming system 72% of energy use
was due to feed production. These results are discussed in relation to regional contexts of production and
focus attention on the sensitivity of each aquatic environment and the use of energy carriers.
Ó 2008 Elsevier Ltd. All rights reserved.
0959-6526/$ – see front matter Ó 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.jclepro.2008.08.008
J. Aubin et al. / Journal of Cleaner Production 17 (2009) 354–361 355
- a farm in which sea-bass (Dicentrarchus labrax) are reared in measurement or assessment of real farms for one normal year of
marine-based cages in Greece, in which the water resource is production, eventually corrected by average data from previous
not limiting, and production cycles.
- a land-based re-circulating system for turbot (Scophtalmus
maximus) production located in France in which high tech-
nology compensates for local constraints. 3. Main characteristics of each production system
The environmental impacts are evaluated using life cycle We can define the three studied systems by a combination of
assessment (LCA), an integrative methodology that provides indi- factors: fish species and their growth characteristics, fish size at
cators of impact at regional and global scales. sale, feed efficiency, facilities, technologies used, and management
practices. The following descriptions are based on the running of
three real farms during one year of production. The reference year
2. Methods, goal and scope
was 2000 for the turbot and trout farms and 2003 for the sea-bass
farm. Each farm’s energetic background (i.e., electricity produc-
Life cycle assessment is a standardised method [1,2] conceived
tion), which influenced the emissions considered in this study, was
to assess potential impacts associated with a product, by quanti-
based on its reference year and location.
fying and evaluating the resources consumed and emissions to the
The rainbow trout flow-through farm is located in the Aquitaine
environment at all stages of its life cycle, from the raw-material
region of South-western France. It is an inland farm that contains
extraction to the disposal of waste [3]. For each environmental
raceway tanks. The farm specialises in raising trout from the
impact taken into account in the LCA, a characterisation model is
fingerling stage (60 g) and produces annually three sizes of fish:
used to convert the inventory data (resources use and emitted
pan-sized trout (270 g) in 6 months, large trout (1250 g) in 12
pollutants) that contribute to this impact into potential-impact
months, and very large trout (3500 g) in 18 months of rearing. The
estimates. This is done by multiplying the resources used and
farm’s annual calculated biomass gain reached 330 ton, provided
emissions by a characterisation factor for each impact category to
mainly by the production of very large trout (63%) and large trout
which it may contribute.
(31%). Because the farm focuses on producing large-size fish, we are
The environmental impact assessment was conducted accord-
able to compare its LCA results to those of farms producing other
ing to Papatryphon et al. [5], using SimaProÒ 6.0 software and its
large salmonids (e.g., Atlantic salmon in [22]). The economic feed
databases, original data collected on farms, and data provided by
conversion ratio (FCR ¼ kg of feed distributed /kg of live weight fish
previous studies [4–7]. The systems studied included production of
produced) equalled 1.21. The average feed composition declared by
fish at the farm level and related processes such as feed production
the manufacturer included 44% protein, 28% lipids, and 1.2%
(including both fisheries and agricultural phases), construction of
phosphorus (Table 2). Water entering the farm is diverted from
infrastructures, and equipment manufacturing. The transportation
a river and had an average flow-rate of 550 l s1. The water oxygen
of material at each step was also taken into account. The slaugh-
supply was supplemented by liquid oxygen use (300 ton year1). At
tering, processing, and sale phases were not included. This study
the outlet of the farm, a drum filter extracts solids from the water.
was limited to the delivery of raw fish at the farm gate (or at the
The extracted solids (sludge) are exported as manure for agricul-
shore, for sea cages). The hatchery phase was not included in this
tural crops and were not taken into account in the environmental
study as it has an entirely different production concept for which
analysis.
data were not available. All environmental impacts were expressed
The sea-bass cages are located on the Evoikos Gulf, north of
for the production of 1 ton of live fish weight.
Athens, Greece. The farm is dedicated to growing European sea-
Local emissions of nutrients associated with fish growth were
bass from 2 to 350 g in approximately 16 months. It consisted of 12
estimated by using nutrient-balance modelling [8]. This modelling
circular net-cages, each 1100 m3 in volume, arrayed around a plat-
method has been adapted and validated for different fish species
form used for equipment handling. The depth of water under the
[9–11]. It has been used previously to establish an emissions
cages was 25 m, and the average water current was 3 cm s1. The
inventory of fish-production systems [5–7].
farm is equipped with boats and land-based facilities for feed and
The calculation of N, P, and solids emissions was based on the
material stocking and net cleaning. The annual biomass gain was
difference between the amount of nutrients provided to fish via the
256 ton. The economic FCR equalled 1.77. The average composition
feed and the amount assimilated as fish weight-gain. The nutrient
digestibility of the feed, fish body-composition, and the non-
ingested part of the distributed feed (estimated at 5%, expert
Table 2
opinion) were used to calculate solid and dissolved fractions of
Mean feed composition in three fish farm systems
emitted N and P. The quantity of sludge extracted by water-cleaning
devices and spread on fields was subtracted from the nutrient Chemical composition (%) Trout feed Sea-bass feed Turbot feed
release. The evaluation of nutrient release and theoretical oxygen Protein 44.0 45.0 55.0
Lipids 28.0 12.0 12.0
demand (calculated using the chemical oxygen demand of protein,
Phosphorus 1.2 1.3 1.6
carbohydrates, lipids, ash and fibre emitted) per ton of fish for each Components (%)
farm is shown in Table 1. The inventory phase was based on direct Fish meal, Norway 22.5 42.0 40.0
Fish meal, Peru 22.5 23.5
Fish oil, Denmark 24.0 5.0
Table 1 Fish oil, Norway 8.0
Local emission of nutrients to water and the theoretical oxygen demand for three Winter wheat, France 11.1 13.5
fish farms, calculated for 1 ton of fish Winter wheat, Greece 22.0
Wheat gluten meal, France 7.0
Local emission to water (kg ton1 of fish) Trout farm Sea-bass farm Turbot farm Wheat gluten meal, Greece 8.0
Dissolved N 57.8 83.2 68.8 Maize gluten meal, France 4.5
Solid N 7.2 18.5 12.7 Soy cake, Brazil 12.0 9.5
Dissolved P 4.0 4.7 2.6 Soybean meal, USA 15.0
Solid P 6.0 12.0 8.0 Additive premix and amino acids, France 0.9 4.0
Theoretical O2 demand 135.7 347.7 147.8 Additive premix and amino acids, Greece 5.0
356 J. Aubin et al. / Journal of Cleaner Production 17 (2009) 354–361
of feeds declared by the manufacturer included 45% protein, 12% factors (potential effect at a 100-year time horizon) used by the
lipids, and 1.3% phosphorus (Table 2). International Panel on Climate Change (IPCC) [13] and is expressed
The re-circulating turbot production system is located in the in kg CO2-equivalents. Net primary production use refers to the
Brittany region of North-western France. It is dedicated to raising amount of net primary production (biomass produced by photo-
15-g fingerlings into 1200-g fish of commercial size. The annual synthesis) required as a biotic resource input, in the sense of it
biomass gain was 70 ton. The facilities are composed of 32 concrete being unavailable for other purposes. It was calculated according to
tanks (each with a volume of 30 m3) supplied by re-circulated Papatryphon et al. [5] and is expressed in kg of carbon. Energy use
water. Water coming from a river estuary is thermoregulated to refers to the consumption of fossil fuels, wood, uranium and
17 C and recycled through mechanical filters that extract solids hydropower electricity. It was calculated by using the lower heating
coming from the rearing tanks and biological filters that change values available in SimaPro 6.0 method [14] and is expressed in MJ.
ammonia into nitrate [7]. The sludge extracted by filters is stocked An additional impact category is proposed: water dependence.
and exported as manure for land crops. The fate of sludge applied to This novel impact category refers to the water input relative to fish
crops was not taken into account in the environmental analysis. biomass production at the farm level. It stems from the occupation
Estimating 50% water replacement per day, the annual water input concept developed for impacts of land use [3]. In land-based fish
reached 339,012 m3. The feeds used in the farm contained a high production systems, water dependence equals the total water input
level of fish-based ingredients as recommended by the Red Quality diverted or pumped from a river, divided by the fish growth
Label (‘‘Label Rouge’’). The Red Quality Label indicates a high measured during the reference period. It is important to note that
standard of quality based on sensorial and chemical characteristics water is not considered to be consumed in the rearing system. For
of the fish produced. This label assumes that higher inclusion rates sea-cage systems the water input takes into account the rearing
of fish-based ingredients in feeds increase the quality of fish meat. volume and the average water current. It is expressed in cubic-
The declared feed composition contained 55% protein, 12% lipids, meters. All of the environmental indicators are calculated on the
and 1.6% phosphorus (Table 2). The observed economic FCR basis of a single functional unit: the production of 1 ton of live fish
equalled 1.23. weight upon harvest.
To understand the contribution of system parts to the impact
categories, the following system components were identified: (1)
4. Impact categories fish production, i.e. the emissions induced by the biological trans-
formation of feeds at the farm level; (2) feeds, including production
The impact categories considered in this study are: eutrophi- of feed ingredients, processing and transportation; (3) equipment,
cation, acidification, climate change, net primary production use including its manufacturing, transport, and use; (4) infrastructure
(NPPU), and energy use (EU). Eutrophication refers to the potential (raceways, cages, buildings), including material production and
impacts of high levels of nutrients in the environment, in particular transportation, construction, and use; (5) chemicals, including
N and P. This indicator is calculated using the factors proposed by liquid oxygen, medicines and disinfectants, and their production
Guinée et al. [3], adding the theoretical oxygen demand calculation and transportation; and (6) energy carriers used on the farm
for solid wastes from fish farms. It is expressed in kg PO4-equiva- (electricity, gasoline, natural gas), including their production and
lents. Acidification refers to the negative effects on soils, ground transportation.
and surface water, and ecosystems of acidifying pollutants. It was
calculated using the average European acidification potential
factors according to Huijbregts [12] and is expressed in kg SO2- 5. Results
equivalents. Climate change is defined as the potential impact of
gaseous emissions on heat-radiation absorption in the atmosphere. The contribution of each system component to each impact
It was calculated according to the global warming potential 100 category was estimated for the three systems studied (Table 3). Fish
Table 3
Environmental impacts per 1000 kg of live fish weight (last column) and the percentage of each impact contributed by system components for three fish farm systems
Fish production Feed production Equipment Infrastructures Chemicals Energy carriers Total
Eutrophication (kg PO4-eq)
Trout flow-through 60.80 (92%) 4.67 (7%) 0.01 (0%) 0.14 (0%) 0.08 (0%) 0.20 (0%) 65.91 (100%)
Sea-bass cages 101.00 (93%) 7.64 (7%) 0.01 (0%) 0.10 (0%) 0.02 (0%) 0.09 (0%) 108.85 (100%)
Turbot re-circulating 69.90 (91%) 3.88 (5%) 0.08 (0%) 0.18 (0%) 0.01 (0%) 2.92 (4%) 76.97 (100%)
Climate Change (kg CO2-eq)
Trout flow-through 0 (0%) 2020 (73%) 33 (1%) 211 (8%) 83 (3%) 406 (15%) 2753 (100%)
Sea-bass cages 0 (0%) 3090 (86%) 22 (1%) 286 (8%) 40 (1%) 163 (5%) 3601 (100%)
Turbot re-circulating 0 (0%) 1920 (32%) 163 (3%) 232 (4%) 32 (1%) 3670 (61%) 6017 (100%)
Acidification (kg SO2-eq)
Trout flow-through 0.00 (0%) 13.10 (68%) 1.03 (5%) 1.58 (8%) 0.67 (4%) 2.79 (15%) 19.17 (100%)
Sea-bass cages 0.00 (0%) 20.20 (80%) 0.19 (1%) 1.55 6% 2.14 (8%) 1.22 (5%) 25.30 (100%)
Turbot re-circulating 0.00 (0%) 13.90 (29%) 6.71 (14%) 2.35 (5%) 0.20 (0%) 25.12 (52%) 48.28 (100%)
Net Primary Production Use (kg C)
Trout flow-through 0 (0%) 62,200 (100%) 0 (0%) 0 (0%) 0 (0%) 0 (0%) 62,200 (100%)
Sea-bass cages 0 (0%) 71,400 (100%) 0 (0%) 0 (0%) 0 (0%) 0 (0%) 71,400 (100%)
Turbot re-circulating 0 (0%) 60,900 (100%) 0 (0%) 0 (0%) 0 (0%) 0 (0%) 60,900 (100%)
Energy Use (MJ)
Trout flow-through 0 (0%) 31,100 (40%) 1354 (2%) 3470 (4%) 5173 (7%) 37,132 (47%) 78,229 (100%)
Sea-bass cages 0 (0%) 39,400 (72%) 325 (1%) 5184 (9%) 557 (1%) 9191 (17%) 54,656 (100%)
Turbot re-circulating 0 (0%) 30,900 (11%) 3120 (1%) 4100 (1%) 2856 (1%) 250,010 (86%) 290,986 (100%)
Water dependence (m3)
Trout flow-through 52.6 (100%) 0.0 (0%) 0.0 (0%) 0.0 (0%) 0.0 (0%) 0.0 (0%) 52.6 (100%)
Sea-bass cages 48,782.2 (100%) 0.0 (0%) 0.0 (0%) 0.0 (0%) 0.0 (0%) 0.0 (0%) 48,782.2 (100%)
Turbot re-circulating 4.8 (100%) 0.0 (0%) 0.0 (0%) 0.0 (0%) 0.0 (0%) 0.0 (0%) 4.8 (100%)
J. Aubin et al. / Journal of Cleaner Production 17 (2009) 354–361 357
(Table 1) is similar to data obtained by other authors. It lies within pumping, and filtering) at the farm level, and despite frequent
the range of values observed by Axler et al. [15] for trout farming: shipping from the cages to the shore. The distant origin of some
47–87 g N kg1 of fish and 4.8–18.7 g P kg1 of fish. For turbot, our feed inputs (i.e., fish meal; Table 2) has been identified as a poten-
results are higher than those of Mallekh et al.[10] (51 g N kg1 of tial source of energy waste, yet the energy cost of transporting raw
fish and 8.7 g P kg1 of fish), probably because their study was materials to feed plants contributes only 4–10% of the energy
conducted in tanks, which induced a low FCR (0.96). For sea-bass consumption of feed production (Table 4) in our study. Nonethe-
nitrogen-loading our results are similar to those (91 g N kg1 of less, other impact categories are affected by raw material trans-
fish) that occurred with a low-lipid diet observed by Boujard et al. portation, especially acidification (10–17% of acidifying emissions
[16]. induced by feed production).
Considering relative homogeneity in the diets of these carniv- The trout flow-through system requires more energy due to the
orous species and high digestibility of the components, the major use of liquid oxygen and water-cleaning drum filters at the outlet. It
factor of variation is the feed conversion ratio, as shown by Papa- reaches the higher level observed by Papatryphon et al. [6]:
tryphon et al. [5] in trout farms in France. The FCR is determined by 78,400 MJ ton1. Energy consumption in the trout and sea-bass
the combination of species characteristics (including their size) and farms is somewhat lower than estimates for salmon (Salmo salar)
the ration management. In particular, the FCR increases with fish farming (94,000 MJ ton1 [23], 99,000 MJ ton1 [25]). In contrast,
size, which leads to a minimisation of the value for the sea-bass the re-circulating system depends more on energy
sold at 350 g, compared to the trout sold mainly at 1250–3500 g in (291,000 MJ ton1) due to technical choices that influence energy
our study. consumption [7]. In the re-circulating system, thermoregulating
Rearing systems in tanks permit easier management of stocks water at 17 C throughout the year using a heat pump requires
(controlling both number of fish and their growth) and allows a high water-replacement rate (around 50% per day) and contin-
a more precise rationing. Thus, FCRs reaching 1.7 for sea-bass cages uous pumping. In the turbot farm, climate change and acidification
can be explained by poor estimation of stocks and low feed intake are determined mainly by electricity generated from fossil fuels
control. Previous studies have observed a FCR of 1.45 for sea-bass at (both on and off the farm). In France, 82% of off-site electric energy
marketable size [17]. The eutrophication results for 1 ton of product is provided by nuclear plants [26], which produce fewer green-
obtained for the trout farm in our study (66 kg PO4-eq) are in house gases and acidifying substances than electric plants that use
accordance with previous results for large trout production from fossil fuels. Locating a re-circulating system in Greece, where 95% of
Papatryphon et al. [6] (58–72 kg PO4-eq) and greater than those electricity is produced by fossil fuel combustion [27], would
obtained from a calculation based on Seppala et al. [18] and Silve- increase the climate change and acidification impacts of this system
nius and Grönroos [19] for trout sea cages (53.7 kg PO4-eq). by a factor of four. This point clearly illustrates the influence of the
The difference in eutrophication between the trout flow- energy context on the environmental impact of a production
through and the turbot re-circulating systems is due not to their system. These reasons could partially explain why the climate
FCR (1.21 vs. 1.23, respectively), but to the protein content of the change estimate observed for this turbot farm (6017 kg CO2-eq) is
diets (44% and 55%, respectively). In this particular case, a high similar to the salmon production estimate proposed by Tyedmers
protein content of marine origin is required by the Red Quality [23]: 6470 kg CO2-eq ton1. Research into renewable energy sour-
Label for turbot production. This feed characteristic induces ces such as solar or wind power, or co-location with other industrial
a higher release of N and P by the turbot, which is partially balanced facilities, could help to decrease climate change and acidification..
by the water refinement facilities of the re-circulating system. In Myrvang [28], describing an innovative turbot farm co-located with
2002, 46% of global fish meal production and 81% of fish oil a refinery, observed an acidification impact 75% lower (12.1 kg SO2-
production were used for aquaculture [20]. eq) and a climate change impact 63% lower (2209 kg CO2-eq) than
Because fish meal and fish oil production have levelled out at 6– those of the turbot farm in our study. The estimate of Seppala et al.
7 million and 1 million tonnes per year, respectively (International [18] for trout sea-cage culture climate change impacts (2780 kg
Fishmeal and Fish Oil Organisation, 2006, in [21]), the increasing CO2-eq ton1) is close to those of the trout and sea-bass farms in
production of aquaculture products (þ8.8% year1 1950–2004 [21]) this study. In developed countries to date, energy for production
makes diversification of protein and lipid sources an important systems has not been a major limiting factor; consequently, tech-
challenge. Raising carnivorous fish with feeds containing high nical choices have mainly been driven by cost factors. Nevertheless,
levels of wild-fish protein (e.g., 50–69% in this study, Table 2) is fossil fuel availability is of great concern. Energy conservation at the
greatly debated in aquaculture, particularly when fish farming is farm scale and replacement of fossil fuels by renewable sources
compared to commercial fishing [22–24]. Research is underway to must be supported.
identify plant-based alternatives to fish meal and oil [21], and there
are great hopes to find high-quality inputs to replace them. The use
7.3. Water characteristics and requirements
of fish by-products seems an interesting way to complement the
nutritional inputs of plant-based diets. It decreases dependence on
The characteristics of the water resource and the sensitivity of
the wild stocks that supply fish meal and seems to offer the best
aquatic ecosystems to eutrophication are another examples of the
environmental compromise [5]; however, these by-products are
influence of geographic location on fish-production system
available only in certain markets, which, considering fish farming
impacts. The trout farm studied requires fresh water of high-
expansion, does not make them a global solution. Nonetheless, at
quality, a resource that is relatively limited, is subject to annual
the farm level, this trend should induce a decrease of wild-fish
protein and oil inputs in feeds and an increase of feed efficiency; in
addition, these choices have a positive economic effect for fish Table 4
farms. Percentage contribution of raw-material transportation to the environmental
impacts of three fish feeds
7.2. Energy consumption and its origin Potential impact Trout feeds (%) Sea-bass feeds (%) Turbot feeds (%)
Eutrophication 5.2 6.9 3.8
The three rearing systems show clear differences in their energy Climate change 7.3 9.4 5.2
requirements. The sea-bass cage system uses the least energy, due Acidification 14.4 17.4 9.5
Energy use 6.4 9.9 4.3
to the absence of equipment for water management (oxygenating,
J. Aubin et al. / Journal of Cleaner Production 17 (2009) 354–361 359
variations, has a multitude of legal constraints, and is closely weighting of the impacts could be developed with stakeholder
monitored by other users [29]. The water requirement of this type collaboration to compare the environmental profiles of rearing
of farm can decrease water flow along a section of river sufficiently systems in different environmental contexts.
to modify the river ecosystem and lower fish migration rates. For the trout flow-through system, two main objectives have to
Moreover, the effluent composition of trout farms is particularly be considered: limitation of both water input and nutrient output.
controlled, as river ecosystems are sensitive to nutrient and sus- Limiting water input decreases the competition for freshwater
pended-solid loadings, which can cause visible changes in the resources with other water users (e.g., cities, anglers, agriculture).
aquatic biotope. Its position in relation to water availability and the Reducing its dependence on water is a key sustainability issue for
perceived value of the natural context by stakeholders is essential this system, especially under fluctuating water regimes. In parallel,
for the continuity of trout production. solid and nutrient loading into the river ecosystem must be limited
The sea-cage farm is located in a gulf of the Mediterranean, to reduce ecological and visible impacts. Management solutions
where the water resource is essentially non-limited, and the (i.e., feeding management) applied to decrease the FCR are forth-
oligotrophic characteristic of the sea induces a low sensitivity to coming. On the other hand, adapting the amount of fish stocked
eutrophication around fish farms [30]. The solid and nutrient to the carrying capacity of the river is a common request by
emissions of the cages mainly have a benthic effect [31]. An impact ecological organizations. The efficiency of cleaning systems at farm
indicator focused on the surface of damaged sea-bottom would give outlets has to be improved. For instance, the installation of re-
a good idea of the visible effect of the cage system. Nevertheless, circulating techniques in trout farms has been studied in Denmark
since the effects of nutrient loading on sensitive ecosystems such as [36]. For the trout flow-through system located in France, the
sea-grass meadows (Posidonia oceanica) have been confirmed [32], control of impact indicators eutrophication, water dependence, and
eutrophication seems to be a relevant impact category, given its NPPU is a priority (Table 5).
integrative feature. This point is often disputed, and sometimes In the sea-bass cage system, the main objective is the limitation of
eutrophication as an impact category is not taken into account nutrient loading by reducing feed spill through better feeding
because these farms do not overwhelm the carrying capacity of the management. In addition, reduction of nutrient and solid loading of
local aquatic environment [22]. Moreover, Machias et al. [33] show cage systems by management or technical solutions (e.g., stock
a positive effect of fish farming under these environmental condi- assessment, feed intake control) is an important objective on envi-
tions by an increase in sea productivity as measured by an increase ronmental (preserving the oligotrophic ecosystem) and economic
in the biomass of the fish community. grounds. The high dependence on water of this rearing system is
In contrast, the turbot farm is located on the shore of an estuary in balanced by the essentially non-limited nature of this resource; thus,
Brittany. In this coastal area, the high phosphorus concentration in this indicator is not well adapted for this kind of system. For the sea-
sediments due to agricultural and other human activities induces bass cage system in Greece, the reduction of net primary production
a high risk of eutrophication when nitrogen is available and thermal use and, secondly, eutrophication is a main objective (Table 5).
conditions are favourable [34]. Consequently, this fish-production The turbot farm has to pursue two main objectives: decreasing its
system competes with other animal production systems in the energy dependence and nutrient loading. These objectives are re-
region (pig production for example) for the environmental carrying enforced due to changes in the economic context (energy cost) or legal
capacity. The primary goal of the re-circulating system was to context (environmental regulation) and could lead to the search for
decrease the water dependence of the farm by moving it away from other types of technical or management solutions. Due to the sensi-
shore (where land competition is high). This goal has been partially tivity of the local aquatic ecosystem to eutrophication, nutrient releases
fulfilled in the studied turbot farm, as lower water replacement is must be limited. This goal could be carried out by technical solutions
now technically possible in such systems [35]. The water replace- such as biofilter design or management choices such as using a lower-
ment limitation tends to balance its increased energy consumption nutrient fish diet. Consequently, the impacts to be considered a priority
due to the thermoregulation and circulation of water. in this system are primarily energy use (and the related impacts climate
change and acidification), as well as net primary production use and, of
less importance, eutrophication (Table 5).
7.4. Specific environmental objectives depending on local and
global contexts 8. Conclusion
Depending on the local context, the major environmental A monoculture of carnivorous fish is not the best form of
constraints are different and can modulate the conclusions given by aquaculture from an environmental point of view. These species
the impact indicators. Table 5 proposes a semi-objective hierarchy require a high level of protein and oil, to date provided by wild-fish-
of impact relevance facing each system according to its character- stocks. Consequently, they release a great amount of nitrogen and
istics and environmental context. A more comprehensive phosphorus into the aquatic environment because of the high feed
quality. Their metabolism requires a high level of oxygen, which
Table 5 requires either great water availability or costly techniques such as
Estimated relative importance of environmental impacts for each of the three re-circulating systems or liquid oxygen use. Nevertheless, carniv-
production systems studied, given their environmental contexts, from more orous fish species are demanded by consumers, and farms raising
important (1) to less important (3). Environmental impacts on a given system may them are constantly expanding all over the world.
have equal importance
Solutions for more environmentally friendly systems are being
Potential Trout flow- Sea-bass cages Turbot researched, due to a fish farming sector in quick evolution. Novel
impact through system in re-circulating fish-production systems are evolving towards more efficient water-
system in Greece system in
France France
filtering techniques to limit water input and decrease nutrient and
solids release [36–39] or towards more integrated systems based
Eutrophication 1 2 2
Climate change 2 3 1 on the reuse of nutrients, water, and energy at the farm level
Acidification 3 3 1 through the associated culture of plants or animals at a lower
Net primary production use 1 1 1 trophic level [40–42]. These two approaches are not exclusive and
Non renewable energy use 2 3 1 can be combined. The evolution of a production system is complex
Water dependence 1 3 3
and involves many constraints such as:
360 J. Aubin et al. / Journal of Cleaner Production 17 (2009) 354–361
- improvement of one impact must not induce pollution transfer [3] Guinée JB, Gorrée M, Heijungs R, Huppes G, Kleijn R, de Koning A, et al.
Handbook on life cycle assessment. An operational guide to the ISO standards.
to other impacts. For instance, trade-offs between eutrophi-
Dordrecht, The Netherlands: Kluwer Academic Publishers; 2002.
cation and climate change (risk of N2O emissions to the air [4] Basset-Mens C, van der Werf HMG. Scenario-based environmental assessment
associated with NO3 reduction in water during denitrification) of farming systems – the case of pig production in France. Agriculture,
are poorly documented in aquatic production systems [7]. Ecosystems, and Environment 2005;105(1–2):127–44.
[5] Papatryphon E, Petit J, Kaushik SJ, van der Werf HMG. Environmental impact
- avoiding a real or perceived decrease in the quality of the assessment of salmonid feeds using life cycle assessment. Ambio
product, which would influence the marketing strategy. This 2004a;33(6):316–23.
could result from using indoor farms or by changing feed [6] Papatryphon E, Petit J, van der Werf HMG. The development of life cycle
assessment for the evaluation of rainbow trout farming in France. In:
components. Halberg N, editor. Life cycle assessment in the agri-food sector. Proceedings
- using techniques that are economically viable considering from the 4th international conference. Horsens, Denmark: Danish Institute of
production costs and fish market-values. Agricultural Sciences; 2004b. p. 73–80.
[7] Aubin J, Papatryphon E, Van der Werf HMG, Petit J, Morvan YM. Character-
isation of the environmental impact of a turbot (Scophthalmus maximus) re-
Life cycle assessment is a relevant tool for qualifying and circulating production system using Life Cycle Assessment. Aquaculture
studying the evolution of production systems, permitting the 2006;261(4):1259–68.
[8] Cho CY, Kaushik SJ. Nutritional energetics in fish: energy and protein utiliza-
characterisation of production systems based on their potential
tion in rainbow trout (Salmo gairdneri). World Reviews in Nutrition and Die-
environmental impact. Its multi-goal approach covering different tetics 1990;61:132–72.
scales offers a wide view of environmental impacts of an activity [9] Kaushik SJ. Nutritional bioenergetics and estimation of waste production in
non salmonids. Aquatic Living Resources 1998;11(4):211–7.
and allows one to compare different systems or alternative strate-
[10] Mallekh R, Boujard T, Lagardère JP. Evaluation of retention and discharge of
gies. In particular, LCA is a good tool to make an inventory of all nitrogen and phosphorus by farmed turbot (Scophthalmus maximus). North
product pollutant emissions and resources used in a system. Thus, American Journal of Aquaculture 1999;61:141–5.
LCA automatically takes into account a part of the regional context [11] Bureau DP, Gunther S, Cho CY. Chemical composition and preliminary theo-
retical estimates of waste outputs of rainbow trout reared in commercial cage
of system location, through the distances of material transportation culture operations in Ontario. North American Journal of Aquaculture
and the energetic background. In our study, we have seen the 2002;65:33–8.
influence of feed raw-material origins and electricity components. [12] Huijbregts MAJ. Life-cycle impact assessment of acidifying and eutrophying air
pollutants. Calculation of characterisation factors with RAINS-LCA. Amster-
Nevertheless, the relevance of an impact is linked to at least two dam, The Netherlands: Interfaculty Department of Environmental Science,
other factors: Faculty of Environmental Science, University of Amsterdam; 1999.
[13] Houghton JT, Meira Filho LG, Callander BA, Harris N, Kattenberg A, Maskell K.
Climate change 1995: The science of climate change. Cambridge, United
- the sensitivity of the environmental target Kingdom: Cambridge University Press; 1996.
- the amount of social and economic constraints on this target. [14] Consultants PRé. SimaPro 2 method, Database manual. The Netherlands:
Amersfoort Pré Consultants B.V.; 1997.
[15] Axler RP, Tikkanen C, Henneck J, Schuldt J, McDonald ME. Characteristics of
In this study, taking these factors into account lead us to
effluent and sludge from two commercial rainbow trout farms in Minnesota.
emphasize the importance of eutrophication in trout and turbot Progressive Fish Culturist 1997;59:161–72.
farms in France and to minimise its importance for sea cages in [16] Boujard T, Gélineau A, Covès D, Corraze G, Dutto G, Gasset E, et al. Regulation
of feed intake, growth, nutrient and energy utilisation in European sea bass
Greece. Considering the worldwide attention paid to global
(Dicentrarchus labrax) fed high fat diets. Aquaculture 2004;231(1–4):529–45.
warming and the availability of fish resources, climate change and [17] Kaushik SJ, Covès D, Dutto G, Blanc D. Almost total replacement of fish meal by
net primary production use are considered important impacts in all plant protein sources in the diet of a marine teleost, the European seabass,
these systems. Dicentrarchus labrax. Aquaculture 2004;230(1–4):391–404.
[18] Seppala J, Silvenius F, Gronroos J, Makinen T, Silvo K, Storhammar E. Rainbow
Consequently, the turbot re-circulating system, less dependent trout production and the environment. Helsinki: Finnish Environmental
on water, invisible from the shore, but requiring a high level of Institute; 2002.
energy, is more relevant in France, where electricity is not a major [19] Silvenius F, Grönroos J. Fish farming and the environment, Result of inventory
analysis. Helsinki: Finnish Environment Institute; 2003.
limiting factor and emits relatively little greenhouse gases [20] Tacon A. State of information on salmon aquaculture feed and the environ-
(important where stakeholders closely watch environmental ment. Salmon Dialog Report, World Wildlife Fund; 2005.
issues). In contrast, the sea cages of sea-bass located in Greece [21] FAO (Food and Agriculture Organisation). State of world aquaculture: 2006.
Fisheries technical paper no. 500. Rome: FAO; 2006.
benefit from essentially unlimited water and a low sensitivity to [22] Ellingsen H, Aanondsen SA. Environmental impacts of wild caught cod and
eutrophication (due to the low availability of phosphorus), com- farmed salmon – a comparison with chicken. International Journal of Life
plemented by a lower social pressure regarding fish farm impacts. Cycle Assessment 2006;11(1):60–5.
[23] Tyedmers, P. Salmon and sustainability: The biophysical cost of producing
These examples must lead us to improve consideration of envi-
salmon through the commercial salmon fishery and the intensive salmon
ronmental, social, and economic contexts to balance the conclu- culture industry. PhD Thesis. Vancouver, Canada: University of British
sions of environmental studies on aquatic production systems. Columbia; 2000.
[24] Folke C, Kautsky N, Troell M. Salmon farming in context: response to Black,
et al. Journal of Environmental Management 1997;50(1):95–103.
Acknowledgements [25] Troell M, Tyedmers P, Kautsky N, Rönnbäck P. Aquaculture and energy use. In:
Encyclopedia of energy, vol. 1. Elsevier, Inc.; 2004. 97–108.
[26] Electricité de France. Rapport annuel du groupe EDF 2000, http://www.edf.fr/
The authors wish to warmly thank the following persons: Yann
html/fr/RA2000/, 2004.
Marchand from Le Gouessant cooperative, Nikos Liberis from [27] DEI (Dimosia Epicheirisi Ilektrismou). Public Power Corporation S.A., Annual
Perseus S.A., Yves-Marie Morvan from France Turbot S.A., and report 2003. Athens; 2004.
[28] Myrvang M. Life Cycle Assessment of a marine farm co-located with a refinery.
Viviers de France S.A., for the time they spent with us and the
Masters Thesis, Trondheim, Norway: Norwegian University of Science and
collection of precise data used in this study. Michael Corson is Technology, Program for Industrial Ecology; 2006.
thanked for his helpful comments on two drafts of this paper. [29] Papatryphon E, Petit J, Van der Werf HMG, Kaushik SJ, Kanyarushoki C.
Nutrient balance modeling as a tool for environmental management in
aquaculture: the case of trout farming in France. Environmental Management
References 2005;35(2):161–74.
[30] Pitta P, Apostolaki ET, Giannoulaki M, Karakassis I. Mesoscale changes in the
[1] ISO (International Organization for Standardisation). International Standard water column in response to fish farming zones in three coastal areas in the
14040. Environmental management – Life cycle assessment – Principles and Eastern Mediterranean Sea. Estuarine, Coastal and Shelf Science
framework. Geneva: International Organisation for Standardisation; 1997. 2005;65(3):501–12.
[2] ISO (International Organization for Standardisation). International Standard [31] Karakassis I, Tsapakis M, Hatziyanni E, Papadopuolou KN, Plaiti W. Impact of
14042. Environmental management – Life cycle assessment – Life cycle impact cage farming of fish on the seabed in three Mediterranean coastal areas. ICES
assessment. Geneva: International Organisation for Standardisation; 2000. Journal of Marine Science 2000;57:1462–71.
J. Aubin et al. / Journal of Cleaner Production 17 (2009) 354–361 361
[32] Marbà N, Santiago R, Diaz-Almela E, Alvarez EM, Duarte C. Seagrass (Posidonia [37] Blancheton JP. Developments in recirculation systems for Mediterranean fish
oceanica) vertical growth as an early indicator of fish farm-derived stress. species. Aquacultural Engineering 2000;22(1–2):17–31.
Estuarine, Coastal and Shelf Science 2006;67(3):475–83. [38] Ebeling JM, Welsh CF, Rishel KL. Performance evaluation of an inclined belt
[33] Machias A, Karakassis I, Labropoulou M, Somarakis S, Papadopoulou KN, filter using coagulation/flocculation aids for the removal of suspended solids
Papaconstantinou C. Changes in wild fish assemblages after the establishment and phosphorus from microscreen backwash effluent. Aquacultural Engi-
of a farming zone in an oligotrophic marine ecosystem. Estuarine, Coastal and neering 2006;35(1):61–77.
Shelf Science 2004;60(4):771–9. [39] Van Rijn J, Tal Y, Schreier HJ. Denitrification in recirculating systems: Theory
[34] Ménesguen, Aminot A, Belin C, Chapelle A, Guillaud J-F, Joanny M, Lefebvre A, and applications. Aquacultural Engineering 2006;34(3):364–76.
Merceron M, Piriou J-Y, Souchu P. L’eutrophisation des eaux marines et sau- [40] Lefebvre S, Bacher C, Meuret A, Hussenot J. Modelling nitrogen cycling in
mâtres en Europe, en particulier en France. Rapport IFREMER DEL/EC/01.02 a mariculture ecosystem as a tool to evaluate its outflow. Estuarine Coastal
pour la Commission Européenne – DG.ENV.B1. Brest, France; 2001. and Shelf Science 2001;52(3):305–25.
[35] Gelfand I, Barak Y, Even-Chen Z, Cytryn E, Rijn J, van Krom MD. A novel zero [41] Neori A, Chopin T, Troell M, Buschmann AH, Kraemer GP, Halling C, et al.
discharge intensive seawater recirculating system for the culture of marine Integrated aquaculture: rationale, evolution and state of the art emphasizing
fish. Journal of the World Aquaculture Society 2003;34(3):344–58. seaweed biofiltration in modern mariculture. Aquaculture 2004;231(1–4):
[36] Lareau S, Champagne R, Ouellet G, Gilbert E, Vandenberg G. Rapport sur les 361–91.
missions d’évaluation de la technologie Danoise pour l’élevage en eau douce [42] Schneider O, Sereti V, Eding EH, Verreth JAJ. Analysis of nutrient flows in
des salmonidés. Québec: Société de Recherche et de Développement en integrated intensive aquaculture systems. Aquacultural Engineering 2005;
Aquaculture Continentale (SORDAC), Inc.; 2004. 32(3–4):379–401.