This is the published version of a paper published in Environmental Management.
Citation for the original published paper (version of record):
Hornborg, S., Främberg, A. (2019)
Carp (Cyprinidae) Fisheries in Swedish Lakes: A Combined Environmental Assessment
Approach to Evaluate Data-limited Freshwater Fish Resources as Food
Environmental Management, 65(2): 232-242
https://doi.org/10.1007/s00267-019-01241-z
Access to the published version may require subscription.
N.B. When citing this work, cite the original published paper.
Permanent link to this version:
Carp (Cyprinidae) Fisheries in Swedish Lakes: A Combined
Environmental Assessment Approach to Evaluate Data-limited
Freshwater Fish Resources as Food
Sara Hornborg 1●Anton Främberg2
Received: 1 July 2019 / Accepted: 4 December 2019 / Published online: 19 December 2019 © The Author(s) 2020
Abstract
The role of aquatic resources to food security is both promising and constrained since the global seafood consumption is increasing while marinefisheries approach the limit of what it can produce. In Sweden, the seafood consumption per capita is higher than the European and world average but the current dietary advice is to increase consumption. Freshwaterfisheries have in general been paid less attention in food security discussions. Carpfishes (Cyprinidae) in Sweden have lost their historical value and are currently, both understudied and underutilized. Here we use a combined environmental assessment approach to examine the environmental sustainability of current and potential cyprinid fisheries. We found that current commercialfisheries for Swedish cyprinids in lakes have an average carbon footprint of 0.77 kg CO2e per kg of edible
product, substantially smaller than most of the popular marine and terrestrial protein sources consumed in Sweden today. This could be even lower if cyprinid resources were better utilized than currently. The cyprinids however exhibited different vulnerability tofishing pressure and are today associated with data deficiencies. Hence, it is currently uncertain how much food for human consumption they can contribute to. Improved consumer interest and management attention is needed, but to the Swedish diet, cyprinids offer a promising opportunity for future more sustainable and nutritious food systems. Keywords Environmental assessment● Life cycle assessment● Productivity susceptibility analysis●Freshwaterfisheries●
Cyprinidae● Food security
Introduction
The need to shift towards more sustainable food systems advocate for new and innovative approaches (Gordon et al. 2017). Examples include improving resource use efficiency and finding solutions to the current distance between con-sumers and producers from globalized markets. Aquatic resources could play an important role to future food and
nutrition security (Godfray et al. 2010; Tacon and Metian 2013). However, current supply has high demand and is extensively traded, affecting e.g. affordability (FAO2018). Growth of aquaculture has been vital, considering that marine capture fisheries has not increased since the 1980s; on the other hand, further increase in production may lead to competition of feed resources (Troell et al.2014). Even so, seafood consumption is often recommended to increase due to important nutritional characteristics—but it is unclear how this could be done sustainably (Thurstan and Roberts2014).
Freshwater fisheries production is one component of aquatic resources that has traditionally been paid less attention to in terms of contribution to food security (McIntyre et al.2016; Deines et al.2017; Lynch et al.2017) and could potentially add to resource availability. In parti-cular in industrialized countries, that often import the majority of seafood consumed from countries with under-nourishment (Smith et al. 2010), freshwater resources are not given adequate consideration as a food resource in terms * Sara Hornborg
Sara.Hornborg@ri.se
1 RISE- Research Institutes of Sweden, PO Box 5401, 402 29
Gothenburg, Sweden
2 Department of Biological and Environmental Sciences, University
of Gothenburg, Gothenburg, Sweden
Supplementary information The online version of this article (https:// doi.org/10.1007/s00267-019-01241-z) contains supplementary material, which is available to authorized users.
123456789
0();,:
123456789
of management systems needed (Arlinghaus et al. 2002; Dudgeon et al. 2006). There is thus little information available to quantitatively determine sustainable exploita-tion levels (Lorenzen et al. 2016), and risk-based assess-ments such as productivity susceptibility analysis have almost exclusively been applied in a marine context (Hor-dyk and Carruthers2018). Small-scale fisheries with pas-sive gears, such as those often used in freshwaters, are also understudied in terms of energy use per kilo landing, while the data available exhibit large variability (Parker and Tyedmers2015). Fuel use during fishing is the key driver behind the carbon footprint of products from capture fish-eries (Ziegler et al.2016; Avadí and Fréon 2013). There-fore, including freshwater products in assessments of healthy and sustainable diets is overall compromised (e.g., Hallström et al.2019; Tilman and Clark 2014).
Sweden represents an interesting case study related to addressing the role of freshwater fisheries in future food systems. Swedish consumption of seafood is higher than the global and European average (Borthwick et al.2019), but below the dietary advice of two to three portions per week recommended by the National Food Agency (NFA) in Sweden (NFA Sweden2019). To reach the recommended level (2.5 portions a week), an increase of 3.75 kg per person and year is required (assuming a portion size of 125 g). However, around 78% of products consumed in Sweden today is imported, most from marinefisheries and aquaculture from around the globe, even if the Swedish government has a food strategy to increase level of self-sufficiency (Borthwick et al. 2019; GoV 2016). Swedish aquaculture production is marginal and has today chal-lenges with strict environmental legislation and high pro-duction costs, with only a few commercial actors. However, over 40,000 km2 is covered by freshwater (SCB 2019a) while freshwater fisheries represented only 0.7% of Swe-den’s reported commercial landings in 2018 (SwAM 2018, 2019a). What role could increased utilization of freshwaterfisheries in Sweden play for improved food and nutrition security?
Current freshwater landings in Sweden mainly comprise of a few popular freshwater fishes, such as salmonids (Salmonidae), pikeperch Sander lucioperca, perch Perca fluviatilis and pike Esox lucius, which are fully utilized or even in need of recovery according to the yearly account of status of Swedish fisheries resources (SwAM 2019b). However, there are also populations of virtually unutilized carp species (Cyprinidae). Latest advice recommends an increase infishing for bream Abramis brama, the only only carp species included in the yearly resource assessment, but with caution. Several species of carp were consumed in Sweden up until the middle of the 20th century but have over time been more or less abandoned in the process of industrialization, urbanization and expansion of marine
fisheries (Bonow et al. 2013). They are still an important resource in many other countries, including neighboring countries. Today, they are in Sweden mainly being fished for other reasons than for human consumption, such as bait fish and reduction fisheries in lakes to remove nutrients as a remedy against eutrophication (Bonow et al.2013; Hansson et al. 1998). This implies that there is very limited data available to inform on stock status and sensitivity tofishing pressure (Östman et al.2016).
To fully identify the net effect of different options for future sustainable diets, or‘rewiring the food system’ (sensu Gordon et al. 2017), an integration of environmental assessments is needed that is capable of including both potential local ecological risks and offstage effects (Pascual et al. 2017). As an example, while carbon footprints (esti-mates of greenhouse gas emissions through life cycle assessment (LCA)) are commonly used as proxies for sus-tainability (e.g. Tilman and Clark 2014), other tools are needed to address local ecological risks offishing (Ziegler et al. 2016), such as ecological risk assessments (Hobday et al.2011). The two tools may be seen as complementary to each other in assessing pressures from capturefisheries. This has been demonstrated for the Australianfisheries for Patagonian toothfish Dissostichus eleginoides, where local ecological risks were assessed as low whereas the carbon footprint was comparatively high (Hornborg et al. 2018); more case studies may be useful to map potential conflicts between different aspects of sustainability and contribute to discussions on sustainable diets.
The objective of this study is to evaluate Swedish fish-eries of cyprinids as a food resource compared to other more popular food items consumed in Sweden today. This is done by assessing the carbon footprint combined with the potential risks posed for the different carp species based on vulnerability tofishing pressure, i.e., a combined approach to environmental assessment. The overall aim is to form a basis for discussing the potential of cyprinids in sustainable food systems.
Materials and Methods
Life Cycle Assessment
The carbon footprint was quantified through a LCA approach, a tool commonly used to quantify the environ-mental pressures and resource use of different products throughout its lifecycle (Ness et al.2007). The goal was to assess the carbon footprint of the current and potential (defined as 100% retainment of discarded cyprinids) Swedish freshwater cyprinid fishing to enable comparison with (1) other fisheries, based on Parker et al. (2018); (2) other top seafood items consumed in Sweden (Borthwick
et al.2019); and (3) Swedish beef, pork, chicken and brown beans based on RISE Climate Database (2018). The func-tional unit (FU), i.e., the unit to which all of the inputs and outputs are related to, was defined as 1 kg of edible carp fish meat to be comparable with RISE (2018). The edible yield was collected from FAO (FAO1989) which states the same yield for all cyprinids (54% of live-weight, based on that of common carp Cyprinus carpio).
The system boundaries, i.e., the processes included in the study, covered the production and combustion of fuel for current commercial and recent reduction fisheries in Swedish lakes, up to the point of landing. Transport of the fish to market is excluded. This may add more to emissions than what is generally seen (low contribution of the trans-ports) because the supply chain is today not optimized due to the small-scale nature of the current fisheries. On the other hand, thefishery in at least one of the lakes (Bolmen) has no addition from the supply chain since it is locally prepared and consumed. Only emissions from fuel pro-duction and combustion during fishing were considered, while production of boats,fishing gear and packaging was excluded. This was motivated from that the fuel used during thefishing phase generally dominates the carbon footprint of seafood from capture fisheries (Ziegler et al. 2016). Allocation of emissions between species in landings was based on mass; alternative allocation such as economy-based is difficult since the market for cyprinids to a large extent is potential rather than existing. Results will also be less influenced by markets over time. In processing, all emissions are allocated to the edible part of the fish (no utilizations of by-products are known). Cleaning and fil-leting of thefish were assumed to be done by hand due to the small scale of the freshwater fisheries, and hence no additional resource use and emissions was assigned to this stage.
Emission factors from fuel combustion was collected from the Swedish fuel and biofuel institute (SPBI2010). It is assumed that all motors fueled by petrol use “Engine petrol without ethanol” (emission factor 2.36 kg CO2e/L)
whereas by diesel“Diesel – Mk1” (emission factor 2.54 kg CO2e/L). Emissions from upstream production of the fuel
came from the Ecoinvent database in the LCA software tool SimaPro. This background data accounts for emissions from production of fuel, such as oilfield exploration, crude oil production, long distance transportation, oil refining and regional distribution. Emissions were converted from being stated per kg to liters using densities corresponding to fuels at SPBI (2010).
To compare the carbon footprints to globalfisheries, the data collected had to be treated according to the method used by Parker et al. (2018). Non-fuel related inputs (such as production of boats and gear) were added as a theoretical estimate based on earlier studies (25% of the total
GHG-emissions up to landing of the fish). Fuel use per kg live weight was used, thus making the carbon footprint calcu-lated corresponding to all freshwaterfishes landed using the evaluatedfishing method.
Data on the current commercial pond net fishery (the fishing method contributing to 88% of cyprinid landings of mainly bream; SwAM 2019a) was collected through semi-structured interviews. A first suggestion of fishers was obtained from the Swedish Lake Fishery Association, adding further fishers through “snow-balling technique” where respondents were asked to suggest others to inter-view. Two of the Swedish big lakes were included, Vänern and Mälaren, and one smaller lake, Bolmen. In the two larger lakes efforts were made to obtain answers from fishers distributed throughout the lakes’ areas. Two large lakes were excluded, namely Vättern and Hjälmaren. Vät-tern was originally considered as a potential candidate but was excluded since the nutrient poor lake holds a relatively small cyprinid population, with all the commercial catches already being used in the lakes’ crayfish fishery (J. Fogel, County Administrative Board Stockholm, Pers. Comm.). A total of 14 commercialfishers provided data for one year of fishing (2018). Furthermore, data from five reduction fish-ing occasions in two lakes was extracted from published reports (Sandström 2011; Annadotter et al. 2013). Addi-tional data on reductionfisheries was obtained from inter-views with three reductionfishers from 20 fishing occasions in eight different lakes (between 2010 and 2018). The lakes included were located in the southern part of Sweden, more specifically the regions Skåne (Finjasjön, Häckebergasjön, Oppmannsjön, Ringsjön and Sövdesjön), Småland (Södra Bergundasjön, Trummen and Växjösjön) and Uppland (Vallentunasjön and Östhammarfjärdarna). The samples are seen as representative for the fishing practices since the percentage of cyprinids in landings in the commercial fisheries were approximately the same in the sample and the wholefleet, and data on reduction fisheries covered 40% of all the recent reductionfisheries granted funds from SwAM as Local Water Management Projects (SwAM 2019c). A Kruskal–Wallis one-way analysis of variance (Kruskal and Wallis 1952) was done to determine if there were any sig-nificant differences between the carbon footprint of fish from different lakes orfishing methods.
Productivity Susceptibility Analysis
Productivity-susceptibility analysis (PSA) is a screening analysis to assess the potential and relative vulnerability of a species to afishery based on a set of attributes related to its productivity (based on life history parameters such as life span) and susceptibility to impacts from fishing activities (based on parameters such as overlap offishing effort and species distribution), with different approaches used today
around the world (Hordyk and Carruthers 2018). The assessment is performed by categorizing a number of attributes of the species andfishery into low/medium/high through a score from 1 to 3. The two average values for the species productivity and susceptibility to thefishery are then combined to form an overall vulnerability V by calculating the Euclidean distance from the origin of the two numbers: V ¼pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiP2þ S2 ð1Þ
where productivity P is the average risk score for all productivity attributes and susceptibility S is the average of the multiplicative risk scores for all susceptibility S attributes (ranging between 1.41 and 4.24, where V < 2.64 equals to low risk and V > 3.18 high respectively). The method does not provide insights on the actual status of a species but gives an idea of how potentially vulnerable it is tofishing relative to other species based on its biological properties, useful to determine which ones should be prioritized to be investigated further.
The PSA method used here was based on Hobday et al. (2007, 2011). This method utilizes a precautionary approach to data-deficiency in the scoring of attributes, where an attribute is associated to high-risk if data is lacking or of poor quality. Species included were all native Swedish cyprinid species based on the Encyclopedia of the Swedish Flora and Fauna (Kullander et al.2012), i.e., excludingfive introduced species. This was motivated from that they do not belong to the native fauna and is thus of less con-servation concern; in total 16 species were included. Three main target species of the current commercial pound net fisheries were also included in the analysis for comparison: pike Esox lucius, perch Perca fluviatilis and pikeperch Sander lucioperca. The same set of productivity attributes were used as in Hobday et al. (2011; Table1). Cut-offs for low-medium-high productivity was based on the distribu-tion of all native freshwater species registered in Swedish lakes (37 species found at the Swedish Species Initiative 2019, excluding sub-species), following the recommended approach. The productivity attributes were primarily
collected from the Encyclopedia of the Swedish Flora and Fauna (Kullander et al.2012). When data was lacking (on for example trophic level, maturity size) FishBase (2019) was used. When ranges were provided, the value resulting in the lowest productivity score (i.e., highest risk) was used in order to be precautionary. For details on all life history parameters and risk levels used see Online resource 1. The total productivity score for each cyprinid was calculated as the average risk value for all attributes.
Susceptibility scores were based on the data collected from fishers during the interviews (fishing depths, post capture mortality), together with information found on FishBase (2019) on species depth range and fishing reg-ulations by SwAM (HVMFS2014:22) on mesh sizes (Table 2). In a few cases, data was extracted from reports published after finished reduction fishery efforts (Sandström 2011; Annadotter et al. 2013). Further information on the fish-eries, such as additional methods utilized only in reduction fisheries, was collected from interviews with fishers. Since freshwaterfish is restricted by geographic barriers (different lakes) there is a high chance of occurrence of local stocks. The use of a stock structure proxy (as presented by Hobday et al. (2007) was thus deemed most appropriate for the Availability score (set to 3, or high risk). There was no information found on depth ranges for the cyprinids, nor habitat overlap of species and fisheries, thus giving the attribute Encounterability a value of 3 (high risk). Cut-off values for selectivity was following the approach for gillnet, i.e., based on mesh size relative to size for maturity/total size (e.g. high risk equals to sizes >2 × mesh size). The mesh size of the pound nets is generally very small, but passive entrapment gear is by law required to have escape openings with a diameter of 60 mm (HVMFS 2014:22). This diameter was considered the mesh size of the com-mercial pound nets. Post capture mortality cut-off attribute was set to 3 (high risk) in the currentfisheries if the species represented a considerable part (more than 10%) of the total pound net landings (SwAM 2018). Other species were given a score of 2 (medium risk) since they are discarded alive and have a high chance for survival based onfishing Table 1 Life history attributes
and productivity cut-offs used based on Swedish
freshwaterfishes
Productivity attribute Low productivity (high risk)
Medium productivity (medium risk)
High productivity (low risk)
Age at maturity (years) >6 4–5 <4
Maximum age (years) >21 12–20 <10
Fecundity (eggs per year) <8000 8000–80,000 >100,000 Maximum size (cm) >60 30–60 <30 Size at maturity (cm) >30 14–30 <14 Trophic level >3.7 3.2–3.7 <3.2
method. A multiplicative approach was used for calculating the average susceptibility score following Hobday et al. (2011), except for the reduction fishery where the overall susceptibility score for the cyprinids to was assumed to be 3 (high) since thefishing aims at removing as much cyprinid biomass as possible.
Evaluation of a Potential Future Carp Fishery
For environmental assessment of a potential increased uti-lization of cyprinids within the currentfishery operations in terms of carbon footprint, ecological risks and available volumes it was assumed that:(1) landing the today discarded cyprinids would not require morefishing trips or more fuel per trip; (2) there are no re-catches of the released cyprinids; and (3) all are of sizes of commercial interest.
The risk score for post capture mortality was adjusted to 3 (high risk) for all cyprinid species in the PSA, and the current fuel use was attributed to higher landing volumes than current (100% retainment of today discarded cyprinids) for the carbon footprint assessment.
To scale up to estimate the potential biomass of cyprinids available to futurefisheries from all lakes, cyprinid catches was collected from the Swedish lake monitoring database (NORS2019) and converted to g of cyprinids per m2of net. The largest nets, Bss (‘översiktsnät typ Stora Sjöarna’), were excluded from the estimate, since these were only used in lake Vänern and are disproportionally ineffective in catching cyprinids compared to smaller nets. The abun-dance from the geographically closest monitoring fishery site was assumed to be the same as on the stated fishing location. Abundance data from 2015 to 2017 was used for the commercial fisheries, whereas for the reduction fish-eries, data from the year before the reductionfishing took place was used if available. When not available, the data
from the year closest to thefishing occasion was used (two cases dating back to 1997).
Current commercial use was also collected from Fish-Base (2019) to map commercial interest, as well as status on the 2015 Swedish IUCN Red List assessment (Swedish Species Initiative2019) as an additional source on potential vulnerability.
Results
All cyprinids are categorized as least concern (LC) on the Swedish IUCN Red List except for two which are red-listed (near threatened, NT): asp Aspius aspius and vimba bream Vimba vimba (Table 3). The global commercial use of the cyprinid species varies considerably, with bream Abramis brama and crucian carp Carassius carassius being highly commercial, while belica Leucaspius delineatus, common dace Leuciscus leuciscus and gudgeon Gobio gobio being of no commercial interest today (FishBase2019).
Cyprinid landing ratio relative to other species was found to vary considerably between the different lakes, from 0.8% in Mälaren to 57.5% in Bolmen (Table4), but also between fishers within the lakes. The annual total landings of all fish and total fuel consumption does however not vary as much between lakes. This results in a relatively homogenous average fuel consumption per kgfish landed, between 0.06 and 0.18 liter of fuel per kg live weight (Table4). Based on current discard rates, cyprinid landings have the potential to increase substantially, which in turn would lead to improved fuel efficiency per kilo landed.
The reduction fisheries all target primarily cyprinids (Table 5). Small perch were occasionally included in the landings from pound nets and seines, making up the missing few percent. The seine and pound net reduction fisheries were all conducted with petrol driven engines, but the trawlers utilized diesel. The carbon footprint of the current commercial fishery in Swedish lakes up to the point of Table 2 Susceptibility attributes and cut-offs for low-medium-high-risk based on Hobday et al. (2011) and assumptions made in this study
Attribute Options This study
High susceptibility Medium susceptibility
Low susceptibility
Availability Two alternatives, (1) overlap of species range with
fishery (in %) or (2) global distribution Endemic/sub-population North hemisphere Global Encounterability Two alternatives, (1) habitat- or (2) depth overlap
withfishery (%)
High overlap with
fishing gear Medium overlapwithfishing gear
Low overlap with fishing gear Selectivity Vary by gear type, based on size (at maturity/
maximum)
Species > 2 times mesh size Species 1–2 times mesh size
Species < mesh size
Post-capture mortality
Vary by gear and species Retained species, or majority dead when released
Released alive Evidence of post-capture release and survival
landing was on average 0.77 kg CO2e/kg edible, which may
decrease by 32% in a futurefishery by landing all cyprinids that are discarded today (Fig. 1). The carbon footprint varied between the lakes in the commercialfishery but was
not statistically significant (Kruskal–Wallis H = 3.51, p = 0.17, df= 2). Emissions from reduction fisheries vary sig-nificantly depending on gears (Kruskal–Wallis H = 18.28, p= 0.0001, df = 2), where trawl is similar to commercial pound netfisheries (0.71 kg CO2e/kg edible), but seine and
pound nets have considerably lower emissions (0.03 kg and 0.10 kg CO2e/kg edible respectively).
The carbon footprint of the cyprinids is substantially lower than that of most popular seafoods consumed in Sweden today. Cyprinids caught in the current commercial fisheries contribute to only about a third of the greenhouse gas emissions compared to popular seafood products like Atlantic cod Gadus morhua, Saithe Pollachius virens, and farmed Atlantic salmon Salmo salar (Fig.2). Herring Clu-pea harengus products are the only ones with a smaller Table 3 Native cyprinids in
Sweden, IUCN Red List status and commercial utilization globally
Common name Species Maximum size IUCN status Commercial use
Asp Aspius aspius 120 cm NT Yes
Belica Leucaspius delineatus 12 cm LC No
Bleak Alburnus alburnus 16 cm LC Minor
Bream Abramis brama 80 cm LC Highly
Chub Squalius cephalus 45 cm LC Minor
Common dace Leuciscus leuciscus 35 cm LC No
Crucian carp Carassius carassius 64 cm LC Highly
Eurasian minnow Phoxinus phoxinus 12 cm LC Minor
Gudeon Gobio gobio 13 cm LC No
Ide Leuciscus idus 24 cm LC Yes
Roach Rutilus rutilus 50 cm LC Yes
Rudd Scardinius erythrophtalmus 35 cm LC Minor
Tench Tinca tinca 60 cm LC Yes
Vimba bream Vimba vimba 50 cm NT Minor
White bream Blicca bjoerkna 45 cm LC Minor
Zope Ballerus ballerus 45 cm LC Minor
Table 4 Inventory results from the commercialfishery with standard deviation, separated by lake Lake % Cyprinids in
landings
Total landing per fisher (tonnes)
Cyprinid landing per fisher (tonnes)
Potential cyprinid landing perfisher (tonnes)
Fuel use per fisher (m3)
Fuel efficiency (liter/ live-weight)
Vänern 19.0 18.2 ± 12.6 3.5 ± 6.0 18.8 ± 14.6 2.2 ± 1.2 0.13 ± 0.07
Mälaren 0.8 20.7 ± 6.7 0.2 ± 0.4 8.0 ± 6.2 3.5 ± 1. 4 0.18 ± 0.06
Bolmen 57.5 40 23 32 2.2 0.06
Average 14.1 21.2 ± 10.2 3.0 ± 6.9 13.5 ± 11.9 3.0 ± 1. 4 0.15 ± 0.07
Table 5 Inventory results from the reductionfisheries with standard deviations, divided into groups based on gear type
Gear % Cyprinids in landings
Total landing per lake (tonnes)
Cyprinid landing per lake (tonnes)
Fuel use per lake (m3) Fuel efficiency (liter/live-weight) Pound nets 98% 16.3 ± 15.6 15.9 ± 15.4 0.22 ± 0.14 0.02 ± 0.01 Seine 99% 23.7 ± 19.1 23.4 ± 19.2 0.10 ± 0.06 0.005 ± 0.004 Trawl 100% 65.5 ± 22.3 65.5 ± 22.3 7.58 ± 3.01 0.12 ± 0.05 Average 99% 28.8 ± 25.9 28.5 ± 25.9 1.65 ± 3.27 0.04 ± 0.05 0.00 0.20 0.40 0.60 0.80 1.00
Current Potential Pound net Seine Trawl
Commercial pound net Reduction fisheries
CO
2
e/kg edible
Fuel combustion Fuel production
Fig. 1 Average carbon footprint of commercial and reductionfisheries (separated by gear) with contribution of fuel combustion and pro-duction to the total calculated average carbon footprint
carbon footprint compared to the cyprinids from the current commercial fishery, whereas brown beans has a similar carbon footprint to current commercial cyprinidfisheries.
In comparison to otherfisheries for which there is fuel data, Swedish freshwater pound net and reductionfisheries are amongst the most fuel-efficient fisheries (Fig. 3), with only fisheries of small pelagic fish (<30 cm) being more efficient.
The PSA indicated different risk levels of the species, but none had a higher vulnerability score than current target species (Table 6). Cyprinids had an average productivity score of 1.93 (SD 0.41) whereas the targeted freshwater species had an average productivity score of 2.81 (SD 0.08), i.e. higher risks than any of the individual cyprinid species. The cyprinids had an average vulnerability score of 2.95 (SD 0.48) in the currentfisheries, increasing to 3.40 (SD 0.52) in a potential futurefishery landing all cyprinids as a result of the increased susceptibility score (post-capture mortality). These estimates are however associated with a lot of uncertainties. Even if most productivity attributes were known (except for size at maturity for two cyprinid species), several susceptibility attributes were given a high-risk based on data deficiency. Out of the 16 considered cyprinid spe-cies, four species are considered having low vulnerability, eight medium and four high respectively to the current freshwater pound netfishery. In a future fishery with 100% retainment of cyprinids, the vulnerability scores increase for all species in the family.
Discussion and Conclusion
The low carbon footprint of current commercial fisheries catching cyprinids in Swedish lakes clearly indicates that it is a promising commodity in this sense compared to most of the commonly consumed protein sources in Sweden today— even compared to most other globalfisheries. An even lower carbon footprint per kilo may be achieved simply by
utilizing more of the cyprinids that are discarded today. The cyprinids caught in reduction fisheries with seine have an even lower carbon footprint, lower than any other fishery type studied Parker et al. (2018). However, since thisfishery is aiming at eradicating fish, the catch efficiency and asso-ciated carbon footprint may not be seen as a long-term sustainable food production but rather a underutilized resource; the biomass caught should be further investigated in terms of usefulness as feed or food depending on quality and supply chains enabled.
The carbon footprint does not however, convey infor-mation on potential local ecological risks with thefisheries. Jointly considering the carbon footprint results with those of the productivity susceptibility analysis (PSA) is important since the PSA indicated that the different cyprinid species may have different vulnerability to fishing pressure. As a result, even if the carbon footprint of current fisheries (which primarily target bream) would decrease in afishery where more cyprinids are utilized, potential risks increase, advocating that expansion offishing should be accompanied by suitable monitoring. PSA is designed to have a precau-tionary approach and is only an initial screening of relative risks for species to support prioritization of further effort by management and research (Hobday et al. 2011). The assessment was associated with uncertainties and could potentially be improved by further method adaptation to freshwater conditions since life history cues for extinction risk may not be correlated for marine and freshwaterfishes (Reynolds et al. 2005; Olden et al. 2007). There is to our knowledge only one example of when PSA has been used to assess the vulnerability in a freshwater context (between populations of Arctic char Salvelinus alpinus in Canada; Roux et al.2011). Even so, none of the species was found to have a potentially higher vulnerability than current target species in Swedish lakefisheries.
The combined carbon footprint and ecological risk-based approach utilized here can illustrate potential trade-offs in sustainability when evaluating new seafood resources from
0 5 10 15 20 25 30 Beef Pork Chicken Brown beans Cod
Herring Saithe Shrimp
Farmed salmon
Commercial
Potential
Agriculture Seafood Cyprinid
fishing
kg CO
2
e/kg edible
Fig. 2 Average carbon footprints for the cyprinids based on this study in relation to other more popular protein sources consumed in Sweden today (based on data from RISE Climate Database2018)
0.00 2.00 4.00 6.00 8.00 10.00 kg CO 2 e/kg
Fig. 3 Average carbon footprint for one kg of live-weight fish, including theoretical estimations for upstream production of boats, gear etc., for the Swedish freshwaterfisheries examined in this study (labeled as Freshwater fish), relative to marine seafood categories presented by Parker et al. (2018)
capturefisheries (e.g., freshwater resources, bycatches), and identify different improvement potentials and associated constraints. Combining elements of the two tools, which both are used to support decision making in environmental man-agement, has gained increase interest in recent years (Harder et al.2015). In a fisheries context, this is the second study utilizing the combined approach, the other being a study on a vastly differentfishery in the form of the Australian fisheries for Patagonian toothfish Dissostichus eleginoides (Hornborg et al.2018). The toothfish study showed a development of the fishery from high to low ecological risks through management actions taken, but the opposite for carbon footprint develop-ment. The latter is not considered by management authorities but of interest to industry and other societal actors. These results are both opposite and similar to this study. Similar in the sense of being cases of exploitation of new resources, with potential ecological risks higher at start (especially the reduction fisheries) but carbon footprint lower. Opposite, since the assessment indicate low carbon footprint of Swedish cyprinids but higher potential ecological risks due to data deficiency, the latter requiring management effort.
Cyprinids as a New Resource in the Swedish Food
System
Improved utilization of current cyprinid resources offers an opportunity to improve food and nutrition security, given
proper measures are taken to minimize potential risks for the species with increased exploitation. From a nutritional perspective, seafood from marine capturefisheries provides important contributions of micronutrients as e.g. shown recently by Hicks et al. (2019). Freshwater fishes were excluded in the analysis due to data deficiencies, but there is no evidence for them having in general lower nutritional value. They may rather have benefits compared to marine species, such as a higher ability to produce longer-chain omega 3 fatty acids out of vegetable oils when farmed (Khalili Tilami, Sampels 2018). Comparing some of the most important nutrients between the freshwater bream (based on the Swedish National Food agency (NFA)) and popular marinefish species (based on Hornborg et al.2019) show that the cyprinid has higher concentration of vitamin B12, selenium, and niacin than both Atlantic cod Gadus morhua and farmed Atlantic salmon Salmo salar. The concentration of polyunsaturated fatty acids in bream is lower than in Atlantic salmon, but on the other hand, sub-stantially higher than in Atlantic cod. Therefore, increased utilization of cyprinids in the Swedish seafood diet may improve nutritional security.
In a recent report on nutritional and climate aspects of Swedish seafood consumption (Hornborg et al. 2019), cyprinids were not included as they were deemed irrelevant due to the negligible consumption. In 2017, 52 tonnes of carpfishes were available for consumption in Swedish retail Table 6 PSA results for the sixteen Swedish cyprinids and three target species in the current commercialfishery and a potential land-all scenario Group Species Productivity Susceptibility,
current Vulnerability, current Risk category, current Susceptibility, potential Vulnerability, potential Risk category, potential
Carps Asp 2.43 2.33 3.36 High 3.00 3.86 High
White bream 1.86 2.33 2.98 Med 3.00 3.53 High
Eurasian minnow 1.43 1.43 2.02 Low 1.65 2.18 Low
Zope 2.00 2.33 3.07 Med 3.00 3.61 High
Chub 2.14 2.33 3.16 Med 3.00 3.69 High
Belica 1.57 2.33 2.81 Med 3.00 3.39 High
Ide 2.71 2.33 3.57 High 3.00 4.05 High
Bleak 1.29 1.88 2.27 Low 2.33 2.66 Med
Roach 2.00 2.33 3.07 Med 3.00 3.61 High
Gudeon 1.29 1.88 2.27 Low 2.33 2.66 Med
Rudd 1.71 1.88 2.54 Low 2.33 2.89 Med
Vimba bream 2.14 2.33 3.16 Med 3.00 3.69 High
Bream 2.29 3.00 3.77 High 3.00 3.77 High
Crucian carp 2.29 2.33 3.26 High 3.00 3.77 High
Common dace 1.71 2.33 2.89 Med 3.00 3.46 High
Tench 2.00 2.33 3.07 Med 3.00 3.61 High
Target Perch 2.71 3.00 4.05 High 3.00 4.05 High
Pike 2.86 3.00 4.14 High 3.00 4.14 High
(Borthwick et al.2019), or 10 grams per person and year (based on the Swedish population in 2019; SCB2019b). A future fishery when all cyprinids are retained in the com-mercialfishery and extrapolated to all active fishers in the lakes included in the study could tentatively provide almost 228 tonnes edible product. This would still only represent less than 1% of the increase in fish consumption recom-mended by the Swedish NFA. If the average current discard level of cyprinids (22%) was extrapolated to all freshwater pound netfishers, including other Swedish lakes, the total landing could be 350 tonnes edible product. This may cause competition between bait and food, especially in lake Vät-tern, but available volumes could also increase further by including catches from coastal fisheries in the Baltic Sea (were abundance of, e.g., roach is increasing; Ådjers et al. 2006), which were not considered here. Further investiga-tions on potential biomass available and associated risks with exploitation are however required since species com-position varies between lakes (Lehtonen et al.2008), which should also include coastalfisheries.
The potential maximum volume available to human consumption from the reductionfisheries is more difficult to estimate since there is no information on how many lakes that would benefit from reduction fisheries. It would also require more research related to product quality and suitable supply chains to ensure high-quality products. If not suited for human consumption, they should be suitable for other purposes like animal feed, resources much sought after today (Troell et al. 2014). The estimate for potential cyprinid volume available to fishing have in general no information on size- and species-composition, and small individuals may not be of commercial interest. The smallest cyprinids are on the other hand not caught in today’s commercial pound netfisheries, due to the 60 mm escape openings, but are likely caught in reductionfisheries. With these aspects providing uncertainties to the estimates, and by (i) assuming that all Swedish lakes with poor or unsa-tisfying ecological status would benefit from reduction fishing of cyprinids, and (ii) the catch per hectare in the reduction fisheries included in this study can be extra-polated to all lakes of respective ecological status, a total catch of 3050 tonnes (edible yield) could be possible. This would represent around 8% of the increase in fish con-sumption recommended by the Swedish NFA. This is however not a stable yearly catch, as reductionfisheries are meant to reduce the cyprinid densities considerably, but some improved utilization than current (at best used for biogas today) should be possible; higher value is also of interest tofishers.
Iffishing effort on cyprinids were to increase, monitoring of stock status development is needed where the contribu-tion of this study has only been to identify the potential and relative risks between different species tofishing pressure,
i.e., not an assessment on absolute risk. The arguable most important question at this point is however how to re-gain consumer interest in Sweden. This has been lost during the past decades with globalized markets (Bonow et al.2013). Only three of the sixteen cyprinid species assessed here have no commercial interest in a global perspective, but all have today low consumer interest in Sweden. Some species that should be fully avoided include red-listed species (asp and zope), and belica, the latter on which fishery is also forbidden in Sweden (FIFS 2004:37). Examination of potential occurrence of hazardous substances (such as heavy metals or toxins) in cyprinids caught in different lakes are also needed. This is especially true for the cyprinids caught in reduction fisheries, since lakes with a high inflow of nutrients also often can potentially have a high inflow of other substances as well.
Lastly, seafood is today a highly traded, global com-modity (FAO 2018), with pitfalls and promises. The knowledgebase concerning freshwater resources and their vulnerability to effects from climate change differ between countries (Deines et al.2017; Lynch et al.2017), calling for increased attention by managers. This may improve by increased consumer interest. A shift back towards improved utilization of local freshwater resources in a country such as Sweden would be an opportunity for improving the current food system (sensu Gordon et al. 2017). This is perhaps especially needed for fish and shellfish since consumers are today distant from the local ecosystem where it is produced and thus fail to see eco-system effects from demand (Crona et al. 2016). Building resilience in food systems in light of climate change is also urgent, since food production shocks are increasing (Cot-trell et al. 2019). Nordic countries have in a European perspective a more vulnerable seafood production sector today than in the past since they are much reliant on temperature-sensitive species—whereas species such as common carp is more tolerant to higher temperatures compared to e.g. salmonids (Blanchet et al.2019), the top consumed seafood product in Sweden today (Borthwick et al. 2019). Freshwater cyprinids offer an opportunity to Swedish consumers, at this point calling for improved management focus and supply chains.
Acknowledgements The authors want to thank the fishers who pro-vided the data and Jonatan Fogel who assisted in interviewingfishers in lake Mälaren. SH want to acknowledge funding from the Swedish Research Council Formas in the form of two grants, SEAWIN (2016-00227) and mobility grant (2016-00455). Open access funding pro-vided by RISE Research Institutes of Sweden.
Compliance with Ethical Standards
Conflict of Interest The authors declare that they have no conflict of interest.
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://crea tivecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.
References
Annadotter H, Forssblad J, Hagenrud K (2013) Utveckling av fångstmetoder och vård av fisksamhället i Finjasjön (Report 2013-08-v01). Vittsjö, Regito AB (in Swedish).https://www.ha ssleholm.se/download/18.23f29b73169d88cef7b1cb24/
1554270281297/2013%2011%2015%20bilaga%202%20Bosarpa sj%C3%B6n.pdf
Ådjers K, Appelberg M, Eschbaum R et al. (2006) Trends in coastal fish stocks of the Baltic Sea. Boreal Environ Res 11(1):13–25 Arlinghaus R, Mehner T, Cowx IG (2002) Reconciling traditional
inland fisheries management and sustainability in industrialized countries, with emphasis on Europe. Fish Fish 3(4):261–316 Avadí A, Fréon P (2013) Life cycle assessment offisheries: a review
forfisheries scientists and managers. Fish Res 143:21–38 Blanchet MA, Primicerio R, Smalås A et al. (2019) How vulnerable is
the European seafood production to climate warming? Fish Res 209:251–258
Bonow M, Svanberg I (2013) Karpfiskarnas tillbakagång i svenskt kosthåll. In: Från matproduktion till gastronomi. Rytkönen P, Bonow M, Wrammer P (eds). COMREC Studies in Environment and Development 7. Södertörns Högskola. Huddinge, 2013. p 241 Borthwick L, Bergman K, Ziegler F (2019) Svensk konsumtion av sjömat. RISE report 2019:27. Göteborg: RISE Research Institutes of Sweden AB. https://www.diva-portal.org/smash/get/diva2: 1305161/FULLTEXT01.pdf. Accessed June 2019
Cottrell RS, Nash KL, Halpern BS et al. (2019) Food production shocks across land and sea. Nat Sustainability 2(2):130 Crona BI, Daw TM, Swartz W et al. (2016) Masked, diluted and
drowned out: how global seafood trade weakens signals from marine ecosystems. Fish Fish 17(4):1175–1182
Deines AM, Bunnell DB, Rogers MW et al. (2017) The contribution of lakes to global inlandfisheries harvest. Front Ecol Environ 15 (6):293–298
Dudgeon D, Arthington AH, Gessner MO et al. (2006) Freshwater biodiversity: importance, threats, status and conservation chal-lenges. Biol Rev 81(2):163–182
FIFS (2004:37) Fiskeriverkets föreskrifter omfiske i sötvattensområ-dena. (in Swedish). https://www.havochvatten.se/hav/va gledning–lagar/foreskrifter/register-fiskereglering/fiske-i-sotva ttensomradena-fifs-200437.html
FAO (1989) Yield and Nutritional Value of the Commercially More Important Fish Species. FAO Fisheries and Aquaculture Department
FAO (2018) The State of World Fisheries and Aquaculture 2018 -Meeting the sustainable development goals. Rome
FishBase (2019) Creative Commons Attribution-Noncommercial 3.0 Unported License.https://www.fishbase.se/search.php. Accessed Mar–Apr 2019
Godfray HCJ, Beddington JR, Crute IR et al. (2010) Food security: the challenge of feeding 9 billion people. Science 327 (5967):812–818
Gordon LJ, Bignet V, Crona B et al. (2017) Rewiring food systems to enhance human health and biosphere stewardship. Environ Res Lett 12(10):100201
GoV (2016) Prop. 2016/17:104 En livsmedelsstrategi för Sverige − fler jobb och hållbar tillväxt i hela landet (in Swedish).
https://www.regeringen.se/490897/contentassets/256cc25ab5a 84db7a76730abb9cc3773/en-livsmedelsstrategi-for-sverige- fler-jobb-och-hallbar-tillvaxt-i-hela-landet-prop-2016-17-104.pdf
Hallström E, Bergman K, Mifflin K et al. (2019) Combined climate and nutritional performance of seafoods. J Clean Prod 230:402–411 Hansson LA, Annadotter H, Bergman E et al. (1998) Biomanipulation
as an application of food-chain theory: constraints, synthesis, and recommendations for temperate lakes. Ecosystems 1(6):558–574 Harder R, Holmquist H, Molander S et al. (2015) Review of envir-onmental assessment case studies blending elements of risk assessment and life cycle assessment. Environ Sci Technol 49 (22):13083–13093
Hicks CC, Cohen PJ, Graham NA et al. (2019) Harnessing globalfisheries to tackle micronutrient deficiencies. Nature 574(7776):95–98 Hobday AJ, Smith ADM, Webb H et al. (2007) Ecological Risk
Assessment for the Effects of Fishing: Methodology. Report R04/ 1072 for the Australian Fisheries Management Authority, Can-berra. http://www.fish.gov.au/reports/Documents/2014_refs/Hob day%20et%20al%202007%20ERA.pdf
Hobday AJ, Smith ADM, Stobutzki IC et al. (2011) Ecological risk assessment for the effects offishing. Fish Res 108(2–3):372–384 Hordyk AR, Carruthers TR (2018) A quantitative evaluation of a qualitative risk assessment framework: examining the assump-tions and predicassump-tions of the productivity susceptibility analysis (PSA). PLoS ONE 13(6):e0198298
Hornborg S, Hobday AJ, Ziegler F et al. (2018) Shaping sustainability of seafood from capturefisheries integrating the perspectives of supply chain stakeholders through combining systems analysis tools. ICES J Mar Sci 75(6):1965–1974
Hornborg S, Hallström E, Ziegler F, et al. (2019) Frisk medfisk utan risk? Betydelsen av svensk konsumtion av sjömat för hälsa och miljö. RISE Rapport 2019:38. Göteborg: RISE Research Insti-tutes of Sweden AB. https://ri.diva-portal.org/smash/get/diva2: 1305158/FULLTEXT02.pdf
HVMFS (2014:22) Havs- och vattenmyndighetens föreskrifter om ändring i Fiskeriverkets föreskrifter (FIFS 2004:37) om fiske i sötvattensområdena (HVMFS 2014:22). (in Swedish)https://www. havochvatten.se/hav/vagledning –lagar/foreskrifter/register-fiskereglering/fiske-i-sotvattensomradena-fifs-200437.html
Khalili Tilami S, Sampels S (2018) Nutritional value of fish: lipids, proteins, vitamins, and minerals. Rev Fish Sci Aqua 26(2):243–253 Kruskal W, Wallis WA (1952) Use of ranks in one-criterion variance
analysis. J Am Stat Assoc 47(260):583–621
Kullander SO, Nyman L, Jilg K, Delling B (2012) Encyclopedia of the Swedish Flora and Fauna. Rayfinned fishes. Arctinopterygii. Artdatabanken, SLU, Uppsala. p 517
Lehtonen H, Rask M, Pakkasmaa S, Hesthagen T (2008) Freshwater fishes, their biodiversity, habitats and fisheries in the Nordic countries. Aquat Ecosyst Health Manag 11(3):298–309 Lorenzen K, Cowx IG, Entsua-Mensah REM et al. (2016) Stock
assessment in inlandfisheries: a foundation for sustainable use and conservation. Rev Fish Biol Fish 26(3):405–440
Lynch AJ, Cowx IG, Fluet-Chouinard E et al. (2017) Inland fisheries–Invisible but integral to the UN Sustainable Develop-ment Agenda for ending poverty by 2030. Glob Environ Change 47:167–173
McIntyre PB, Liermann CAR, Revenga C (2016) Linking freshwater fishery management to global food security and biodiversity conservation. PNAS 113(45):12880–12885
Ness B, Urbel-Piirsalu E, Anderberg S, Olsson L (2007) Categorising tools for sustainability assessment. Ecol Econ 60(3):498–508 NFA Sweden (2019) National Food Agency, Sweden. Fish and
shellfish – advice. https://www.livsmedelsverket.se/en/food-ha bits-health-and-environment/dietary-guidelines/adults/ fisk-och-skaldjur—rad.Accessed 20 May 2019
NORS (2019) Nationellt Register över Sjöprovfisken – NORS 2019. Sveriges lantbruksuniversitet (SLU), Institutionen för akvatiska resur-ser.http://www.slu.se/sjoprovfiskedatabasen. Accessed 24 Apr 2019 Olden J, Hogan ZS, Zanden MJV (2007) Smallfish, big fish, red fish,
bluefish: size‐biased extinction risk of the world's freshwater and marinefishes. Glob Ecol Biogeogr 16(6):694–701
Östman Ö, Beier U, Ragnarsson Stabo H, et al. (2016) Förvaltningsmål för nationellt förvaltadefiskbestånd (No. 2016:10) (in Swedish).
https://pub.epsilon.slu.se/14134/8/ostman_o_et_al_170329.pdf
Parker RW, Tyedmers PH (2015) Fuel consumption of globalfishing fleets: current understanding and knowledge gaps. Fish Fish 16(4):684–696 Parker R. W., Blanchard J. L., Gardner C. et al. (2018) Fuel use and greenhouse gas emissions of worldfisheries. Nat Clim Change 8 (4):333
Pascual U, Palomo I, Adams WM et al. (2017) Off-stage ecosystem service burdens: a blind spot for global sustainability. Environ Res Lett 12(7):075001
Reynolds JD, Webb TJ, Hawkins LA (2005) Life history and ecolo-gical correlates of extinction risk in European freshwaterfishes. Can J FAS 62(4):854–862
RISE Climate Database (2018) Öppna listan– ett utdrag från RISE klimatdatabas för livsmedel v 1.5.https://www.ri.se/sv/v%C3% A5ra-ber%C3%A4ttelser/klimatdatabas-smartare-ma
tkonsumtion. Accessed 5 May 2019
Roux MJ, Tallman RF, Lewis CW (2011) Small‐scale Arctic charr Salvelinus alpinus fisheries in Canada’s Nunavut: management challenges and options. J Fish Biol 79(6):1625–1647
Sandström O (2011) Reduktionsfiske som metod för att minska över-gödningen i Östhammarsfjärdarna. SKUTAB & Upplands-stiftelsen. (in Swedish) http://www.upplandsstiftelsen.se/ UserFiles/Archive/5417/Rapporter/2011_2_Reduktionsfiske.pdf
SCB (2019a) Statistics Sweden. Land- och vattenareal per den 1 januari efter region och arealtyp. År 2012–2019.http://www.sta tistikdatabasen.scb.se/pxweb/sv/ssd/START__MI__MI0802/
Areal2012/?rxid=55ac2603-a882-49a2-8d81-72ade05d2451. Accessed 20 May 2019
SCB (2019b) Statistics Sweden. Sveriges befolkning (The population of Sweden). https://www.scb.se/hitta-statistik/sverige-isiffror/ma nniskorna-i-sverige/sveriges-befolkning. Accessed 20 May 2019 Smith M, Roheim C, Crowder LB et al. (2010) Sustainability and
global seafood. Science 327(5967):784–786
SPBI (2010) Swedish Petroleum and Biofuels Institute SPBI. Ener-giinnehåll, densitet och koldioxidutsläpp. https://spbi.se/uppsla gsverk/fakta/berakningsfaktorer/energiinnehall-densitet-och-koldioxidemission/Accessed 4 Apr 2019
SwAM (2018) Fångststatistik för yrkesfisket. https://www.havochva tten.se/hav/samordning –fakta/data–statistik/fangststatistik-yrkesfisket.html(Catch statistics). Accessed 15 May 2019 SwAM (2019a) Statistics on Swedish freshwater catches andfishing
efforts. Requested and received in March 2019.
SwAM (2019b) Fisk- och skaldjursbestånd i hav och sötvatten 2018 Resursöversikt. Göteborg: Havs och Vattenmyndigheten (Over-view offish resources in sea and lakes 2018). (in Swedish)www. havsochvattenmyndigheten.se
SwAM (2019c) Projekt för en bättre havs- och vattenmiljö (Projects for improved sea- and freswater environment, in Swedish).
https://projektkatalog.havochvatten.se/db04p01/f?p=108:1. Accessed 30 May 2019
Swedish Species Initiative (2019) Artportalen.https://www.artdataba nken.se/sok-art-och-miljodata/artportalen/Accessed Apr 2019 Tacon AG, Metian M (2013) Fish matters: importance of aquatic foods
in human nutrition and global food supply. Rev Fish Sci 21 (1):22–38
Tilman D, Clark M (2014) Global diets link environmental sustain-ability and human health. Nature 515(7528):518
Thurstan RH, Roberts CM (2014) The past and future offish con-sumption: can supplies meet healthy eating recommendations? Mar Pol Bull 89(1–2):5–11
Troell M, Naylor RL, Metian M et al. (2014) Does aquaculture add resilience to the global food system? PNAS 111 (37):13257–13263
Ziegler F, Hornborg S, Green BS et al. (2016) Expanding the concept of sustainable seafood using life cycle assessment. Fish Fish 17 (4):1073–1093
