Environmental Research Letters
LETTER • OPEN ACCESS
Indonesian aquaculture futures—identifying interventions for reducing
environmental impacts
To cite this article: Patrik John Gustav Henriksson et al 2019 Environ. Res. Lett. 14 124062
 
View the article online for updates and enhancements.
This content was downloaded from IP address 42.189.140.170 on 24/04/2020 at 07:05
Environ. Res. Lett. 14 (2019) 124062 https://doi.org/10.1088/1748-9326/ab4b79
LETTER
Indonesian aquaculture futures—identifying interventions for
OPEN ACCESS reducing environmental impacts
RECEIVED
26April 2019 Patrik JohnGustavHenriksson1,2,3,7 , LaurenKBanks2,4 , SharonKSuri2,5 , Trini YPratiwi2,6,
REVISED NurulhudaAhmad Fatan2 andMaxTroell1,3
3October 2019
1 StockholmResilience Centre, StockholmUniversity, Kräftriket 2B, SE-10691, Sweden
ACCEPTED FOR PUBLICATION 2
7October 2019 WorldFish, Jalan BatuMaung, 11960 Penang,Malaysia
3 Beijer Institute of Ecological Economics, The Royal SwedishAcademy of Science, Box 50005, 104 05 Stockholm, Sweden
PUBLISHED 4 Department ofGeography,WesternUniversity &CanadianRivers Institute, London, Ontario, Canada
16December 2019
5 Department of Anthropology andDepartment ofGeography, Planning and International Development Studies, University of
Amsterdam,NieuweAchtergracht 166, 1018WVAmsterdam, TheNetherlands
Original content from this 6 IDHThe Sustainable Trade Indonesia,MDPlace Tower 2, 3rd FloorUnit C, Jl. Setiabudi SelatanNo. 7, Jakarta 12910, Indonesia
workmay be used under 7 Author towhomany correspondence should be addressed.
the terms of the Creative
CommonsAttribution 3.0 E-mail: patrik.henriksson@beijer.kva.se
licence.
Any further distribution of Keywords: seafood, LCA, sustainability, land use, coral,mangrove
this workmustmaintain
attribution to the Supplementarymaterial for this article is available online
author(s) and the title of
thework, journal citation
andDOI.
Abstract
Indonesia is theworld’s second largest producer and third largest consumer of seafood. Fish is
therefore essential to the nation, bothfinancially and nutritionally. Overfishing and the effects of
climate changewill, however, limit future landings of capturefisheries, so any increases in future
seafood productionwill need to come from aquaculture. The ecological effects of aquaculture are
dependent upon the choice of species,management, andwhere it is sited. In the present studywe use
life cycle assessment (LCA) to evaluate howpossible interventions and innovations canmitigate
environmental impacts related to the aquaculture sector’s growth. Themitigation potential of six
interventions were also quantified, namely (1) FCR reductions for whiteleg shrimp, carp, and tilapia;
(2) sustainable intensification ofmilkfish andAsian tiger shrimp polyculture; (3) shifting groupers
fromwholefish diets to pellets; (4) favoring freshwaterfinfish over shrimp; (5) renewable electricity;
and (6) reduced foodwaste and improved byproduct utilization. If all six interventions are
implemented, we demonstrate that global warming, acidification, eutrophication, land occupation,
freshwater use, and fossil energy use could be reduced by between 28%and 49%per unit offish. The
addition ofmany innovations that could not be quantified in the present study, including innovative
feed ingredients, suggest that production could double within the current environmental footprint.
This does not, however, satisfy the expected 3.25-fold increase under a business-as-usual scenario,
neither does it satisfy the government’s growth targets.We therefore also explore possible
geographical areas across Indonesia where aquaculture expansions and ecological hotspotsmay
conflict. Conclusively, we advocatemore conservative production targets and investment inmore
sustainable farming practices. To accelerate the implementation of these improvements, it will be
central to identify themost cost-effective aquaculture interventions.
1. Introduction demand and increase exports (Directorate General of
Aquaculture 2017). This is while Indonesia is already
The Indonesian Ministry of Marine Affairs and Fish- the third largest seafood consuming nation, after
eries (KKP/MMAF) has set ambitious growth targets China and Japan (Guillen et al 2019), and fourth
for most aquaculture species of around 8.5% growth largest exporter of shrimp, after India, Ecuador, and
per annum up to 2030 (IDH 2018), to satisfy national Argentina (comtrade.un.org accessed 12-Sep-2019). At
©2019TheAuthor(s). Published by IOPPublishing Ltd
Environ. Res. Lett. 14 (2019) 124062
present, over 80% of capture fisheries and 95% of aquaculture sector to increase production with less
aquaculture production in Indonesia is consumed additional stress on the environment. Reflections will
domestically (Belton et al 2017). Despite this, both also be made in terms of provincial differences, with
macro- and micronutrient deficiencies remain com- regards to ecosystems, infrastructure, and additional
mon in Indonesia (Usfar and Fahmida 2011). sectors (Guilmoto and Jones 2016).
The consequent ‘nationally recommended diet’ The main research question is whether better
advises increased consumption of both meat and fish farming practices and innovations will allow Indone-
(Behrens et al 2017). sia to meet production targets up to 2030 without jeo-
Most seafood consumed in Indonesia still origi- pardizing functions of existing ecological systems. In
nates from capture fisheries (FAO 2018), but many order to address this question, we have updated the
fish stocks are fully or overexploited, with no potential dataset by Henriksson et al (2017a, 2017b) and short-
for further increase. Future fish catches may even listed possible interventions and innovations for more
decline by over 30% due to predicted climate change sustainable farming practices. The 2030 production
and other anthropogenic stressors (Cheung et al target was assumed to be in line with the business-as-
2016). This shifts emphasis toward aquaculture pro- usual scenario (BAU) from Tran et al (2017), project-
duction, which is already well established across the ing a 3.3-fold increase in production volumes by 2030
country. However, competition for suitable land and based upon the AsiaFish model. We discuss our
resources is intensifying, posing conservation chal- results, giving consideration to spatial planning and
lenges in a country with numerous biodiversity hot- important ecosystems.
spots (Gaither and Rocha 2013, Murray et al 2015),
many of which have already been degraded (Maynard
et al 2010, Abood et al 2015). In planning for sustain- 2.Material andmethods
able growth of aquaculture, consideration will there-
fore need to be given to how different species and 2.1. Primary data
variations of production systems relate to resources Building upon the lifecycle inventory dataset from
and ecosystems, both at the national and local levels. Henriksson et al (2017a, 2017b), supplementary data
Indonesia’s aquaculture industry is today domi- (available online at stacks.iop.org/ERL/14/124062/
nated by freshwater finfish species, including tilapia, mmedia) were collected for shrimp, milkfish polycul-
clarias catfish, carp, and pangasius catfish (3 277 kMT ture, and carp farming in ponds during the spring of
in total in 2017) (BPS 2018). This is followed by brack- 2018. Nineteen farms were visited on Java during this
ish water farming of shrimp and milkfish (1 621 k fieldwork, nine of whichwere common carp (Cyprinus
MT). The contribution from mariculture remains carpio) pond farms, five were whiteleg shrimp farms,
marginal (around 78 k MT), when excluding seaweed and five were Asian tiger shrimp farms (Penaeus
farming (10 547 k MT) and bivalves (50 k MT). Of monodon). Consequently, farming in Java, Sumatra,
these, shrimps are the most frequently exported Lombok and Sulawesi were represented, representing
farmed seafood commodity, both by volume and grow-out cycles between 2014 and 2017. Individual
value, followed by tilapia (FAO 2019). Whiteleg farmers were identified through local partner net-
shrimp production (Litopenaeus vannamei) was also works, with an overrepresentation of farmers prox-
the aquaculture species for which production has imal to urban centers. These data were averaged with
increased the fastest over the last five years, more than the previous dataset to create the unit process dataset
doubling in output (FAO2019). for eight species and ten systems. Bivalves farming
Previous work explored possible growth scenarios systems were not evaluated as their contribution
for the Indonesian aquaculture industry by using remainsmarginal.
growth projections from the AsiaFish model (Dey et al This dataset detail eleven farming system: tilapia in
2016, Tran et al 2017) and environmental con- cages (n=5) and in ponds (n=4); carp in cages
sequences from life cycle assessment (LCA) modeling (n=6) and in ponds (n=8); pangasius in cages
(Henriksson et al 2017b). Both these studies explored (n=6) and in ponds (n=3); clarias in ponds
six scenarios, including business as usual (BAU), stag- (n=5); milkfish in ponds (n=3); milkfish andAsian
nant capture fisheries, export-oriented aquaculture, tiger shrimp in ponds (n=8); whiteleg shrimp
domestic-oriented aquaculture, slow aquaculture (n=11); and grouper (n=5). Groupers were inclu-
growth, and a disease outbreak scenario in shrimp and ded as a representative for marine finfish, even though
carp (Henriksson et al 2017b). The major conclusions they currently make up only 1.4% of overall produc-
from these studies showed that none of the six alter- tion volume (BPS 2018), as mariculture is expected to
native scenarios explored could satisfy the MMAF expand proportionately faster than other aquaculture
growth targets, andmost would result in serious envir- sectors (Tran et al 2017). The unit process dataset was
onmental consequences both locally and globally. All averaged using the protocol for horizontal averaging
these projections built upon current aquaculture by Henriksson et al (2014), including dispersion esti-
farming practices. In this article, we evaluate possible mates around inventory flows. Asian tiger shrimp and
innovations and interventions that would allow the milkfish polyculture exhibited the largest overall
2
Environ. Res. Lett. 14 (2019) 124062
dispersions, suggesting more variable farming prac- with industry partners and academics, a list of
tices. The updated unit process data are presented in possible aquaculture interventions and innovations
the supportingmaterial (SM). (AqIs) was identified. A stakeholder workshop with
industry, NGOs, and KKP/MMAF was also orga-
2.2. Secondary data nized on 24 January 2019 in Jakarta to discuss which
Geographical data were sourced fromGiri et al (2011), of the interventions were of relevance for Indonesian
Global Forest Watch (2014), United Nations Environ- aquaculture (see SM for a complete list of interven-
ment Programme and World Conservation Monitor- tions identified during the workshop). Many inter-
ing Centre (UNEP-WCMC) (2010), and UNEP- ventions were identified, but our requirements were
WCMC (2017) comprising landuse,mangrove forests, that each intervention should: (1) be quantifiable
coral reefs, and seagrass beds. Land use was classified using empirical data; (2)make a serious contribution
as forested, including protected areas and regions of to either of the six LCA impact categories under
logging moratorium (up to 2017) or land concessions, study; and (3) should be viable to operationalize
including logging and palm oil concessions, and before 2030. Subsequently, six interventions were
plantations. The condition of mangrove forests, coral, explored: (1) FCR reductions for whiteleg shrimp,
and seagrass beds were classi ed based on an assess- carp, and tilapia (Oreochromis spp.); (2) sustainable
fi
ment by BPS-Statistics Indonesia and Subdirectorate intensification ofmilkfish (Chanos chanos) and Asian
of Enviroment Statistics (2018), and maps were tiger shrimp polyculture; (3) shifting groupers (Epi-
assembled usingArcGIS v10.5. nephelus spp.) from whole fish diets to pellets; (4)
The unit process dataset for the LCA was supple- favoring freshwater finfish over shrimp; (5) renew-
mented with methane emissions from freshwater able electricity; and (6) improved byproduct utiliza-
ponds, assuming 533 kg ha−1 yr−1 (coef cient of var- tion and reduced food waste. Novel feed ingredients
fi
iation (CV)=0.4; lognormal distribution) (Astudillo were only evaluated in theory, but not quantified as
et al 2015). Emissions resulting from land use and an intervention, as they remain highly diverse and
land-use change of mangroves were derived from rapidly evolving, subsequently LCA data remain
Järviö et al (2018), assuming 129 tonnes CO eq. ha−1 unsatisfactory. Given variable costs for implementa-
2
yr−1 (CV=0.441, lognormal distribution), but not tion and economy of scale, we choose not to estimate
included in the global warming estimates due to the monetary value or number of jobs as socioeconomic
dif culty of assigning land use and land-use change to indicators in this study, as was done in Henriksson
fi
speci c species. et al (2017a, 2017b).
fi
2.4.1. AqI1: improved feed efficiency in whiteleg shrimp,
2.3. Life cycle assessment carp, and tilapia by 20%
LCA is an ISO-certified (ISO 2006) environmental Feed conversion ratio (FCR) can be a good indicator
assessment framework that allows whole production of how efficient animals in different systems are at
chains to be evaluated. In this study, however, impacts converting feed into weight gain, although changed
were only evaluated up to farm gate, including impacts composition of feed ingredients also need to be
related to producing feed and feed raw materials, accounted for (Fry et al 2018). The reported FCRs for
electricity generation, extraction and refining of diesel, whiteleg shrimp, carp, and tilapia farms visited for
and transportation. Impacts related to processing, cold this study and by Henriksson et al (2017a, 2017b)
storage, packaging, and consumption were not con- were 1.45, 1.5–1.8, and 1.6–1.7, respectively. These
sidered, but could be assumed to be similar among are markedly higher than those reported in China
products and generally have a marginal influence on (Cao et al 2015). Thus, through a combination of
overall performance (Henriksson et al 2015). Edible improved genetic strains, better environments,
yields and nutritional content were not considered water monitoring, feeding practices, and best farm
either, but would have a stronger influence on conclu- management practices (Henriksson et al 2017a,
sions. The LCA matrix was structured and calculated Ullman et al 2019), we assumed that FCRs for these
using the CMLCA v5.2 software, relying upon ecoin- species could be lowered by 20% by 2030 using
vent v2.2 for background processes. Overall disper- existing feeds. This equates to FCRs of 1.16 for
sions were calculated over 1 000 Monte Carlo whiteleg shrimp, 1.33 for tilapia, and 1.5 for com-
simulations. Coproduct allocation was solved based mon carp; these values are on average still higher
upon mass, with results using economic allocation than actual FCRs measured by Cao et al and thus
available in the SM. For a complete goal and scope, considered conservative estimates.
please seeHenriksson et al (2017a, 2017b).
2.4.2. AqI2: sustainable intensification of milkfish and
2.4. Aquaculture interventions (AqIs) Asian tiger shrimp
Building upon a nonstructured review of peer- Integrated milkfish and Asian tiger shrimp farming
reviewed and grey literature, alongside discussions relies upon extensive farming practices, yielding less
3
Environ. Res. Lett. 14 (2019) 124062
than two tonnes per hectare at two crops a year. is generated by burning coal, but Indonesia is also the
Farming of these species has expanded in mangroves third-largest producer of electricity from geothermal
and resulted in biodiversity loss, and land use and power plants and holds an estimated 40% of the
land-use change emissions. Moreover, extensive world’s geothermal potential, the equivalent of 28 000
brackish water farming consumes large volumes of megawatts (Hasan et al 2012, Semedi et al 2017).
fresh water for diluting marine water. Thus, by Electricity from large-scale (>100MW) geothermal
increasing the use of high quality extruded feed pellets plants is cheaper than from coal power plants, but the
to an FCR of 1.4 alongside better feeding practices, initial costs and risk of return on investment are higher
improved genetic strains, and overall better manage- (Clauser and Ewert 2018). Indonesia also has great
ment practices, the stocking density could be increased potential to expand both hydro and wind power. In
to only require half the area for the same amount of 2016, however, only 4% of all electricity generated in
shrimp andfish (Rimmer et al 2013,White et al 2018). Indonesia came from geothermal sources and 13%
from renewable sources (iea.org; accessed 8 March
2.4.3. AqI3: shifting from low-cost fish to feed pellets as 2019). In AqI5 we consequently aspire to all electricity
feed for groupers in Indonesia originating from renewable sources by
The majority of grouper farms visited used low-cost 2030. This means that all farms, feed mills, fishmeal
fish as the main feed source, with 18 kg of whole fish factories, and crop mills in Indonesia would run on
used per kg grouper on average. Low-cost fish renewable electricity.
generally results in poor FCRs as it breaks up and is
partially lost from cages (Sim et al 2005). Using whole
sh is also problematic because of its short shelf-life 2.4.6. AqI6: improved byproduct utilization and waste
fi
and variable availability (Sim et al 2005) and diverts a reduction
source of nutritionally-rich food away from direct The most efficient and sustainable way to increase
human consumption domestically (Buchary 2010, food availability is to reduce waste. Seafood is perish-
Thilsted et al 2016). In response, we propose a shift able, so it has been estimated that 35% of fish and
towards pelleted feeds for groupers and other carni- seafood is wasted globally, compared to 20% of meats
vorous marine fish as an intervention to lower direct or 30% of cereals (fao.org/save-food/en/; accessed 8
use of fish and reduced eutrophication. The reported March 2019). Byproduct utilization, through reduc-
FCRs for groupers being fed formulated feeds range tion into fishmeal or other useful commodities, could
between 1.5 and 3.1 (Sim et al 2005, Hasan 2012, also be improved in Indonesia as seafood is primarily
Bunlipatanon et al 2014). From this range, we assumed sold whole, as in most parts of Asia (Newton et al
that groupers by 2030 could be entirely fed by pelleted 2014). AqI6 consequently expected that food waste
feeds at an FCR of 2. Even though these pelleted feeds could be halved in Indonesia by 2030; down from 35%
contain 60% fishmeal and fish oil (Henriksson et al to 17.5% thanks to better access to cold storage, more
2017b), and roughly 5 kg fish are needed per kg efficient supply chains, consumer awareness, and
fishmeal (Parker and Tyedmers 2012), it is still a better byproduct utilization. The additional energy
reduction of 16 kgwild fish per kg grouper. needed for refrigeration was not considered in this
scenario, butwouldmost likely be negligible (Henriks-
2.4.4. AqI4: favoring omnivorous finfish species over son et al 2015).
shrimp
Henriksson et al (2017a, 2017b) concluded that
omnivorous finfish species were associated with lower 3. Results
environmental impacts than shrimp and groupers.
This as they required less fishmeal in feeds, had higher 3.1. LCA results of current production systems
productivity per ha, and were less susceptible to The environmental impacts per tonne of product at
disease. In AqI4 we consequently assumed that half of farm gate are presented in table 1, with ranges of
both the shrimp and grouper volumes (963 tonnes) Monte Carlo simulations detailed in the SM.Groupers
would be replaced by tilapia, carp, clarias, pangasius, had the largest global warming impact, followed by
andmilkfish frommonoculture systems. common carp, shrimps, and milkfish from polycul-
ture systems. Groupers were also associated with the
2.4.5. AqI5: renewable electricity largest eutrophication and fossil energy use impacts.
On-farm energy use varied greatly among farms and Common carp, in the meantime, emitted the most
farming systems, from 68 208MJ of petrol per tonne acidifying agents. Milkfish and Asian tiger shrimp
of grouper, to 31MJ of electricity per tonne of polyculture occupied the largest land areas and
panagsius. Nine of the eleven production systemswere consumed most freshwater. If these systems had been
directly reliant on electricity from the Indonesian grid, established on converted mangroves, they would also
and all indirectly through supporting processes ran- be the largest greenhouse gas emitters, causing an
ging fromnitrogen fixation to feedmaterial processing additional 184 000–297 000 kg CO2-eq. t
−1 shrimp or
(see SM for details). At present, most of this electricity milkfish.
4
Environ. Res. Lett. 14 (2019) 124062
Table 1. Lifecycle environmental impacts from the production systems under study usingmass allocation, scaled to one tonne of live animal
at farm gate.
Global Land Fossil
warming Acidification Eutrophication occupation Fresh-water use energy use
Species System kgCO2-eq. kg SO2-eq. kg PO4-eq. m2a m3 MJ
A. tiger shrimp Poly 10 500 75.1 96.7 21 300 47 900 104 000
Whiteleg Ponds 9430 80.8 120 7130 9700 108 000
shrimp
Milkfish Ponds 6600 69.8 103 14 300 3620 58 200
Poly 10 500 75.1 96.7 21 300 47 900 104 000
Clarias Ponds 5260 58.1 51.0 4680 241 52 800
Pangasius Ponds 5060 52.7 50.9 4930 427 50 400
Cages 5370 57.1 58.6 5070 132 55 200
Common carp Ponds 9260 102 91.1 8210 265 97 400
Cages 9530 104 106 8260 242 101 000
Tilapia Ponds 8450 91.8 82.5 7430 369 88 300
Cages 7950 87.3 89.4 7100 207 83 000
Groupers Cages 14 300 66.6 341 437 33.6 160 000
Others 8518 76.7 107 9170 9255 88 525
3.2. Environmental impacts by 2030 after competition with mangrove forests in coastal areas
aquaculture interventions (figure 1). Extensive logging on Sumatra and Kaliman-
The environmental impacts from the BAU scenario up tan has also had repercussions on air quality from fires,
to 2030, with the updated LCA impacts, resulted in precipitation patterns, and freshwater quality (Marlier
between three- and four-fold increases in environ- et al 2015, Wang-Erlandsson et al 2018). Conse-
mental impacts, proportionally higher than the 3.25- quently, land concessions lead to a negative feedback
fold increase in fish volume (table 2). Improved feed loop for the aquaculture industry itself, due to
efficiency for whiteleg shrimp, common carp and deteriorated ecosystems. Thus, apart from the large
tilapia (AqI1) witnessed major reductions for most carbon emissions, the arguments for zero-tolerance
impact categories, apart from freshwater consump- toward further land concessions are overwhelming,
tion. Freshwater consumption could, however, be and central to the success of the aquaculture industry.
reduced by intensifying milkfish and Asian tiger In the meantime, most of the systems evaluated have
shrimp farming systems (AqI2). This shift towards potential for intensification, but this needs to be
pelleted feeds unfortunately also resulted in an carried out with minimal negative trade-offs, such as
increased demand for wild fish as a feed ingredient. deforestation of the Amazon for soybean production
Greater use of pelleted feed rather than low-cost fish (Henriksson et al 2018).
(commonly small pelagic fish or bycatch) in grouper
production (AqI3) would, however, more than coun-
teract the demand for wild 5.Discussion
fish. Further reductions
would be achieved by shifting shrimp production
towards omnivorous species (AqI4), alongside reduc- Indonesia harbors some of the most important
tions in freshwater consumption. However, a trans- terrestrial and marine ecosystems worldwide. It is also
ition toward renewable electricity (AqI5) had only a a nation whose population is strongly reliant on
modest effect on most impact categories, apart from seafood. The present research indicates that six inter-
fossil energy use. This is because the bulk of the global ventions could reduce environmental impacts with
warming and acidifying emissions originated from between 28% and 49% per volume of fish. BAU
agricultural elds for feed ingredients, and nitrous growth would result in a 3.25-fold increase by 2030,
fi
oxide and methane from ponds and manure manage- while the Indonesian government aim for a 2.5-fold
ment. Not surprisingly, the most ef cient and prob- increase for most species using an additional
fi
ably cheapest way to reduce all environmental impacts 12 million hectares (IDH 2018). Both of these scenar-
from the seafood industry would be to reduce food ios would consequently imply additional encroach-
waste and better utilize byproducts (AqI6). ments and degradation of ecosystems, even with the
implementation of the AqIs. In addition, the present
footprint might already be exceeding the carrying
4. Spatial planning for aquaculture growth capacity in many parts of the country. Thus, the KKP/
MMAF should reevaluate their production targets and
Shrimp farming, mainly practiced on Java, Sumatra take environmental assessments, like the present one,
and Sulawesi, has resulted in extensive ecological into consideration. They also need to shift their focus
degradation due to the large areas needed and their away from high value species such as shrimp and
5
Environ. Res. Lett. 14 (2019) 124062
6
Table 2.Mitigation potential of six aquaculture interventions (AqIs) by 2030, and their cumulative effect in total and per tonne offish.
AqI1: 20% lower FCR for AqI2: BMPmilkfish&
Business-as- whiteleg shrimp, carp, and Asian tiger shrimp AqI3: Grouper AqI4: Shift to AqI5: Renewable AqI6: Reduce Aq1–6Cumulative AqI1–6: Per
Impact category usual tilapia polyculture pellets, FCR 2 omnivorous species energy foodwaste effect tonne fish
Fish quantity 325% 325% 325% 325% 325% 325% 325% 325% 100%
Global warming 339% 317% 350% 341% 327% 339% 267% 222% 68%
Acidification 330% 305% 345% 333% 329% 304% 260% 230% 71%
Eutrophication 345% 317% 347% 335% 329% 341% 272% 227% 70%
Land occupation 359% 342% 316% 360% 348% 360% 283% 234% 72%
Freshwater use 419% 419% 265% 419% 343% 419% 330% 166% 51%
Fossil energy use 339% 318% 368% 343% 325% 305% 267% 220% 68%
Wildfish use 355% 327% 334% 304% 312% 355% 279% 184% 57%
Environ. Res. Lett. 14 (2019) 124062
Figure 1.Ecological status of key habitat types and aquaculture production of individual species across Indonesia. Percentages of
species by location refer to the percentage of national production for that species in themajor aquaculture producing provinces. For
example, 41%of the total national tilapia production comes from Java. The total national production is shown in the bottomportion
of the graphic. For example, total national production of tilapiawas 947 000 tonnes in 2016, 78% from freshwater ponds and 24%
from floating cages. The condition ofmangrove forests, coral, and seagrass bedswere classified based on an assessment by BPS-
Statistics Indonesia and Subdirectorate of Enviroment Statistics (2018). Land concessions are from all activities, not only aquaculture.
Data from:UNEP-WCMC (2010), Giri et al (2011), Global ForestWatch (2014), and BPS-Statistics Indonesia and Subdirectorate of
Enviroment Statistics (2018), UNEP-WCMC (2017).
grouper, and instead acknowledge the importance of yield gap and therefore the best prospects for improv-
omnivorous finfish species that are consumed domes- ing. It is one of few omnivorous finfish species that can
tically. Environmental stewardship would also gener- be farmed in brackish and marine environments.
ate monetary returns through better fish yields and Thus, if production could be sustainably intensified,
tourism, and hopefully soon through international this species holds potential to utilize a wider set of feed
compensation schemes for ecosystem services, such as resources than its carnivorous counterparts. Mean-
carbon credits (Malik et al 2015, Rao et al 2015, Alongi while, the traditional tambak system, where milkfish
et al 2016). are extensively farmed alongside Asian tiger shrimp
Feed was the major contributor to most impact and often seaweed in brackish-water ponds, con-
categories. Domestic fishmeal and fish oil are espe- stituted the major competitor for valuable coastal
cially controversial ingredients as they often compete areas due to large area requirements and low produc-
with human food and most domestic fish stocks tion volumes. Two of the largest milkfish producing
destined for reduction are already overexploited provinces, Java and South Sulawesi, were also the pro-
(Buchary et al 2011). Many promising substitutes vinces with themost degraded coastal andmarine eco-
have, in the meantime, entered the market over the systems. Thus, given that aquaculture is still a major
last decade, including derivatives from agricultural driver for deforestation in Indonesia (Richards and
products, macroalgae, microalgae, single cell pro- Friess 2016), these farmers need to transition to more
teins, and insects. Apart from their scalability, it will intensive farming systems. Intensification, however,
also be important to evaluate their respective foot- goes hand in handwith an increased risk of disease and
prints in order to avoid environmental trade-offs consequently needs to be supported by better spatial
(Pelletier et al 2018). Feed used and species composi- planning, rapid alert systems, veterinary support, and
tions also influence the nutritional profile of seafood. disease resistant strains.
All these aspects need to be evaluated in more depth Getting farmers, processors, and consumers to adopt
when setting national production goals and promot- better production and consumption practices and invest
ing specific farming systems. in favorable innovations remain the largest challenges.
Milkfish displayed the most variable production Ideally, savings gained by reducing feed inputs, improv-
performance, which suggests that it has the largest ing edible yields, and/or reducing food waste would
7
Environ. Res. Lett. 14 (2019) 124062
offset the costs of implementation, but initiating these temperature ranges, shifting precipitation patterns, and
types of changes has proved difficult in practice due to less availability of wild fish (Cheung et al 2016, Klinger
limited financial and farmer buy-in (Yi et al 2018). A et al 2017, Supari et al 2017, Barange et al 2018). Subse-
cost/benefit analysis would therefore be useful to iden- quently, the Indonesian aquaculture industry does not
tify the largest gains per unit of investment. Resources only need to reduce its emissions, it also needs to adopt
would also be needed for disseminating information, to the effects of climate change.
establishing pilot projects, providing extension services,
imposing and enforcing tighter regulations, and evaluat- Acknowledgments
ing progress. These resources could be sourced nation-
ally, by government institutions, or internationally,
In memoriam of Dr Nur Bambang Priyo Utomo who
through foreign aid or nongovernmental organizations.
made this research possible.
Influencing consumer choices would be equally hard,
We are grateful to the Walton Family Foundation
especially since the Indonesian archipelago exhibits a
for funding this research. This work was undertaken as
large cultural heterogeneity. This calls for an adaptive
part of the CGIAR Research Programs on Fish Agri-
strategywhen evaluating anddisseminating information.
Food Systems (FISH) led by WorldFish and on Climate
In summary, if the AqIs explored were implemented in
Change, Agriculture and Food Security (CCAFS). These
combinationwithmore novel technologies that have not
programs are supported by contributors to the CGIAR
yet been benchmarked or are only at small-scale produc-
Trust Fund. PJGHandMTare partially funded by FOR-
tion, including novel feed ingredients (e.g. Simon et al
MAS SeaWin project (2016-00227). We thank Martha
2019, Couture et al 2019), we believe that aquaculture
Mamora for her assistance in the field and our collea-
production could be doubled within its current environ-
gues at PTHatfield.
mental footprint in Indonesia. However, the challenges
for successful implementation of individual interven-
tions are many-including economic viability, political Data availability statement
support,market demand, and ethical considerations.
Ecological heterogeneity also needs to be accounted Any data that support the findings of this study are
for and enforced in zoning maps. While figure 1 pro- includedwithin the article.
vides a rough national outline, PT Hatfield has devel-
oped a more detailed map for the suitability of shrimp ORCID iDs
farms that also accounts for urban sprawl (https://
aquaculture.ipb.ourecosystem.com/interface/). Com- Patrik JohnGustavHenriksson https://orcid.org/
plementing assessments should be carried out for fresh- 0000-0002-3439-623X
water lakes, as many of these already have exceeded LaurenKBanks https://orcid.org/0000-0001-
carrying capacity (Fukushima et al 2017). Competition 6197-4245
for resources with other sectors also needs to be further SharonK Suri https://orcid.org/0000-0003-
developed, as everyone has their own ambitious targets. 3064-7991
The Indonesian Ministry of Agriculture, for example, NurulhudaAhmad Fatan https://orcid.org/0000-
intends to increase the production of charcoal by 15%, 0002-3599-9892
sugar cane by 20%, and palm oil by 12% by 2019 MaxTroell https://orcid.org/0000-0002-7509-8140
(Directorate General of Estate Crops 2014, Directorate
of Energy andMineral Resources 2014).
Beyond environmental considerations, species selec- References
tion and production practices also influence food and
nutrition security, economic growth, job opportunities, Abood SAA, Lee J SHSH, Burivalova Z,Garcia-Ulloa J and
Koh L PP 2015Relative contributions of the logging, fiber, oil
and equity (Béné et al 2016), variables that were not con- palm, andmining industries to forest loss in Indonesia
sidered in the present study. This is because prices and Conserv. Lett. 8 58–67
labor are challenging to predict following the interven- AlongiDM et al 2016 Indonesia’s blue carbon: a globally significant
tions identified. Undoubtedly, however, the proposed and vulnerable sink for seagrass andmangrove carbonWetl.
Ecol.Manag. 24 3–13
interventions and novel feed ingredients will influence
AstudilloMF, Thalwitz G andVollrath F 2015Modern analysis of
the cost of production and ultimately consumer prices, an ancient integrated farming arrangement: life cycle
either negatively or positively. This, in turn, will decide assessment of amulberry dyke and pond system Int. J. Life
who can afford to consume the seafood produced, and Cycle Assess. 20 1387–98
whowill gain nancially. BarangeM, Bahri T, BeveridgeMCM,CochraneK L,
fi
Funge-Smith S and Poulain F 2018 Impacts of climate change
Consequences of climate change were not con- onfisheries and aquaculture. synthesis of current knowledge,
sidered in the modeling, but small-scale farms and low- adaptation andmitigation options FAOTechnical Report: 627
latitude countries such as Indonesia are expected to be (Rome: FAO)
especially vulnerable (Barange et al 2018). For Indonesia, Behrens P, Kiefte-de Jong J C, Bosker T, Rodrigues J FD,
deKoningA andTukker A 2017Evaluating the
these consequences would include more frequent environmental impacts of dietary recommendations Proc.
extremeweather events, rising sea levels, shifts in optimal Natl Acad. Sci. 114 13412–7
8
Environ. Res. Lett. 14 (2019) 124062
BeltonB, Bush S R and Little DC 2017Not just for the wealthy: GuilmotoCZ and JonesGW2016ContemporaryDemographic
rethinking farmedfish consumption in theGlobal South Transformations in China, India and Indonesia (Cham:
Glob. Food Sec. 16 85–92 Springer International Publishing)
BénéC et al 2016Contribution offisheries and aquaculture to food HasanMR2012Transition fromLow-value Fish to Compound Feeds
security and poverty reduction: assessing the current evidence inMarine Cage Farming inAsia: Technical PaperNo 573
WorldDev. 79 177–96 (Rome: FAOFisheries andAquaculture)
BPS 2018Badan Pusat Statistik (SatuDataKelautan dan Perikanan) HasanMH,Mahlia TM I andNurH2012A review on energy
(https://bps.go.id/subject/56/perikanan.html#subjek scenario and sustainable energy in IndonesiaRenew. Sustain.
ViewTab4) Energy Rev. 16 2316–28
Buchary EA 2010 In Search of Viable PolicyOptions for Responsible Henriksson P JG,Guinée J B,Heijungs R,DeKoningA and
use of Sardine Resources in the Bali Strait, Indonesia PhDThesis GreenDM2014Aprotocol for horizontal averaging of unit
(Institute for Resources, Environment and Sustainability, process data—including estimates for uncertainty Int. J. Life
University of British Columbia) (https://doi.org/10.14288/ Cycle Assess. 19 429–36
1.0069259) Henriksson P JG et al 2015Comparison of Asian aquaculture
Buchary EA, Pitcher T J and Sumaila UR 2011Unaccounted values products by use of statistically supported life cycle assessment
—under-reporting sardine catches as a strategy against Environ. Sci. Technol. 49 14176–83
poverty in the Bali Strait, IndonesiaWorld Fisheries—A Social- HenrikssonP JG,DicksonM,AllahAN,Al-KenawyDandPhillipsM
Ecological Analysis edREOmmer et al (Oxford:Wiley- 2017aBenchmarking the environmental performanceof best
Blackwell) pp 203–23 managementpractice andgenetic improvements inEgyptian
Bunlipatanon P, SongseechanN,KongkeoH, AberyNWand aquacultureusing life cycle assessmentAquaculture46853–59
De Silva S S 2014Comparative efficacy of trash fish versus Henriksson P JG, TranN,MohanCV,ChanCY, RodriguezU-P,
compounded commercial feeds in cage aquaculture of Asian Suri S,Mateos LD,UtomoNBP,Hall S and PhillipsM J
seabass (Lates calcarifer) (Bloch) and tiger grouper 2017b Indonesian aquaculture futures—Evaluating
(Epinephelus fuscoguttatus) (Forsskål)Aquac. Res. 45 373–88 environmental and socioeconomic potentials and limitations
Cao L,Naylor R L,Henriksson P JG, LeadbitterD,MetianM, J. Clean. Prod. 162 1482–90
TroellM andZhangW2015China’s aquaculture and the Henriksson P JG, Belton B,Murshed-E-JahanK andRicoA 2018
world’s wild fisheries Science 347 133–5 Measuring the potential for sustainable intensification of
CheungWWL, ReygondeauG and Frölicher T L 2016 Large aquaculture in Bangladesh using life cycle assessment Proc.
benefits tomarine fisheries ofmeeting the 1.5 °Cglobal Natl Acad. Sci. USA 115 2958–63
warming target Science 354 1591–4 IDH2018 Investment Guideline for Sustainable Aquaculture in
Clauser C and EwertM2018The renewables cost challenge: Indonesia (The Sustainable Trade Initiative, Utrecht, The
levelized cost of geothermal electric energy compared to Netherlands) (https://idhsustainabletrade.com/
other sources of primary energy—review and case study publication/investment-guideline-for-sustainable-
Renew. Sustain. Energy Rev. 82 3683–93 aquaculture-in-indonesia-2018/)
Couture J L, Geyer R,Hansen JØ, Kuczenski B, ØverlandM, ISO 2006 International Organization for Standardization 14044.
Palazzo J, SahlmannC and LenihanH2019 Environmental EnvironmentalManagement—Life Cycle Assessment—
benefits of novel nonhuman food inputs to salmon feeds Requirements andGuidelines (Geneva, Switzerland)
Environ. Sci. Technol. 53 1967–75 JärviöN,Henriksson P JG andGuinée J B 2018 IncludingGHG
DeyMM,RosegrantMW,GoshK,ChenOL and emissions frommangrove forests LULUC in LCA: a case
Valmonte-Santos R 2016Analysis of the economic impact of study on shrimp farming in theMekongDelta, Vietnam Int. J.
climate change and climate change adaptation strategies for Life Cycle Assess. 23 1078–90Online: (http://link.springer.
fisheries sector in Pacific coral triangle countries:model, com/10.1007/s11367-017-1332-9)
estimation strategy, and baseline resultsMar. Policy 67 KlingerDH, Levin SA andWatson J R 2017The growth offinfish in
156–63 global open-ocean aquaculture under climate changeProc. R.
Directorate General of Estate Crops 2014 Strategic Plan 2015–2019 Soc. B Biol. Sci. 284 20170834
No 19,HK.140 / 4/ 2015 (Jakarta:Ministry of Agriculture) MalikA, Fensholt R andMertzO 2015 Economic valuation of
Directorate of Energy andMineral Resources 2014 Final Report: Mangroves for comparisonwith commercial aquaculture in
Review of Target DMOCoal Production 60%by 2019 pp 115 south Sulawesi, Indonesia Forests 6 3028–44
Directorate General of Aquaculture 2017Roadmap, Aquaculture MarlierME,DeFries R S, KimP S, Koplitz SN, JacobD J,
Development 2017–2021 (Jakarta) Mickley L J andMyers S S 2015 Fire emissions and regional air
FAO2018The State ofWorld Fisheries and Aquaculture 2018 quality impacts from fires in oil palm, timber, and logging
FAO2019 Fishery andAquaculture Statistics. (Rome: FAOFisheries concessions in Indonesia Environ. Res. Lett. 10 085005
andAquacultureDepartment) Maynard JA,AnthonyKRN,Afatta S,Dahl-TacconiNand
Fry J P,MaillouxNA, LoveDC,MilliMC andCao L 2018 Feed Hoegh-GuldbergO2010Making amodelmeaningful to coral
conversion efficiency in aquaculture: dowemeasure it reefmanagers in a developingnation: a case study ofoverfishing
correctly?Environ. Res. Lett. 13 024017 and rock anchoring in indonesiaConserv. Biol.241316–26
FukushimaT,Matsushita B, Subehi L, Setiawan F andWibowoH Murray J P, Grenyer R,Wunder S, RaesN and Jones J PG2015
2017Will hypolimnetic waters become anoxic in all deep Spatial patterns of carbon, biodiversity, deforestation threat,
tropical lakes? Sci. Rep. 7 45320 andREDD+ projects in IndonesiaConserv. Biol. 29 1434–45
GaitherMR and Rocha L A 2013Origins of species richness in NewtonR, Telfer T and Little D 2014 Perspectives on the utilization
the Indo-Malay-philippine biodiversity hotspot: evidence of aquaculture coproduct in Europe andAsia: prospects for
for the centre of overlap hypothesis J. Biogeogr. 40 value addition and improved resource efficiencyCrit. Rev.
1638–48 Food Sci. Nutr. 54 495–510
Giri C,Ochieng E, Tieszen L L, ZhuZ, SinghA, LovelandT, Parker RWRandTyedmers PH2012Uncertainty andnatural
Masek J andDukeN 2011 Status and distribution of variability in the ecological footprint offisheries Ecol. Indic.
mangrove forests of theworld using earth observation 16 76–83
satellite dataGlob. Ecol. Biogeogr. 20 154–9 PelletierN, Klinger DH, SimsNA, Yoshioka J R andKittinger JN
Global ForestWatch 2014OpenData Portal: (http://data. 2018Nutritional attributes, substitutability, scalability, and
globalforestwatch.org/) (Accessed 10October 2018) environmental intensity of an illustrative subset of current
Guillen J,Natale F, CarvalhoN,Casey J, Hofherr J, Druon JN, and future protein sources for aquaculture feeds: joint
FioreG,GibinM, Zanzi A andMartinsohn J T 2019Global consideration of potential synergies and trade-offs Environ.
seafood consumption footprintAmbio. 48 111–22 Sci. Technol. 52 5532–44
9
Environ. Res. Lett. 14 (2019) 124062
RaoN S,Ghermandi A, Portela R andWangX 2015Global values of demand in Indonesia to 2030 and role of aquaculture using
coastal ecosystem services: a spatial economic analysis of the AsiaFishmodelMar. Policy 79 25–32
shoreline protection values Ecosyst. Serv. 11 95–105 UllmanC, RhodesM,Hanson T, ClineD andDavis DA 2019 Effects
RichardsDR and FriessDA 2016Rates and drivers ofmangrove of four different feeding techniques on the pond culture of
deforestation in Southeast Asia, 2000–2012Proc. Natl Acad. pacificwhite shrimp, Litopenaeus vannamei J.World Aquac.
Sci. 113 344–9 Soc. 50 54–64
RimmerMA, SugamaK, Rakhmawati D, RofiqR andHabgoodRH UNEP-WCMC2010Global distribution ofwarm-water coral reefs,
2013A review and SWOTanalysis of aquaculture compiled frommultiple sources (listed in ‘Coral_Source.
development in IndonesiaRev. Aquac. 5 255–79 mdb’), and including IMaRS-USF and IRD (2005), IMaRS-
Semedi JM,Willemen L,NurlambangT, van derMeer F and USF (2005) and Spalding et al (2001) (Cambridge: UNEP
Koestoer RH2017Developing a framework for assessing the WorldConservationMonitoringCentre) (http://data.unep-
impact of geothermal development phases on ecosystem wcmc.org/datasets/13)
services IOPConf. Ser.: Earth Environ. Sci. 103 012003 UNEP-WCMC2017Global distribution of seagrasses (version 6.0).
Sim S-Y, RimmerMA,WilliamsK, Toledo JD, SugamaK, Sixth update to the data layer used inGreen and Short (2003)
Rumengan I and PhillipsM J 2005APractical Guide to Feeds (Cambridge: UNEnvironmentWorldConservation
and FeedManagement for CulturedGroupers MonitoringCentre) (http://data.unepwcmc.org/datasets/7)
(Bangkok:NACA) Usfar AA and FahmidaU 2011Do Indonesians follow its dietary
SimonC J, BlythD, Ahmad FatanN and Suri S 2019Microbial guidelines?—evidence related to food consumption, healthy
biomass (NovacqTM) stimulates feeding and improves the lifestyle, and nutritional status within the period 2000–2010
growth performance on extruded low to zero-fishmeal diets Asia Pac. J. Clin. Nutr. 20 484–94
in tilapia (GIFT strain)Aquaculture 501 319–24 Wang-Erlandsson L, Fetzer I, Keys PW,VanDer Ent R J,
Supari T F, Juneng L andAldrian E 2017Observed changes in SavenijeHHGandGordon L J 2018Remote land use
extreme temperature and precipitation over Indonesia Int. J. impacts on riverflows through atmospheric teleconnections
Climatol. 37 1979–97 Hydrol. Earth Syst. Sci. 22 4311–28
Thilsted SH, Thorne-LymanA,Webb P, Bogard J R, SubasingheR, White PG, ShiptonTA, Bueno PB andHasanMR2018Better
PhillipsM J andAllison EH2016 Sustaining healthy diets: the Management Practices for Feed Production andManagement of
role of capture fisheries and aquaculture for improving Nile Tilapia andMilkfish in the Philippines Technical Paper
nutrition in the post-2015 era Food Policy 61 126–31 No 614 (Rome: FAOFisheries andAquaculture) pp 98
TranN, RodriguezU-P, ChanCY, PhillipsM J,MohanCV, YiD, ReardonT and Stringer R 2018 Shrimp aquaculture
Henriksson P JG, Koeshendrajana S, Suri S andHall S 2017 technology change in Indonesia: are small farmers included?
Indonesian aquaculture futures: an analysis offish supply and Aquaculture 493 436–45
10