Content uploaded by Jaime Romero
Author content
All content in this area was uploaded by Jaime Romero on Nov 11, 2015
Content may be subject to copyright.
1 2 3 3 4 5
6 7 8 8 9 10
1,†11 11 1 8
1
Norwegian College of Fishery Science, Faculty of Biosciences, Fisheries and Economics, UiT The Arctic University of
Norway, Tromsø, Norway;
2
Key Laboratory for Feed Biotechnology of the Ministry of Agriculture, Feed Research Insti-
tute, Chinese Academy of Agricultural Sciences, Beijing, China;
3
EWOS Innovation LTD, Dirdal, Norway;
4
Laborato-
rio de Biotecnolog
ıa, Instituto de Nutrici
on y Tecnolog
ıa de los Alimentos (INTA), Universidad de Chile, Macul,
Santiago, Chile;
5
Department of Basic Sciences and Aquatic Medicine, Faculty of Veterinary Medicine and Biosciences,
Norwegian University of Life Sciences, Oslo, Norway;
6
Institute of Biology, Norwegian University for Science and Tech-
nology, Trondheim, Norway;
7
Arkadios Dimitroglou R&D Department, Nireus Aquaculture SA, Koropi-Attica, Greece;
8
Plymouth University, Plymouth, UK;
9
Skretting LTD, Stavanger, Norway;
10
Primex Ehf, Siglufjordur, Iceland;
11
Laboratory of Aquaculture and Artemia Reference Center, Ghent University, Ghent, Belgium
It is well known that healthy gut microbiota is essential to
promote host health and well-being. The intestinal micro-
biota of endothermic animals as well as fish are classified as
autochthonous or indigenous, when they are able to colonize
the host’s epithelial surface or are associated with the micro-
villi, or as allochthonous or transient (associated with digesta
or are present in the lumen). Furthermore, the gut micro-
biota of aquatic animals is more fluidic than that of terres-
trial vertebrates and is highly sensitive to dietary changes. In
fish, it is demonstrated that [a] dietary form (live feeds or pel-
leted diets), [b] dietary lipid (lipid levels, lipid sources and
polyunsaturated fatty acids), [c] protein sources (soybean
meal, krill meal and other meal products), [d] functional gly-
comic ingredients (chitin and cellulose), [e] nutraceuticals
(probiotics, prebiotics, synbiotics and immunostimulants), [f]
antibiotics, [g] dietary iron and [h] chromic oxide affect the
gut microbiota. Furthermore, some information is available
on bacterial colonization of the gut enterocyte surface as a
result of dietary manipulation which indicates that changes
in indigenous microbial populations may have repercussion
on secondary host–microbe interactions. The effect of diet-
ary components on the gut microbiota is important to inves-
tigate, as the gastrointestinal tract has been suggested as one
of the major routes of infection in fish. Possible interactions
between dietary components and the protective microbiota
colonizing the digestive tract are discussed.
KEY WORDS: antibiotics, aquatic animals, dietary
components, intestine, microbiota
Received 29 December 2014; accepted 2 June 2015
Correspondence: E. Ringø, Norwegian College of Fishery Science,
Faculty of Biosciences, Fisheries and Economics, UiT The Arctic Univer-
sity of Norway, N-9037 Tromsø, Norway. E-mail: Einar.Ringo@uit.no
†
Present address: Marine Harvest, Thoning Ovesensgt. 28, 7044
Trondheim, Norway
Numerous studies have reviewed the structure and function
of the fish gut in relation to the diet (Bakke et al. 2011;
Buddington et al. 1997; Olsen & Ringø 1997; Ray & Ringø
2014). As the gut microbiota has not been included in these
papers, a more specific discussion is needed. Therefore, the
main objective of the present review was to summarize the
available information regarding the effect of dietary com-
ponents on the gastrointestinal (GI) microbiota of fish.
Until the 1970s, controversy existed about the role, and
even the existence, of an indigenous gut microbiota in fish.
However, it is now generally accepted that fish and other
aquatic animals have a microbiota in the GI tract (for
review see; Yoshimizu & Kimura 1976; Horsley 1977;
Cahill 1990; Ringø et al. 1995; Hansen & Olafsen 1999;
Ringø & Birkbeck 1999; Ringø et al. 2003; Ringø 2004;
Austin 2006; Izvekova et al. 2007; Merrifield et al. 2010a,
...................................................................... ........................
ª2015 The Authors. Aquaculture Nutrition Published by John Wiley & Sons Ltd.
This is an open access article under the terms of the Creative Commons Attribution-NonCommercial License,
which permits use, distribution and reproduction in any medium, provided the original work is properly cited and
is not used for commercial purposes.
2015 doi: 10.1111/anu.12346
....................................................................... ...................
Aquaculture Nutrition
2011a; Nayak 2010a; P
erez et al. 2010; Ringø et al. 2010a;
Lauzon & Ringø 2012; Nya & Austin 2013; Llewellyn et al.
2014) which in turn has increased interest in their diversity
and functional relationship (Clements et al. 2014; Zhou
et al. 2014). However, the gut microbiota is modulated by
dietary manipulations (Table 1) as well as by seasonal vari-
ations, stress, individual variations and different regions of
the GI tract, cultured versus wild, triploid versus diploid,
day-to-day variations, male versus female, developmental
stages/life cycle, microbial aspects of live feed, fast versus
slow growing, hierarchy formation, starvation, migration
from fresh water into sea water and migration from sea
water back to fresh water, water quality [pseudo-green
water versus clear water, recirculation versus conventional
flow-through, fish farms within a restricted area, environ-
mental and ecological factors, and host ecology and envi-
ronment (Table 2)]. To avoid the individual level variations
of the gut microbiota, several studies have analysed the
samples of pooled individuals (e.g. Hovda et al. 2007;
Ringø et al. 1995; Roeselers et al. 2011; Spanggaard et al.
2000; Sullam et al. 2012; Zarkasi et al. 2014).
The intestinal microbiota of fish, as is the case of mam-
mals, is classified as autochthonous (indigenous) or
allochthonous bacteria (Kim et al. 2007; Ringø & Birkbeck
1999; Ringø et al. 2003). The autochthonous bacteria are
those able to colonize the host’s gut epithelial surface or
are associated with the microvilli (Fig. 1), while the
allochthonous bacteria are transient, associated with food
particles or present in the lumen. In this context, it is
important to evaluate the effect of dietary components on
the intestinal microbiota of fish, as the gastrointestinal (GI)
tract is one of the major ports of entry for some pathogens
(Birkbeck & Ringø 2005; Burbank et al. 2011; Groff &
LaPatra 2000; Harikrishnan & Balasundaram 2005; Ringø
et al. 2007a,b).
In the 1970s, 1980s and 1990s, numerous investigations
were conducted to determine the dietary effects on the
intestinal microbiota, and the majority of these studies
were based on culture-dependent techniques and the use of
physiological and biochemical properties to characterize
the gut microbiota. However, from 2000 to 2006, there was
a shift to use molecular methods to characterize culturable
gut bacteria (e.g. Huber et al. 2004; Jensen et al. 2002;
Ringø et al. 2006a,b,c; Spanggaard et al. 2000), but nowa-
days culture-independent methods have become more com-
mon (Table 3). These recent investigations have widened
our knowledge about the intestinal microbiota of fish and
demonstrate that the microbial diversity of the fish gut is
more complex than previously believed.
Table 1 Overview of studies investigated the effect of diet on gut
microbiota of aquatic animals
Dietary
component
used Aquatic animals References
Levels of dietary
lipid
Rainbow trout Lesel et al. (1989)
Arctic charr Ringø & Olsen
(1999)
Different dietary
lipid sources
Arctic charr Ringø et al. (2002)
Gilthead sea bream Montero et al.
(2006)
Atlantic salmon E. Ringø, R.E. Olsen &
S. Sperstad
(unpublished data
Table 4)
Rainbow trout Ringø et al.
(unpublished data
–Table 5)
Grass carp Huang (2008)
Dietary
polyunsaturated
fatty acids
Arctic charr Ringø (1993a);
Ringø et al. (1998)
Brown trout Manzano et al.
(2012)
Thymus vulgaris
essential oil
Rainbow trout Navarrete et al.
(2010b)
Levels of dietary
fish protein
hydrolysates
European sea bass Kotzamanis et al.
(2007)
Different diets Gold fish Sugita et al. (1988a)
Atlantic cod Strøm & Olafsen
(1990)
Arctic charr Ringø & Olsen
(1994)
Rainbow trout Mansfield et al.
(2010)
Puffer fish Yang et al. (2007)
Yellow grouper Feng et al. (2010)
Rohu Ramachandran
et al. (2005)
Senegalese sole Makridis et al.
(2005)
Tilapia Kihara & Sakata (1997)
Winter flounder Seychelles et al. (2011)
Common carp Li et al. (2013)
Threespine
stickleback and
Eurasian perch
Bolnick et al. (2014)
Abalone Tanaka et al. (2003,
2004)
Different binding
agents
Plaice Gilmour et al.
(1976)
Marine protein
hydrolysates
Sea bass Delcroix et al.
(2015)
Microalga
(Scenedesmus
alimeriensis)
Gilthead sea bream Vizca
ıno et al. (2014)
Soybean meal
products
Atlantic cod Refstie et al. (2006),
Ringø et al.
(2006b)
..................................................................... .........................
Aquaculture Nutrition ª2015 The Authors. Aquaculture Nutrition Published by John Wiley & Sons Ltd.
Table 1 (Continued)
Dietary
component
used Aquatic animals References
Atlantic salmon Bakke-McKellep
et al. (2007), Ringø
et al. (2008), Green
et al. (2013),
Navarrete et al.
(2013)
Gilthead seabream Dimitroglou et al.
(2010)
Gilthead seabream de Paula Silva et al.
(2011)
Goldfish de Paula Silva et al.
(2011)
Grass carp Huang (2008)
Rainbow trout Heikkinen et al.
(2006), Merrifield
et al. (2009a)
Silver crucian carp Cai et al. (2012)
Different fish
species
Merrifield et al.
(2011a)
Plant-based diets Rainbow trout Desai et al. (2012)
Invertebrate meals Mirror carp Z.Y. Wan, S. Davies &
D.L. Merrifield
(unpublished data)
Casing meal Grass carp Huang (2008)
Rapeseed –and
cottonseed meal
Grass carp Huang (2008)
Wheat middling
and corn meal
Grass carp Huang (2008)
Wheat germ roots
and distillers dried
grains
Grass carp Huang (2008)
Different protein
sources
Atlantic salmon Hartviksen et al.
(2014, 2015)
Atlantic krill Atlantic salmon Ringø et al. (2006c)
Chitin Red sea bream Sera & Kimata
(1972), Kono et al.
(1987)
Atlantic cod Zhou et al. (2013a)
Atlantic salmon Askarian et al.
(2012)
Japanese eel Kono et al. (1987)
Giant fresh water
prawn
Kumar et al. (2006)
Chitosan Gibel carp Chen et al. (2014)
Cellulose Atlantic salmon Ringø et al. (2008)
Cellulase Grass carp Zhou et al. (2013b)
Xylanase Jian carp Jiang et al. (2014)
Iron Sea bass Gatesoupe et al.
(1997)
Copper Nile tilapia Hu et al. (2007)
Metal-nanoparticle Zebra fish Merrifield et al.
(2013)
Chromic oxide
(Cr
2
O
3
)
Arctic charr Ringø (1993b,c, 1994)
Inositol Jian carp Jiang et al. (2009)
Table 1 (Continued)
Dietary
component
used Aquatic animals References
Acidic calcium
sulphate
Pacific white shrimp Anuta et al. (2011)
Sodium butyrate Tropical catfish Owen et al. (2006)
Common carp Liu et al. (2014)
Organic acid blend
(formic-, lactic-,
malic-, tartaric-
and citric acid)
Red hybrid tilapia Koh et al. (2014)
Organic acid
(Mera
TM
Cid)
Rainbow trout Jaafar et al. (2013)
Salt European sea bass Sun et al. (2013)
Alginic acid Tilapia Merrifield et al.
(2011c)
Poly-b-
hydroxybutyrate
Siberian sturgeon Najdegerami et al.
(2012, 2015)
Giant fresh water
prawn
Nhan et al. (2010)
Vitamin C Jian carp Liu et al. (2010)
Methionine Jian carp Tang et al. (2009)
Valine Jian carp Dong et al. 2013;
Inositol Jian carp Jiang et al. (2009)
Pantothenic acid Jian carp Wen et al. (2010)
Biotin Jian carp Zhao et al. (2012b)
Thiamine Jian carp Feng et al. (2011)
Phosphorus Jian carp Xie et al. (2011)
Betaine Hybrid tilapia He et al. (2012)
Probiotics Different fish
species
Ringø (2004), Gatlin
et al. (2006), Ringø
et al. (2005),
Merrifield et al.
(2010a,b,c),
Dimitroglou et al.
(2011), Lauzon &
Ringø 2012,
Daniels et al.
(2013), Hoseinifar
et al. (2014b),
Cordero et al.
(2015)
Prebiotics Different fish
species
Gatlin et al. (2006),
Merrifield et al.
(2010a), Ringø
et al. (2010b),
Dimitroglou et al.
(2011), Daniels
et al. (2013),
Hoseinifar et al.
(2014a,b), Ringø
et al. (2014a)
Synbiotics Different species Abid et al. (2013),
Daniels et al.
(2013), Hoseinifar
et al. (2014a,b),
Ringø & Song
(2015)
Yeast culture Hybrid tilapia Zhou et al. (2009c)
...................................................................... ........................
Aquaculture Nutrition ª2015 The Authors. Aquaculture Nutrition Published by John Wiley & Sons Ltd.
Even though the traditional culture-based technique pos-
sesses rather low sensitivity for measuring the composition,
structure and stability of bacteria colonizing the digestive
tract of fish, it is able to indicate differences due to minor
dietary alterations as observed by Refstie et al. (2006) in
gut microbiota of Atlantic cod (Gadus morhua L.) fed stan-
dard or bioprocessed soybean meal (BPSBM) (Fig. 2).
The gut microbiota may function to prevent pathogens
from colonization; it is likely that the gut microbiota might
be of vital importance with regard to fish health. The
objective of the present paper, to review present informa-
tion on how dietary supplements affect the population level
of gut bacteria and composition, is relevant. This is
strengthened by the fact that during the last decade, the
aquaculture industry is increasingly demanding sustainable
alternative lipid and protein sources to reduce the use of
fish meal (FM) and fish oil (FO) (Tacon & Metian 2008;
Hemre et al. 2009; Merrifield et al. 2011a; Morais et al.
2012; Olsen & Hasan 2012; Hansen & Hemre 2013).
This review firstly presents a short overview of the GI
tract of fishes and the techniques most often used for the
study of GI microbiota as a background for the succeeding
chapters covering impact of the nutrients sources, probi-
otics, prebiotics and antibiotics. The results cited in the
present review include works published in peer-reviewed
scientific journals, open access peer-reviewed scientific jour-
nals, books as well as minimally circulated investigations
available as short communications, and abstracts presented
in books from international conferences. The latter is done
in order to indicate that there are numerous interesting
investigations ongoing albeit not yet been published in sci-
entific journals. Furthermore, in order to give the reader
satisfactory information on dietary effect on the gut micro-
biota, the present authors include some information from
endothermic animals when needed.
The key function of the alimentary tract is its ability to dis-
solve foodstuffs and process nutrients to make them suit-
Table 1 (Continued)
Dietary
component
used Aquatic animals References
Inactive brewer‘s
yeast
Beluga Hoseinifar et al.
(2011b)
Common carp Omar et al. (2012)
Potassium
diformate
Hybrid tilapia Zhou et al. (2009d)
Immunostimulants Atlantic cod Gildberg &
Mikkelsen (1998),
Skjermo et al.
(2006)
Atlantic salmon Liu et al. (2008)
Red tilapia Merrifield et al.
(2010c)
Hybrid tilapia He et al. (2010)
Different antibiotics Different species For review see
the present study
Table 2 Overview of studies investigated the effect on gut micro-
biota of fish
Factors References
Seasonal variations Sugita et al. (1981, 1983), Ringø
(2000), Al-Harbi & Naim Uddin
(2004), Hagi et al. (2004);
Naviner et al. (2006), Hovda
et al. (2012)
Stress Olsen et al. (2005), Ringø et al.
(2000, 2014b)
Individual variations and
different regions of the
GI tract
Sugita et al. (1990), Ringø et al.
(1995), Spanggaard et al.
(2000), Ringø et al. (2001b);
Hovda et al. (2007),
McDonald et al. (2012),
Fjellheim et al. (2012);
Star et al. (2013)
Cultured versus wild Bacanu & Oprea (2013),
Kim & Kim (2013)
Triploid versus diploid Cantas et al. (2011)
Day-to-day variations Sugita et al. (1987, 1990)
Male versus female Iehata et al. (2015)
Different fish species fed
similar diet
Li et al. (2014b)
Developmental stages/life cycle Cao et al. (2012), Zhao et al.
(2012a), Huang et al. (2014)
Microbial aspects of live feed Conceic
ß
~
ao et al. (2010),
Bakke et al. (2013)
Fast versus slow growing fish Sun et al. (2009)
Hierarchy formation Ringø et al. (1997)
Starvation Xia et al. (2014)
Migration from fresh water to
sea water and Migration from
sea water back to fresh water
Ringø (2004)
Water quality Gatesoupe et al. (2013)
Recirculation versus
conventional flow-through
Attramadal et al. (2012)
Fish farms Diler et al. (2000)
Environmental and ecological
factors
Sullam et al. (2012)
Host ecology and environments Wong & Rawls (2012)
..................................................................... .........................
Aquaculture Nutrition ª2015 The Authors. Aquaculture Nutrition Published by John Wiley & Sons Ltd.
able for absorption by various transport mechanisms in the
wall of the GI sections. Besides hydrolytic reactions catal-
ysed by endogenous enzymes secreted by the pancreas and
cells in the gut wall, considered to play the major roles in
digestion, fermentation may also play key roles in digestive
processes in fish as in many other monogastrics. The role
of fermentation in fish is unclear, as research on microbiota
in fish intestine is still in its early stages. However, its role
is considered to be of minor quantitative importance for
nutrient supply in cold-water species. The importance of
the intestinal microbiota is highly significant for normal
functioning of the immune apparatus of the GI tract and
the general resistance of the fish towards pathogens and
other foreign factors constantly influencing the fish via the
intestine.
The characteristics of the microbiota, products of meta-
bolism, etc. depend greatly on the conditions of the intes-
tine, determined by species-specific parameters along the
GI tract such as anatomy, endogenous inputs of digestive
secreta, pH, osmolality, redox potential, compartment size
and structure, passage rate and residence time (Ray &
Ringø 2014).
The GI tract is a tube histologically differentiated in dif-
ferent segments that course through the body. This tube
may have a few to several hundred subcompartments in
which microbes may divide and grow. The GI tract is
commonly divided in the following regions: mouth, gill
arch, oesophagus, stomach, pyloric caeca, mid-intestine
(MI), distal intestine (DI) and rectum. The GI tract of
Atlantic cod is illustrated in Fig. 3. Some fish species lack
a typical stomach which in these fish is replaced by a
foregut. Pyloric caeca are finger-like extensions typical of
most teleost fish. They are located in the proximal part of
the intestine, MI, and, when present, number from a few,
as in Atlantic halibut (Hippoglossus hippoglossus L.), to
several hundred as in the Atlantic cod. The structure of
the wall of the GI tract varies along the tract, but has in
common a surface facing the lumen of mucus-producing
(goblet) cells between enterocytes. The latter holds diges-
tive and transport apparatus located in microvilli facing
the lumen, and being responsible for the uptake of nutri-
ents (Fig. 4a). The mucosa lining of the GI tract repre-
sents an interface between the external and internal
environments and, in conjunction with the associated
organs (e.g., pancreas, liver and gall bladder), provides
the functions of digestion, osmoregulation, immunity,
endocrine regulation of GI tract and systemic functions,
as well as the elimination of environmental contaminants
and toxic metabolites.
Just below the mucosa, we find the submucosa which is
a layer of connective tissue, blood vessels and nerves. A
single or double layer of muscles is located outside the
submucosa. The serosa forms the outer layer of the GI
tract. In some fish, the compartments may hardly be dis-
tinguishable macroscopically, while in other the sections
are divided clearly and may be separated by valves or
sphincters. The presence of valves and sphincters between
the subcompartments of the intestine may greatly influ-
ence the residence time of the chyme in the compartment
and hence for the possibilities of the microbiota to
develop.
The oesophagus is, in most fish, short and of small diam-
eter, with the possibilities to expand greatly, and with
numerous goblet cells aiding in food passage. A common
feature of carnivore fish species is great elasticity and
strong musculature in the stomach wall. In some fish spe-
cies, the muscles of the stomach seem to function as a grin-
der. Carnivore species show the shortest GI tract, typically
less than the body length, whereas in herbivore, such as
tilapia, the GI tract may be more than 20 times the body
length (Ray & Ringø 2014). The rectum is usually
separated from the rest of the tract by a sphincter and con-
tains more goblet cells than the more proximal parts, but
also has absorptive cells making the distal intestine of fish
Figure 1 Bacteria associated with the microvilli in the
hindgut chamber of Atlantic cod (Gadus morhua L.). After Løvmo
(2007).
...................................................................... ........................
Aquaculture Nutrition ª2015 The Authors. Aquaculture Nutrition Published by John Wiley & Sons Ltd.
phology and growth performance of Caspian white fish (Rutilus
frisii kutum) fry. Fish Shellfish Immunol.,39, 231–236.
Hoseinifar, S.H., Ringø, E., Shenavar Masouleh, A. &
Angeles
Esteban, M. (2014b) Probiotic, prebiotic and synbiotic supple-
ments in sturgeon aquaculture: a review. Rev. Aquacult., doi:
10.1111/raq.12082. [Epub ahead of print].
Hovda, M.B., Lunestad, B.T., Fontanillas, R. & Rosnes, J.T.
(2007) Molecular characterisation of the intestinal microbiota of
farmed Atlantic salmon (Salmo salar L.). Aquaculture,272, 581–
588.
Hovda, M.B., Fontanillas, R., McGurk, C., Obach, A. & Rosnes,
J.T. (2012) Seasonal variations in the intestinal microbiota of
farmed Atlantic salmon (Salmo salar L.). Aquacult. Res.,43,
154–159.
Hu, C.H. & Xia, M.S. (2005) Antibacterial effect of copper-bearing
montmorillonite on aquacultural pathogenic bacteria and its
mechanism. J. Chin. Ceram. Soc.,33, l376–l1380 (in Chinese
with English abstract).
Hu, C.H. & Xia, M.S. (2006) Adsorption and antibacterial effect
of copper-exchanged montmorillonite on Escherichia coli K
88
.
Appl. Clay Sci.,31, 180–184.
Hu, C.H., Xu, Z.R. & Xia, M.S. (2005) Antibacterial effect of
Cu
2+
-exchanged montmorillonite on Aeromonas hydrophila and
discussion on its mechanism. Vet. Microbiol.,109,83–88.
Hu, C.H., Xu, Y., Xia, M.S., Xiong, L. & Xu, Z.R. (2007) Effects
of Cu
2+
–exchanged montmorillonite on growth performance,
microbial ecology and intestinal morphology on Nile tilapia
(Oreochromis niloticus). Aquaculture,270, 200–206.
Huang, G. (2008) The study of intestinal bacterial molecular ecol-
ogy of cultured fishes. Mater thesis, Huazhong Agricultural
University.
Huang, I.-N., Okawara, T., Watanabe, M. et al. (2013) New
screening methods for probiotics with adhesion properties to
sialic acid and sulphate residues inhuman colonic mucin using
Biacore assay. J. Appl. Microbiol.,114, 854–860.
Huang, Z., Li, X., Wang, L. & Shao, Z. (2014) Changes in the
intestinal bacterial community during the growth of white
shrimp, Litopenaeus vannamei.Aquacult. Res., doi: 10.1111/
are.12628. [Epub ahead of print].
Huber, I., Spanggaard, B., Appel, K.F., Rossen, L., Nielsen, T. &
Gram, L. (2004) Phylogenetic analysis and in situ identification
of the intestinal microbial community of rainbow trout (On-
corhynchus mykiss, Walbaum). J. Appl. Microbiol.,96, 117–132.
Kihara, M. & Sakata, T. (1997) Fermentation of dietary carbohy-
drates to short-chain fatty acids by gut microbes and its influ-
ence on intestinal morphology of a detritivorus teleost tilapia
(Oreochromis niloticus). Comp. Biochem. Physiol.,118A, 1201–
1207.
Klappenbach, J.A., Dunbar, J.M. & Schmidt, T.M. (2000) RRNA
operon copy number reflects ecological strategies of bacteria.
Appl. Environ. Microbiol.,66, 1328–1333.
Iehata, S., Inagaki, T., Okunishi, S., Nakano, M., Tanaka, R. &
Maeda, H. (2009) Colonization and probiotic effects of lactic
acid bacteria in the gut of the abalone Haliotis gigantea.Fish.
Sci.,75, 1285–1293.
Iehata, S., Valenzuela, F. & Riquelme, C. (2015) Analysis of bacte-
rial community and nutritional enzyme activity associated with
the digestive tract of wild Chilean octopus (Octopus mimus
Gould, 1852). Aquacult. Res.,46, 861–873.
Iji, P.A., Saki, A.A. & Tivey, D.R. (2001) Intestinal structure and
function of broiler chickens on diets supplemented with a man-
nan oligosaccharide. J. Sci. Food Agric.,81, 1186–1192.
Inglis, G.D., Thomas, M.C., Thomas, D.K., Kalmokoff, M.L.,
Brooks, S.P.J. & Selinger, L.B. (2012) Molecular methods to
measure intestinal bacteria: a review. J. AOAC Int.,95,5–23.
Irianto, A. & Austin, B. (2002) Probiotics in aquaculture. J. Fish
Dis.,25, 633–642.
Ishikawa, M., Nakajima, K., Yanagi, M., Yamamoto, Y. &
Yamasato, K. (2003) Marinilactibacillus psychrotolerans gen. nov.,
sp. nov., a halophilic and alkaliphilic marine lactic acid bacterium
isolated from marine organisms in temperate and subtropical
areas of Japan. Int. J. Syst. Evol. Microbiol.,53, 711–720.
Ishino, R., Iehata, S., Nakano, M., Tanaka, R., Yoshimatsu, T. &
Maeda, H. (2012) Bacterial diversity associated with the rotifer
Brachionus plicatilis sp. complex determined by culture-depen-
dent and –independent methods. Biocontrol Sci.,17,51–56.
Isola, D., Pardini, M., Varaine, F. et al. (2005) A pyrosequencing
assay for rapid recognition of SNPs in Mycobacterium tuberculo-
sis embB306 region. J. Microbiol. Methods,62, 113–120.
Izumi, S. & Aranishi, F. (2004) Relationship between gyrA muta-
tions and quinolone resistance in Flavobacterium psychrophilum
isolates. Appl. Environ. Microbiol.,70, 3968–3972.
Izvekova, G.I., Izvekov, E.I. & Plotnikov, A.O. (2007) Symbiotic
microflora in fishes of different ecological groups. Biol. Bull.,34,
610–618.
Jaafar, R.M., Kania, P.W., Larsen, A.H., Nielsen, D.S., Fouz, B.,
Browdy, C. & Buchmann, K. (2013) Gut microbiota changes in
rainbow trout, Oncorhynchus mykiss (Walbaum), during organic
acid feed supplementation and Yesrinia ruckeri infection. J. Fish
Dis.,36, 599–606.
Jacobs, L. & Chenia, H.Y. (2007) Characterization of integrons
and tetracycline resistance determinants in Aeromonas spp. iso-
lated from South African aquaculture systems. Int. J. Food
Microbiol.,114, 295–306.
Jahn, C.E., Charkowski, A.O. & Willis, D.K. (2008) Evaluation of
isolation methods and RNA integrity for bacterial RNA quan-
tification. J. Microbiol. Methods,75, 318–324.
Jendrossek, D. & Handrick, R. (2002) Microbial degradation of
polyhydroxyalkanoates. Ann. Rev. Microbiol.,56, 403–432.
Jensen, M.A., Webster, J.A. & Straus, N. (1993) Rapid identifica-
tion of bacteria on the basis of polymerase chain reaction-ampli-
fied ribosomal DNA spacer polymorphisms. Appl. Environ.
Microbiol.,59, 945–952.
Jensen, S., Bergh, Ø., Enger, Ø. & Hjeltnes, B. (2002) Use of
PCR-RFLP for genotyping 16SrRNA and characterizing bacte-
ria cultured from halibut fry. Can. J. Microbiol.,48, 379–386.
Jensen, S., Øvre
as, L., Bergh, Ø. & Torsvik, V. (2004) Phylogenetic
analysis of bacterial communities associated with larvae of the
Atlantic halibut propose succession from a uniform normal
flora. Syst. Appl. Microbiol.,27, 728–736.
Jiang, W.-D., Feng, L., Liu, Y., Jiang, J. & Zhou, X.-Q. (2009)
Growth, digestive capacity and intestinal microflora of juvenile
Jian carp (Cyprinus carpio var. Jian) fed graded levels of dietary
inositol. Aquacult. Res.,40, 955–962.
Jiang, T.-T., Feng, L., Liu, Y., Jiang, W.-D., Jiang, J., Li, S.-
H., Tang, L., Kuang, S.-Y. & Zhou, X.-Q. (2014) Effects of
exogenous xylanase supplementation in plant protein-enriched
diets on growth performance, intestinal enzyme activities and
microflora of juvenile Jian carp (Cyprinus carpio var. Jian).
Aquacult. Nutr.,20, 632–645.
J€
oborn, A., Olsson, J.C., Westerdahl, A., Conway, P.L. & Kjelle-
berg, S. (1997) Colonization in the fish intestinal tract and pro-
duction of inhibitory substances in intestinal mucus and faecal
extracts by Carnobacterium sp. strain K1. J. Fish Dis.,20, 383–
392.
..................................................................... .........................
Aquaculture Nutrition ª2015 The Authors. Aquaculture Nutrition Published by John Wiley & Sons Ltd.
the gastrointestinal tract of adult yellow grouper (Epinephelus
awoara) cultured in cages. Aquaculture,286, 184–189.
Zhou, Q., Li, K., Jun, X. & Bo, L. (2009b) Role and functions of
beneficial microorganisms in sustainable aquaculture. Bioresour.
Technol.,100, 3780–3786.
Zhou, Z.-G., He, S., Liu, Y., Shi, P., Huang, G. & Yao, B.
(2009c) The effects of dietary yeast culture or short-chain
fructo-oligosaccharides on the intestinal autochthonous bacterial
communities in juvenile tilapia, Oreochromis niloticus♀9O. au-
reus♂.J. World Aquacult. Soc.,40, 450–459.
Zhou, Z., Liu, Y., He, S., Shi, P., Gao, X., Bin Yao, B. & Ringø,
E. (2009d) Effects of dietary potassium diformate (KDF) on
growth performance, feed conversion and intestinal bacterial
community of hybrid tilapia (Oreochromis niloticus♀9O. au-
reus♂). Aquaculture,291,89–94.
Zhou, Z., Karlsen, Ø., He, S., Olsen, R.E., Yao, B. & Ringø, E.
(2013a) The effect of dietary chitin on the autochthonous gut
bacteria of Atlantic cod (Gadus morhua L.). Aquacult. Res.,44,
1889–1900.
Zhou, Y., Yuan, X., Liang, X.-F., Fang, L., Li, J., Guo, X., Bai, X.
& He, S. (2013b) Enhancement of growth and intestinal flora in
grass carp: the effect of exogenous cellulase. Aquaculture,416–417,
1–7.
Zhou, Z., Yao, B., Romero, J., Waines, P., Ringø, E., Emery, M.,
Liles, M.R. & Merrifield, D.L. (2014) Methodological
approaches used to assess fish gastrointestinal communities. In:
Aquaculture Nutrition: Gut health, Probiotics and Prebiotics
(Merrifield, D. & Ringø, E. eds), pp. 101–127. Wiley-Blackwell
Publishing, Oxford, UK.
Zou, J., Clarke, M. & Secombes, C.J. (2003) Characterisation,
expression and promoter analysis of an interleukin 10 homologue
in the puffer fish, Fugu rubripes.Immunogenetics,55, 325–335.
Zou, J., Bird, S., Truckle, J., Bols, N., Horne, M. & Secombes,
C.J. (2004) Identification and expression analysis of an IL-18
homologue and its alternatively spliced form in rainbow trout
(Oncorhynchus mykiss). Eur. J. Biochem.,271, 1913–1923.
..................................................................... .........................
Aquaculture Nutrition ª2015 The Authors. Aquaculture Nutrition Published by John Wiley & Sons Ltd.
Mucosa-associated lymphoid tissue (MALT) in teleost
fish is subdivided into gut-associated lymphoid tissue
(GALT), skin-associated lymphoid tissue (SALT) and
gill-associated lymphoid tissue (GIALT) (Salinas et al.
2011). GALT which represents an essential part of an
organism’s adaptive defence system is considered to pro-
tect the host against pathogens not only by fighting the
intruding bacteria but also by modulating the composi-
tion of the microbiota. Microbiota stability in animals,
including fish, has been observed (Rawls et al. 2006).
Microbial communities transplanted from mice to gnoto-
biotic zebra fish altered quantitatively in the direction of
the normal biota of the zebra fish species and vice versa.
Antibodies, lysozyme and other antimicrobial components
in mucus secreted from the wall of the GI tract may
play a key role in this apparant stability of the intestinal
microbiota. The function of GALT depends on diet com-
position, such as its content of oligosaccharides, and the
nutritional status regarding essential nutrients, such as
selenium (Sweetman et al. 2010). In addition, GALT
must develop mechanisms to discriminate between patho-
genic and commensal micro-organisms (P
erez et al. 2010;
Suzuki et al. 2007).
It is logical that passage rate and residence time in the vari-
ous sections along the GI tract may influence the microbial
gut community and subsequent host–microbial interactions.
Stomach evacuation rate and passage time through the
intestine have been observed to vary with temperature,
meal size, particle size, feed composition, previous nutri-
tional history, fish size and stress (reviewed by F€
ange &
Grove 1979; Bromley 1994). Diet is also known to affect
passage time (Storebakken et al. 1999) and hence may
affect microbial growth. No information has been reported
on the relationship between growth of the microbiota, gut
passage rate and residence time.
In conclusion, the anatomy and function of the GI tract
and the digestive process, typical for a fish species,
Figure 3 The gastrointestinal tract of Atlantic cod. Note the many
pyloric ceaca which may number several hundred in this species.
The distal intestine is a pouch closed by sphincters in both ends.
3
4
5
6
7
8
9
Stomach Pyloric region Mid intestine Distal intestine
pH
(a)
(b)
Figure 4 (a) Structure of the GI tract wall. A histological presenta-
tion, stained with haematoxylin and eosin, of the wall of the mid-
intestine in Atlantic salmon. A layer of mucus, secreted by the gob-
let cells, covers the mucosal folds. Cells are dying continuously and
released from the top of the folds into the chyme, mixing with
unabsorbed food material as well as components of endogenous
secreta. Photograph: M. Penn. (b) pH in chyme of Atlantic salmon
in sea water (H. Holm &
A. Krogdahl, unpublished data). The data
originate from three feeding experiments, each testing three diets
varying in protein content or amino acid supplementation. Each
circle represents the mean pH of observations in several fish fed the
same diet. Only fish with content in the gut segments were used. No
significant effects of diet on pH were observed within experiment.
..................................................................... .........................
Aquaculture Nutrition ª2015 The Authors. Aquaculture Nutrition Published by John Wiley & Sons Ltd.
represent important conditions for the development and
growth of the microbiota. Variation in external environ-
mental conditions also greatly impacts the conditions for
microbial growth. However, our knowledge of these rela-
tionships is very limited in particular for fish.
Several different molecular methods are today available for
detecting micro-organisms in a given sample and monitor-
ing the change in microbial communities, without culture-
dependent techniques (Zhou et al. 2014). In most cases, a
combinatorial approach may be necessary for determining
the microbial composition (e.g. clone library construction
and PCR-based profiling techniques) and is a prerequisite
to determine the ‘true’ microbial communities in the fish
gut. An area of research that may shed light on the physio-
logical role of bacteria in the fish gut is to investigate the
gene expression of the bacteria and not only determine the
presence/absence of the bacteria, that is studying mRNAs
and not only DNA.
In the early studies investigating the gut microbiota of
fish, conventional culture-based methods, which are time-
consuming and selective, were used (for review, see Cahill
1990; Ringø et al. 1995). Conventional culture-based tech-
niques, even if several different media are used, do not pre-
sent a ‘true’ picture of the bacterial diversity. Therefore, to
present more reliable information about the gut microbiota
of fish, molecular methods are necessary. Culture-indepen-
dent methodologies are useful tools in furthering our
understanding of complex ecosystems and have highlighted
the limitations associated with culture-dependent tech-
niques. Denaturing gradient gel electrophoresis (DGGE),
terminal-restriction fragment length polymorphism
(T-RFLP), automated rRNA intergenic spacer analysis
(ARISA), single-strand conformation polymorphism
(SSCP), 16S rRNA gene clone libraries and the newly
developed 16S rRNA tag pyrosequencing method are
examples of such culture-independent techniques that have
been used to profile bacterial populations in a wide variety
of ecosystems (Lee et al. 1996; McBain et al. 2003; Saka-
moto et al. 2004; Yannarell & Triplett 2004; Van der
Gucht et al. 2005), including gut habitats such as the
rumen (Edwards et al. 2005; Yu & Morrison 2004) and
hindgut (Green et al. 2006; Simpson et al. 1999; Suchodol-
ski et al. 2004). In all of these techniques, extracted com-
munity DNA is amplified using the polymerase chain
reaction (PCR), utilizing the primers specific for conserved
regions of 16S rRNA. Examples of the published papers
using culture-independent methods in studies evaluating the
gut microbiota of aquatic animals are presented in Table 3.
Molecular-based methods to describe the microbial com-
munities in a certain sample can be divided into two
groups: (i) the PCR-based techniques which amplify certain
fragments of DNA or cDNA using user-defined primers,
and (ii) the PCR-independent methods which detect bacte-
ria without any gene- or cDNA amplification. Generally,
the PCR-independent methods are less specific and sensi-
tive than PCR-based techniques, and they are less suitable
for profiling bacterial communities. They are, however,
important tools to visualize bacteria in a spatial scale. Con-
ventional PCR-based methods are qualitative methods
when applied to environmental samples, due to the inher-
ent biasing in PCR amplification (Suzuki & Giovannoni
1996; von Wintzingerode et al. 1997; Polz & Cavanaugh
1998). When amplifying the 16S rRNA gene, which is by
far the most common target gene in studies of gut micro-
biota, the copy number heterogeneity will affect the diver-
sity in the resulting amplicon (Klappenbach et al. 2000).
This bias is particularly important to consider whether
searching for novel bacteria in environmental samples and
extensive sequencing may be required to detect less abun-
dant species. Furthermore, many of the PCR-dependent
techniques amplify only a short region of a particular gene,
and a precise taxonomic affiliation is often difficult. Never-
theless, amplification of selected gene(s) is necessary as at
least a supplementary approach to traditional culture-de-
pendent methods in order to better describe complex bacte-
rial communities, and to better monitor the changes that
occur in these communities across temporal and spatial
scales when influenced by biotic or abiotic factors. We will
here first describe some of the PCR-independent techniques
available, and then the methods which are based on the
PCR technique.
In situ hybridization In situ hybridization involves anneal-
ing of a probe to nucleic acids within the bacteria. The
specificity is defined by the probe sequence (Amann et al.
1995), which usually is in the size of 15–30 nucleotides. A
successful annealing of the probe will yield a visualization
of the bacteria in the spatial space. It requires prior knowl-
edge about the target sequence(s). In situ hybridization is a
powerful but challenging technique which requires opti-
mization of several of the involved steps (Amann & Lud-
wig 2000; Moter & G€
obel 2000). The most common
...................................................................... ........................
Aquaculture Nutrition ª2015 The Authors. Aquaculture Nutrition Published by John Wiley & Sons Ltd.
procedure today is to label the probe with a fluorophore,
called fluorescent in situ hybridization (FISH). This allows
for the simultaneous detection of different micro-organ-
isms, using a set of fluorophores with different excitation
and emission maxima. The probe can be either RNA- or
DNA-oligonucleotides, and the target can be RNA or
DNA. If using DNA as the target, both dead and viable
bacteria will be detected, while RNA as the target will only
reveal viable bacteria. Labelling can be performed directly
with a fluorescently labelled probe, which is the fastest,
cheapest and easiest way. To increase the labelling sensitiv-
ity, which may be relatively low using direct labelling, the
probe can be labelled indirectly by enzymatic signal ampli-
fication (Yamaguchi et al. 1996). The latter approach can
increase the signal intensity by up to ten-fold (Sch€
ohuber
et al. 1997), but the number of positively labelled bacteria
may decrease, probably due to insufficient penetration of
the high-molecular weight enzymes. The lack of automati-
zation makes in situ hybridization unsuitable to conduct
high-throughput analyses, but it is a valuable tool in
detecting bacteria in the spatial space. FISH has been used
successfully to understand the intestinal microbiota of sev-
eral fish species (Asfie et al. 2003; Holben et al. 2002;
Spanggaard et al. 2000; Tanaka et al. 2004).
Immunohistochemistry Instead of using oligonucleotides
for the detection of micro-organisms, bacteria can be
labelled with antibodies which can subsequently be visual-
ized by the use of secondary antibodies. This method has
some similarities with in situ hybridization; the samples are
fixed prior to labelling, the target bacteria can be visualized
in the spatial space, and a lack of automatization makes it
unsuitable for high-throughput analyses. However, whereas
probes for in situ hybridization are user-defined and anneal
to complementary nucleotides, antibodies are usually raised
against whole bacteria and will not necessarily have the
desired specificity. Immunohistochemistry is highly suitable
to follow the infection route of bacterial strains to which a
specific antibody has been raised (Løvoll et al. 2009), and
antibodies can be used both in light microscopy and electron
microscopy studies, depending on the desired degree of mag-
nification. The challenge is to raise a monoclonal antibody
with high specificity and with no cross-reactivity against
other closely related bacteria (Rengpipat et al. 2008), which
is a time-consuming and expensive process. In addition,
bacteria cultured in vitro and used for immunization may
have a slightly different morphology in vivo, considering that
bacteria are affected by the environment in which they grow.
This may result in changes in the antigen morphology
between in vitro and in vivo growth (Jung et al. 2008). At
last, immunohistochemistry does not yield a very high sensi-
tivity. The advantage of using a monoclonal antibody is that
a highly specific antibody can differentiate even between dif-
ferent strains, and it requires less optimization compared
with in situ hybridization.
Transcript analysis with aid of affinity capture New meth-
ods are continuously being developed to more accurately
determine the composition of microbial communities. One
of these is the transcript analysis with aid of affinity capture
(TRAC) method, which is a multiplexed and sensitive
method for relative quantification of bacteria. By solution
hybridization of biotinylated nucleic acids and fluorophore-
labelled oligonucleotides in combination with capillary elec-
trophoresis, it is possible to make a relative quantification
of selected transcripts (Kataja et al. 2006; Satokari et al.
2005). Technically, biotinylated nucleic acid probes anneal
according to their specificities to a pool of mRNA, and the
probes are then captured by streptavidin-coated beads, and
then separated through capillary electrophoresis. The
probes can vary in length from 18 to 41 nucleotides (Rau-
tio et al. 2006; Satokari et al. 2005) up to several hundred
nucleotides (Kataja et al. 2006). The size difference between
the different probes ensures that the fragments are well sep-
arated. This method does not rely on PCR amplification,
although amplification of the probe will increase the sensi-
tivity (Kataja et al. 2006), and can offer reliable and high-
throughput analyses of bacterial communities. Instead of
constructing primers annealing to user-defined target(s),
probes of different sizes with the same fluorophore are con-
structed which anneal to specific targets. After capture, the
probes are eluted and separated, and the peak intensity can
be measured quantitatively. The TRAC method is a profil-
ing technique without the need for PCR amplification and
may offer a more reliable estimate of the bacterial composi-
tion in a given sample than PCR-based methods.
All PCR-based methods consist of three basic steps: (i) nucleic
acid extraction, (ii) amplification of DNA and (iii) analysis
(either quantitatively or qualitatively) of PCR products.
Nucleic acid extraction For investigating the presence or
absence of bacteria, DNA can be extracted and used as
template in either PCR-dependent or PCR-independent
methods. There are, however, different DNA extraction
methods and these may influence the relative composition
..................................................................... .........................
Aquaculture Nutrition ª2015 The Authors. Aquaculture Nutrition Published by John Wiley & Sons Ltd.
of the DNA pool. He et al. (2009) compared the effects of
three different DNA extraction methods (lysozyme diges-
tion, CTAB method and bead mill) on the analysis of dif-
ferent micro-ecological environments in a farming pond
[pond sludge, feed, intestinal contents and the intestinal
wall of grass carp (Ctenopharyngoden idellus)] by DGGE of
16S rDNA V3 region. The results revealed obvious effects
in the DGGE fingerprints of the different micro-ecological
environments by the different DNA extraction methods.
Technically, DNA can be relatively easily extracted from
bacteria and can subsequently be used as templates in
downstream applications. The extraction of RNA, how-
ever, is challenging and requires more awareness than
DNA extraction. This is because RNases are not easily
denaturated, and thus chemicals and equipment have to be
RNase-free through the extraction procedure (Jahn et al.
2008). The clear reasons for studying the expression of
genes, and not purely the genes themselves, is an opportu-
nity to investigate the changes in expression of selected
genes due to biotic and abiotic treatments, and also for
studying the viable portion of the microbiota.
Clone library constructions Constructing clone libraries of
the 16S rRNA gene is probably the most widely used
method to gain sequence information from a given sample.
The clone library construction consists of several steps that
may influence the composition of the clones, for example,
the method used for DNA extraction, the primers used for
gene amplification, and the conditions used to amplify the
gene. As for all PCR-dependent techniques, the user-
defined primers determine the amplicon. Usually, primers
annealing to highly conserved regions of the 16S rDNA are
chosen in order to obtain an amplicon consisting of the
highest diversity as possible. It is also important to amplify
fragment that gives the best phylogenetic resolution, and
thus, primers annealing close to the 50-end and the 30-end
are preferred (e.g. 8F/27F as the forward primer and
1492R as the reverse primer). The construction of clone
libraries is often accompanied by other types of techniques,
such as PCR-DGGE, PCR-TGGE or T-RFLP, all of
which are typical profiling methods.
PCR-DGGE and PCR-TGGE Combining PCR amplifica-
tion with separation of the amplicons with either
denaturing- or temperature-gradient gel electrophoresis
(PCR-DGGE or PCR-TGGE) is a widely used technique
to determine the bacterial communities in fish and crus-
tacean (e.g. Hovda et al. 2007; Zhou et al. 2007; Liu et al.
2008; McIntosh et al. 2008; Merrifield et al. 2009a, 2013;
Li et al. 2012). In PCR-DGGE, the PCR fragments are
separated through a chemical denaturating gradient, while
a temperature gradient is used in TGGE (Muyzer & Smalla
1998). Both techniques are based on the separation of PCR
products of the same size but with different nucleotide
sequences. The different regions in the DNA strand will
denature at different time points when migrating through
an increasing denaturating agent (chemical or temperature).
Depending on the nucleotide composition (specifically the
G-C content), the migration behaviour will change. The
optimal resolution is achieved when the amplicons are not
entirely denaturated, and thus, a so-called GC clamp is
added to one of the primers to ensure that there is not a
complete denaturation (Myers et al. 1985). The bands can
be cut from the gel and sequenced; thus, sequence informa-
tion can be achieved without constructing a clone library
prior to the method. Although these methods are much
used to describe bacterial communities, they have some
limitations that should be noted. The amplicons are rela-
tively small (typically 150–500 bp), and it may be difficult
to ascertain a taxonomical affiliation of sequenced bands.
Furthermore, the GC clamp may produce primer–dimer
formations and overlapping (migration to the same point
in the gel) of phylotypes can occur which can lead to an
underestimation of community diversity and difficulties in
sequencing of bands. Additionally, the sensitivity of stain-
ing with traditional staining is relatively low (Nocker et al.
2007).
Terminal-restriction fragment length polymorphism Similar
to PCR-DGGE and PCR-TGGE, terminal-restriction frag-
ment length polymorphism (T-RFLP) is a profiling tech-
nique used to monitor changes in the bacterial community.
Here, either the forward or reverse primer is labelled with
a fluorophore. A subsequent digestion with restriction
enzyme(s) of the amplicon after PCR amplification creates
fragments of varying length depending on the restriction
site of each sequence. The fragments are then separated
using capillary electrophoresis, creating peaks which repre-
sents terminal-restriction fragments (TRFs). The separation
of the fragments is performed on the basis of differences in
fragment length, which is estimated by the comparison to
one or several DNA standards run simultaneously. This is
a high-resolution technique, which theoretically is able to
distinguish sequences differing by only one base pair. With
a prior knowledge of the microbial diversity of the sample,
for example through clone library sequencing, restriction
enzymes can be chosen for creating the highest numbers of
TRFs. The clear disadvantage compared with the DGGE
...................................................................... ........................
Aquaculture Nutrition ª2015 The Authors. Aquaculture Nutrition Published by John Wiley & Sons Ltd.
and TGGE techniques is that the fragments cannot be
sequenced after separation, and thus, T-RFLP has to be
accompanied by sequencing or in silico analyses. As the
separation is performed by capillary electrophoresis, analy-
ses can be performed much more rapidly than using
PCR-DGGE and PCR-TGGE. Although T-RFLP has
some limitations, it has proven useful for describing
changes in microbial communities (Blackwood et al. 2003;
Nocker et al. 2007; Dave et al. 2011).
Real-time (RT)-PCR RT-PCR is a quantitative PCR-
based approach combining traditional PCR amplification
with detection of fluorescence signals. It offers the possibil-
ity to monitor the abundance of selected gene(s) or tran-
script(s) in environmental samples, either relatively against
a normalization factor (e.g. nucleic acid concentration or
housekeeping gene expression), or absolutely by comparing
the expression to known quantities of the target gene con-
tained in plasmids. Two different reporter systems are most
often used to detect the fluorescence signal, namely the
SYBR green assay (Wittwer et al. 1997) and the TaqMan
probe system (Holland et al. 1991; Livak et al. 1995).
SYBR green binds double-stranded DNA, and the former
assay thus measures accumulating amounts of SYBR green,
and the TaqMan probe system utilizes a 50-end fluores-
cently labelled probe which emits light when separated
from a quencher located at the 30-end. During the anneal-
ing steps, the fluorophore will be released from the probe,
and thus, the latter system measures the fluorescence from
the released fluorophore. The TaqMan probe system is
highly specific, but it is more expensive than the SYBR
green chemistry. In addition, the SYBR green assay only
requires two conserved regions for the two primers to
anneal, in contrast to the probe technology which needs a
third conserved region for the binding of the probe. Tem-
plates for the amplification can be either DNA-extracted
from the sample, for purely detecting the specified gene(s),
or cDNA-synthesized from the mRNA pool to calculate
the gene expression. If using a taxonomic gene, such as 16S
rDNA, it may be possible to calculate the relative amount
of certain microbes. However, using functional genes as the
target(s) makes it possible to link any change in gene
expression to functionality. Using bacterial cDNA as tem-
plate in RT-PCR is still not routinely done, mainly due to
the difficulties in isolating high-quality RNA from bacteria
(Jahn et al. 2008). RT-PCR is, however, a promising
method for investigating bacterial processes in environmen-
tal samples, but the limitations, such as methods for
high-quality RNA isolation and a prior knowledge of the
target sequences, should be noticed (Smith & Osborn
2009).
Next-generation sequencing technologies Pyrosequencing is
another method that is used for high-throughput sequenc-
ing of clone libraries (e.g. Edwards et al. 2006; Jones et al.
2009). In contrast to traditional sequencing using the San-
ger method, pyrosequencing utilizes the released pyrophos-
phate which is used to produce ATP. ATP is then used by
luciferase to convert luciferin to oxyluceferin, and the emit-
ted light from this reaction is measured (Ronaghi et al.
1998; Gharizadeh et al. 2002). The method has now been
scaled up and may determine the composition of hundreds
of thousands of sequences simultaneously. It has a great
potential for high-throughput sequencing to a considerably
lower cost than the Sanger method, but still the sequence
length is relatively short (700 bp) and thus, the taxonomic
resolution is much weaker than traditional sequencing. In
microbiological applications, it has proven useful in
analysis of the microbial community in human intestine
(Dethlefsen et al. 2008), macaque gut (McKenna et al.
2008) and tidal flat sediments (Kim et al. 2008), in bacterial
typing (Jonasson et al. 2002), and analysis of single nucleo-
tide polymorphism (Isola et al. 2005).
Single-strand conformation polymorphism In the single-
strand conformation polymorphism-PCR (SSCP-PCR)
technique (Lee et al. 1996), denaturated PCR products are
separated through either an acrylamide gel or a capillary
array sequencer. In non-denaturating conditions, single-
stranded DNA folds into tertiary structures based on their
nucleotide sequence and the physiochemical environment
(e.g. temperature and ionic strength), and these different
conformations will separate the PCR products through dif-
ferences in migration behaviour. If using acrylamide gels,
the bands can be cut from the gel and sequenced. This is
not possible using a capillary array sequencer. Some major
limitations, such as reannealing of PCR products during
separation and the unpredicted behaviour of the bands,
must be considered if employing this method (Nocker et al.
2007).
rRNA intergenic spacer analysis rRNA intergenic spacer
analysis (RISA) is a method of microbial community
analysis which provides estimates of microbial diversity
and community composition without the bias imposed by
culture-based approaches or the labour involved with
small-subunit rRNA gene clone library construction.
RISA was used originally to contrast diversity in soils
..................................................................... .........................
Aquaculture Nutrition ª2015 The Authors. Aquaculture Nutrition Published by John Wiley & Sons Ltd.
(Borneman & Triplett 1997) and more recently to examine
microbial diversity in the rhizosphere and marine environ-
ments (Acinas et al. 1999; Robleto et al. 1998). The
method involves PCR amplification from total bacterial
community DNA of the intergenic region between the
small (16S) and large (23S) subunit rRNA genes in the
rRNA operon, with oligonucleotide primers targeted to
conserved regions in the 16S and 23S genes. The 16S-23S
intergenic region, which may encode tRNAs depending on
the bacterial species, displays significant heterogeneity in
both length and nucleotide sequence. Both types of varia-
tion have been extensively used to distinguish bacterial
strains and closely related species (Aubel et al. 1997; Jen-
sen et al. 1993; Maes et al. 1997; Navarro et al. 1992;
Scheinert et al. 1996). In RISA, the length heterogeneity
of the intergenic spacer is exploited. The PCR product (a
mixture of fragments contributed by community members)
is electrophoresed in a polyacrylamide gel, and the DNA
is visualized by silver staining. The result is a complex
banding pattern that provides a community-specific pro-
file, with each DNA band corresponding to at least one
organism in the original assemblage. To the author’s
knowledge, this method has only been used in one fish
study (Navarrete et al. 2010a).
From a microbial point of view, the pyloric caeca is of vital
interest as lipid digestion and absorption occur in this
organ (Olsen & Ringø 1997). However, due to its complex
morphology, only some studies have investigated the
microbiota of pyloric caeca in fish (Al-Hisnawi et al. 2014;
Gildberg & Mikkelsen 1998; Gildberg et al. 1997; Lesel &
Pointel 1979; Ransom et al. 1984; Zhou et al. 2009a), and
to our knowledge, no investigation so far has evaluated the
effect of dietary lipid on microbiota of pyloric caeca, a
topic that merits investigations.
In their study with rainbow trout (Oncorhynchus mykiss
Walbaum), Lesel et al. (1989) fed the fish two different
diets, low (50 g kg
1
) and high (160 g kg
1
) lipid levels.
Differences in faecal bacterial composition were observed,
as the faecal microbiota of fish fed low lipid level consisted
of only Acinetobacter spp. and Enterobacteria. In contrast,
Acinetobacter spp., Aeromonas spp., Enterobacteria,
Flavobacterium spp., Pseudomonas spp. and coryneforms
were isolated from faeces of fish fed the high-lipid level.
However, as only 12 isolates from each dietary group were
isolated, no clear conclusion can be drawn.
Earlier diets for cultured salmonids contain high
amounts of carbohydrates (approximately 20% of dry
weight), but in recent years, there has been a tendency
towards decreasing dietary carbohydrate content from
about 20% to 10%, with a subsequent increase in the
level of dietary lipid from <20 to up to 30%. Based on
this tendency, Ringø & Olsen (1999) fed Arctic charr
(Salvelinus alpinus L.) diets containing high (27%) and
low (13%) levels of dietary lipid. In this study, approxi-
mately 190 isolates were identified from each dietary
group. Dietary manipulation modulated the species com-
position of carnobacteria, as Carnobacterium spp.- and
C.mobile-like strains were only isolated from DI of fish
fed low dietary lipid, while C.piscicola-like strains were
isolated from proximal intestine. In contrast to these
results, C.divergens-like isolates were found associated
with the mucus layer of proximal and DI of fish fed high
dietary lipid.
Fish oils were for many years the predominating lipid
source in diets for carnivorous fish species. However, the
increase in aquaculture led to an increased consumption
from 16% of available fish oil in 1988 to 81% in 2002
(Tacon et al. 2006). This was close to full exploitation.
Although studies with substitutes had been done in the
past, the prospect of deficiencies spurred extensive work
into finding replacements. One obvious choice was veg-
etable oils (Dosanjh et al. 1984; Guillou et al. 1995; Hardy
et al. 1987; Montero et al. 2003; Ng et al. 2007; Rosenlund
et al. 2001; Torsteinsen et al. 2004). The main reason was
that the global production is approximately 100 times
higher than that of fish oils [FO] (Bimbo 1990; Tacon et al.
2006) with no prospects of limitations. Secondly, they often
come at compatible prices compared with FO.
As no information was available about how inclusion of
vegetable oils in commercial raw material affects the gut
microbiota of fish, Ringø et al. (2002) investigated the
effect of soybean-, linseed- and marine oils on the hindgut
microbiota of Arctic charr. This study showed clear differ-
ences in the hindgut microbiota of fish fed different oils
(after and prior to challenge with Aeromonas salmonicida
subsp. salmonicida). Carnobacteria were only isolated from
the hindgut region of fish fed soybean oil (SBO) and lin-
seed oil before challenge, while Carnobacterium spp.- and
C.funditum-like strains were isolated from fish fed the
...................................................................... ........................
Aquaculture Nutrition ª2015 The Authors. Aquaculture Nutrition Published by John Wiley & Sons Ltd.
same oils after challenge. Furthermore, the ability of
carnobacteria to inhibit the growth of A.salmonicida ssp.
salmonicida was highest in strains isolated after challenge.
These results might have some interest as Lødemel (2000)
and Lødemel et al. (2001) demonstrated that survival of
Arctic charr after challenge with A.salmonicida subsp.
salmonicida was improved by dietary SBO.
In an unpublished study, E. Ringø, R.E. Olsen & S.
Sperstad fed Atlantic salmon and rainbow trout extruded
diets consisting of lipid-free protein meal supplemented
with either 30% vegetable oil containing either (i) linseed
oil, (ii) rapeseed oil, (iii) sunflower oil or (iv) 30% fish oil.
As a control diet, the fish were fed a commercial diet.
Nine hundred bacterial strains were isolated from the
rearing water, diet and the hindgut of Atlantic salmon. The
number of viable aerobic and facultative aerobic autochtho-
nous bacteria associated in the hindgut was reduced from
10
5
bacteria per gram wet weight intestinal tissue (fish fed a
commercial diet, prior to experimental start) to approxi-
mately 5 910
3
bacteria g
1
wet mass intestinal tissue of
fish fed dietary vegetable oil, after two months of feeding.
Seven hundred and twenty-six autochthonous bacterial
strains were isolated from the digestive tract. Of these, 29
strains died prior to positive identification. A wide range of
bacteria were isolated from the hindgut of the five rearing
groups (four experimental and one commercial). There
was, however, some difference in bacterial composition
between the rearing groups (Table 4). In the DI of fish fed
the commercial diet, Psychrobacter and Staphylococcus
were the dominant bacterial genera, followed by Pseu-
domonas jessenii/fragi-like and Psychrobacter submarinus/
marincola-like strains. In addition, Acinetobacter spp.- and
Carnobacterium mobile-like strains were also isolated. Eight
strains died prior to positive identification.
When Atlantic salmon was fed the sunflower oil diet,
P.jessenii/fragi-like strains were dominant followed by Sta-
phylococcus spp. However, when feeding the fish with a diet
supplemented with rapeseed oil, P.submarinus/marincola-
like strains and Staphylococcus spp. were prevalent. From
this rearing group, Pseudomonas fluorescens-like and
C.mobile-like strains were also isolated. In the hind gut of
fish fed linseed oil, Enterococcus pseudoavium-like strains
were the dominant bacterial species, followed by Acineto-
bacter spp. and Staphylococcus spp. In contrast to these
results, P.submarinus/marincola-like strains were dominant
in the GI tract of fish fed fish oil. Furthermore, dietary
manipulation seemed to modulate the presence of the lactic
acid bacteria (LAB), as 13.6% and 9% of the strains
isolated in the hindgut of fish fed rapeseed- and fish oil
belong to C.mobile, respectively. However, very few (~2%)
C.mobile strains were isolated from the digestive tract of
the other dietary groups.
Table 4 Log total viable counts (log TVC) per gram wet mass, number of isolates and changes in log TVC of bacterial species isolated
from the digestive tract of seven Atlantic salmon fed; sunflower oil (A), rapeseed oil (B), linseed oil (C) and fish oil (D) and from five fish
fed commercial diet (prior to experimental start) (E). After E. Ringø, R.E. Olsen & S. Sperstad (unpublished data)
ABCDE
Log TVC 3.47 3.90 3.70 3.60 5.0
No. of isolates 154 154 154 154 110
Gram-negative
Acinetobacter spp.
1
n.d n.d 2.87 (7) 2.59 (5) 3.56 (2)
Pseudomonas fluorescens
1
1.76 (1) 3.10 (7) 2.41 (5) 2.41 (4) n.d
Pseudomonas jessenii/fragi
1
3.30 (7) n.d n.d 2.74 (7) 4.10 (5)
Pseudoalteromonas agarivorans
1
n.d 2.89 (7) 2.66 (4) n.d n.d
Psychrobacter spp.
1
n.d 2.41 (5) n.d n.d 4.36 (5)
Psychrobacter submarinus/marincola
1
2.07 (4) 3.37 (6) 2.11 (2) 3.19 (7) 4.10 (5)
Gram-positive
Carnobacterium mobile
1
1.29 (1) 3.04 (5) n.d 2.56 (6) 3.56 (2)
Enterococcus spp.
1
2.37 (7) n.d n.d 2.46 (5) n.d
Enterococcus pseudoavium
1
n.d 2.02 (1) 3.41 (7) 2.31 (4) n.d
Staphylococcus spp.
1
2.72 (7) 3.26 (7) 2.79 (6) 2.41 (4) 4.34 (5)
Not identified 4.19
Unknown 1.76 2.56 2.29 2.11 3.86
n.d, not detected.
Unknown –isolates that died prior to positive identification.
The number of fish from which bacteria were isolated is given in brackets.
1
Partial sequence of 16S rRNA were analysed and edited in BioEdit. An initial BLAST-search in GenBank retrieved the taxonomic groups
for which showed highest identities.
..................................................................... .........................
Aquaculture Nutrition ª2015 The Authors. Aquaculture Nutrition Published by John Wiley & Sons Ltd.
Information regarding the antagonistic activity of gut
bacteria against fish pathogens has been investigated in sev-
eral papers during the last decade (Gatesoupe 1999; Ringø
et al. 2005; Balc
azar et al. 2007a; Ringø 2008; P
erez-
S
anchez et al. 2011). Antagonistic activity of intestinal
bacteria to inhibit the growth of three fish pathogenic bac-
teria (A.salmonicida subsp. salmonicida,Vibrio anguillarum
and Vibrio (Aliivibrio)salmonicida) was tested by the micro-
titre plate assay. Of the 692 gut isolates tested, antimicro-
bial activity against A.salmonicida and V.anguillarum was
only observed in six C.mobile strains isolated from Atlan-
tic salmon fed sunflower oil. These six isolates were identi-
fied by random-amplified polymorphic DNA polymerase
chain reaction as described by Seppola et al. (2006).
In an experiment with rainbow trout fed similar diets to
that of the previously discussed Atlantic salmon study, 850
bacterial strains were isolated from the rearing water, diet
and the hindgut (E. Ringø, R.E. Olsen & S. Sperstad,
unpublished data). Of these, 726 autochthonous strains
were isolated from DI of fish fed diets supplemented; sun-
flower-, rapeseed-, linseed- or marine oils and the commer-
cial diet prior to experimental start were characterized.
Thirty-four strains, 4.6% of the total bacterial strains iso-
lated, died prior to positive identification.
The number of viable aerobic and facultative aerobic
heterotrophic bacteria associated with the DI was reduced
from 8 910
4
bacteria per gram wet weight intestinal tissue
(rainbow trout fed a commercial diet, prior to experimental
start) to approximately 4 910
3
bacteria g
1
wet mass
intestine in fish fed the experimental diets, after two
months of feeding (Table 5). These results were similar to
those reported for Atlantic salmon.
The DI microbiota of rainbow trout fed dietary veg-
etable oils was dominated by P.submarinus/marincola-like
strains and strains belonging to Staphylococcus linens/equo-
rum (Table 5). Strains belonging to Acinetobacter,
Table 5 Log total viable counts (log TVC) per gram wet mass, number of isolates and changes in log TVC of bacterial species isolated
from the digestive tract of seven rainbow trout fed; sunflower oil (A), rapeseed oil (B), linseed oil (C) and marine oil (D) and from five fish
fed commercial diet (prior to experimental start) (E). After E. Ringø, R.E. Olsen & S. Sperstad (unpublished data)
ABCDE
Log TVC 3.90 3.70 3.47 3.60 4.90
No. of isolates 154 154 154 154 110
Gram-negative
Acinetobacter johnsonii
1
n.d 1.51 (1) n.d 2.32 (5) 3.86 (3)
Brachybacterium tyrofermentas 2.62 (6) 1.81 (1) 1.89 (2) n.d n.d
Janibacter spp.
1
n.d 2.00 (2) n.d n.d n.d
Pseudomonas fluorescens
1,2
n.d n.d n.d n.d
Pseudomonas jessenii/fragi
1
2.89 (5) 2.41 (6) 2.50 (5) 2.85 (7) 4.51 (5)
Pseudoalteromonas agarivorans
1,2
2.41 (3) n.d n.d n.d n.d
Psychrobacter faecalis
1
n.d 1.81 (2) n.d n.d
Psychrobacter submarinus/marincola
1
3.46 (7) 3.25 (7) 3.00 (7) 2.91 (7) 4.16 (4)
Serratia spp.
1
n.d n.d n.d 1.72 (1) n.d
Shewanella putrefaciens
1
n.d n.d n.d 1.89 (2) n.d
Gram-positive
Arthrobacter rhombi
1
2.56 (4) 2.00 (2) n.d n.d n.d
Bacillus pumilus
1
n.d n.d 2.14 (4) n.d n.d
Brevibacterium casei
1
2.41 (4) n.d n.d n.d n.d
Carnobacterium spp.
1
n.d n.d 1.89 (2) 2.46 (5) 3.56 (3)
Carnobacterium mobile
1
2.89 (5) n.d n.d n.d n.d
Enterococcus faecalis
1
n.d n.d 2.29 (6) n.d n.d
Enterococcus gallinarum
1
2.19 (2) n.d n.d n.d n.d
Marinilactibacillus psychrotolerans
1
n.d 2.36 (4) n.d n.d n.d
Staphylococcus spp.
1
n.d n.d n.d n.d 4.26 (5)
Staphylococcus equorum spp. linens
1
3.19 (7) 3.29 (7) 3.02 (7) 3.24 (7) n.d
Unknown 2.71 2.47 2.24 2.41 3.56
n.d, not detected.
Unknown –isolates that died prior to positive identification.
The number of fish from which bacteria were isolated is given in brackets.
1
Partial sequence of 16S rRNA were analysed and edited in BioEdit. An initial BLAST-search in GenBank retrieved the taxonomic groups
for which showed highest identities.
2
Pseudomonas fluorescens and Pseudoalteromonas agarivorans were also isolated from rearing water.
...................................................................... ........................
Aquaculture Nutrition ª2015 The Authors. Aquaculture Nutrition Published by John Wiley & Sons Ltd.
Arthrobacter,Brachybacterium,Janibacter,Pseudomonas,
Pseudoalteromonas,Psychrobacter,Brevibacterium,Bacillus,
Brevibacterium,Enterococcus,Marinilactibacillus psychrotol-
erans and carnobacteria were also isolated. However, some
changes in microbiota composition seem to occur between
the dietary groups. For example, Enterococcus strains were
not isolated from fish fed dietary rapeseed oil, while
Marinilactibacillus psychrotolerans-like strains were only
isolated from this dietary group. To our knowledge, the
new genus and species, M.psychrotolerans a halophilic and
alkaliphilic marine LAB, was proposed by Ishikawa et al.
(2003).
A difference in gut microbiota between fish fed vegetable
oils and marine oil was observed as Shewanella putrefaciens
and Serratia spp. were only isolated from fish fed marine oil.
On the other hand, the clearest differences in gut microbiota
were detected between dietary groups fed vegetable oils and
marine oil compared with fish fed the commercial diet prior
to the start of the experiment, as the hindgut microbiota of
fish fed the commercial diet comprised of Acinetobacter john-
sonii,Pseudomonas jessenii,Psychrobacter submarinus,
Carnobacterium spp. and Staphylococcus spp.
Fifteen of the 154 strains isolated from the digestive tract
of rainbow trout fed sunflower oil were identified as C.mo-
bile by 16S rRNA gene sequence analysis, but C.mobile
was not isolated from the gut of the other rearing groups.
On the other hand, 4 strains of Carnobacterium spp. were
isolated from the gut of two fish fed linseed oil, 11 strains
from 5 fish fed marine oil and 5 strains from three fish fed
the commercial diet prior the start of the experiment. These
results indicate that carnobacteria are present in the ali-
mentary tract of rainbow trout, but generally at relatively
low population levels and they seem easily affected by diet-
ary manipulation.
During the last decade, some information has become
available about the antagonistic activity of carnobacteria
against fish pathogens (Ringø et al. 2005). Antagonistic
activity against A.salmonicida and V. anguillarum was
observed in 11 of the 15 C.mobile strains isolated from
rainbow trout fed sunflower oil, but antagonistic activity
against the two pathogens was only observed by 4 of the
20 carnobacteria strains isolated from the other rearing
groups. An important question rises based on the antago-
nistic activity of carnobacteria isolated from the distal
intestine of Atlantic salmon and rainbow trout. Why was
antagonistic activity mostly detected in some C.mobile iso-
lated from Atlantic salmon and rainbow trout? This finding
has not been elucidated and merits further investigations.
The low frequency of antimicrobial activity observed in gut
isolates isolated from Atlantic salmon and rainbow trout
contradicts the results reported by Makridis et al. (2005) in
their study on Senegalese sole (Solea senegalensis) fed two
diets, commercial diet or polychaete (Hediste diversicolor),
revealing that the numbers of bacterial strains with antibac-
terial activities towards pathogens increased by feeding the
fish polychaete.
Lauzon et al. (2008) reported that 13.8% of bacteria iso-
lated from cod rearing (not all gut bacteria) revealed inhi-
bitory activity, but only 3.2% of the bacteria were
antagonistic to all three pathogens tested. Similarly, Hjelm
et al. (2004) identified 8% of turbot larval rearing isolates
as inhibitory. Spanggaard et al. (2001) reported 4% of
isolates (from skin, gills and GIT of rainbow trout) to be
inhibitory towards V. anguillarum. When evaluating antag-
onistic activity, one should also pay attention to the
method used to determine this property as medium utilized
may influence outcome (Lauzon et al. 2008) as well as tem-
perature may do (Caipang et al. 2010).
When discussing the antagonistic effect of gut bacteria
towards pathogens, it is worth to mention that in a probi-
otic study with Atlantic salmon, Gram et al. (2001) used
P. fluorescens strain AH2 a strain showing strong in vitro
inhibitory activity towards A. salmonicida. However,
co-habitant infection by A.salmonicida in Atlantic salmon
did not result in any effect on furunculosis-related mortal-
ity. Based on their results, the authors concluded that a
strong in vitro growth inhibition cannot be used to predict
a possible in vivo effect.
Montero et al. (2006) studied the effect of vegetable oils
on the gut microbiota of gilthead sea bream (Sparus aur-
ata) postintraperitoneal challenge with Vibrio alginolyticus.
As this was not the main emphasis of the study, little infor-
mation was reported. However, the authors noted that cul-
turable levels of total Gram-positive bacteria and Vibrio
spp. were present in fish fed a diet containing 60% replace-
ment of fish oil with a blend of rapeseed oil, linseed oil and
palm oil (at a 15 : 60 : 25 ratio) in comparison with diets
containing 100% fish oil or 100% vegetable oil blend. The
authors concluded that this was a result of modified fatty
acid composition of the intestinal epithelial cell walls. How-
ever, as these preliminary data were presented with little
information on the methodology, that is number of repli-
cates and number of isolates identified, the data should be
viewed with caution.
In study by Huang (2008), a 1.23% mixture of oil (phos-
pholipid oil: rice bran oil in the ratio 2 : 1) was supple-
mented into a practical diet for grass carp and the
autochthonous gut bacteria was evaluated by 16S rDNA
..................................................................... .........................
Aquaculture Nutrition ª2015 The Authors. Aquaculture Nutrition Published by John Wiley & Sons Ltd.
V3 DGGE after 8-week feeding. The similarity of the
bacterial communities between the two dietary groups was
0.73 by the cluster analysis. Uncultured Proteobacterium-
like organisms were stimulated, while Clostridium mariti-
mum-like were depressed by the dietary mixture oil
compared with the control.
Electron microscopy Attention has been paid to the
importance of scanning electron microscopy (SEM) investi-
gations in gut studies of the microbiota (Ringø & Olsen
1999; Ringø et al. 2001a,b, 2002, 2003, 2007a,b; ; Merri-
field et al. 2009a,b, 2011a,b; Harper et al. 2011). Ringø &
Olsen (1999) demonstrated a correlation between classical
microbial examination and SEM, as both methods showed
predominance of coccoid-shaped bacteria in the hindgut
region of Arctic charr. In a later study, Ringø et al.
(2001a) showed substantial associations of both coccoid-
and rod-shaped bacterial cells with the apices of, and
between, microvilli of the enterocytes in the MI of Arctic
charr fed SO. The reason for this difference in colonization
pattern of the enterocyte surface has not been elucidated,
but Ringø et al. (2001a) suggested three possible causes: [1]
enterocyte ageing, [2] differentiation, or [3] specific recep-
tors for receptor-mediated endocytosis of bacteria. Com-
pared with that reported by Ringø et al. (2001a), clear
differences in bacterial colonization of the enterocyte
surface in the hindgut were observed when Arctic charr
were fed linseed oil (Ringø et al. 2002). These differences
in bacterial colonization of enterocyte surface between fish
fed SO and linseed oil may be related to the different gut
microbiota observed between the two dietary groups
(Ringø et al. 2002). This controversial hypothesis calls for
further investigations.
Although fatty acids are important in fish metabolism, few
studies have evaluated the effect of dietary polyunsaturated
fatty acids on the gut microbiota (Ringø 1993a; Ringø
et al. 1998). In his study on the effect of linoleic acid
(18:2n-6) on intestinal microbiota of Arctic charr, Ringø
(1993a) was not able to isolate LAB in the intestinal con-
tents, but large numbers of Aeromonas spp., Pseudomonas
spp. and Enterobacteriaceae were isolated when 2.5% lino-
leic acid was supplemented to a commercial recipe. In con-
trast to these findings, Lactobacillus spp. accounted for
approximately 10% of the microbiota when the fish were
fed the unsupplemented diet. In a recent study with Arctic
charr fed casein-based diets supplemented with different
fatty acids [18:2n-6, a-linolenic acid (18:3n-3), or a HUFA
mix (20:5n-3 and 22:6n-3)], Ringø et al. (1998) showed no
suppression of LAB (Carnobacterium spp., Carnobacterium
piscicola and Lactobacillus plantarum) in the stomach, PI
and DI. However, a significant increase in both total viable
counts and population level of LAB was observed in DI
and faeces of fish fed 7% 18:3n-3 or 4% HUFA mix. This
was due to a large extent to the increased levels of
Carnobacterium spp. The reason for the increase in LAB in
fish fed 7% linolenic acid and HUFA mix has not been
elucidated, but the authors suggest that dietary fatty acids
influence intestinal membrane composition, function and
fluidity which may affect the attachment sites of the gut
mucosa. Later, this controversial hypothesis was confirmed
by Kankaanp€
a€
aet al. (2001). They demonstrated that cul-
turing of Caco-2-cells with arachidonic acid (20:4n-6)
reduced the Caco-2 cell adhesion of LAB, whereas 18:3
(n-3) did not hinder adhesion of Lactobacillus GG or
Lactobacillus bulgaricus, and promoted the adhesion of
Lactobacillus casei Shirota.
In view of the results observed by Ringø et al. (1998), it
is interesting to note that the ability of C.piscicola-like iso-
lates to inhibit the fish pathogen A.salmonicida subsp.
salmonicida was highest in strains isolated from fish fed
linolenic acid or the HUFA mix (E. Ringø, unpublished
data). Based on these results, it is recommended that
greater attention should be given to the subject of how to
increase the level of intestinal carnobacteria with inhibitory
effect against fish pathogens by dietary manipulation. The
results obtained from fish fed dietary 18:3 (n-3) may lead
to the conclusion that it is desirable to increase the level of
dietary 18:3 (n-3) in commercial diets in order to obtain a
higher population level of intestinal strains of C. piscicola
able to inhibit growth of A. salmonicida subsp. salmonicida.
However, it is worthwhile to note that feeding the charr an
experimental diet with high levels (>15%) of dietary 18:3
(n-3) increased accumulation of lipid droplets in the entero-
cytes and caused cell damage which may increase the risk
of microbial infections (Olsen et al. 1999, 2000).
The gut microbiota of goldfish (Carassius auratus) (Sugita
et al. 1988a), Atlantic cod (Strøm & Olafsen 1990), Arctic
charr (Ringø & Strøm 1994), abalone (Haliotis discus)
(Tanaka et al. 2003), puffer fish (Takifugu obscurus) (Yang
et al. 2007), yellow grouper (Epinephelus awoora Temminck
& Schlegel, 1942) (Feng et al. 2010), gilthead sea bream
and goldfish (de Paula Silva et al. 2011) and transgenic
...................................................................... ........................
Aquaculture Nutrition ª2015 The Authors. Aquaculture Nutrition Published by John Wiley & Sons Ltd.
common carp (Cyprinus carpio L.) (Li et al. 2013) have
been investigated in an attempt to clarify the effect of
different diets on the intestinal microbiota.
In their early study, Sugita et al. (1988a) concluded that
Aeromonas hydrophila and Bacteroides type A were pre-
dominant in almost all goldfish fed on either the pelleted
diets and tubifex worms or pelleted diets and dried daph-
nia. The authors concluded that the gut microbiota was
not influenced by the diets. However, as the fishes were fed
the different diets for only a short period of time (22 days),
no conclusion can be drawn. In contrast to the conclusion
by Sugita et al. (1988a), Strøm & Olafsen (1990) demon-
strat