AAAS

Research
Volume 2020, Article ID 8272705, 15 pages
https://doi.org/10.34133/2020/8272705

Review Article
Sensors, Biosensors, and Analytical Technologies for Aquaculture
Water Quality

Xiaodi Su ,1,2 Laura Sutarlie,1 and Xian Jun Loh 1

1
Institute of Materials Research and Engineering, Agency for Science, Technology and Research, 2 Fusionopolis Way. Innovis #08-03,
Singapore 138634
2
Department of Chemistry, National University of Singapore, Block S8, Level 3, 3 Science Drive 3, Singapore 117543

Correspondence should be addressed to Xiaodi Su; xd-su@imre.a-star.edu.sg and Xian Jun Loh; lohxj@imre.a-star.edu.sg

Received 28 September 2019; Accepted 8 January 2020; Published 17 February 2020

Copyright © 2020 Xiaodi Su et al. Exclusive Licensee Science and Technology Review Publishing House. Distributed under a
Creative Commons Attribution License (CC BY 4.0).

In aquaculture industry, fish, shellfish, and aquatic plants are cultivated in fresh, salt, or brackish waters. The increasing demand of
aquatic products has stimulated the rapid growth of aquaculture industries. How to effectively monitor and control water quality is
one of the key concerns for aquaculture industry to ensure high productivity and high quality. There are four major categories of
water quality concerns that affect aquaculture cultivations, namely, (1) physical parameters, e.g., pH, temperature, dissolved
oxygen, and salinity, (2) organic contaminants, (3) biochemical hazards, e.g., cyanotoxins, and (4) biological contaminants, i.e.,
pathogens. While the physical parameters are affected by climate changes, the latter three are considered as environmental
factors. In this review, we provide a comprehensive summary of sensors, biosensors, and analytical technologies available for
monitoring aquaculture water quality. They include low-cost commercial sensors and sensor network setups for physical
parameters. They also include chromatography, mass spectrometry, biochemistry, and molecular methods (e.g., immunoassays
and polymerase chain reaction assays), culture-based method, and biophysical technologies (e.g., biosensors and nanosensors)
for environmental contamination factors. According to the different levels of sophistication of various analytical techniques and
the information they can provide (either fine fingerprint, highly accurate quantification, semiquantification, qualitative
detection, or fast screening), we will comment on how they may be used as complementary tools, as well as their potential and
gaps toward current demand of real-time, online, and/or onsite detection.

1. Introduction lenges that affect farm productivity and quality. The lack of
understanding in aquaculture nutrition, feed preparation,
1.1. Aquaculture Industry and Challenges. Aquaculture is the and proper feeding management will cause less desirable
farming (breeding, raising, and harvesting) of aquatic organ- water quality in both the land-based and nonland-based
isms, especially for human consumption. It is a global indus- farms due to accumulation of undigested food. Poor disease
try with increased importance in battling the challenges of management, partially due to the slow pathogen identifica-
the food supply in the future [1]. The aquaculture industry tion relying on laboratory culture plate count, and thus
has, in the last four decades, grown at a rate of 7% on average improper usage of drugs, will lead to drug/chemical residual
each year [2], being faster compared to other sectors in ani- deposits in the fish tissue. This will not only pose potential
mal food production industry. The global human population health hazard to humans when consumed, it would also lead
will eat 30 million tons of fish by 2030, according to the to discharge of fish wastes containing such residual chemicals
United Nations Food and Agriculture Organization. Millions into the surrounding water, causing a buildup of antibiotic
of people globally have found income and livelihood in the or drug resistance in the farmed products and surrounding
fisheries and aquaculture sector [2]. ecosystem overtime. In offshore aquaculture, or open ocean
In aquaculture industry, fish, shellfish, and aquatic plants aquaculture, on the other hand, organic contaminants, like
(such as algae, seaweeds) are cultivated in fresh, salt, and polycyclic aromatic hydrocarbons (PHAs) and polychlori-
brackish waters. Feeds and feeding, fish health and disease nated biphenyls (PCBs) etc., formed due to incomplete,
management, good aquaculture practices, etc. are key chal- but high-temperature and short-duration combustions of
2 Research

organic matters including fossil fuels and biomass are another (NH4+) in the water is converted to toxic ammonia (NH3), a
important factor affecting for fish health and quality [3]. compound that is lethal to fishes. On the other hand, when
To ensure successful scaling up of aquatic farming, water the pH level goes below 5, acidic water would leach metals
quality control is a key aspect of fisheries management. Tech- from rocks and sediments. These metals would adversely
nologies capable of rapid, real-time, and automatic monitor- affect fishes’ metabolism rates and their ability to take in
ing of aquaculture environment are in high demand. water through their gills, resulting in a fatality as well.

1.2. Scope of This Review. In this review article, we will discuss 2.2. Wireless Sensor Networks for Physical Parameters. Tradi-
four major areas of aquaculture system monitoring, concern- tionally, water quality fish farms are measured periodically
ing water quality (Figure 1), namely, physical parameters, i.e., onsite using handheld sensors. The aforementioned physical
pH, temperature, dissolved oxygen, and salinity (Section 2), parameters are commonly required to maintain acceptable
organic chemical contaminants (Section 3), biochemical haz- levels for fish growth regardless the type of fishes. While
ard, i.e., cyanotoxins (Section 4), and biological contami- handheld instruments or sensors can provide onsite mea-
nants, i.e., bacteria and virus (Section 5). We will discuss surement by the staff during office hours, the variation of
the conventional and modern analytical technologies that one of the key water parameters beyond a safe level can occur
have been developed and applied for these parameters and out of office hours, unseen by the staff. When the bad situa-
analytes, and their status of commercial exploitation. For tion is persistent, it will lead to undesirable effects, such as
newer biosensors and analytical technologies for the environ- poor growth, undetected disease symptoms, or abnormal
mental contamination factors, we will particularly comment behavior of the fish [12, 13].
on their strengths and limitations, and their suitability for In recent years, advancements in different Information
fast inspection and/or for accurate diagnosis, in either onsite and Communication Technologies (ICT) along with the
manner or real-time. With a summary of future demand of development of low-cost small sensors have increased the
real-time continuous detection, smart sensors, data connec- feasibility of monitoring numerous parameters concurrently
tivity, etc., we hope this review can provide a clear view on through wireless sensor networks (WSN). The applications
the level of readiness of various analytical technologies to of WSN include the monitoring of the three vigor [14],
meet the future demands. greenhouses [15], citrus crops [16], the state of farm animals
This review will complement to earlier review articles of (goats [17] or cows [18]), and fish farms [19–25]. WSN is
relevant scopes, for example, those discuss direct fish health composed of many self-organized sensors deployed in a
monitoring (stress level, prawning prediction, and fish dis- monitoring region that measure, collect, transmit, and pro-
ease) [4, 5], that for aquaculture pathogens [6], that about cess information in real time. The information measured is
sources of monitoring environmental exposure to polycyclic then displayed on a computer or conveyed in the form of a
aromatic hydrocarbons [7], and that about detection of cya- message to farmers for the real-time update. This real-time
notoxins in freshwater [8]. remote monitoring technology allows the streamlining of
the information accumulation process, which conceivably
2. Physical Parameters minimizes human lapses and time delays, hence increasing
the quantity and quality of data on temporal and spatial
2.1. Major Physical Parameters and Their Impact. Physical scales [20].
parameters, including dissolved oxygen (DO), temperature, Numerous studies have been conducted to study and
pH level, salinity, and turbidity, are the basic parameters to develop WSN for use in aquaculture [21–25]. Most of these
be monitored and controlled in an aquaculture system [9]. technologies boost the ability to measure the majority of
Fluctuations in these parameters will directly affect the health the important physical parameters in real-time and the capa-
of the animals, feed utilization, growth rates, and carrying bility of informing the necessary people at the facility when-
capacities [10]. ever a problem arises, allowing for quick and immediate
The temperature of the water affects the feeding pattern resolution. Espinosa-Faller and Redón-Rodríguez [22] pre-
and the growth of fish. When the temperature is chronically sented a WSN-based water monitoring system that transmits
near their maximum tolerance or fluctuates suddenly, fishes the gathered information and stores them in a database. The
would generally experience stress and disease breakout. system is also able to measure temperature, pressure, and DO
Moreover, warm water carries less DO than cool water. The throughout the day. When a problem was detected, an SMS
level of DO in water, and hence the amount of oxygen or an E-mail was forwarded to alert the person responsible
consumed, is directly linked to the size of fish, feeding rate, for the facility. Zhang et al. [23] proposed another WSN-
activity level, and pond temperature [11]. The DO level based water monitoring system that was able to measure
would decrease as temperature increases, and when salinity pH, water temperature, water level, and DO. The data col-
increases. Not only the optimal level of DO is essential for lected was forwarded to a database that provided the infor-
fish respiration, it is also imperative for the survival of phyto- mation to the software to be monitored in real time. In
plankton—an organism that breaks down toxic ammonia their system, the software used was able to separate logic, dis-
into harmless forms. This organism thus plays an indirect play, and data layers to improve scalability and reusability.
part in maintaining the pH level of the water as well. The Warnings could also be forwarded via SMS to the users.
acceptable range of pH for fish culture is usually between Lastly, Huang et al. [24] presented a WSN-based system that
pH 6.5 and 9.0. When pH level is higher than 9, ammonium gathered data on pH, temperature, and DO. It contains a
Research 3

Aquatic water quality (fresh water, salt water, brackish water)

Physical parameters Organic Biological

Biochemical hazards
(pH, temp, DO, etc) contaminants contaminants
(cyanotoxins)
(Pathogens & virus)

Environmental contaminations

Nanosensors
Electrical sensors Bacteria culture
Commercial sensors
HPLC/MS ELISA
Sensor networks PCR
SERS

Optical sensors Electrochemical sensors

Sensors, biosensors, and analytical techniques

Figure 1: Scopes of this review. Four major water quality parameters and analytical technologies involved. SERS: surface-enhanced
Raman spectroscopy; PCR: polymerase chain reaction; ELISA: enzyme-linked immunosorbent assay; LPS: lipopolysaccharides; DO:
dissolved oxygen.

real-time interface that displays the data numerically and isms. They also interfere with the food safety of aquaculture
graphically. More recently, Luo et al. [25] reported a real- products. For example, DDT (dichlorodiphenyltrichloroeth-
time remote monitoring system for aquaculture water quality ane) and aldrin (examples of OCPs) have the strongest ten-
using solar cells and lithium cells for power supply. In their dency to bioaccumulate in fish fatty tissue, eventually
system, they have integrated the commercial YCS-2000 DO reaching to high levels at the end of the food chain [28]. Stud-
sensor, pH electrode, Pt1000 temperature sensor, and ammo- ies have reported the presence of OCPs in human lipid tissues
nia nitrogen sensor. and human breast milk, of which fish consumption would be
Not only for water quality monitoring, Parra et al. [26] a possible source [29–31].
proposed a WSN for monitoring the water quality and fish
behavior in aquaculture tanks during the feeding process
(Figure 2). The system is composed of three sensor nodes 3.2. Laboratory-Based Techniques for Organic Contaminant
in each tank that send the information though the local area Detection. Sensitive and reliable determination of various
network to a database on the Internet. A smart algorithm is contaminants in aquaculture systems, particularly in food-
also included in the system that detects abnormal values stuffs (fish or shellfish), has been largely relied on laboratory
and sends alarms when they happen. It is a low-cost system based techniques, including liquid (LC) or gas chromatogra-
(<90€) including the sensors and nodes. phy (GC) combined with mass spectrometry (MS). GC is
preferred over LC due to its higher selectivity and resolution
for complex matrices such as aquaculture samples [32]. Con-
3. Organic Contaminants
figurations, such as GC coupled to tandem MS with a triple
Apart from physical parameters that require constant moni- quadrupole (QqQ), successfully detected 16 PAHs which
toring to optimize water quality in aquaculture, various con- were present in the Environmental Protection Agency
taminants (including chemical, biochemical, and biological (EPA) priority pollutant list. GC–ECD (electron capture
hazards) from the environment are also of serious concerns detection) was used to measure PCBs [33], while OCPs were
of water quality. In this section, we will discuss organic con- measured by GC-MS. Finally, UHPLC-MS/MS was widely
taminants and their analysis. used to simultaneously determine 41 antibiotics [34].
Table 1 is a summary of these studies, from which one can
3.1. Organic Contaminants and Health Risks. Organic com- have a clear view on the specific sample types and the perfor-
pounds, such as polycyclic aromatic hydrocarbons (PAHs), mance of these methods. For organic contamination analysis
polychlorinated biphenyls (PCBs), organochlorinated pesti- in solid aquatic samples (e.g., fish body, tissue, fish fillet, and
cides (OCPs), brominated flame retardants (BFRs), and fish oil), microwave-assisted extraction (MAE), conventional
residues of veterinary drugs and antibiotics, can enter aqua- solvent microextraction, and ultrasound-assisted extraction
culture systems via feed, which can be potentially transferred (UAE) techniques have been largely developed and applied
to organisms [27]. These contaminants are able to bioaccu- to extract the analytes [34, 35]. In two most recent review
mulate as they go across the food chain, causing detrimental articles, MAE and UAE techniques have been discussed for
effects on human health upon ingesting contaminated organ- their working principle, efficiency, and applications for both
4 Research

Internet

Box nodes
Router
Wi-Fi link
Switch

Access point Ethernet link

Server
Serial link

Other devices Internet

(a)

Database
t
sen
t ion Local offices
ma Local area network
for
In
ion
at Box node 1
m
n for est Light sensor
I qu Alarm
re Data from
data sensors Switch Access point Presence Oil
sensor sensor
Router Level
sensor
External supervisors Plexiglas tube
Internet
Smart algorithm

Fish presence
Box node 2 sensor

Worker
Pellet sensor
Antenna Box node 3

(b)

Figure 2: A low-cost sensor network for monitoring the water quality and fish behavior in aquaculture tanks. (a) Network topology. (b)
Architecture of proposed system [26] (Open Access).

solid and liquid sample extractions in food and environmen- antibodies to capture respective target analytes in microplate
tal samples [36, 37]. wells. After a few steps of incubation and washing, immuno-
complexes are formed inside the wells. The analytes are
3.3. Biochemistry Methods. The enzyme-linked immunosor- detected by either colorimetric or fluorescent signals gener-
bent assay (ELISA) is a very popular analytical biochemistry ated by the enzyme-conjugated antibody in the complexes
method. It has been widely developed for various analytes in the presence of enzyme substrates.
from macromolecules (e.g., proteins and DNA) to small- There are a range of ELISA kits available for PAHs, PCBs,
molecular-weight drugs and other organic compounds of and antibiotic residues. Table 2 shows a few examples that are
health and environmental concerns. ELISA mostly relies on mostly for water and soil analysis. Since these analytes are
Research 5

Table 1: Laboratory-based analytical techniques for organic contaminants in aquaculture samples.

Performance
Target Analytical
Matrix Dynamic Ref.
analyte technique LOD
range
PAHs Fish fillet from gilthead sea bream (S. aurata) GC/QqQ-MS/MS 0.2-200 ng/ml 0.02 μg/kg-0.1 μg/kg [32]
PCBs
PCBs Fish muscle from rainbow trout GC–ECD 1–200 ng/ml 0.1 (PCB 105)–1.4 (PCB 153) ng/g [35]
(Oncorhynchus mykiss)
Muscle tissues of five fish species
OCPs (O. mossambicus, L. parsia, E. suretensis, GC-MS 5–200 ppb 0.7–18.2 ng/ml [38]
C. striata, and S. wynaadensis)
Fish muscle from gilthead sea
Antibiotics UHPLC-MS/MS 50–300 μg/kg NIL [39]
bream (S. aurata)

Table 2: Commercial ELISA kits for PHAs, PCBs, and antibiotic residues.

Kit/company Principle LOD Sample matrix

PAH RaPID Assay Competition ELISA Groundwater, surface water, well
Total PHAs in ppb level
(Strategic Diagnostics Inc.) on magnetic beads water
Total petroleum
Soil: 20, 50, 100, 200 ppm as diesel fuel
hydrocarbon (TPH) Soil and water
Water: 2, 5, 10, 20 ppm as diesel fuel
(HACH)
Competitive
Polychlorinated biphenyls Semiquantitative screening based on thresholds
colorimetric ELISA Water
(PCBs) (HACH) for PCB
assay
Competitive
MaxSignal® Florfenicol Human samples (urine and serum)
colorimetric ELISA 0.2-1.0 ppb
ELISA Test Kit foods (milk, meat, egg, honey, etc.)
assay
Competitive
RaPID Assay® Soil: 0.2 ppm to 5 ppm as phenanthrene
colorimetric ELISA Soil and water
ModernWater (UK) Water: 2.66 ppb to 66.5 ppb as phenanthrene
assay
Competition ELISA
Aviva PAH ELISA Kit LOD 10 ng/ml Environmental PAH samples
microplate
Abraxis Tetracyclines Competition ELISA 4.0 ppb in honey; 4.0 ppb in milk; 8.0 ppb in
Food and water
ELISA microplate meat; 4.0 ppb in shrimp; 0.11 ppb in water

mostly small molecules, it is difficult to use the sandwiched copy (SERS) is a powerful technique to provide fingerprint
assay principle, but more of competitive principles. The information for trace chemicals. It has been largely applied
assays can be in either microplate format or using magnetic for identification and detection of organic contaminants in
beads. Antibodies against the chemical and chemical groups water, particularly PAHs, e.g., phenanthrene and fluorene
are used as the major affinity ligand to bind with the [40], naphthalene and pyren [41], anthracene and pyrene
chemicals. [42], and the mixture of PAHs [43, 44]. In these studies,
ELISA is considered more portable than GC-MS and metallic nanostructures were functionalized by hydrophobic
HPLC, due to the reliance of less sophisticated plate reader. films (e.g., glycidyl methacrylate-ethylene dimethacrylate,
But they still require hours of incubation and multiple 10-decanethiol monolayer, or β-cyclodextrin) to allow pre-
washing steps. Moreover, since antibodies often have the concentration of nonpolar molecules on the nanostructure
cross reactivity to chemicals of similar structure, the assays surface. The chemical identities and concentration of each
usually cannot differentiate individual PHAs, for example, chemical are analyzed by their distinct Raman signatures,
as stated by the RaPID Assay® and Aviva PAH ELISA Kit. amplified under the plasmonic effect. Figure 3 shows the
Sensitivity wise, ELISA offers similar sensitivity as GC-MS. example where β-CD dimer on silver nanoparticles embed-
ELISA can serve the purpose of fast screening. Those ded with silica nanoparticles (Ag@SiO2 NPs) structure is fab-
ELISA-positive samples are usually sent for HPLC and ricated for detecting PAHs [44]. A thioether-bridged dimeric
GC-MS for confirmation. β-CD was immobilized on Ag surface to capture PAHs. The
β-CD dimer@Ag@SiO2 SERS sensor can detect perylene in a
3.4. Biosensors and Nanosensors. Nanomaterials and nano- wide linearity range of 10−7 M to 10−2 M with a low detection
technologies have been largely applied in analytical sciences. limit of 10−8 M (1000 times lower than that without β-CD
Metal nanoparticle-based surface-enhanced Raman spectros- dimer). Furthermore, this β-CD dimer@Ag@SiO2 SERS
6 Research

7 35
Pyrene
6 30 1240 cm–1
PAHs

Norm intensity (a.u.)

Norm intensity (a.u.)

5 25
4 20
3 15

2 10
5
1
0
0
–5
Silica NP 400 600 800 1000 1200 1400 1600 1200 1300 1400 1500 1600 1700
Silver NP Raman shift (cm–1) Raman shift (cm–1)

𝛽-CD dimer 1:1 Parylene

1:0.1 Pyrene

(a) (b) (c)

Figure 3: (a) Schematic of β-CD dimer-immobilized Ag assembly with embedded silica NPs (β-CD dimer@Ag@SiO2 NPs) for SERS
detection of PAHs. (b) SERS spectra of four PAHs. (c) SERS spectra of perylene and a fixed concentration of pyrene [44] (Copyright ©
2020, Springer Nature).

sensor and another gold nanoparticle/β-CD sensor [43] can are chromatography tandem mass spectrometry [51–53].
detect multiple PAH compounds in a mixture by exploiting For example, an ultra-performance liquid chromatography
their distinct SERS bands. In another gold nanoparticle- tandem mass spectrometry (UPLC-MS/MS) method was
based SERS study, its ppb level sensitivity for PHAs and the developed to measure nine cyanotoxins in fish muscle tissue,
potential for field test has been demonstrated through off- and free cyanotoxin levels in 34 fish (muscle tissue) and 17
shore experiments [41]. livers and eggs from five pond-based aquaculture farms in
Southeast Asia [46]. The results indicated the presence of
4. Biochemical Hazards cyanotoxins. The levels are lower than those in Zaria, North-
ern Nigeria, and China, but similar to those reported from a
4.1. Biochemical Hazards and Health Risks. With increasing freshwater lake in Mexico. While being highly accurate and
global climate change and eutrophication, there has been a sensitive, these technologies require sophisticated laboratory
rise in Harmful Algal Blooms (HABs), exponential blooming equipment and are not suitable for onsite rapid analysis.
of algae, in freshwater and marine ecosystems [45–47]. Some
cyanobacterial species of HABs produce cyanotoxins that 4.3. Biochemistry Methods for Biochemical Hazards. There is
usually target human nervous systems (neurotoxins), livers a diverse range of biochemistry methods that are considered
(hepatotoxins), or skin (dermatoxins). The chemical struc- as rapid tests to detect and identify cyanobacteria cells and
ture of cyanotoxins falls into three broad groups: cyclic pep- cyanotoxins in water, including enzyme-linked immunosor-
tides, alkaloids, and lipopolysaccharide (LPS) endotoxins. bent assays (ELISA), protein phosphatase inhibition assay
They are nature-occurring organic pollutants. (PPIA), protein synthesis inhibition assay (PSI), conven-
Cyanobacterial hepatotoxins, such as a group of toxic tional polymerase chain reaction (PCR), quantitative real-
cyclic peptides called microcystins (MCs) with ~80 conge- time PCR (qPCR), and microarrays/DNA chips [54]. These
ners, nodularin (NODs), and cylindrospermopsin (CYN), methods can vary greatly in their degree of sophistication
are known to cause liver failure [48]. MCs are reported in and the information they provide. For example, the PPIA is
numerous regions such as Asia, Europe, North Africa, North a cost-effective assay using p-nitrophenyl phosphate to
America, and Scandinavian countries, while NODs are con- quickly assess water and rumen content samples [55].
fined within Australia, New Zealand, and the Baltic Sea Researchers have compared this rapid assay with LC-MS.
[49]. Evidence of accumulation of these toxins in humans They have concluded that this rapid assay can be used in
from fish in MC-contaminated waters is documented by a combination with LC-MS, where after the PPIA, subsequent
study conducted on fishermen subsisting from Lake Chaohu analysis using liquid chromatography mass spectrometry
in China, where MCs were detected in the serum of fisher- (LC-MS/MS) can provide more detailed information about
men [50]. microcystin congeners -LR, -LA, -RR, and -LF. They have
Lipopolysaccharide (LPS) endotoxins are structural com- commented on the advantages of using this rapid functional
ponents of the outer membrane of cyanobacteria and other assay in combination with LC-MS/MS. Similar to the PPIA,
gram-negative bacteria. They are responsible for a wide range PSI quantify cyanotoxins by nonphosphatase-related protein
of infections in humans, including sepsis and septic shock. inhibition. Froscio et al. conducted PSI for cyanobacterial
LPS endotoxins are found in all types of water including toxin cylindrospermopsin (CYN) quantification [56]. The
freshwater, saline surface water, and groundwater. results were compared to quantifications obtained by HPLC
and HPLC-tandem mass spectrometry (HPLCMS-MS). They
4.2. Conventional Analytical Technologies for Biochemical found that the results correlate well with both HPLCMS-MS
Hazards. Conventional methods for cyanotoxins analysis (r 2 = 0:99) and HPLC (r 2 = 0:97) quantifications. While PSI
Research 7

and PPIA are mostly for cyanotoxins detection, PCR and sample matrix. Thus, a careful design of the sensor aggrega-
qPCR are nucleic acid-based technologies mostly for identi- tion strategy to avoid fault aggregation and a careful study
fying cyanobacteria (more PCR-related bacteria detection is of the sample matrix effects on the aggregation are essential
in Section 5). [73–77].
In a comparative cell numbers and toxin concentrations To achieve an ultrahigh sensitivity, Xie et al. developed a
measured using ELISA, it is found that the results do not nec- colorimetric aptasensor based on DNA hybridization chain
essarily correlate and that enumeration of potentially toxic reaction (HCR) in microplates (Figure 4(b)) [78]. Briefly,
cyanobacteria by microscopy. The concentrations of certain two complementary biotinylated DNA hairpins coexisted in
cyanotoxins (e.g., saxitoxins) quantified by ELISA were sig- the assay solution. They will not hybridize until the introduc-
nificantly different than those measured by LC-MS. On the tion of a detection probe. The detection probe consists of
other hand, results were comparable in both assays for other three regions, namely, LPS-binding aptamer, a spacer, and
cyanotoxins (e.g., microcystin and cylindrospermopsin). HCR initiator. In the presence of LPS, this detection probe
With all these facts, people have reached to a conclusion that triggered a hybridization chain reaction cascade in the micro-
there is no “gold standard” technique for the detection of cya- plate, where the LPS were captured by the ethanolamine
notoxins. The choice of a detection assay depends on cost, aptamer attached to the reaction well surface. Under the opti-
practicality, and reliability of the results that can indicate mal conditions, the increase in the LPS concentration led to
toxin exposure potential [54]. increase of the optical density value generated by the HRP-
catalyzed color reaction. The sensor has a LOD of 1.73 ng/ml
4.4. Biosensors for Cyanotoxins and Cyanobacteria. Biosen- and a linear response range of 1–10 ng/ml.
sors are compact analytical devices and have been largely As mentioned in Section 2, portable sensors for water
developed for toxin and bacteria analysis [57–59]. They physical parameters have been largely available. But for envi-
could be more appealing for onsite and real-time mea- ronmental contaminants, sensor portability has been less
surement [60]. Among various biosensor formats, optical well addressed. Dickman et al. recently reported a portable
and electrochemical sensors are mostly developed for biosensor system for rapid detection of freshwater cyanotox-
aquatic water analysis. Cunha et al. reported an optical ins [79]. It is a planar waveguide optical sensor that delivers
sensor that exploits DNA aptamer and quantum dots quantitative fluorescent competitive immunoassay results in
(QDs) for real-time onsite detection of aquatic toxins pro- a disposable cartridge. They supply a duplex microcystin
duced by marine and freshwater microorganisms (cyano- (MC)/cylindrospermopsin (CYN) assay cartridge that relies
bacteria, dinoflagellates, and diatoms) [61]. This sensor on a combination of fluorophore-conjugated monoclonal
exploits fluorescence resonance energy transfer (FRET) antibodies (Figure 5). More of this type of portable biosensor
between a quencher-labeled DNA aptamer and QDs, in is in great demand in aquatic industry.
the presence of target toxin. In addition to the detection of cyanobacteria toxins,
Lebogang et al. [62] reported an electrochemical, particu- direct detection of cyanobacteria is also useful for water qual-
larly capacitive immunosensor for broad-spectrum detection ity assessment. A nucleic acid biosensor assay has been
of the group of toxic cyclic peptides, called microcystins (∼80 described to detect cyanopeptolin coding region of one of
congeners). The sensor can detect at very low concentration the cyanobacteria (Planktothrix agardhii NIVA-CYA 116)
range (1 × 10−13 M to 1 × 10−10 M) closer to the MC-LR stan- genome for monitoring of the fresh water resources [78].
dard, with a limit of detection of 2:1 × 10−14 M. Yu et al. [63] This is an electrochemical sensor integrated to a microflui-
presented an electrochemical sensor for microcystin-LR that dics system. A real-time amperometric measurement leads
exploits a quantum dot/antibody (QD/Ab) probe for signal to a detection limit of 6 × 10−12 M target DNA (calibration
amplification. A qualitative analysis for microcystin-LR was curve r 2 = 0:98).
achieved using the specific peak potential of the anodic volt-
ammogram at −0:6 ± 0:05 V. The microcystin-LR analyte
can be detected in a dynamic range of 0.227 to 50 μg/l with 5. Biological Contaminants
a limit of detection of 0.099 μg/l.
Due to the wide range concern of endotoxin contamina- 5.1. Fish Pathogens. A pathogen is defined as an organism
tion, LPS biosensors have been largely developed, including that causes diseases to its host. Pathogens are widely diverse,
optical fiber-based plasmonic sensor [64], colorimetric sen- including mainly bacteria and viruses [80]. They damage ani-
sors [65, 66], electrochemical sensors [8, 67, 68], and nano- mal tissues or cells during replication, usually by generating
material sensors [69]. The nanomaterial-based LPS sensors toxins, allowing pathogens to enter new tissues or exit the
have exploited versatile sensing principles, where the nano- cells inside which they replicated. Fish and fish product-
particle either being used as direct signal transducer, i.e., col- related bacteria and virus are diverse. Based on the extent
orimetric sensing based on their aggregation (Figure 4(a)) or of infection, some of them only infect or kill fishes without
as signal amplifier in electrochemical sensors [62, 68]. Metal infecting human, e.g., Vibrio harveyi, Vibrio anguillarum,
nanoparticle aggregation-based colorimetric sensors are easy and Aeromonas salmonicida [81, 82], but some of them can
to use (mostly with simple mixing, followed by either visual infect both fish and human such as Vibrio vulnificus, Myco-
inspection of color change or UV-vis measurement of the bacterium marinum, and Streptococcus iniae [83]. There are
absorption spectrum) [70–72]. But they are prone to generate also pathogens that may not necessarily infect the fish, but
fault-positive results as they tend to aggregate in complex infect human through consumption of contaminated fish
8 Research

n
AuNPs

u
SH
LBP (A)
LPS
(B) n TMB+H2O2

Wash
SH n

AuNPs
(C) n

Aptamer of LPS
from E coli 0111.84
spacer sequence Detection probe H1 H2 Biotin
initiator of HCR

Streptavidin-HRP Streptavidin Aptamer of ethanolamine LPS

(a) (b)

Figure 4: (a) Colorimetric LPS sensor exploiting gold nanoparticle aggregation [66]. (b) Hybridization chain reaction-based aptasensor for
LPS [76] (Open Access).

products, e.g., Vibrio cholerae, Escherichia coli, and Salmo-

nella spp.
Among many bacteria pathogens, Vibrio spp., a genus of
gram-negative bacteria ubiquitous in many aquacultures and
marine habitats, are the most common and serious patho-
gens in fish and shellfish aquaculture worldwide [81]. Many
of their species, such as Vibrio vulnificus, Vibrio harveyi, Vib-
rio anguillarum, Virbio alginolyticus, Vibrio parahaemolyti-
(a) cus, and Vibrio salmonicida, cause mass death and vibriosis
(marked by infection on skin and other organs) in many cul-
Dried Fluid
input Outlet
tured fish, shrimps, and shellfish [81, 84]. Among the Vibrio
reagent Cartridge upper
spp., around 12 species of them cause infections in humans,
which can be classified into cholera (by Vibrio cholerae)
and noncholera Vibrio infections (by Vibrio parahaemolyti-
Printed protein cus, Vibrio alginolyticus, and Vibrio vulnificus) [85].
Integrated Region imaged or capture
Laser input coupling lens by istrument antibody array
Analytical methods for quantitative and qualitative
pathogen detection in aquaculture include culture-based
Liquid channel
methods, molecular methods, biosensors, and microscopy
observation methods. The first three methods will be dis-
cussed in the following sessions due to their popularity
(e.g., culture methods are considered as current standard),
Evanescent Control Toxin Toxin Negative Comer
field antibody conjugate conjugate control marker
power of quantitative and qualitative analysis (e.g., molecular
1 2 methods), and high potential for onsite and real-time analy-
Waveguide substrate sis (e.g., biosensors).

5.2. Culture-Based Methods. Culture-based methods are con-

Imaging lens
and filter sidered as current “standard” for bacteria pathogen detec-
tion. In culture methods, bacteria samples from aquaculture
Camera water or part of the infected fish are incubated on agar
medium and the bacteria colonies grown on the agar are enu-
(b) merated. Various agar media have been formulated to culture
fish pathogens, including general media to culture a broad
Figure 5: A portable planar waveguide optical sensor for rapid
detection of freshwater cyanotoxins. (a) The proposed MBio range of potential pathogens, and specialized media (selecti-
reader and cartridge. (b) Schematic of LightDeck technology ve/differential) to target specific pathogens. The specialized
elements [77] (Copyright © 2020, American Chemical Society). media are typically formulated with selective agents to pro-
mote growth of target bacteria, inhibitor agents to stop the
growth of other bacteria, and color indicator to differentiate
Research 9

target bacteria [86]. A few commercial media for fish patho- on the sensor surface to capture target analyte of viral RNA
gens are shown in Figure 6. or bacteria cells, respectively. While biosensors are consid-
Culture-based methods have limitation as they are time- ered more suitable for onsite application, the RNA detection
consuming (more than 1 day to get results), not suitable to poses two major limitations, namely, (1) requiring sample
grow bacteria in viable but nonculturable state, and require preparation to extract the RNA and (2) involving PCR pro-
lab setting. In addition, specialized media may not be able cesses. On the other hand, direct detection of bacteria cells
to completely identify the target pathogens at species and by biosensors does not require extraction step and could be
strain level, and further pathogen identification would more suitable for onsite application so long as the sensitivity
require more specific molecular methods or biosensors [87]. can fulfill the requirement. Lateral flow test can be considered
as the most portable form of biosensor that combines sample
5.3. Molecular Assays-PCR. Bacteria and virus identification separation, interaction, and detection in one chromato-
and detection can be at cell level or molecular level. DNA is graphic strip. For Vibrio cholerae O1, for example, a lateral
currently the best molecule for bacteria detection as it pre- flow test can reach to a LOD of 107 cfu/ml. To push to a lower
sents evident biological information. Polymerase chain reac- detection limit of 102 cfu/ml, 6 hours of culture enrichment is
tion (PCR) is a popular technique extensively employed to needed. The SPR sensor work can be appreciated by its effort
amplify a single or several copies of a specific DNA sequence of screening aptamer for Vibrio parahaemolyticus and the
present in a heterogeneous population to millions of copies decent selectivity of the aptamer to this specific bacteria, rel-
with high precision in a short duration. It can detect a large ative to a few others, E. coli, L. monocytogenes, V. fischeri, and
variety of bacteria. Several PCR variants have been devel- S. soneii. However, most of the SPR equipment is still a bulky
oped, such as reverse transcriptase PCR, real-time PCR laboratory-based setup that is not suitable for onsite applica-
(RT-PCR), real-time revers transcription PCR (qRT-PCR), tions [100–102]. In the context of using aptamer as sensing
and multiplex PCR and nested PCR to identify bacteria and probe, Zhao et al. developed a potentiometric aptasensor
viruses with high accuracy. Among all variants, RT-PCR is for Vibrio alginolyticus, involving DNA nanostructure-
most powerful. It allows simultaneous monitoring of the modified magnetic beads [98]. In this design (Figure 7), the
formed product as amplification proceeds. Fluorescent dyes bacteria cells compete the DNA aptamer with the DNA-
or fluorescent probes are utilized to visualize the amplified coated magnetic bead, causing disassembly of the DNA
product. For example, Norovirus detection by qRT-PCR nanostructures on the bead. The change in the charge or
[88] shows improved specificity and sensitivity. It covers all DNA concentration on the magnetic beads is chronopo-
known human NoV genotypes with the use of only four for- tentiometrially detected by a solid-contact polycation-
ward primers, two probes, and one reverse primer. sensitive electrode using protamine as an indicator based
The power of PCR for simultaneous identification of on the electrostatic interaction between DNA and prot-
multiple marine fish pathogens (i.e., the multiplex PCR) has amine. With this method, Vibrio alginolyticus can be
been reported. For example, multiplex- (m-) PCR-based pro- detected within a linear range of 10–100 cfu/ml and with
tocol was designed for the simultaneous detection of the a LOD of 10 cfu/ml. This proposed strategy can be used
main marine bacterial pathogens in Chilean salmon farms: for the detection of other microorganisms by changing
Streptococcus phocae, Aeromonas salmonicida, Vibrio anguil- the aptamers in the DNA nanostructures.
larum, and Piscirickettsia salmonis [89]. Each of the 4 oligo- In general, among the methods discussed in Section 5
nucleotide primer pairs exclusively amplified the target (culture-based method assays, novel molecular methods,
gene of the specific bacterial pathogen. With the adequate and biosensors), the culture-based methods are the “gold
sensitivity for the multiple pathogen (50 pg μl−1 for V. angu- standard” in bacteria detection. Molecular-based techniques
illarum, 500 fg μl−1 for P. salmonis, and 5 pg μl−1 for S. phocae have a higher sensitivity, but they require specific and expen-
and A. salmonicida), this technique is considered as an alter- sive equipment. Fast, reliable, easy to use, sensitive, and spe-
native to culture-based methods for the diagnosis of infec- cific systems are always in great demand for field use to
tions in fish. In general, PCR technique can identify and predict outbreaks and during the outbreak. Biosensors would
detect bacteria directly from a water sample with or without have such potential, subjected to performance improvements.
preenrichment. It is faster (hours) than the culture plate
count (days). 6. Challenges and Future Perspectives
5.4. Biosensors. Biosensors have been developed for detection The increasing demand of aquatic products has stimulated
of pathogens related to livestock and poultry [90] and for the rapid growth of aquaculture industries. However, a
Vibrio cholerae [91]. Table 3 is a summary of various biosen- healthy growth of this industry sector requires strong tech-
sors for detecting fish pathogens. They include quartz crystal nology innovation. Water quality is one of the key determin-
microbalance (QCM), the microcantilever sensor, ampero- ing factors. Freshwater tank high-density farming, seawater
metric sensor, potentiometric sensor, and surface plasmon offshore farming, and pond farming all face unique chal-
resonance (SPR) biosensor, and lateral flow tests, targeting lenges in relation to respective water quality factors, e.g., dif-
either viral RNA [92–94] or the bacteria cells [95–99]. Bio- ferent types of contaminants and the source of contaminants.
sensors are typically designed to detect known bacteria or Among all four water quality aspects, sensing and detec-
virus, but less so for identification of unknown. In a typical tion technologies for basic physical parameters have all been
biosensor design, DNA probe or antibodies are immobilized very well developed and available in the market. A quick
10 Research

(a) (b) (c)

Figure 6: Plate culture for isolation and detection of (a) Vibrio using CHROMagar™ Vibrio, (b) Vibrio using HiCrome Vibrio Agar (Merck),
and (c) Pseudomonas spp. using CHROMagar™ Pseudomonas.

Table 3: Molecular sensors and bacteria cell sensors for fish pathogens.

Biosensor format Pathogen Analyte Performance Ref.

Viral haemorrhagic
Quartz crystal microbalance (QCM) Viral RNA LOD 1.6 nM [92]
septicaemia (VHS)
Electrochemistry with gold nanoparticle LOD 0.5 fM of linear target DNA
Aphanomyces invadans Viral RNA [93]
for signal amplification 1 fM for PCR product
Lateral flow with gold nanoparticle Nervous necrosis virus (NNV) Viral RNA LOD 270 pg of PCR product [94]
Dynamic range 1 × 103 -1 × 107 CFU/ml
Microcantilever Vibrio cholerae O1 Cells [95]
LOD ∼ 1 × 103 CFU/ml
LOD 107 cfu/ml
Lateral flow with AuNPs Vibrio cholerae O1 and O139 Cells 3 [96]
LOD 10 cfu/ml after 6 h culture enrichment
LOD 8 cfu/ml in seawater
Amperometric immunosensors Vibrio cholerae O1 Cells [97]
80 cfu/ml in sewer water and tap water
Potentiometric aptasensing involving Dynamic range:10–100 cfu/ml
Vibrio alginolyticus Cells [98]
magnetic beads LOD 10 cfu/ml
Surface plasmon resonance spectroscopy Vibrio parahaemolyticus Cells Not specified [99]

Google search of aquaculture water quality monitoring will one example, researchers have developed an immunomag-
lead to the appearance of many companies that provide com- netic separation-coupled laboratory PCR or flow cytometry
mercial solutions, including YSI (a xylem brand, USA), Nile- to reach ultrahigh sensitivity for a fish pathogen [103, 104].
bot (Maadi, Cairo, Egypt), and Aquasend (USA), among How to couple such or similar sample enrichment techniques
many others. Most of the products claim features of Multipa- with portable sensors for onsite application with sufficient
rameter, Online Analyzer, Intelligent Water Quality Control- sensitivity for early warming would be an area that requires
ler, etc. The actual products indeed are capable of online further R&D. For pathogen detection, PCR method domi-
monitoring of several different parameters by selecting corre- nates the practice, but it requires sample preparation to
sponding sensors upon different requests, which usually extract DNA or RNA, and its performance heavily depends
includes temperature, pH, conductivity, DO, turbidity, on reagent quality (DNA primers and enzymes etc.) and
sludge concentration, chlorophyll, and blue-green algae. PCR apparatus.
These systems have all aimed to address the demand of mul- To bridge the gaps of multiple parameter systems cover-
timode sensors, creating data connectivity, and feedback loop ing a wider range of parameters (physical, chemical, bio-
to trigger actions. A reliable sensor network should come chemical, and biological), firstly the chemical, biochemical,
with optimal sampling strategies, to ensure the reliability of and biological sensors must reach to the desired level of por-
the data. tability, and then advanced WSN technology would be
In comparison to low-cost portable physical sensors, sen- needed to cover not only the physical parameters but also
sors for organic chemical, biochemical, and biological haz- those hard-to-measure chemical contaminants and biohaz-
ards are still behind in their feasibility for onsite application ards. For all parameters, the sensor technologies must be
without scarifying performance as laboratory methods. In highly robust because the aquatic environment varies from
Research 11

Solid Membrane Aqueous

GCE

n
n contact phase phase
A

n
n
n
Control n

Potential
n

n
B
B A

Target Time

Magnetic bead Aptamer

Current direction
H2 DNA V. alginolyticus

Capture DNA H1 DNA MWCNTs DNNS-TDDA

Protamine IL Protamine

(a) (b)

Figure 7: A potentiometric aptasensing of V. alginolyticus based on DNA nanostructure-modified magnetic beads. (a) Schematic illustration
of the principle. (b) Schematic diagram of the polycation-sensitive electrode based on a MWCNT-IL composite as a solid contact for the
chronopotentiometric detection of protamine [98] (Open Access).

time to time. The uncontrollable environmental factor may application of various technologies. Thus, not one technol-
pose challenges to the reliability of the sensing technologies. ogy can master it all, and knowledge of the advantages and
Also, the sensors and sensor physics must be compatible disadvantages of different technologies available is vital.
with seawater for their background and unstable chemical Some future outlooks include miniaturization of an auto-
composition. Future prospects of the monitoring system mated system that can simultaneously measure various con-
include improving current aquaculture systems and equip- taminants and parameters.
ment. This can be carried out by miniaturization of the
highly robust and accurate autonomous system, but at the Conflicts of Interest
same time allowing contaminants and physical parameters
to be measured simultaneously at low costs. Such a technol- The authors declare that they have no conflicts of interest.
ogy would probably require the combination of nanotech-
nology, microelectronics, and microfluidics to achieve this Acknowledgments
cost-effective production. With this advance system, aqua-
culture industries can get a more comprehensive data/para- This work is supported by the A∗ STAR RIE2020 Advanced
meters to better control and create optimum condition that Manufacturing and Engineering (AME) Programmatic
maximize fish production even at remote areas. The system Grant (Grant No. A18A8b0059). SuX acknowledges NUS
shall have a versatile platform, enabling it to adapt to vari- students (Chan Jun Jie, Kang Ying Yi, Lerh Keng Yeow,
ous industries besides aquaculture, which includes tourism Niklas Ladegaard Andersen, Ng Yan Jie) for collecting partial
and agricultures. In addition, the advance system also will of the references under their project work.
better support sustainable practices such as the prevention
of overfishing and using fewer antibiotics to prevent antimi- References
crobial resistance.
[1] T. Hastein, B. Hjeltnes, A. Lillehaug, J. Utne Skare,
M. Berntssen, and A. K. Lundebye, “Food safety hazards that
7. Conclusion occur during the production stage: challenges for fish farming
In summary, aquaculture is the fastest growing sector of agri- and the fishing industry,” Revue Scientifique et Technique de
l'OIE, vol. 25, no. 2, pp. 607–625, 2006.
culture in the world, due to the world’s growing population
[2] FAO, The State of World Fisheries and Aquaculture 2018:
and demand for seafood. We have discussed the various ana-
Meeting the Sustainable Development Goals, Food and Agri-
lytical technologies and biosensors used within aquaculture culture Organization of the United Nations, Rome, 2018,
to detect and monitor various water parameters (physical, Licence: CC BY-NC-SA 3.0 IGO.
organic chemical, biochemical, and biological) affecting [3] M. Sinaei and A. Mashinchian, “Polycyclic aromatic hydro-
aquaculture cultivations. Numerous analytical techniques carbons in the coastal sea water, the surface sediment and
are available in literatures and in the markets, with different Mudskipper Boleophthalmus dussumieri from coastal areas
levels of sophistication and cost. However, the monitoring of the Persian Gulf: source investigation, composition pattern
and control of aquacultures is a complex task, which requires and spatial distribution,” Journal of Environmental Health
continuous development of technologies and combinatorial Science and Engineering, vol. 12, no. 1, p. 59, 2014.
12 Research

[4] H. Endo and H. Wu, “Biosensors for the assessment of fish [19] S. B. Chandanapalli, “Design and deployment of aqua moni-
health: a review,” Fisheries Science, vol. 85, no. 4, pp. 641– toring system using wireless sensor networks and IAR-Kick,”
654, 2019. Journal of Aquaculture Research & Development, vol. 5, no. 7,
[5] V. Venugopal, “Biosensors in fish production and quality 2014.
control,” Biosensors and Bioelectronics, vol. 17, no. 3, [20] H. B. Glasgow, J. M. Burkholder, R. E. Reed, A. J. Lewitus, and
pp. 147–157, 2002. J. E. Kleinman, “Real-time remote monitoring of water qual-
[6] A. Adams and K. D. Thompson, “Development of diagnostics ity: a review of current applications, and advancements in
for aquaculture: challenges and opportunities,” Aquaculture sensor, telemetry, and computing technologies,” Journal of
Research, vol. 42, pp. 93–102, 2011. Experimental Marine Biology and Ecology, vol. 300, no. 1-2,
pp. 409–448, 2004.
[7] K. Srogi, “Monitoring of environmental exposure to polycy-
clic aromatic hydrocarbons: a review,” Environmental Chem- [21] G. Cario, A. Casavola, P. Gjanci, M. Lupia, C. Petrioli, and
istry Letters, vol. 5, no. 4, pp. 169–195, 2007. D. Spaccini, “Long lasting underwater wireless sensors net-
work for water quality monitoring in fish farms,” in OCEANS
[8] V. Vogiazi, A. de la Cruz, S. Mishra, V. Shanov, W. R.
2017 - Aberdeen, Aberdeen, UK, June 2017.
Heineman, and D. D. Dionysiou, “A comprehensive review:
development of electrochemical biosensors for detection of [22] F. J. Espinosa-Faller and G. E. Rendón-Rodríguez, “A ZigBee
cyanotoxins in freshwater,” ACS Sensors, vol. 4, no. 5, wireless sensor network for monitoring an aquaculture recir-
pp. 1151–1173, 2019. culating system,” Journal of Applied Research and Technol-
ogy, vol. 10, no. 3, pp. 380–387, 2012.
[9] W. J. S. Mwegoha, M. E. Kaseva, and S. M. M. Sabai, “Math-
[23] M. Zhang, D. Li, L. Wang, D. Ma, and Q. Ding, “Design and
ematical modeling of dissolved oxygen in fish ponds,” African
development of water quality monitoring system based on
Journal of Environmental Science and Technology, vol. 4,
wireless sensor network in aquaculture,” in Computer and
no. 9, pp. 623–638, 2010.
Computing Technologies in Agriculture IV, D. Li, Y. Liu, and
[10] A. D. M. Africa, J. C. C. A. Aguilar, C. M. S. Lim, P. A. Y. Chen, Eds., pp. 629–641, Springer, 2011.
Pacheco, and S. E. C. Rodrin, “Automated aquaculture sys-
[24] J. Huang, W. Wang, S. Jiang, D. Sun, G. Ou, and K. Lu,
tem that regulates pH, temperature and ammonia,” in 2017
“Development and test of aquacultural water quality moni-
IEEE 9th International Conference on Humanoid, Nanotech-
toring system based on wireless sensor network,” Transac-
nology, Information Technology, Communication and Con-
tions from the Chinese Society of Agricultural Engineering,
trol, Environment and Management (HNICEM), pp. 1–6,
vol. 29, no. 4, pp. 183–190, 2013.
Manila, Philippines, December 2017.
[25] H. Luo, G. Li, W. Peng, J. Song, and Q. Bai, “Real-time remote
[11] S. Chumkiew, M. Jaroensutasinee, F. W. Tina, monitoring system for aquaculture water quality,” Interna-
K. Jaroensutasinee, and P. Koad, “Physical factors affecting
tional Journal of Agriculture & Biological Engineering,
oyster diversity and distribution in Southern Thailand,” Jour-
vol. 8, no. 6, pp. 136–143, 2015.
nal of Environmental Biology, vol. 40, no. 1, pp. 03–08, 2019.
[26] L. Parra, S. Sendra, L. García, and J. Lloret, “Design and
[12] Y. Shifeng, K. E. Jing, and Z. Jimin, “Wireless monitoring deployment of low-cost sensors for monitoring the water
system for aquiculture environment,” in 2007 IEEE Interna- quality and fish behavior in aquaculture tanks during the
tional Workshop on Radio-Frequency Integration Technology, feeding process,” Sensors, vol. 18, no. 3, p. 750, 2018.
pp. 274–277, Rasa Sentosa Resort, Singapore, December 2007.
[27] C. I. L. Justino, K. R. Duarte, A. C. Freitas, T. S. L. Panteleitchouk,
[13] S. Yang and Y. Li, “Dissolved oxygen remote monitoring A. C. Duarte, and T. A. P. Rocha-Santos, “Contaminants in
system based on the internet,” Electronic Measurement aquaculture: overview of analytical techniques for their deter-
Technology, vol. 7, pp. 88–90, 2011. mination,” TrAC Trends in Analytical Chemistry, vol. 80,
[14] S. Kameoka, S. Isoda, A. Hashimoto et al., “A wireless sensor pp. 293–310, 2016.
network for growth environment measurement and multi- [28] T. Hastein, B. Hjeltnes, A. Lillehaug, J. Utne Skare,
band optical sensing to diagnose tree vigor,” Sensors, M. Berntssen, and A. K. Lundebye, “Peligros para la inocui-
vol. 17, no. 5, p. 966, 2017. dad de los alimentos que surgen durante la fase de produc-
[15] J. Wang, X. Niu, L. Zheng, C. Zheng, and Y. Wang, “Wireless ción: problemas de la piscicultura y la industria piscícola,”
mid-infrared spectroscopy sensor network for automatic Revue Scientifique et Technique, vol. 25, no. 2, pp. 607–625,
carbon dioxide fertilization in a greenhouse environment,” 2006.
Sensors, vol. 16, no. 11, pp. 1941–1960, 2016. [29] M. Muñoz-de-Toro, H. R. Beldoménico, S. R. García et al.,
[16] S. Sawant, S. S. Durbha, and A. Jagarlapudi, “Interoperable “Organochlorine levels in adipose tissue of women from a lit-
agro-meteorological observation and analysis platform for toral region of Argentina,” Environmental Research, vol. 102,
precision agriculture: a case study in citrus crop water no. 1, pp. 107–112, 2006.
requirement estimation,” Computers and Electronics in Agri- [30] H. Tsukino, T. Hanaoka, H. Sasaki et al., “Fish intake and
culture, vol. 138, pp. 175–187, 2017. serum levels of organochlorines among Japanese women,”
[17] E. S. Nadimi, V. Blanes-Vidal, R. N. Jørgensen, and Science of the Total Environment, vol. 359, no. 1-3, pp. 90–
S. Christensen, “Energy generation for an ad hoc wireless sen- 100, 2006.
sor network-based monitoring system using animal head [31] M. Someya, M. Ohtake, T. Kunisue et al., “Persistent organic
movement,” Computers and Electronics in Agriculture, pollutants in breast milk of mothers residing around an open
vol. 75, no. 2, pp. 238–242, 2011. dumping site in Kolkata, India: specific dioxin-like PCB levels
[18] R. Handcock, D. Swain, G. Bishop-Hurley et al., “Monitoring and fish as a potential source,” Environment International,
animal behaviour and environmental interactions using vol. 36, no. 1, pp. 27–35, 2010.
wireless sensor networks, GPS collars and satellite remote [32] J. Nácher‐Mestre, R. Serrano, T. Portolés‐Nicolau,
sensing,” Sensors, vol. 9, no. 5, pp. 3586–3603, 2009. F. Hernández, L. Benedito‐Palos, and J. Pérez‐Sánchez, “A
Research 13

reliable analytical approach based on gas chromatography Industrial Microbiology Biotechnology, vol. 30, no. 7,
coupled to triple quadrupole and time&hyphen;of&hyphen;- pp. 383–406, 2003.
flight mass analyzers for the determination and confirmation [46] B. Greer, R. Maul, K. Campbell, and C. T. Elliott, “Detection
of polycyclic aromatic hydrocarbons in complex matrices of freshwater cyanotoxins and measurement of masked
from aquaculture activities,” Rapid Communications in Mass microcystins in tilapia from Southeast Asian aquaculture
Spectrometry, vol. 23, no. 13, pp. 2075–2086, 2009. farms,” Analytical and Bioanalytical Chemistry, vol. 409,
[33] M. Perugini, M. Manera, T. Tavoloni, C. Lestingi, I. Pecorelli, no. 16, pp. 4057–4069, 2017.
and A. Piersanti, “Temporal trends of PCBs in feed and die- [47] R. P. Rastogi, R. P. Sinha, and A. Incharoensakdi, “The
tary influence in farmed rainbow trout (Oncorhynchus cyanotoxin-microcystins: current overview,” Reviews in Envi-
mykiss),” Food Chemistry, vol. 141, no. 3, pp. 2321–2327, ronmental Science and Biotechnology, vol. 13, no. 2, pp. 215–
2013. 249, 2014.
[34] A. Freitas, S. Leston, J. Rosa et al., “Multi-residue and [48] T. Boopathi and J.-S. Ki, “Impact of environmental factors on
multi-class determination of antibiotics in gilthead sea the regulation of cyanotoxin production,” Toxins, vol. 6,
bream (Sparus aurata) by ultra high-performance liquid no. 7, pp. 1951–1978, 2014.
chromatography-tandem mass spectrometry,” Food Addi- [49] M. M. Gehringer and N. Wannicke, “Climate change and reg-
tives & Contaminants: Part A, vol. 31, no. 5, pp. 817– ulation of hepatotoxin production in Cyanobacteria,” FEMS
826, 2014. Microbiology Ecology, vol. 88, no. 1, pp. 1–25, 2014.
[35] A. M. Wilkowska and M. Biziuk, “Rapid method for the [50] J. Chen, P. Xie, L. Li, and J. Xu, “First identification of the
determination of organochlorine pesticides and PCBs in fish hepatotoxic microcystins in the serum of a chronically
muscle samples by microwave-assisted extraction and analy- exposed human population together with indication of hepa-
sis of extracts by GC-ECD,” The Journal of AOAC Interna- tocellular damage,” Toxicological Sciences, vol. 108, no. 1,
tional, vol. 93, no. 6, pp. 1987–1994, 2010. pp. 81–89, 2009.
[36] B. Albero, J. L. Tadeo, and R. A. Pérez, “Ultrasound-assisted [51] L. A. Lawton and C. Edwards, “Chapter 23, conventional
extraction of organic contaminants,” TrAC Trends in Analyt- laboratory methods for cyanotoxins,” Cyanobacterial Harm-
ical Chemistry, vol. 118, pp. 739–750, 2019. ful Algal Blooms: State of the Science and Research Needs,
[37] S. Moret, C. Conchione, A. Srbinovska, and P. Lucci, “Micro- vol. 3, pp. 513–537, 2008.
wave-based technique for fast and reliable extraction of [52] A. Chia, S. J. Oniye, Z. Ladan, Z. Lado, A. P. V. O. Inekwe,
organic contaminants from food, with a special focus on and J. Mmerole, “A survey for the presence of microcystins
hydrocarbon contaminants,” Food, vol. 8, no. 10, p. 503, in aquaculture ponds in Zaria, Northern-Nigeria: possible
2019. public health implication,” African Journal of Biotechnology,
[38] R. Bhuvaneshwari and R. B. Rajendran, “GCMS determina- vol. 8, no. 22, pp. 6282–6289, 2009.
tion of organochlorine pesticides (OCPs) in fish from River [53] L. Peng, Y. Liu, W. Chen, L. Liu, M. Kent, and L. Song,
Cauvery and Veeranam Lake,” Journal of Chemistry, vol. 9, “Health risks associated with consumption of microcystin-
Article ID 395707, 8 pages, 2012. contaminated fish and shellfish in three Chinese lakes: sig-
[39] P. Kestemont, S. Jourdan, M. Houbart et al., “Size heterogene- nificance for freshwater aquacultures,” Ecotoxicology and
ity, cannibalism and competition in cultured predatory fish Environmental Safety, vol. 73, no. 7, pp. 1804–1811, 2010.
larvae: biotic and abiotic influences,” Aquaculture, vol. 227, [54] V. Gaget, M. Lau, B. Sendall, S. Froscio, and A. R. Humpage,
no. 1-4, pp. 333–356, 2003. “Cyanotoxins: which detection technique for an optimum
[40] X. Yan, X. Shi, J. Yang, X. Zhang, W. Jia, and J. Ma, “Towards risk assessment?,” Water Research, vol. 118, pp. 227–238,
field detection of polycyclic aromatic hydrocarbons (PAHs) 2017.
in environment water using a self-assembled SERS sensor,” [55] C. Moore, J. Juan, Y. Lin, C. Gaskill, and B. Puschner, “Com-
in AOPC 2017: Optical Spectroscopy and Imaging,, Beijing, parison of protein phosphatase inhibition assay with LC-
China, 2017. MS/MS for diagnosis of microcystin toxicosis in veterinary
[41] O. Péron, E. Rinnert, M. Lehaitre, P. Crassous, and Cases,” Marine Drugs, vol. 14, no. 3, p. 54, 2016.
C. Compère, “Detection of polycyclic aromatic hydrocarbon [56] S. M. Froscio, A. R. Humpage, P. C. Burcham, and I. R. Fal-
(PAH) compounds in artificial sea-water using surface- coner, “Cell-free protein synthesis inhibition assay for the
enhanced Raman scattering (SERS),” Talanta, vol. 79, no. 2, cyanobacterial toxin cylindrospermopsin,” Environmental
pp. 199–204, 2009. Toxicology, vol. 16, no. 5, pp. 408–412, 2001.
[42] X. Gu, S. Tian, Q. Zhou et al., “SERS detection of polycyclic [57] N. Li, X. Su, and Y. Lu, “Nanomaterial-based biosen-
aromatic hydrocarbons on a bowl-shaped silver cavity sub- sors using dual transducing elements for solution phase
strate,” RSC Advances, vol. 3, no. 48, pp. 25989–25996, 2013. detection,” The Analyst, vol. 140, no. 9, pp. 2916–2943,
[43] Y. Xie, X. Wang, X. Han et al., “Selective SERS detection of 2015.
each polycyclic aromatic hydrocarbon (PAH) in a mixture [58] L. Sutarlie, S. Y. Ow, and X. Su, “Nanomaterials-based bio-
of five kinds of PAHs,” Journal of Raman Spectroscopy, sensors for detection of microorganisms and microbial
vol. 42, no. 5, pp. 945–950, 2011. toxins,” Biotechnology Journal, vol. 12, no. 4, 2017.
[44] E. Hahm, D. Jeong, M. G. Cha et al., “β-CD Dimer- [59] A. P. Das, P. S. Kumar, and S. Swain, “Recent advances in bio-
immobilized Ag Assembly Embedded Silica Nanoparticles sensor based endotoxin detection,” Biosensors and Bioelec-
for Sensitive Detection of Polycyclic Aromatic Hydrocar- tronics, vol. 51, pp. 62–75, 2014.
bons,” Scientific Reports, vol. 6, no. 1, article 26082, 2016. [60] M. Weller, “Immunoassays and biosensors for the detection
[45] K. G. Sellner, G. J. Doucette, and G. J. Kirkpatrick, “Harmful of Cyanobacterial toxins in water,” Sensors, vol. 13, no. 11,
algal blooms: causes, impacts, and detection,” Journal of pp. 15085–15112, 2013.
14 Research

[61] I. Cunha, R. Biltes, M. G. F. Sales, and V. Vasconcelos, plexes, and its Application for colorimetric DNA detection,”
“Aptamer-based biosensors to detect aquatic phycotoxins ACS Nano, vol. 3, no. 9, pp. 2751–2759, 2009.
and cyanotoxins,” Sensors, vol. 18, no. 7, article 2367, 2018. [77] R. Kanjanawarut and X. Su, “Colorimetric detection of
[62] L. Lebogang, M. Hedström, and B. Mattiasson, “Develop- DNA using unmodified metallic nanoparticles and peptide
ment of a real-time capacitive biosensor for cyclic cyanotoxic nucleic acid probes,” Analytical Chemistry, vol. 81, no. 15,
peptides based on Adda-specific antibodies,” Analytica Chi- pp. 6122–6129, 2009.
mica Acta, vol. 826, pp. 69–76, 2014. [78] P. Xie, L. Zhu, X. Shao, K. Huang, J. Tian, and W. Xu, “Highly
[63] H.-W. Yu, J. Lee, S. Kim, G. H. Nguyen, and I. S. Kim, “Elec- sensitive detection of lipopolysaccharides using an aptasensor
trochemical immunoassay using quantum dot/antibody based on hybridization chain reaction,” Scientific Reports,
probe for identification of cyanobacterial hepatotoxin micro- vol. 6, no. 1, article 29524, 2016.
cystin-LR,” Analytical and Bioanalytical Chemistry, vol. 394, [79] S. R. Bickman, K. Campbell, C. Elliott et al., “An innovative
no. 8, pp. 2173–2181, 2009. portable biosensor system for the rapid detection of freshwa-
[64] H. Manoharan, P. Kalita, S. Gupta, and V. V. R. Sai, “Plas- ter Cyanobacterial algal bloom toxins,” Environmental Sci-
monic biosensors for bacterial endotoxin detection on biomi- ence and Technology, vol. 52, no. 20, pp. 11691–11698, 2018.
metic C-18 supported fiber optic probes,” Biosensors and [80] M. Crane and A. Hyatt, “Viruses of fish: an overview of signif-
Bioelectronics, vol. 129, pp. 79–86, 2019. icant pathogens,” Viruses, vol. 3, no. 11, pp. 2025–2046, 2011.
[65] T. Liu, L. Gao, J. Zhao, Y. Cao, Y. Tang, and P. Miao, “A poly- [81] S. Chatterjee and S. Haldar, “Vibrio related diseases in aqua-
myxin B–silver nanoparticle colloidal system and the applica- culture and development of rapid and accurate identification
tion of lipopolysaccharide analysis,” Analyst, vol. 143, no. 5, methods,” Journal of Marine Science: Research & Develop-
pp. 1053–1058, 2018. ment, vol. S1, 2012.
[66] C. Lei, Z. Qiao, Y. Fu, and Y. Li, “Colorimetric detection of [82] P. S. Sudheesh, A. Al-Ghabshi, N. Al-Mazrooei, and
lipopolysaccharides based on a lipopolysaccharide-binding S. Al-Habsi, “Comparative pathogenomics of bacteria caus-
peptide and AuNPs,” Analytical Methods, vol. 8, no. 45, ing infectious diseases in fish,” International Journal of Evolu-
pp. 8079–8083, 2016. tionary Biology, vol. 2012, Article ID 457264, 16 pages, 2012.
[67] A. Zuzuarregui, D. Souto, E. Pérez-Lorenzo et al., “Novel [83] O. L. M. Haenen, J. J. Evans, and F. Berthe, “Bacterial infec-
integrated and portable endotoxin detection system based tions from aquatic species: potential for and prevention of
on an electrochemical biosensor,” Analyst, vol. 140, no. 2, contact zoonoses,” Revue Scientifique et Technique, vol. 32,
pp. 654–660, 2015. no. 2, pp. 497–507, 2013.
[68] N. Wang, H. Dai, L. Sai, H. Ma, and M. Lin, “Copper ion- [84] M. Y. Ina-Salwany, N. Al-saari, A. Mohamad et al., “Vibriosis
assisted gold nanoparticle aggregates for electrochemical sig- in fish: a review on disease development and prevention,”
nal amplification of lipopolysaccharide sensing,” Biosensors Journal of Aquatic Animal Health, vol. 31, no. 1, pp. 3–22,
and Bioelectronics, vol. 126, pp. 529–534, 2019. 2019.
[69] G. Ahn, S. S. Sekhon, Y. E. Jeon et al., “Detection of endo- [85] C. Baker-Austin, J. D. Oliver, M. Alam et al., “Vibrio spp.
toxins using nanomaterials,” Toxicology and Environmental infections,” Nature Reviews Disease Primers, vol. 4, no. 1,
Health Sciences, vol. 9, no. 5, pp. 259–268, 2017. 2018.
[70] R. M. Pallares, N. T. K. Thanh, and X. Su, “Quantifying the [86] C. E. Starliper, “General and specialized media routinely
binding between proteins and open chromatin-like DNA employed for primary isolation of BACTERIAL pathogens
sequences with gold nanorods,” Chemical Communications, of fishes,” Journal of Wildlife Diseases, vol. 44, no. 1,
vol. 55, no. 100, pp. 15041–15044, 2019. pp. 121–132, 2008.
[71] R. M. Pallares, N. T. K. Thanh, and X. Su, “sensing of circu- [87] A. Huq, B. J. Haley, E. Taviani, A. Chen, N. A. Hasan, and
lating cancer biomarkers with metal nanoparticles,” Nano- R. R. Colwell, “Detection, isolation, and identification of Vib-
scale, vol. 11, no. 46, pp. 22152–22171, 2019. rio choleraefrom the environment,” Current Protocols in
[72] R. M. Pallares, M. Bosman, N. T. K. Thanh, and X. Su, “A Microbiology, vol. 26, no. 1, 2012.
Plasmonic multi-logic gate platform based on sequence- [88] I. Glowacka, G. Harste, J. Witthuhn, and A. Heim, “An
specific binding of estrogen receptors and gold nanorods,” improved one-step real-time reverse transcription-PCR assay
Nanoscale, vol. 8, no. 48, pp. 19973–19977, 2016. for detection of norovirus,” Journal of Clinical Microbiology,
[73] R. M. Pallares, S. L. Kong, T. Hui Ru, N. T. K. Thanh, vol. 54, no. 2, pp. 497–499, 2016.
Y. Lu, and X. Su, “A plasmonic nanosensor with inverse [89] D. Tapia-Cammas, A. Yañez, G. Arancibia, A. E. Toranzo,
sensitivity for circulating cell-free DNA quantification,” and R. Avendaño-Herrera, “Multiplex PCR for the detection
Chemical Communications, vol. 51, no. 77, pp. 14524– of Piscirickettsia salmonis, Vibrio anguillarum, Aeromonas
14527, 2015. salmonicida and Streptococcus phocae in Chilean marine
[74] R. M. Pallares, N. T. K. Thanh, and X. Su, “Tunable farms,” Diseases of Aquatic Organisms, vol. 97, no. 2,
plasmonic colorimetric assay with inverse sensitivity for pp. 135–142, 2011.
extracellular DNA quantification,” Chemical Communica- [90] J. Vidic, M. Manzano, C.-M. Chang, and N. Jaffrezic-Renault,
tions, vol. 54, no. 80, pp. 11260–11263, 2018. “Advanced biosensors for detection of pathogens related to
[75] S. Y. New, K. M. M. Aung, G. L. Lim et al., “Fast screening of livestock and poultry,” Veterinary Research, vol. 48, no. 1,
ligand-protein interactions based on ligand-induced protein p. 11, 2017.
stabilization of gold nanoparticles,” Analytical Chemistry, [91] F. Cecchini, L. Fajs, S. Cosnier, and R. S. Marks, “Vibrio
vol. 86, no. 5, pp. 2361–2370, 2014. cholerae detection: Traditional assays, novel diagnostic tech-
[76] X. Su and R. Kanjanawarut, “Control of metal nanoparticles niques and biosensors,” TrAC Trends in Analytical Chemis-
aggregation and dispersion by PNA and PNA-DNA com- try, vol. 79, pp. 199–209, 2016.
Research 15

[92] S.-R. Hong, H.-D. Jeong, and S. Hong, “QCM DNA biosensor
for the diagnosis of a fish pathogenic virus VHSV,” Talanta,
vol. 82, no. 3, pp. 899–903, 2010.
[93] G. C. Kuan, L. P. Sheng, P. Rijiravanich et al., “Gold-nanopar-
ticle based electrochemical DNA sensor for the detection of
fish pathogen Aphanomyces invadans,” Talanta, vol. 117,
no. 3, pp. 312–317, 2013.
[94] D. K. Toubanaki, M. Margaroni, and E. Karagouni, “Dual
enhancement with a nanoparticle-based lateral flow biosen-
sor for the determination of DNA,” Analytical Letters,
vol. 49, no. 7, pp. 1040–1055, 2015.
[95] U. Sungkanak, A. Sappat, A. Wisitsoraat, C. Promptmas, and
A. Tuantranont, “Ultrasensitive detection of Vibrio cholerae
O1 using microcantilever-based biosensor with dynamic
force microscopy,” Biosensors and Bioelectronics, vol. 26,
no. 2, pp. 784–789, 2010.
[96] C. Y. Yu, G. Y. Ang, A. L. Chua et al., “Dry-reagent gold
nanoparticle-based lateral flow biosensor for the simulta-
neous detection of Vibrio cholerae serogroups O1 and
O139,” Journal of Microbiological Methods, vol. 86, no. 3,
pp. 277–282, 2011.
[97] M. K. Sharma, A. K. Goel, L. Singh, and V. K. Rao, “Immuno-
logical biosensor for detection of Vibrio cholerae O1in envi-
ronmental water samples,” World Journal of Microbiology
and Biotechnology, vol. 22, no. 11, pp. 1155–1159, 2006.
[98] G. Zhao, J. Ding, H. Yu, T. Yin, and W. Qin, “Potentiometric
aptasensing of vibrio alginolyticus based on DNA
nanostructure-modified magnetic beads,” Sensors, vol. 16,
no. 12, article 2052, 2016.
[99] J. Y. Ahn, K. A. Lee, M. J. Lee et al., “Surface plasmon
resonance aptamer biosensor for discriminating pathogenic
BacteriaVibrio parahaemolyticus,” Journal of Nanoscience
and Nanotechnology, vol. 18, no. 3, pp. 1599–1605, 2018.
[100] H. Y. Song, X. Zhou, J. Hobley, and X. Su, “Comparative
study of random and oriented antibody immobilization as
measured by dual polarization interferometry and surface
plasmon resonance spectroscopy,” Langmuir, vol. 28, no. 1,
pp. 997–1004, 2011.
[101] W. Y. X. Peh, E. Reimhult, H. F. Teh, J. S. Thomsen, and
X. Su, “Understanding Ligand Binding Effects on the Confor-
mation of Estrogen Receptor α-DNA Complexes: A Combi-
national Quartz Crystal Microbalance with Dissipation and
Surface Plasmon Resonance Study,” Biophysical Journal,
vol. 92, no. 12, pp. 4415–4423, 2007.
[102] S. J. Neo, X. Su, and J. S. Thomsen, “Surface Plasmon Resonance
study of Cooperative interactions of estrogen receptor α and
Transcriptional factor Sp1 with composite DNA Elements,”
Analytical Chemistry, vol. 81, no. 9, pp. 3344–3349, 2009.
[103] U. Ryumae, K. Hibi, Y. Yoshiura, H. Ren, and H. Endo,
“Rapid and highly sensitive detection of Flavobacterium psy-
chrophilum using high gradient immunomagnetic separation
with flow cytometry,” Aquaculture, vol. 309, no. 1-4, pp. 125–
130, 2010.
[104] U. Ryumae, K. Hibi, Y. Yoshiura, H. Ren, and H. Endo, “Ultra
highly sensitive method for detecting Flavobacterium psy-
chrophilum using high-gradient immunomagnetic separation
with a polymerase chain reaction,” Aquaculture Research,
vol. 43, no. 6, pp. 929–939, 2012.