Characterization of bigeye tuna habitat in the Southern Waters off Java–Bali using remote sensing data

Abstract

Bigeye tuna (Thunnus obesus) habitat was investigated based on catch data and environmental satellite data, such as sea surface temperature (SST), sea surface chlorophyll (SSC), and sea surface height deviation (SSHD) data in the Southern Waters off Java and Bali. First, we obtained daily fish catch data and monthly satellite data for SST, SSC, and SSHD for 2006–2010. Then, we analyzed the relationship between daily catch data and satellite data by combining the statistical method of generalized additive model (GAM) and geographic information system (GIS). Seven GAM models were generated with the number of bigeye tuna as a response variable, and SST, SSC, and SSHD as predictor variables. All of the predictors of SST, SSC, and SSHD were highly significant (P < 0.001) to the number of bigeye tuna. Values of SST, SSHD, and SSC in bigeye tuna habitat ranged from 24.8 to 28.7 °C, −3 to 7 cm, and 0.05 to 0.17 mg/m3, respectively. Validation of the predicted number of bigeye tuna with the observed value was significant (P < 0.05, r2 = 0.56). SST was the most important environmental variable to the number of bigeye tuna caught, followed by SSHD and SSC.

Full text

Characterization of bigeye tuna habitat in the Southern Waters
off Java–Bali using remote sensing data
Martiwi Diah Setiawati
a,b,1
, Abu Bakar Sambah
a,c
, Fusanori Miura
a
, Tasuku Tanaka
b,1
,
Abd. Rahman As-syakur
b,d,
⇑
a
Graduate School of Science and Engineering, Department of Environmental Science and Engineering, Yamaguchi University, 2-16-1 Tokiwadai, Ube
755-8611, Japan
b
Center for Remote Sensing and Ocean Science (CReSOS), Udayana University, Sudirman Campus, Post Graduate Building (3rd Fl), Jl. P.B.
Sudirman, Denpasar-Bali 80232, Indonesia
c
Marine Science Department, Faculty of Fisheries and Marine Science, Brawijaya University, Jl. Veteran, Malang 65145, Indonesia
d
Marine Science Department, Faculty of Marine and Fisheries, Udayana University, Bukit Jimbaran Campus, Bali 80361, Indonesia
Received 23 August 2013; received in revised form 7 October 2014; accepted 9 October 2014
Abstract
Bigeye tuna (Thunnus obesus) habitat was investigated based on catch data and environmental satellite data, such as sea surface tem-
perature (SST), sea surface chlorophyll (SSC), and sea surface height deviation (SSHD) data in the Southern Waters off Java and Bali.
First, we obtained daily fish catch data and monthly satellite data for SST, SSC, and SSHD for 2006–2010. Then, we analyzed the rela-
tionship between daily catch data and satellite data by combining the statistical method of generalized additive model (GAM) and geo-
graphic information system (GIS). Seven GAM models were generated with the number of bigeye tuna as a response variable, and SST,
SSC, and SSHD as predictor variables. All of the predictors of SST, SSC, and SSHD were highly significant (P < 0.001) to the number of
bigeye tuna. Values of SST, SSHD, and SSC in bigeye tuna habitat ranged from 24.8 to 28.7 °C, 3 to 7 cm, and 0.05 to 0.17 mg/m
3
,
respectively. Validation of the predicted number of bigeye tuna with the observed value was significant (P < 0.05, r
2
= 0.56). SST was the
most important environmental variable to the number of bigeye tuna caught, followed by SSHD and SSC.
Ó 2014 COSPAR. Published by Elsevier Ltd. All rights reserved.
Keywords: Bigeye tuna; SST; SSC; SSHD; GAM; Remote sensing
1. Introduction
According to the Indian Ocean Tuna Commission
(IOTC), Indonesia is the fourth leading tuna-fishing
nations in the Indian Ocean after Spain, Sri Lanka, and
Maldives (Gillet, 2013). The rate of tuna exported from
Indonesia was valued at approximately US$224 million
in 2000 (Simorangkir, 2003) and increased to US$750 mil-
lion in 2012. The waters of the Indian Ocean, between
Indonesia and Australia, are known as important spawning
grounds for commercial tuna and tuna-like species
(Nishikawa et al., 1985). Furthermore, Indonesian fishing
fleets are of major importance to management assessments
of Indian Ocean stocks (Proctor et al., 2003). The Southern
Waters off Java and Bali, part of the Indian Ocean, were
identified as a potential fishing ground for large pelagic fish
http://dx.doi.org/10.1016/j.asr.2014.10.007
0273-1177/Ó 2014 COSPAR. Published by Elsevier Ltd. All rights reserved.
⇑
Corresponding author at: Center for Remote Sensing and Ocean
Science (CReSOS), Udayana University, Sudirman Campus, Post
Graduate Building (3rd Fl), Jl. P.B. Sudirman, Denpasar-Bali 80113
Indonesia. Tel./fax: +62 361256162.
E-mail addresses: s503wf@yamaguchi-u.ac.jp (M.D. Setiawati),
r501wf@yamaguchi-u.ac.jp (A.B. Sambah), miura@yamaguchi-u.ac.jp
(F. Miura), tttanaka@yamaguchi-u.ac.jp (T. Tanaka), ar.assyakur@
pplh.unud.ac.id (A.R. As-syakur).
1
Tel./fax: +62 361256162.
www.elsevier.com/locate/asr
Available online at www.sciencedirect.com
ScienceDirec t
Advances in Space Research xxx (2014) xxx–xxx
Please cite this article in press as: Setiawati, M.D., et al. Characterization of bigeye tuna habitat in the Southern Waters off Java–Bali using
remote sensing data. Adv. Space Res. (2014), http://dx.doi.org/10.1016/j.asr.2014.10.007

(Bailey et al., 1987; Osawa and Julimantoro, 2010). Biolog-
ical and environmental data from the Indian Ocean are
needed to understand the preferred hab itat for sustainable
management of bigeye tuna resources.
Bigeye tuna are highly migratory species, moving both
horizontally and vertically. Physical adaptations of bigeye
tuna allow for tolerance of large temperature changes,
but maintenance of muscle temperature is required (Brill
et al., 2005). Therefore, temperature will affect bigeye tuna
movement in relation to controlling thermoregulation pro-
cesses (Howell et al., 2010). Bigeye tuna are considered to
be opportunistic feeders and visual predators and their for-
age base consists of a mixture of organisms such as fish,
crustaceans, squid, and gelatinous creatures (Sund et al.,
1981; Blackburn et al., 1968; and Bertrand et al., 2002).
Because of this, bigeye tuna prefer to remain in clear waters
to increase the efficiency of visual hunting and to select
appropriate targets. Clear water is nutrient-poor, meaning
there are low chlorophyll-a concentrations in the water
(Sund et al., 1981). The other environmental parameter
that influences bigeye tuna is the current system. Bigeye
tuna migration is influenced by ocean currents; the fish
move along prevailing currents, utilizing them as foraging
habitats (Uda, 1973). Moreover, bigeye tuna also prefer
to stay near, and usually below, thermocline and come to
the surface periodi cally (Brill et al., 2005). The biophysical
environment plays an important role in controlling tuna
distribution and abundance (Zainuddin et al., 2006),
including those of bigeye tuna. The near-real-time data of
biophysical environment by global coverage can be derived
from satellite remote sensing. In recent decades, satellite
remote sensing has become an instrumental ecology for
environmental monitoring (Chassot et al., 2011) and is
used to manage sustainable levels of fisheries (Klemas,
2013).
Satellite remote sensing data provide reliable global
ocean coverage of sea surface temperature (SST), sea surface
height (SSH), surface winds, and sea surface chlorophyll
(SSC), with relatively high spatial and temporal resolution
(Polovina and Howell, 2005). Application of satellite remote
sensing in fisheries is increasing worldwide (e.g. Laurs et al.,
1984; Laurs, 1986; Stretta, 1991; Lehodey et al., 1997;
Santos, 2000; Zag aglia et al., 2004; Zainuddin et al., 2006;
Druon, 2010; Yen et al., 2012; Perez et al., 2013; Kamei
et al., 2014). Oceanographic phenomena are often used to
understand preferred habitat and to estimate the potential
of fishing grounds (Mohri, 1999; Mohri and Nishida,
1999; Lennert-Cody et al., 2008; Song et al., 2009; Osawa
and Julimantoro, 2010). However, there have been relatively
few tuna fisheries ecology studies using satellite remote sens-
ing data in the Southern Waters off Java and Bali (Osawa
and Julimantoro, 2010; Syamsuddin et al., 2013). Osawa
and Julim antoro (2010) reported that environmental vari-
ables had no significance to the abundance of bigeye tuna,
while Syamsuddin et al. (2013) reported that El Nino gave
the significant effect to bigeye tuna abundance in the South-
ern Waters off Java–Bali.
Tuna has preferred biophysical living environment.
Hence, we assumed that fish catch statistics should have
some correlation with ocean environmental variables. In
this study, some environmental variables were used to dis-
tinguish tuna habitat such as SST, SSH, an d SSC. SST has
been used to investigate productive frontal zones
(Hanamoto, 1987; Andrade and Garcia, 1999; Lu et al.,
2001), while SSH can be used to infer oceanic features such
as current dynamics, fronts, eddies, and convergences
(Polovina and Howell, 2005). In addition, SSC can also
be used as a valuable indicator of water mass boundaries
and may identify upwelling, which can influence tuna dis-
tribution in a region(Song et al., 2009; Song and Zhou,
2010). We used satellite oceanographic and fish catch
time-series data to examine characteristic of bigeye tuna
habitat.
Tuna habitat characteristics can be analyzed by using
generalized additive models (GAMs) from catch data and
environmental satellite data (Zagaglia et al., 2004; Song
et al., 2008; Valavanis et al., 2008; Druon et al., 2011).
GAMs can explain the fisheries data and environmental
variables and enhance our understanding of ecological sys-
tems. A GAM is a semi-parametric extension of a general-
ized linear model, which has the smoot h components of the
explanatory variables (Guisan et al., 2002). The additional
value of GAMs is their capacity to express highly nonlinear
and non-monotonic relationships between the response and
explanatory variables (Lizarazo, 2012). Statistical models
and geographic information systems (GIS) have the ability
to improve species habitat studies. Given this background,
this work attempted to investigate the characteristics of
bigeye tuna habitat in the Southern Waters off Java and
Bali by utilizing satellite data and bigeye tuna catch data
during 2006–2010. GAM and GIS data were combined to
understand the characteristics of bigeye tuna habitat.
2. Materials and Methods
2.1. Study Area
The Southern Waters off Java and Bali, part of the
Indian Ocean, are selected as a study area and are located
between 10° S and 18° S latitude and 110° E and 118° E
longitude as shown in Fig. 1. Five dominant waves and
current systems pass the study area: Indonesian Through-
flow (ITF), Indian Ocean South Equatorial Current
(SEC), South Java Current (SJC), Indian Ocean Kelvin
Waves (IOKW), and Rossby Waves (RW). The ITF trans-
fers high heat content from the Pacific Ocean to the Indian
Ocean through a series of straits along the Indonesian
archipelago, and initially accumulates the heat in the
region between northwestern Australia and Indonesia
(Qu and Meyers, 2005; Gordon et al., 2010). The SEC
flows wes tward from the west side of Australia, and is a
large-scale current in this region, wher e mass and property
exchange (Talley an d Sprintall, 2005). In the northern part
of the study area, the SJC and the IOKW flow near the
2 M.D. Setiawati et al. / Advances in Space Research xxx (2014) xxx–xxx
Please cite this article in press as: Setiawati, M.D., et al. Characterization of bigeye tuna habitat in the Southern Waters off Java–Bali using
remote sensing data. Adv. Space Res. (2014), http://dx.doi.org/10.1016/j.asr.2014.10.007

Sumatra–Java c oast (Sprintall et al., 2010). The SJC
changes direction twice each year when the IOKW associ-
ates with the SJC near this coast. The RW propagates west-
ward at 12° S–15° S(Gordon, 2005).
The study area is characterized by a tropical monsoon
climate that results from the Asian–Australian monsoon
wind systems that change direction seasonally. During
July–September, the prevailing southeast monsoon favors
upwelling along the coast of Java and Bali and Sumatra
(Du et al., 2008; Ningsih et al., 2013). These conditions
are reversed during the northwest monsoon from Novem-
ber to April and create warm SSTs (Susanto et al., 2006;
Manessa and As-syakur, 2011). The Southern W aters off
Java and Bali are not only forced by intense annually
reversing monsoonal winds but also influenced by variabil-
ity in throughflow currents (mainly the ITF) (Feng and
Wijffels, 2002).
2.2. Fisheries data and classification
Data sets for bigeye tuna catch from Janu ary 2006 to
December 2010 were used to investigate potential bigeye
tuna habitat in the Southern Waters off Java and Bali. In
this study, in situ bigeye tuna catch data were obtained from
19 longline fishing logbooks provided by PT Perikanan
Nusantara, an incorporated company of the Indonesian
government, at Benoa, Bali, Indonesia. The majority of
the fishing operations were conducted using medium-sized
vessels (100 gross tonnes). Each month, 19–20 vessels were
in operation; the vessels used the same fishing gear (longline
sets) and similar fishing techniques (Syamsuddin et al.,
2013). The data sets consisted of geographic positions (lat-
itude and longitude) of the fishing activities, the operational
days, vessel numbers, and the number of tuna caught per
day during the period. We digitized and compiled all data
into a monthly da tabase. The unit of daily catch data
referred to the number of bigeye tuna caught. Although
most researchers have used catch per unit effort (CPUE)
as an index of fish abundance (Zagaglia et al., 2004;
Zainuddin et al., 2008; Lan et al., 2011; Mugo et al.,
2010), we used number of bigeye tuna caught as a represen-
tation of fish abundance because of limited in-situ data.
Classification of fisheries data is very useful for inferring
the type of fish catch data and for determining the opti-
mum range of oc eanographic parameters and the highest
catch period (Andrade and Garcia, 1999; Zainuddin
et al., 2008). According to Andrade and Garcia (1999),
we divided the fish catch data into three groups: (i) null
catches (0), (ii) positive catches (1–3), and (iii) high catches
(P4). The high catch number of four was determined based
on the lower limit of the uppe r quartile (Q3). The Q3 was
obtained from 7751 observational data.
2.3. Remote sensing data
As an environmental database, monthly SST, SSC, and
SSHD were used in this study. The units of SST, SSC, and
SSHD were °C, mg m
3
and m, respectively. SST and SSC
Level-3 Standard Mapped Images (SMI) with 4-km spatial
resolution from 2006 to 2010 were downloaded from Aqua
MODIS satellite data (http: //oceancolor.gsfc.nasa.gov/). A
correction for SSC data was performed to eliminate noise
that was mainly due to clouds (Maritorena et al., 2010),
eliminating unexpected values of SSC concentration (<0
and> 10 mg/ m
3
)(Abbott and Letelier, 1999). We used the
Environmental Data Connector (EDC) to download
monthly SSHD satellite images from the Archiving, Vali-
dation and Interpretation of Satellite Oceanographic data
Fig. 1. The study area in the Southern Waters off Java–Bali. This area has been passed by five dominant waves and current systems, namely, South Java
Current (SJC), Indonesia Through Flow (ITF), Indian Ocean Kelvin Waves (IOKW), Rossby Waves (RW), and the Indian Ocean South Equatorial
Current (SEC). (Modified from Syamsuddin et al., 2013).
M.D. Setiawati et al. / Advances in Space Research xxx (2014) xxx–xxx 3
Please cite this article in press as: Setiawati, M.D., et al. Characterization of bigeye tuna habitat in the Southern Waters off Java–Bali using
remote sensing data. Adv. Space Res. (2014), http://dx.doi.org/10.1016/j.asr.2014.10.007

(AVISO). The EDC is compatible with ArcGIS software
and can be downloaded free from the National Oceanic
and Atmospheric Administration (NOAA) website
(http://www.pfeg.n oaa.gov/products/edc/). The SSHD
data were global images with 0.25° spatial resolution that
were re-sampled to fit the SST and SSC resolution and sub-
set to the study area. Monthly values of SST, SSC, and
SSHD data were extracted from each pixel corresponding
to the location of fishing activities. The result was a full
matrix of the number of bigeye tuna and the environmental
variables. The full matrix was used in the GAM analysis.
2.4. Generalized additive model
GAM models were used in this study to assess the influ-
ence of environmental variables on potential bigeye tuna
habitat. This statistical method has been commonly used
to predict the habitat and fishing grounds of tuna in the
Pacific and Atlantic oceans (Zagaglia et al., 2004;
Zainuddin et al., 2006, 2008; Mugo et al., 2010), but has
rarely been applied in the current study area. The advan-
tage of this statistical model is that it allows for analysis
of nonparametric relationships and extends the use of addi-
tive models to data sets that have non-Gaussian distribu-
tions, such as binomial, Poisson, and gamma distributions.
GAM model was created in R version 3.0.2 software,
using the gam function of the mgcv package (Wood,
2006), with the number of bigeye tuna as a response vari-
able and SST, SSC, and SSHD as predictor variables.
GAM models in the form of Eq. (1) were applied:
g l
i
ðÞ¼a
0
þ f
1
x
1i
ðÞþf
2
x
2i
ðÞþf
3
x
3i
ðÞþ...f
n
ðx
ni
Þð1Þ
where g is the link function, l
i
is the expected value of the
dependent variable (number of bigeye tuna), a
0
is the
model constant, and f
n
is a smoothing function of the x
n
(which corresponds to the environmental variable in this
study) (Wood, 2006).
The number of bigeye tuna caught data distribution was
right skewness. Hence, to reduce right skewness, logarith-
mic transformation was applied. Logarithmic transforma-
tion gave strong transformation effect on distribution
shape and it is likely to be more symmetrically distributed
(Box and Cox, 1964). The number “ 1 ” was added to the
number of bigeye tuna caught before log-transformatio n
to avoid the singularity of zero values for bigeye tuna
(Zagaglia et al., 2004). The number of bigeye tuna caught
could be predicted using the predict.gam function in the
mgcv package using similar covariates as were used to build
the model. Zagaglia et al. (2004) and Mugo et al. (2010)
employed this a pproach.
In this study, seven models were constructed from the
simplest form by using only one independent variable
(i.e., SST, SSC, and SSHD) and combinations of variables
(i.e., SST + SSC, SST + SSHD, SSC + SSHD, and
SST + SSC + SSHD) as listed in Table 1. For example,
x
1i
correspond to SST in model 1; in model 7, x
Ii
corre-
sponds to SST, x
2i
corresponds to SSC, and x
3i
corre-
sponds to SSHD. These models were evaluated based on
the significance level of predictors (P-value), deviance
explained (DE), and the Akaike information criterion
(AIC) value (Mugo et al., 2010). AIC and DE were used
to determine the best model. The smallest value of AIC
and the highest value of DE were selected as the best
model. As a reference, the parameters of the respect ive
degrees of freedom (EDF) are also listed in Table 1. The
predicted number of bigeye tuna was compared with the
observed number using linear models. The optimal values
of each predictor variable (SST, SSC and SSHD) deter-
mined by GAM were used as main parameters to predict
bigeye tuna habita t.
2.5. Habitat suitability index
Habitat suitability index (HSI) is a numerical index that
represents the capacity of a given habita t to support a
selected species (Oldham et al., 2000). An HSI is a
numerical index between 0 and 1, where 0 indica tes unsuit-
able habitat and 1 represents an optimal habitat. We used
raster calculator function in the spatial analysis tools in
ArcGIS 10.1 to processed HSI. Combining the habitat
factors based on GAMs and accomplished by an additive
priority function P, as shown in Eq. (2),(Store and
Jokimaki, 2003)
Table 1
GAM models used in this study and obtained values for P-value, percent DE, AIC value, and DF, respectively (N = 7751).
No Model Variable P-value DE AIC DF
1 SST SST <2  10
16
5.38% 2759.216 6.783
2 SSC SSC <2  10
16
2.90% 2962.338 8.055
3 SSHD SSHD <2  10
16
3.34% 2927.313 8.402
4 SST + SSC SST <2  10
16
5.94% 2724.76 6.168
SSC 2.28  10
7
6.517
5 SST + SSHD SST <2  10
16
8.03% 2551.88 5.345
SSHD <2  10
16
7.973
6 SSC + SSHD SSC <2  10
16
5.65% 2754.828 7.776
SSHD <2  10
16
8.036
7 SST + SSC + SSHD SST <2  10
16
8.39% 2531.947 4.155
SSC 0.000117 6.468
SSHD <2  10
16
7.968
4 M.D. Setiawati et al. / Advances in Space Research xxx (2014) xxx–xxx
Please cite this article in press as: Setiawati, M.D., et al. Characterization of bigeye tuna habitat in the Southern Waters off Java–Bali using
remote sensing data. Adv. Space Res. (2014), http://dx.doi.org/10.1016/j.asr.2014.10.007

P ¼
X
m
i¼1
ai ... ð2Þ
where P is HIS, m is number of factors, and ai is the rela-
tive important factor of “i”. In this case, “i” is SST, SSHD,
and SSC where “a” is the weight value of each variable.
Weight value was calculated based on the proportion of
important habitat predictor for bigeye tuna according to
GAM result.
3. Results
3.1. Classification of fisheries data and average temporal
variability
The frequency of fishing days in relation to SST, SSC,
SSHD, and month is shown in Fig. 2. For the SST, SSC,
and SSHD high catches, positive catches and null catches
had similar patterns, except that positive catch was the pre-
dominant group of bigeye tuna catch in the Southern
Waters off Java and Bali from 2006 to 2010. The average
null catch during this 5-year period was almost 19% and
the highest was approximately 30% in 2010. The average
positive catch frequency was approximately 53% and the
frequency of high catches was approxim ately 28%. The
average SST values of the null, positive, and high catches
were 28.4 ± 1.3, 28.1 ± 1.3, and 27.8 ± 1.2 °C, respectively;
the average SSC values of the null, positive, a nd high
catches were 0.1 ± 0. 06, 0.11 ± 0. 05, and 0.11 ± 0.
05 mg m
3
, respectively. In addition, the average SSHD
values of the null, positive, and high catches were
0.08 ± 0.06, 0.08 ± 0.07, and 0.08 ± 0.08 m, respect ively.
Judging from the distribution of high catches data, the opti-
mum ranges of SST, SSC, and SSHD were 26.5–28.7 °C,
0.05– 0.12 mg m
3
, and 0.05 to 1.3 m, respectively.
By using high catches data, the preferable time to catch
bigeye tuna can be determined. High catches can be found
from January to December (i.e., year round), with the high-
est frequency in July and the lowest in March (Fig. 2d).
The distribution of high catches data was significantly dif-
ferent from that of the other distributions (positive and null
catches), which was confirmed using a Student’s t-test with
significance level of 95%.
Fig. 3 shows the average temporal variability of number
of bigeye tuna ca tches in the northwest and southeast mon-
soon from 2006 to 2010. The tempor al variability of num-
ber of bigeye tuna caught throughout the year was very
similar. The numbers of bigeye tuna caught tended to be
high and stable during July–October when southeast mon-
soon occurred. The rectangle symbol explained the starting
and closing number of bigeye tuna caught in each month,
that is, in July the rectangle size was small and it indicated
that in the early and the last of July the number of bigeye
tuna caught was almost similar. In addition, null catches
and high catches were found throughout the year.
Fig. 2. Frequency of fishing days in relation to (a) SST, (b) SSC, (c) SSHD
and (d) month from 2006 to 2010. They were grouped according to the
way used by Andrade and Garcia (1999).
M.D. Setiawati et al. / Advances in Space Research xxx (2014) xxx–xxx 5
Please cite this article in press as: Setiawati, M.D., et al. Characterization of bigeye tuna habitat in the Southern Waters off Java–Bali using
remote sensing data. Adv. Space Res. (2014), http://dx.doi.org/10.1016/j.asr.2014.10.007

According to the fisheri es data classification, null catches
were found in the southeast monsoon season, approxi-
mately 15%, and in the northwest monsoon season
approximately 22% of the time. Moreover, high catches
in the southeast monsoon were higher (34%) than those
in the northwest monsoon (24%). Thus, the southeast mon-
soon season was more productive than the northwest
monsoon.
3.2. Distribution of number of bigeye tuna caught and
environmental data
The distribution of number of bigeye tuna caught and
the three environmental variables in the Southern Waters
off Java and Bali from 2006 to 2010 are shown in Fig. 4.
The distribution of the number of bigeye tuna caught
was asymmetrical (Fig. 4a). A log transformation of the
number of bigeye tuna caught indicated Poisson distribu-
tion (Fig. 4b). Bigeye tuna were caught at SST between
24.8 and 30.8 °C, with the highest frequency at 28.5 °C
(Fig. 4c). The range of SSC for the fishing sets was 0.02–
0.46 mg m
3
and the preferable concentration ranged from
0.05–0.17 mg m
3
(Fig. 4d). The SSHD ranged from 20
to 30 cm and value of 5 to 15 cm was preferable for the
fishing sets, with the peak at 10 cm (Fig. 4e). The prefer able
environmental factors for fishing sets can be distinguished
using the histograms shown in Fig. 4.
3.3. Analysis of habitat characteristics for bigeye tuna by
using GAM
Prior to examining the relationship between the bigeye
tuna catches and environmental variables, we examined
the relationship between number of bigeye tuna caught
Fig. 3. Average temporal variability of number of bigeye tuna catches
from 2006 to 2010.
Fig. 4. Histograms of number of bigeye tuna and environmental data: (a) distribution of number of bigeye tuna, (b) distribution of log-transformed
number of bigeye tuna, (c) SST, (d) SSC, (e) SSHD.
6 M.D. Setiawati et al. / Advances in Space Research xxx (2014) xxx–xxx
Please cite this article in press as: Setiawati, M.D., et al. Characterization of bigeye tuna habitat in the Southern Waters off Java–Bali using
remote sensing data. Adv. Space Res. (2014), http://dx.doi.org/10.1016/j.asr.2014.10.007

and environmental variables. Table 1 lists the Model vari-
able, P-value, DE, AIC, and DF for some models. The pre-
dictor variables were highly significant (P < 0.001) for all of
the models. High significance was indicated from the lowest
AIC and the highest DE. DE has the same meaning as the
determination value in the linear regression. SST showed
the highest DE among the single-parameter models. Mod-
els developed from three parameters (i.e., model 7) had the
lowest AIC values and the highest DE, which indicated
that the combination of three parameters generated the
best models.
Fig. 5 shows GAM plots developed to interp ret the indi-
vidual effect of each predictor variable on the number of
bigeye tuna. The effect of SST, SSC, and SSHD on the
number of bigeye tuna is shown in Fig. 5a–c, respectively.
A negative effect of SST on the number of bigeye tuna was
observed at temperature >28.7 °C. There was a positive
effect of temperature on the number of bigeye tuna, which
was from 24.5 to 28.7 °C. Bigeye tuna appeared to prefer
cooler waters, but the number of sets performed at tempe r-
atures <25 °C was low. As a result, the confidence interval
was wider for SST <25 °C. There was an indication of
greater number of bigeye tuna caught at lower SSTs, but
the number of data points in the lower temperature range
declined and the confidence level also declined. For SSC,
a positive effect on the number of bigeye tuna occurred
between 0.07 and 0.22 mg m
3
(Fig. 5b). From
0.19 mg m
3
, a decline occurred towards the highest SSC
value. A GAM plot of SSHD showed a positive effect of
this variable on the number of bigeye tuna caught between
3 and 7 cm in the region of high confidence level (Fig. 5c).
3.4. Model vali dation and bigeye tuna habitat prediction
A scatter plot of randomly selected value of in-situ fish-
eries data and predicted values generated by the GAM
using observational explanatory variables as input is pre-
sented in Fig. 6. Out of the sample pool, 90 samples were
randomly selected from the fisheries data classification.
We used stratified random sampling to determine the sam-
ple size and data for each stratification. The data were
stratified based on the results of fisheries data classification
(i.e., the sample sizes of null, positive, an d high catches are
18, 47, and 25, respectively). The data sample was selected
for each stratification using the random sample function in
Microsoft Excel. The adjusted simple linear regression line
was significant (P < 0.05, r
2
= 0.56) (Fig. 6). Therefore,
although not as well as expected, we consider that the num-
ber of bigeye tuna generated from the GAM can estimate
the observed value in the region.
HIS maps of bigeye tuna in 2009 are presented in Fig. 8.
The red color indicates the most suitable habitat for bigeye
Fig. 5. Effect of three oceanographic variables on the number of bigeye tuna (a) SST, (b) SSC and (c) SSHD. Tick marks at abscissa axis represent the
observed data points. Full line is the GAMs function. Dashed dot lines indicate the 95% confidence level.
M.D. Setiawati et al. / Advances in Space Research xxx (2014) xxx–xxx 7
Please cite this article in press as: Setiawati, M.D., et al. Characterization of bigeye tuna habitat in the Southern Waters off Java–Bali using
remote sensing data. Adv. Space Res. (2014), http://dx.doi.org/10.1016/j.asr.2014.10.007

tuna, and blue color indicates unsuitable habitat for bigeye
tuna. Most of the fishing activity was done in the yellow
color (HSI = 0.6–0.7). According to HSI map, bigeye tuna
preference area changed depending on the month as can be
seen in Fig. 8, but the fishing location seemed constant. As
shown in Fig. 8, the high preferable bigeye tuna habitat
tended to wes tward. This result was supported by the pre-
vious research on the eastern coast of Australia, where fish
move from east to the west (Gunn et al., 2005).
4. Discussion
In general, fisheries data are abundant for developed
countries, but the data are limited in terms of study area.
That is why we used the number of bigeye caught tuna as
an index of fish abundance. Because we used a different
unit of fish abundance, classification of fisheries data was
performed to define the type of fish catch data and was
used as preliminary investigation to determine the best
method of statistical analysis for our data. Here, we exam-
ined our results and their inherent relevance as an environ-
mental indicator of bigeye tuna habitat.
Identification of bigeye tuna habitat in the Southern
Waters off Java and Bali is a challenge because the distribu-
tion of habitat is variable over time. Tuna resources remain
under pressure globally (Sunoko an d Huang, 2014). For
that reason, identification of bigeye tuna habitat character-
istic using remote sensing of biophysical environment
parameters would be especially important to predict stocks’
responses to externalities such as climate change, illegal
fishing, and fish ing pressure.
Bigeye tuna catch rate varied as time and environmental
variables changed (Fig. 2). During a year, the fishery oper-
ated between 10° Sand18° S(Fig. 7). Based on Fig. 7, the
spatial distribution of fishing activity did not change signif-
icantly, but the number of bigeye tuna caught changing by
the time. The highest fishing activity was from June to
October because of low null catches and rich high catches
(Fig. 2d). Most of the null catches occurred during the
northwest monsoon season, especially from February to
April (Fig. 2d). This condition imposed high costs on fish-
ermen. According to the classification of fisheries data, the
average frequency of null catches over 5 years was 19% and
it reached almost 30% in 2010, when a strong La Nina
event was observed (Feng et al., 2013). During February
to April, the numbers of bigeye tuna caught tend to
decrease from 2006 to 2010 (Fig. 3). These two factors
(the decline in number of bigeye tuna and the La Nina
event) caused a reduction in fishing activity around the
Southern Waters off Java and Bali, and a corresponding
move to the Pacific Ocean in the eastern part of Indonesia
since 2011.
The effect of environmental conditions, de duced from
GAMs, indicated that environmental variables strongly
influenced the numbers of bigeye tuna caught. SST was
more important than SSC or SSHD in the study area. This
was indicated by SST having the highest DE and lowest
AIC in all models. In addition, the Pacific Ocean influences
the transfer of heat energy to the Indian Ocean by ITF (Lee
et al., 2001), which causes changes in SST. During south-
east monsoon, the reduction of heat transfer caused SST
to be lower (approximately 26.7 °C). Furthermore, SST is
higher when the Intertropical Convergence Zone (ITCZ)
occurs because of weak winds and high relative humidity
that result in reduced evaporative cooling of SST (Farrar
and Weller, 2003). Bigeye tuna catches increased in areas
with relatively low SST (24.5–28.7 °C) and decreased in
the areas with SST >28.7 °C. This was supported by previ-
ous research (e.g. Gunn et al., 2005; Howell et al., 2010;
Syamsuddin et al., 2013). Furthermore, bigeye tuna pre-
ferred to remain in lower-temperature areas. Our finding
seems to agree wi th the result of Brill et al. (1994), who
explained that bigeye tuna move towards the cooling hab-
itat to prevent overheating.
Fig. 6. A Scatter plot between the observed values and GAM model predicted ones.
8 M.D. Setiawati et al. / Advances in Space Research xxx (2014) xxx–xxx
Please cite this article in press as: Setiawati, M.D., et al. Characterization of bigeye tuna habitat in the Southern Waters off Java–Bali using
remote sensing data. Adv. Space Res. (2014), http://dx.doi.org/10.1016/j.asr.2014.10.007

Temperature limits horizontal and vertical distribution
of bigeye tuna and this varies by region and size (Miyabe
Naozumi, 1993; Brill et al., 2005; Howell et al., 2010).
Lehodey et al. (2010) reported that natural mortality of
older stages of bigeye tuna in the Pacific Ocean increased
due to too warm surface temperature and decreasing oxy-
gen concentration in the sub-surface caused by global
warming. Howell et al. (2010) reported that tagged bigeye
tuna in the central Nort h Pacific Ocean showed daily verti-
cal movement, where they spent much of the time (61%)
Fig. 7. The spatial distribution of SSC and bigeye tuna catches in Southern Waters off Java–Bali in 2009.
M.D. Setiawati et al. / Advances in Space Research xxx (2014) xxx–xxx 9
Please cite this article in press as: Setiawati, M.D., et al. Characterization of bigeye tuna habitat in the Southern Waters off Java–Bali using
remote sensing data. Adv. Space Res. (2014), http://dx.doi.org/10.1016/j.asr.2014.10.007

near the surface layer and above the thermocline layer
during nighttime, but less time (39%) during daytime.
Nighttime depth ranged from the surface to 100 m and
where daytime dive beyond 500 m. Bigeye tuna regularly
expose themselves to temperature change up to 20 °C (from
25 °C surface layer temperature to 5 °C at 500 m depth
during their daily vertical movement). Bigeye tuna occa-
sionally makes an upward excursion into the mixed layer
water to warm their muscles ( Brill et al., 2005). Such tagging
experiments are important for understanding bigeye tuna’s
Fig 7. (continued)
10 M.D. Setiawati et al. / Advances in Space Research xxx (2014) xxx–xxx
Please cite this article in press as: Setiawati, M.D., et al. Characterization of bigeye tuna habitat in the Southern Waters off Java–Bali using
remote sensing data. Adv. Space Res. (2014), http://dx.doi.org/10.1016/j.asr.2014.10.007

vertical habitat utilization. Our results indicated that few
fishing sets (8%) occurred at temperatures <25 °C(Fig 6a).
SSHD was the second most significant oceanographic
predictor of bigeye tuna in the study area. We used SSHD
to understand oceanic variability, such as current dynam-
ics, eddies, convergences, and divergences, which can be
used as proxies for the potential location of tuna catches
(Polovina and Howell, 2005). Our study showed that big-
eye tuna preferred areas with SSHD values of –3 to 7 cm
(Fig 5c). Actually, the negative extreme values of SSHD
had a positive effect on the number of bigeye tuna caught,
but the number of observations was low and the confidence
interval was wide. This finding indicated that bigeye tuna
foraged in areas with negative SSHD and low of SSHD.
Fig. 8. Habitat suitability index for bigeye tuna from January to December 2009 overlaid with bigeye tuna fishing location (continue to the next page).
M.D. Setiawati et al. / Advances in Space Research xxx (2014) xxx–xxx 11
Please cite this article in press as: Setiawati, M.D., et al. Characterization of bigeye tuna habitat in the Southern Waters off Java–Bali using
remote sensing data. Adv. Space Res. (2014), http://dx.doi.org/10.1016/j.asr.2014.10.007

Negative SSHD will push the thermocline upward near the
surface layer and the elevation of thermocline will allow
bigeye tuna from below to become accessible to lon gline
gear. The upward movement of thermocline layer causes
the temperature in the surface layer to become cooler.
According to Arrizabalaga et al. (2008), only for very neg-
ative SSHD, bigeye tuna in shallow waters is only attracted
by the thermocline when this is closer to the surface. This
phenomenon was reported by Syamsuddin et al. (2013)
during the El-Nino event in 1997, where extreme minus
SSHD with many observation points occurred and gave
the positive effect to the abundance of bigeye tuna.
Among three environmental predictors used in the
model, SSC was the least important, but was still
Fig 8. (continued)
12 M.D. Setiawati et al. / Advances in Space Research xxx (2014) xxx–xxx
Please cite this article in press as: Setiawati, M.D., et al. Characterization of bigeye tuna habitat in the Southern Waters off Java–Bali using
remote sensing data. Adv. Space Res. (2014), http://dx.doi.org/10.1016/j.asr.2014.10.007

statistically significant (P < 0.001). As a biological compo-
nent available for satellite remote sensing, SSC is an index
of phytoplankton biomass that provides valuable informa-
tion about trophic interaction in marine ecosystem (Wilson
et al., 2008). Chlorophyll-a data are a valuable proxy for
water mass boundaries and upwelling events. High value
of SSC was concentrated along the southern coast of Java
(7–9° S) (Fig. 7). Susanto et al. (2001) and Ningsih et al.
(2013) reported that the seasonal appearance of chlorophyll
front and the yearly upwelling phenomenon occurred in the
Southern Waters off Java and Bali, especially in the coastal
area (Fig. 7). Upwelling areas are potential convergence
zone for plankton aggregation, attracting larger predators,
such as tuna (Lehodey et al., 1997). Bigeye tuna is a visual
predator where water clarity is important (Brill et al.,
2005). The open ocean provides the optically clearest aqua-
tic habitat (Jerlov, 1976). Hence, in the open ocean bigeye
tuna can forage the prey optimally. Yearly upwelling
occurred in the study area, especially in the coastal zone,
so that SSC did not directly affect the abundance of bigeye
tuna. Overall, SST and SSHD mainly influenced bigeye
tuna catch. In this study, the fishermen used the same fish-
ing gear with similar fishing techniques. Therefore, we
assumed that differences in fishing gear did not affect the
catchability of bigeye tuna.
Spatial mapping of bigeye tuna habitat was conducted
by the HSI approach. The HSI map from January to
December is shown in Fig. 8. It explains that most fishing
activities were located when HSI was 0.6–0.7, but in Sep-
tember fishing activities were located in the most suitable
habitat (HSI = 1). HSI showed concurrence with actual
fishing location for the September to December period.
This is also the period that showed low null catches fre-
quencies (Fig. 2d). However, the model appears to have
difficulties in predicting higher catch rates, that is, in July
and August. The prediction of bigeye tuna by GAM
showed a significant relationship with the observed value
with a confidence level of 95% (r
2
= 0.56) (Fig. 6).
Zagaglia et al. (2004) also reported the significant relation-
ship between observed CPUE and predicted CPUE from
GAM (r
2
= 0.51) for yellowfin tuna in the equatorial
Atlantic Ocean. Mugo et al. (2010) also applied GAM to
skipjack tuna in the western part of the North Pacific
Ocean and found a significant relationship between
observed CPUE and predicted CPUE from GAM
(r
2
= 0.64). Our results cannot correctly predict the number
of bigeye tuna caught as in Mugo et al. (2010). This is
because we used daily catch data as numbers of bigeye tuna
caught and this was difficult when we predict null catches.
Nevertheless, our model explained 8.39% (Table 1) of var-
iability in bigeye tuna abundance based on environmental
variables only; the model generated by Mugo et al.
(2010) explained 13.3% of variability. This indicates that
our method is useful. Environmental variables are impor-
tant to predicting the bigeye tuna habitat, but are probably
not only the factors that influence fishing locations for this
species. In addition, data that have a high temporal
resolution and more years are likely to generate a better
model to predict bigeye tuna habitat in the study area.
5. Conclusions
Characterization of bigeye tuna habitat in the Southern
Waters off Java and Bali using a remot e sensing approach
has been performed. Daily in-situ fish catch data from PT
Perikanan Nusantara and monthly remotely sensed envi-
ronmental data of SST, SSC, and SSHD for the period
2006–2010 were used here. The GAM statistical method
and GIS were combined. Seven GAM models were gener-
ated with the number of bigeye tuna caught as a response
variable, and SST, SSC, SSHD as predictor variables.
The results showed that SST was the mo st important hab-
itat predictor for bigeye tuna migration in the Southern
Waters off Java and Bali, followed by SSHD and SSC.
The spatial pattern of bigeye tuna habitat characteristics
gave typical low SST, negative to low SSHD, and low to
moderate SSC. Thermocline layer or depth is the important
feature to predict the vertical migration of bigeye tuna and
SSHD seems to be a good parameter to forecast the ther-
mocline depth.
Knowledge of habitant location would guide fishermen
to productive areas and they could thus reduce the costs
of operating their boats. The results revealed that fisher-
men still obtained null catches with a frequency of 19%
over the 5-year period, which indicated suboptimal success
in identifying favorable bigeye tuna habitat. Meanwhile,
the El Nin
˜
o–Southern Oscillation (ENSO) also might affect
the number of null catches, as indicated by an increa se dur-
ing a La Nin
˜
a event. The use of GAMs showed that most
of fishing activity was located in medium-potential habitat
(HSI = 0.6–0.7). How ever, bigeye tuna preference area
changed depending on the month (tended to westward)
and fishermen did not understand that condition, which
may lead to still obtained null catch.
In future work, increasing the number of predictor envi-
ronmental variables with the high temporal resolution may
improve the model of bigeye tuna habitat. Furthermore,
utilizing the predictive habitat maps would help fisheries
managers to establish decision-making criteria regarding
quotas, and would inform regulations for ecosystem and
habitat protection.
Acknowledgments
We would like to thank PT. Perikanan Nusantara,
Benoa, Bali, Indonesia for providing fisheries data. We
thank DIKNAS (ministry of education of Indonesia),
LPDP (Indonesia Endowment Fund of Ministry of
Finance), and JAXA (Japan Aerospace Exploration
Agency) for financial suppo rt. We also gratefully acknowl-
edge NASA for the ocean color AQUA-MODIS SST and
chlorophyll-a data that were downloaded from ocean-color
homepage and the use of altimetry data for SSH datasets
downloaded from the AVISO homep age.
M.D. Setiawati et al. / Advances in Space Research xxx (2014) xxx–xxx 13
Please cite this article in press as: Setiawati, M.D., et al. Characterization of bigeye tuna habitat in the Southern Waters off Java–Bali using
remote sensing data. Adv. Space Res. (2014), http://dx.doi.org/10.1016/j.asr.2014.10.007

References
Abbott, M.R., Letelier, R.M. 1999. Chlorophyll fluorescence (MODIS
product number 20). NASA Algorithm Theoretical Basis Document.
(http://modis.gsfc.nasa.gov/data/atbd/atbd_mod22.pdf).
Andrade, H.A., Garcia, C.A., 1999. Skipjack tuna fishery in relation to sea
surface temperature off the southern Brazilian coast. Fish. Oceanogr.
8, 245–254
.
Arrizabalaga, H., Pereira, J.G., Royer, F., Galuardi, B., Goni, N.,
Artetxe, I., Arregi, I., 2008. Lutcavage Fishery Oceanogr. 17, 74–83
.
Bailey, C., Dwiponggo, A., Marahudin, F., 1987. Indonesia Marine
Capture Fisheries. International Center for Living Aquatic Resources,
Manila.
Bertrand, A., Josse, E., Bach, P., Gros, P., Dagorn, L., 2002. Hydrological
and trophic characteristics of tuna habitat: consequences on tuna
distribution and longline catchability. Can. J. Fish. Aquat. Sci. 59,
1002–1013
.
Blackburn, M., 1968. Micronekton of the eastern tropical Pacific Ocean:
family composition, distribution, abundance and relation to tuna.
Fishery Bull. US 67, 71–115.
Box, G.E., Cox, D.R., 1964. An analysis of transformations. J Roy.
Statist. Soc. Ser. B (Methodological) 26, 211–252
.
Brill, R.W., Dewar, H., Graham, J.B., 1994. Basic concepts relevant to
heat transfers in fishes, and their use in measuring the psychological
thermoregulatory abilities of tuna. Environ. Biol. Fishes 40, 109–124
.
Brill, R.W., Bigelow, K.A., Musyl, M.K., Fritsches, K.A., Warrant, E.J.,
2005. Bigeye tuna (Thunnus obesus) behaviour and physiology and
their relevance to stock assessments and fishery biology. Collective
Vol. Sci. Pap. ICCAT 57, 142–161
.
Chassot, E., Bonhommeau, S., Reygondeau, G., Nieto, K., Polovina, J.,
Huret, M., Dulvy, N.K., Demarcq, H., 2011. Satellite remote sensing
for an ecosystem approach to fisheries management. ICES J. Mar. Sci.
68, 651–666
.
Druon, J.N., 2010. Habitat mapping of the Atlantic bluefin tuna derived
from satellite data: its potential as a tool for the sustainable
management of pelagic fisheries. Mar. Policy 34, 293–297
.
Druon, J.N., Fromentin, J.M., Aulanier, F., Heikkonen, J., 2011.
Potential feeding and spawning habitats of atlantic bluefin tuna in
the mediterranean sea. Mar. Ecol. Prog. Ser. 439, 223–240.
Du, Y., Qu, T., Meyers, G., 2008. Interannual variability of sea surface
temperature off java and sumatra in a global GCM. J. Clim. 21, 2451–
2465.
Farrar, T., Weller, R., 2003. Where the trade winds meet: air-sea coupling
in the inter-tropical convergence zone. Maryland-USA: NOAA
Research Archive of Spotlight Feature Articles, Office of Oceanic
and Atmospheric Research “In the spotlight” internet article (http://
www.oar.noaa.gov/spotlite/archive/spot_pacs.html).
Feng, M., Wijffels, S., 2002. Intraseasonal variability in the south
equatorial current of the east Indian Ocean. J. Phys. Oceanogr. 32,
265–277.
Feng, M., McPhaden, M.J., Xie, S.P., Hafner, J., 2013. La Nin
˜
a forces
unprecedented Leeuwin Current warming in 2011. Sci. Rep. 3, 1–9.
Gillet, R., Tuna for tomorrow, 2013, Working Paper 011, Indian Ocean
Commission, SmartFish Programme; Port Louis-Mauritius.
Gordon, A.L., 2005. Oceanography of the Indonesian seas and their
throughflow. Oceanography 18, 14–27.
Gordon, A.L., Sprintall, J., Van Aken, H.M., Susanto, D., Wijffels, S.,
Molcard, R., Ffield, A., Pranowo, W., Wirasantosa, S., 2010. The
Indonesian throughflow during 2004–2006 as observed by the
INSTANT program. Dyn. Atmos. Oceans. 50, 115–128
.
Guisan, A., Edwards Jr, C.T., Hastie, T., 2002. Generalized linear and
generalized additive models in studies of species distributions: setting
the scene. Ecol. Modell. 157, 89–100
.
Gunn, J., Hampton, J., Evans, K., Clear, N., Patterson, T., Bigelow, K.,
Langley, A., Leroy, B., Williams, P., Miyabe, N., Sibert, J., Bestley, F.,
Hartmann, K., 2005. Migration and habitat preferences of bigeye tuna,
Thunnus obesus, on the east coast of Australia – a project using
archival and conventional tags to determine key uncertainties in the
species stock structure, movement dynamics and CPUE trends.
Canberra-Australia: Project no. 1999/109, CSIRO Marine Research.
Hanamoto, E., 1987. Effect of oceanographic environment on bigeye tuna
distribution. Bull. Japn. Soc. Fish. Oceanogr. 51, 203–216
.
Howell, E.A., Hawn, D.R., Polovina, J.J., 2010. Spatiotemporal variabil-
ity in bigeye tuna (Thunnus obesus) dive behaviour in the central north
pacific Ocean. Prog. Oceanogr. 86, 81–93.
Jerlov, N.G., 1976. Marine Optics. Elsevier, Amsterdam.
Kamei, G., Felix, J.F., Shenoy, L., Shukla, S.P., Devi, H.M., 2014.
Application of remote sensing in fisheries: role of potential fishing zone
advisories. In: Sundaresan, J., Santosh, K.M., De
´
ri, A., Roggema, R.,
Singh, R. (Eds.), Geospatial Technologies and Climate Change.
Springer International Publishing, Amsterdam-Netherlands, pp. 175–
186
.
Klemas, V., 2013. Fisheries applications of remote sensing: an overview.
Fish. Res. 148, 124–136.
Lan, K.W., Lee, M.A., Lu, H.J., Shieh, W.J., Lin, W.K., Kao, S.C., 2011.
Ocean variations associated with fishing conditions for yellowfin tuna
(Thunnus albacares) in the equatorial Atlantic Ocean. ICES J. Mar.
Sci. 68, 1063–1071
.
Laurs, R.M., 1986. Applications of satellite remote sensing to fisheries.
Soc. Airborne Satell. Phys. Fishery Oceanogr. 8, 139–150.
Laurs, R.M., Fiedler, P.C., Montgomery, D.R., 1984. Albacore tuna catch
distributions relative to environmental features observed from satel-
lites. Deep-Sea Res. 31, 1085–1099
.
Lee, T., Fukumori, I., Minemenlis, D., Xing, Z., Fu, L.L., 2001. Effects of
the indonesian throughflow on the pacific and Indian Oceans. Phys.
Oceanogr. 32, 1404–1429
.
Lehodey, P., Bertignac, M., Hampton, J., Lewis, A., Picaut, J., 1997. El
nino southern oscillation and tuna in the western pacific. Nature 389,
715–718.
Lehodey, P., Senina, I., Sibert, J., Bopp, L., Calmettes, B., Hampton, J.,
Murtugudd, R., 2010. Progress In Oceanography. doi:10.1016/
j.pocean.2010.04.021
Lennert-Cody, C.E., Roberts, J.J., Stephenson, R.J., 2008. Effects of gear
characteristics on the presence of Bigeye tuna (Thunnus obesus) in the
catches of the purse-seine fishery of the eastern Pacific Ocean. ICES J.
Mar. Sci. 65, 970–978
.
Lizarazo, I., 2012. Quantitative land cover change analysis using fuzzy
segmentation. Int. J. Appl. Earth Obs. Geoinf. 15, 16–27.
Lu, H.-J., Lee, K.-T., Lin, H.-L., Liao, C.-H., 2001. Spatio-temporal
distribution of yellowfin tuna Thunnus albacares and bigeye tuna
Thunnus obesus in the Tropical Pacific Ocean in relation to large-scale
temperature fluctuation during ENSO episodes. Fish. Sci. 67, 1046–
1052
.
Manessa, M.D.M., As-syakur, A.R., 2011. Observation of SST variability
and their relationship with ENSO over Coral Triangle region using
satellite data. Proceeding of the International Seminar on Marine:
Implication of Climate Change in Coral Triangle Initiative (CTI)
Region (pp. 199–212). Denpasar-Indonesia: Institute for Marine
Research and Observation, Indonesian Ministry of Marines Affairs
and Fisheries.
Maritorena, S., d’Andon, O.H.F., Mangin, A., Siegel, D.A., 2010. Merged
satellite ocean color data products using a bio-optical model: charac-
teristics, benefits and issues. Remote Sens. Environ. 114, 1791–1804.
Miyabe, Naozumi., 1993. A review of the biology and fisheries for bigeye
tuna, Thunnus obesus, in the Pacific Ocean. FAO Fisheries Technical
Papers 336, 2, 207–244.
Mohri, M., 1999. Seasonal change in bigeye tuna fishing areas in relation
to the oceanographic parameters in the Indian Ocean. J. Nat. Fish.
Univ. 47, 43–54
.
Mohri, M., Nishida, T., 1999. Distribution of bigeye tuna and its
relationship to the environmental conditions in the Indian Ocean based
on the Japanese longline fisheries information. IOTC Proc. 2, 207–220.
Mugo, R., Saitoh, S.-I., Nihira, A., Kuroyama, T., 2010. Habitat
characteristics of skipjack tuna (Katsuwonus pelamis) in the Western
North Pacific. Fish. Oceanogr. 19, 382–396.
14 M.D. Setiawati et al. / Advances in Space Research xxx (2014) xxx–xxx
Please cite this article in press as: Setiawati, M.D., et al. Characterization of bigeye tuna habitat in the Southern Waters off Java–Bali using
remote sensing data. Adv. Space Res. (2014), http://dx.doi.org/10.1016/j.asr.2014.10.007

Ningsih, N.S., Rakhmaputeri, N., Harto, A.B., 2013. Upwelling variabil-
ity along the southern coast of Bali and in Nusa Tenggara waters.
Ocean Sci. J. 48, 49–57
.
Nishikawa, Y., Honma, M., Ueyanagi, S., Kikawa, S., 1985. Average
Distribution of Larvae of Oceanic Species of Scombroid Fishes, 12. S
series/Far Seas Fisheries Research Laboratory, 99 p.
Oldham, R.S., Keeble, J., Swan, M.J.S., Jeffcote, M., 2000. Evaluating the
suitability of habitat for the Great Crested Newt (Triturus cristatus).
Herpetological J. 10, 143–155
.
Osawa, T., Julimantoro, S., 2010. Study of fishery ground around
indonesia archipelago using remote sensing data. In: Sumi, A.,
Fukushi, K., Hiramatsu, A. (Eds.), Adaptation and Mitigation
Strategies for Climate Change. Springer, Tokyo, pp. 57–69
.
Perez, J.C., Alvarez, M.A., Heikkonen, J., Guillen, J., Barbas, T., 2013.
The efficiency of using remote sensing for fisheries enforcement:
application to the mediterranean bluefin tuna fishery. Fish. Res. 147,
24–31
.
Polovina, J.J., Howell, E.A., 2005. Ecosystem indicators derived from
satellite remotely sensed oceanographic data for the North Pacific.
ICES J. Mar. Sci. 62, 319–327.
Proctor, C.H., Merta, I.G.S., Sondita, M.F.A., Wahuju, R.I., Davis,
T.L.O., Gunn, J.S., Andamari, R., 2003. A review of Indonesia’s
Indian Ocean tuna fisheries. Jakarta-Indonesia: CSIRO Marine
Research, Research Institute of Marine Fisheries, and Bogor Agricul-
tural University.
Qu, T., Meyers, G., 2005. Seasonal characteristics of circulation in the
southeastern tropical Indian Ocean. J. Phys. Oceanogr. 35, 255–267.
Santos, A.M.P., 2000. Fisheries oceanography using satellite and airborne
remote sensing methods: a review. Fish. Res. 49, 1–20
.
Simorangkir, S., 2003. Status Perikanan Tuna Nasional. In: Paper
presented to workshop: Lokakarya Pengkajian Stok Sumberdaya
Ikan Nasional. Jakarta-Indonesia: Indonesian Ministry of Marine
Affairs and Fisheries.
Song, L., Zhou, Y., 2010. Developing an integrated habitat index for
bigeye tuna Thunnus obesus in the Indian Ocean based on longline
fisheries data. Fish. Res. 105, 63–74
.
Song, L., Zhou, J., Zhou, Y., Nishida, T., Jiang, W., Wang, J., 2009.
Environmental preferences of Bigeye tuna, Thunnus obesus, in the
Indian Ocean: an application to a longline fishery. Environ. Biol.
Fishes 85, 153–171
.
Song, L., Zhang, Y., Xu, L., Jiang, W., Wang, J., 2008. Environmental
preferences of longlining for yellowfin tuna (Thunnus albacares) in the
tropical high seas of the Indian Ocean. Fish. Oceanogr. 17, 239–253.
Sprintall, J., Wijffels, S., Molcard, R., Jaya, I., 2010. Direct evidence of the
south java current system in ombai strait. Dyn. Atmos. Oceans 50,
140–156.
Store, R., Jokimaki, J., 2003. A GIS based multi-scale approach to habitat
suitability modeling. Ecol. Model. 169, 1–15
.
Stretta, J.M., 1991. Forecasting models for tuna fishery with aerospatial
remote sensing. Int. J. Remote Sens. 12, 771–779.
Sund, P.N., Blackburn, M., William, F., 1981. Tunas and their environ-
ment in the Pacific Ocean: a review. Oceanogr. Mar. Biol. Annu. Rev.
19, 443–512
.
Sunoko, R., Huang, H.W., 2014. Indonesia tuna fisheries development
and future strategy. Mar. Policy 43, 174–183
.
Susanto, R.D., Gordon, A.I., Zheng, Q., 2001. Upwelling along the
Coasts of Java and Sumatra and Its Relation to ENSO. Geophys. Res.
Lett. 28, 1599–1602
.
Susanto, R.D., Moore II, T.S., Marra, J., 2006. Ocean color variability in
the Indonesian Seas during the SeaWiFS era. Geochem. Geophys.
Geosyst. 7, Q05021
.
Syamsuddin, M.L., Saitoh, S.-I., Hirawake, T., Bachri, S., Harto, A.B.,
2013. Effects of El Nin
˜
o-Southern Oscillation events on catches of
Bigeye Tuna (Thunnus obesus) in the eastern Indian Ocean of Java.
Fishery Bull. 111, 175–188
.
Talley, L.D., Sprintall, J., 2005. Deep expression of the Indonesian
Throughflow: Indonesian intermediate water in the South Equatorial
Current. J. Geophys. Res. oceans 110, C10009.
Uda, M., 1973. Pulsative fluctuation of oceanic fronts in association
with the tuna fishing grounds and fisheries. J. Mar. Sci. Technol. 7,
245–264.
Valavanis, V.D., Pierce, G.J., Zuur, A.F., Palialexis, A., Saveliev, A.,
Katara, I., Wang, J., 2008. Modelling of essential fish habitat
based on remote sensing, spatial analysis and GIS. Hydrobiologia
612, 5–20
.
Wilson, C., Morales, J., Nayak, S., Asanuma, I., Feldman, G., 2008.
Ocean-color radiometry and fisheries. In: Platt, T., Hoepffner, N.,
Stuart, V., Brown, C. (Eds.), Why Ocean Colour? pp. 47–58.
Dartmouth-Canada: The Societal Benefits of Ocean-Color Technol-
ogy. Reports of the International Ocean-Color Coordinating Group 7.
Wood, S.M., 2006. Generalized Additive Models, an Introduction with R.
Chapman and Hall, London-UK
.
Yen, K.W., Lu, H.J., Chang, Y., Lee, M.A., 2012. Using remote-sensing
data to detect habitat suitability for yellowfin tuna in the Western and
Central Pacific Ocean. Int. J. Remote Sens. 33, 7507–7522
.
Zagaglia, C.R., Lorenzzetti, J.A., Stech, J.L., 2004. Remote sensing data
and longline catches of yellowfin tuna (Thunnus alalunga) in the
Equatorial Atlantic. Remote Sens. Environ. 93, 267–281
.
Zainuddin, M., Kiyofuji, H., Saitoh, K., Saitoh, S.-I., 2006. Using multi-
sensor satellite remote sensing and catch data to detect ocean hot spots
for albacore (Thunnus alalunga) in the northwestern North Pacific.
Deep-Sea Res. II 53, 419–431
.
Zainuddin, M., Saitoh, K., Saitoh, S.-I., 2008. Albacore (Thunnus
alalunga) fishing ground in relation to oceanographic condition in
the Western North Pacific Ocean using remotely sensed satellite data.
Fish. Oceanogr. 17, 61–73
.
M.D. Setiawati et al. / Advances in Space Research xxx (2014) xxx–xxx 15
Please cite this article in press as: Setiawati, M.D., et al. Characterization of bigeye tuna habitat in the Southern Waters off Java–Bali using
remote sensing data. Adv. Space Res. (2014), http://dx.doi.org/10.1016/j.asr.2014.10.007