Blackwell Science, LtdOxford, UKFISFisheries Science0919 92682005 Blackwell Science Asia Pty LtdDecember 200571613321340Original ArticleFeeding ecology of dominant

fish at a beachM Yamamoto and O Tominaga

FISHERIES SCIENCE 2005; 71: 1332–1340

Feeding ecology of dominant demersal fish species

Favonigobius gymnauchen, Repomucenus spp. and
Tarphops oligolepis at a sandy beach where larval
Japanese flounder settle in the Seto Inland Sea, Japan
Masayuki YAMAMOTO1* AND Osamu TOMINAGA2

1
Kagawa Prefectural Fisheries Experimental Station, Takamatsu, Kagawa 761-0111 and 2Faculty of
Biotechnology, Fukui Prefectural University, Obama, Fukui 917-0003, Japan

ABSTRACT: Feeding habits of Favonigobius gymnauchen, Repomucenus spp. and Tarphops oli-
golepis were examined during the period from May to August in 1999, 2000 and 2001 at a sandy
beach in the central Seto Inland Sea, Japan. The three species of fish accounted for more than 60%
of all fish collected in terms of number of individuals. All three species mainly consumed small crus-
taceans. However, the major prey of F. gymnauchen and Repomucenus spp. differed from that of
T. oligolepis. F. gymnauchen and Repomucenus spp. mainly consumed mysids and small crangonid
shrimp (<12 mm in body length), which predominated in the study area. T. oligolepis actively selected
only epifaunal mysids Nipponomysis ornata and avoided crangonid shrimp and gammarids. The
frequencies of occurrence of fish in the guts of the three dominant species were very low, and larval
and juvenile Japanese flounder were not observed in any of the three fish species. The diet of juvenile
Japanese flounder was similar to that of the three species. From these results, it appears that these
three fish are competing species for the flounder.

KEY WORDS: demersal fish, Favonigobius gymnauchen, feeding habits, food web, Repo-
mucenus spp., sandy beach, Seto Inland Sea, Tarphops oligolepis.

INTRODUCTION the density of mysids was low and shrimp were

abundant.12,23
Investigations of demersal fish have revealed that In contrast, Favonigobius gymnauchen,6,24 Repo-
flounder, gobies and dragonets predominate at mucenus spp.6,25–27 and Tarphops oligolepis6,28,29
sandy beaches in the temperate waters of Japan.1–12 mainly consumed small crustaceans (mysids,
Because Japanese flounder Paralichthys oliva- copepods, gammarids and decapods) and poly-
ceus, a dominant demersal fish, is extremely chaetes, but also preyed on newly settled Japanese
important to commercial fisheries, a large amount flounder at beaches where the flounder are abun-
of information on the feeding habits of this species dant.6,7 Daily growth rate of flounder is influenced
has accumulated.2,12–23 At sandy beaches of Japan by density of mysids.17 The results suggest that
Sea, the East China Sea,2,13–20 and the Pacific these three dominant species might influence the
Ocean21,22, where mysids are abundant, the domi- growth rate and mortality of larval and juvenile
nant item of prey in the diet of settled small flounder. However, scant information on the feed-
Japanese flounder (smaller than approximately ing ecology of these dominant fish species because
50 mm in total length) is mysids. However, at of less commercial importance makes further anal-
beaches in the central Seto Inland Sea, flounder fed yses difficult.
chiefly on mysids and crangonid shrimp, because The present paper describes the feeding ecol-
ogy of F. gymnauchen, Repomucenus spp. and
T. oligolepis at a sandy beach in Hiuchi-Nada, the
*Corresponding author: Tel: 81-87-843-6511. central Seto Inland Sea, Japan. In addition, the
Fax: 81-87-841-8133. Email: ky0554@pref.kagawa.lg.jp interspecific relationships among the dominant
Received 24 December 2004. Accepted 20 June 2005. demersal fish are discussed.
Feeding ecology of dominant fish at a beach FISHERIES SCIENCE 1333

MARERIALS AND METHODS measured to the nearest 0.1 mg (AEX-200B, Shi-

madzu Corporation, Japan). Body lengths (mysids
Field sampling and shrimp: from the level of the posterior margin
of the orbit to the tip of the telson; gammarids:
Ohama Beach (34°13.6′N, 133°36.5′E), which is from the rostrum to the tip of the tellson) of undi-
approximately 300 m in length, is located on the gested mysids, crangonid shrimp and gammarids
island of Shikoku, Japan, facing westward to the were measured to the nearest 0.1 mm under a
central Seto Inland Sea. The bottom sediment stereomicroscope.
consists of fine sand (median particle diameter:
0.2 mm).30 The surface water temperature from
May to August (the investigation period) ranged Data analysis
from approximately 17–31°C.23 Further details of
the sampling site are given in Yamamoto et al.12 The relative importance (%IRI)31 of a prey item in
Sampling surveys were carried out for 15 days the diet of the three dominant fish species was
during the period from May to August in 1999, 2000 determined from the product of percentage fre-
and 2001.30 Samples were collected with a beam quency of occurrence, percentage number and
trawl (net mouth 0.3 m high and 2 m wide, mesh percentage weight. An IRI value for each prey item
size 2.1 mm).30 The beam trawl was towed by a boat was calculated as follows:
along 200 m of the shoreline at water depths of
IRIi = (%nni + %wwi) × %Fi,
between 1.1 and 4.9 m during the day. The samples
were immediately fixed in 10% formalin. The sam- where %nni is the number of each prey item i as a
pling procedure is described in more detail in percentage of the total number of all prey items
Yamamoto et al.23 identified, %wwi is the percentage in wet weight of
each prey item i to the total wet weight of all prey
items identified, and %Fi is the frequency of occur-
Sample treatments rence for each prey item i in the total number of
guts examined. The IRI of each prey item was stan-
Fish, mysids and crangonid shrimp were sorted dardized to %IRI:
from the samples in the laboratory. Numbers of
%IRIi = 100 × IRIi/ΣIRIi.
individuals and wet weights in each species were
measured and, subsequently, these samples were Diet breadth was analyzed with the Simpson diver-
preserved in 70% ethanol. Two species of dragon- sity index,32 because this index, 1 − D, is not only
ets, Repomucenus richardsonii and Repomucenus unbiased, but also has a small standard deviation
beniteguri, were treated as Repomucenus spp. in small samples.33 The Simpson diversity of each
because of difficulty in the identification of small predator was computed as follows:
fish.
1 − D = 1 − ΣNi(Ni − 1)/N(N − 1),
Total length (TL) and upper jaw length (UJL: lin-
ear distance from anteriormost point of premaxilla where Ni is the number of each prey item i and N is
to posteriormost point of maxilla) or mouth width the total number of all prey items.
(MW: linear distance between right and left most To assess preference, a selective index, Chesson’s
posterior points of maxilla) of F. gymnauchen, α selectivity for main prey items, was calculated for
Repomucenus spp. and T. oligolepis were measured each day.34 The selectivity, αi, was as follows:
to the nearest 0.1 mm with calipers. When more
αi = (ni/ri)/Σ(ni/ri), i = 1, . . . m,
than 10 individuals of F. gymnauchen and Repo-
mucenus spp. were obtained in a day, we measured where ni and ri are the proportions of the number
10 individuals (>20 mm in TL) chosen at random of prey i in the diet and in the environment, respec-
from the samples. Additionally, total length and tively. The proportions of prey items in the
upper jaw length of Japanese flounder P. olivaceus environment were calculated using the samples
collected with these three fish were measured to collected by a sledge-net (net mouth 0.3 m high
compare mouth size between flounder and and 0.6 m wide, mesh size 0.76 mm).23 This index
T. oligolepis. varies between 0 and 1 with values greater than 1/
Gut contents (F. gymnauchen and T. oligolepis: m indicating a positive preference, and those less
stomach; Repomucenus spp. fore quarter of gut) than 1/m indicating a negative preference. The
of the specimens were classified into major taxa value of αi that represents neutral preference is 0.20
and mysids were identified to the species level for m = 5.
under a stereomicroscope. Numbers of each prey Dietary overlaps among dominant species in-
item were counted and their wet weights were cluding Japanese flounder at Ohama Beach were
1334 FISHERIES SCIENCE M Yamamoto and O Tominaga

measured for each day with Pianka’s O index,35 Of which, F. gymnauchen, Repomucenus spp.,
which was as follows: P. olivaceus and T. oligolepis comprised 64.7% of
the total number and 95.6% of the total weight
O12 = Σp1ip2iΣ(p1i2p2i2)(1/2),
of fish collected. Because sizes of R. ercodes and
where p1i and p2i are the proportions of the number H. japonica were small (major size: <15 mm in
of prey item i in predator 1 and predator 2, respec- TL), the stomachs of two species were not
tively. This index ranges from 0 (no overlap) to 1 examined. Although large dragonets (larger
(complete overlap). Here, we used the stomach than approximately 40 mm in TL) were composed
contents of the Japanese flounder (9.80–75.95 mm of R. richardsonii and R. beniteguri (mostly
TL) collected with the three dominant species.23 R. beniteguri), smaller individuals were not identi-
fied. Developmental stages of all of specimens
of F. gymnauchen, Repomucenus spp. and
RESULTS T. oligolepis were juveniles or adults.
Besides fish, 32,235 mysids (wet weight: 166.1 g)
Abundance of fish, mysids and crangonid shrimp and 120,491 crangonid shrimp Crangon uritai
(3591.9 g) were also collected. The dominant
A total of 4460 individual fish, including more than mysids were Iiella ohshimai and Nipponomysis
25 species in 15 families, were collected by 35 ornata.
operations of the 2-meter beam trawl at Ohama
Beach from May to August in 1999, 2000 and
2001 (Table 1). In terms of individual numbers, Gut contents
F. gymnauchen (percentage in number: 50.7%),
Rudarius ercodes (26.1%), Repomucenus spp. Of 150 individuals of F. gymnauchen examined,
(7.0%), Heteromycteris japonica (4.7%), P. olivaceus which ranged from 20.7 to 71.4 mm in TL, 76.7%
(4.3%) and T. oligolepis (2.7%) were dominant. contained food (Table 2). In Repomucenus spp.,

Table 1 Demersal fish collected by a 2-meter beam trawl at Ohama Beach from May to August in 1999, 2000 and
2001
Family Species Number of individuals Weight (g)
Clupeidae Konosirus punctatus 3 0.1
Syngnathidae Syngathus schlegeli 3 0.2
Hippocampus mohnikei 3 0.4
Scorpaenidae Sebastes inermis 1 1.8
Inimicus japonicus 2 6.8
Congiopodidae Hypodytes rubripinnis 18 3.9
Platycephalidae Platycephalidae spp. 80 3.2
Cottidae Pseudoblennius cottoides 2 1.6
Sparidae Pagrus major 7 0.5
Acanthopagrus schlegeli 31 2.2
Labridae Halichoeres poecilopterus 3 94.2
Blenniidae Scartella emarginata 2 0.1
Callionymidae Repomucenus spp. 313 214.7
Gobiidae Sagamia genetonema 5 2.0
Favonigobius gymnauchen 2262 5612.7
Tridentiger trigonocephalus 2 0.1
Gobiidae spp. 24 0.4
Paralichthyidae Paralichthys olivaceus 190 73.2
Tarphops oligolepis 122 72.4
Soleidae Heteromycteris japonica 211 122.0
Pseudaesopia japonica 1 –
Monacanthidae Rudarius ercodes 1166 33.9
Thamnaconus modestus 1 0.6
Tetraodontidae Takifugu niphobles 2 0.1
Unidentified 6 0.7
Total 4460 6247.9

–, less than 0.05 g.

Table 2 Feeding incidence, percentage in number (%nn), percentage in weight (%ww), percentage frequency of occurrence (%F) and percentage in index of rel-
ative importance (%IRI) of prey items of the three dominant fish species collected at Ohama Beach in 1999, 2000 and 2001
Favonigobius gymnauchen Repomucenus spp. Tarphops oligolepis
Number 150 90 122
Range of total length (mm) 20.7–71.4 20.1–107.4 9.0–75.4
Feeding Incidence (%) 76.7 84.4 87.7
Contents %nn %ww %F %IRI %nn %ww %F %IRI %nn %ww %F %IRI
Polychaeta 1.0 9.3 3.3 0.9
Ostracoda [Vargula hilgendorfii] 0.9 0.1 1.3 –
Feeding ecology of dominant fish at a beach

Copepoda 2.3 0.4 5.7 0.2

Mysidae 69.2 48.8 49.3 67.9 68.4 44.0 46.7 71.1 84.3 71.5 67.2 92.6
(Iiella ohshimai) 30.2 23.0 18.7 30.5 18.6 14.1 13.3 11.3 18.8 17.5 24.6 14.0
(Nipponomysis ornata) 22.0 16.3 17.3 20.4 11.3 7.5 12.2 5.9 54.4 47.6 45.9 73.4
(Anisomysis ijimai) 1.8 1.4 2.0 0.2 1.3 1.3 3.3 0.2 2.5 2.0 6.6 0.5
(Mysidae fragments) 15.3 8.2 23.3 16.8 37.2 21.1 35.6 53.7 8.5 4.4 23.8 4.8
Gammaridea 5.3 2.3 8.0 1.9 7.0 1.4 10.0 2.2 5.2 3.7 9.8 1.4
Caprellidea 0.6 0.3 1.3 – 2.0 0.5 1.1 0.1
Flabelifera 0.7 0.2 2.2 –
Tanaidacea 0.6 0.0 0.7 –
Cumacea 0.3 0.4 0.8 –
FISHERIES SCIENCE

Penaeidea 0.3 0.2 0.7 –

Crangon uritai 14.1 41.2 15.3 26.0 11.0 27.2 16.7 16.5 3.7 15.5 12.3 3.7
Callianassidae 0.3 0.6 1.1 – 2.0 3.5 9.0 0.8
Brachyura 1.8 1.3 2.0 0.2 1.3 0.8 4.4
Pisces [teleostei] 1.2 1.6 2.7 0.1 0.7 4.6 2.2 0.2 0.3 1.0 1.6 –
(Repomucenus spp.) 0.3 1.5 1.1 0.1 0.1 0.1 0.8 –
(Favonigobius gymnauchen) 0.9 1.3 2.0 0.1 0.3 3.1 1.1 0.1 0.1 0.9 0.8 –
(Tarphops oligolepis) 0.3 0.3 0.7 –
Egg of Hemiramphus sajori 0.6 0.4 0.7 – 1.7 1.0 4.4 0.3
Unidentified 5.6 4.0 12.7 3.7 6.0 10.4 20.0 8.5 2.0 4.1 13.1 1.3

–, less than 0.05%.

1335
1336 FISHERIES SCIENCE M Yamamoto and O Tominaga

84.4% of 90 individuals examined contained food. of T. oligolepis at the same total length of larger
Although this species (n = 313) ranged from 20.1 to than 15.0 mm.
107.4 mm in TL, few individuals ranging from 30 to
40 mm were present (n = 5). The feeding incidence
(percentage of stomach with food) of 122 individ- Diet shift with growth
uals of T. oligolepis ranging from 9.0 to 75.4 mm TL
was 87.7%. The feeding incidences of these three The reverse trends of diet shift with growth were
species were not significantly correlated with body observed in mysids and crangonid shrimp in
size (Spearman’s rank correlation coefficient test, F. gymnauchen (Fig. 1). That is, the weight com-
P > 0.3). position of mysids decreased with total length,
The main prey items in the three dominant whereas that of crangonid shrimp increased. The
fish, F. gymnauchen, Repomucenus spp., and diet of small Repomucenus spp. was composed
T. oligolepis, were small crustaceans (mostly of mysids and gammarids; and with growth,
mysids, gammarids and decapods) (Table 2). Set- the percentage of crangonid shrimp and fish
tled larval and juvenile Japanese flounder were increased. The stomach contents of T. oligolepis
not observed in the guts of any fish examined. consisted primarily of mysids of all sizes, and no
F. gymnauchen mainly consumed mysids (%IRI: ontogenetic shift in the feeding habits was
67.9%) and crangonid shrimp (26.0%). The %IRI observed.
of I. ohshimai (30.5%) was higher than that of The size relationships among the three fish and
N. ornata (20.4%). Other prey items were gam- their major prey items are shown in Figure 2. In
marids, brachyurans, fish and eggs of Hemiram- F. gymnauchen up to 40 mm, Repomucenus spp.
phus sajori, but the %IRI of these prey items were up to 50 mm and T. oligolepis up to 30 mm, the
very low. In the gut of Repomucenus spp., mysids body length (BL) of mysids consumed increased
and crangonid shrimp predominated, and sand with the total length of the predators. The maxi-
grains were often observed (frequency of occur- mum prey size of F. gymnauchen, Repomucenus
rence: approximately 40%). T. oligolepis mainly spp. and T. oligolepis were 9.3 mm (mean: 5.9 mm),
consumed mysids, particularly N. ornata. 10.0 mm (6.4 mm) and 8.3 mm (4.7 mm), respec-
The dietary diversity (Simpson diversity index, tively. The maximum size of mysids (approximately
1 − D) values of F. gymnauchen and Repomucenus 9 mm BL) was consistent with that of mysids in the
spp., of 0.81 and 0.80, respectively, were higher environment.23 The maximum size of crangonid
than that of T. oligolepis, 0.66. The dietary diversity shrimp increased with the growth of the predators,
of Japanese flounder P olivaceus calculated (not and the mean body lengths of shrimp observed in
tabled) was 0.77. the guts of F. gymnauchen, Repomucenus spp. and
T. oligolepis were 8.3 mm (range: 5.2–11.8 mm),
8.5 mm (5.0–11.3 mm) and 7.7 mm (3.8–10.2 mm),
Relationship between mouth size and fish length respectively. In gammarids, it was not possible to
analyze the size relationship between predator and
Mouth size (upper jaw length and/or mouth width) their prey because of lack of prey body length.
was significantly correlated with total length However, the mean body length of the gammarids
(Table 3). The slopes of the regression lines were eaten by F. gymnauchen, Repomucenus spp. and
significantly different between P. olivaceus and T. oligolepis were 2.3 mm (range: 0.8–3.8 mm),
T. oligolepis (analysis of covariance, P < 0.01), and 1.3 mm (0.8–1.8 mm) and 2.2 mm (0.8–4.2 mm),
the mouth size of P. olivaceus was larger than that respectively.

Table 3 Relationships between upper jaw length (UJL: in mm)/mouth width (MW: in mm) and total length (TL: in mm)
of Favonigobius gymnauchen, Repomucenus spp., Tarphops oligolepis and Paralichthys olivaceus. Regression equations
and significance levels were shown
Equation Degrees of freedom r2 F P
Favonigobius gymnauchen UJL = 0.0523 TL + 0.209 49 0.81 207 <0.01
MW = 0.0913 TL − 0.00068 49 0.83 251 <0.01
Repomucenus spp. MW = 0.0646 TL + 0.149 49 0.94 766 <0.01
Tarphops oligolepis UJL = 0.0678 TL + 0.600 29 0.90 273 <0.01
Paralichthys olivaceus UJL = 0.103 TL − 0.0704 29 0.97 1034 <0.01
Feeding ecology of dominant fish at a beach FISHERIES SCIENCE 1337

Prey selection than neutral preference (Wilcoxon signed-ranks

test, P < 0.05), although those for the other mysids,
The selectivity, Chesson’s αi, of the major prey I. ohshimai (0.303) and Anisomysis ijimai (0.402)
items is shown in Table 4. In F. gymnauchen, only and crangonid shrimp (0.221) were greater than
the selectivity of N. ornata was significantly higher neutral preference (0.20). Repomucenus spp.
selected N. ornata significantly (P < 0.05), and con-
sumed other prey items at random (P > 0.1).
Favonigobius gymnauchen
T. oligolepis had both active selectivity of N. ornata
0 13 25 35 33 9 0
100 and negative selectivity of gammarids and
Composition (%)

80 crangonid shrimp (P < 0.01). The trends of the

selectivity of Repomucenus spp. and F. gym-
60
nauchen were similar to each other, whereas
40 T. oligolepis differed from the other two fish, with a
20 higher preference for N. ornata. T. oligolepis had
the most selective and specialized diet among the
0
0 20 30 40 50 60 70 72 three fish.
Repomucenus spp.
0 14 3 1 20 20 18
100 Diet overlap
Composition (%)

80
60 All of the mean overlap values, Pianka’s O index, for
pair species from F. gymnauchen, Repomucenus
40
spp. T. oligolepis and P. olivaceus ranged from
20 0.48 to 0.66, although the individual values greatly
0 fluctuated (Fig. 3). Those values for pairs of
0 20 30 40 50 60 70 76 F. gymnauchen and Repomucenus spp., T. oligolepis
Tarphops oligolepis and P. olivaceus were relatively higher.
42 25 17 13 4 4 2
100
Composition (%)

80 DISCUSSION
60
Gut contents and prey selection
40
20 The feeding incidences of the three dominant
0 fish, F. gymnauchen, Repomucenus spp. and
0 20 30 40 50 60 70 110 T. oligolepis, varied between 76.7 and 87.7% during
Total length (mm) the day. The major prey items were the epifaunal
Mysidae Callianassdae mysids N. ornata and A. ijimai, and the sand-
burrowing mysid I. ohshimai. I. ohshimai, Archae-
Gammaridea Pisces omysis kokuboi and A. japonica, belonging to the
subfamily Gastrosaccinae, burrow under the sand
Crangonidea Others during the day and swim up to the water column
during the night.36,37 The stomach content index
Fig. 1 Wet weight composition of stomach contents (SCI) of many species of juveniles in the surf-zone
by size for Favonigobius gymnauchen, Repomucenus of the Pacific Ocean, which feed on the sand-
spp. and Tarphops oligolepis. Numerals above bars burrowing mysids A. kokuboi, was increased in
indicate the number of specimens examined. the night.22 Imada and Nanba24 observed that
Body length of prey (mm)

Favonigobius gymnauchen Repomucenus spp. Tarphops oligolepis

14 Mysidae
12
10 Crangonidea
Fig. 2 Size relationship between
8 Gammaridea
major prey taxa (body length) 6
and predator (total length) for 4
Favonigobius gymnauchen, Re- 2
0
pomucenus spp. and Tarphops 10 20 30 40 50 60 70 80 10 20 30 40 50 60 70 80 10 20 30 40 50 60 70 80
oligolepis. Total length of predator (mm)
1338 FISHERIES SCIENCE M Yamamoto and O Tominaga

Table 4 Selectivity (Chesson’s αi) of the major food items for Favonigobius gymnauchen, Repomucenus spp. and Tar-
phops oligolepis at Ohama Beach in 1999, 2000 and 2001. The neutral preference was 0.20
Favonigobius
gymnauchen Repomucenus spp. Tarphops oligolepis
Contents Mean (range) Mean (range) Mean (range)
Mysidae
Iiella ohshimai 0.303 (0.004–0.899) 0.242 (0–0.84) 0.156 (0.006–0.558)
Nipponomysis ornata 0.576 (0.083–1) 0.524 (0.247–0.732) 0.747 (0.241–1)
Anisomysis ijimai 0.402 (0.153–0.847) 0.460 (0.180–0.740) 0.203 (0–0.541)
Gammaridea 0.093 (0–0.217) 0.203 (0–1) 0.037 (0–0.178)
Crangon uritai. 0.221 (0–1) 0.150 (0–0.637) 0.015 (0–0.089)

very low.6 These results indicate that this species is

Mean: 0.66 carnivorous and a benthos feeder.
Favonigobius gymnauchen
vs Repomucenus spp. Previous studies on the feeding ecology of drag-
0.50 onets have suggested that the numerical composi-
Favonigobius gymnauchen
vs Traphops oligolepis tion of the gut contents was similar to that of the
0.48
prey in the environment.25–27 Eda reported that the
Repomucenus spp. numerically dominant prey items of adult dragon-
vs Traphops oligolepis
0.56
ets, Repomucenus rechardsonii and Repomucenus
Favonigobius gymnauchen valenciennei, were bivalves and polychaetes in
vs Paralichthys olivaceus
Omura Bay, Nagasaki Prefecture, Japan.26 The main
0.56 foods of R. rechardsonii at Yanagihama Beach6 and
Repomucenus spp.
vs Paralichthys olivaceus in the Seto Inland Sea25 were decapods and poly-
Traphops oligolepis
0.66 chaetes, and cumacea, ostracods and gastropods,
vs Paralichthys olivaceus respectively. In the diet of Calliurichthys japonicus
and R. huguenini in Tosa Bay, Japan, polychaetes
and gammarids were abundant.27 In our study
0 0.2 0.4 0.6 0.8 1 area, Repomucenus spp. mainly consumed mysids
Overlap index and crangonid shrimp, which were the predomi-
Fig. 3 The overlap values, Pianka’s O index, between
nant prey items in the field.23 The present result is
pair of two fish species from Favonigobius gym- consistent with the previous studies.
nauchen, Repomucenus spp., Tarphops oligolepis and Tarphops oligolepis actively selected the epifau-
Paralichthys olivaceus. Open and closed circles indi- nal mysids N. ornata. Yamamoto et al. found that
cate the individual values and the mean values, the Japanese flounder P. olivaceus also preferred
respectively. Data on the Japanese flounder after N. ornata to sand-burrowing mysids.23 The impor-
Yamamoto et al.23 tance of mysids as food for Japanese flounder,
however, was relatively low in this study area com-
pared to the Japan Sea because of the low density
of mysids,12,23 although settled small flounder
F. gymnauchen mainly consumed blue-crab Portu- generally fed chiefly on mysids.13–22 In contrast,
nus trituberculatus at midnight in a laboratory T. oligolepis in the present research field mainly
experiment. Therefore, feeding incidences of the consumed mysids of all sizes. The results indicate
three fish species might be higher in the night than the preference for mysids to be higher among
in the day. T. oligolepis than Japanese flounder irrespective of
Yamamoto et al.23 reported that mysids, the densities of mysids in the environment.
I. ohshimai and N. ornata, and crangonid shrimp
Crangon spp. were abundant in this study area.
F. gymnauchen mainly consumed mysids, Diet overlap
crangonid shrimp and small fish. At Yanagihama
Beach in Nagasaki Prefecture, the eastern East The mean overlap values among F. gymnauchen,
China Sea, the main prey items of F. gymnauchen Repomucenus spp., T. oligolepis and P. olivaceus
were decapods, gammarids, polychaetes and were relatively high, because the four fish fed
mysids, whereas the importance of fish as food was chiefly on small crustaceans (mostly mysids and
Feeding ecology of dominant fish at a beach FISHERIES SCIENCE 1339

Prey Predator values were very low. Newly settled Japanese floun-
Mysid
Goby der, however, was not found in the guts of the fish
Favonigobius gymnauchen
Iiella ohshimai examined and predation pressure on larval and
Dragnet
juvenile Japanese flounder by the three fish might
Repomucenus spp. not be high in the present study field.
Mysid
Nipponomysis ornata
Flounder
Tarphops oligolepis ACKNOWLEDGMENTS
Crangonid shrimp
Crangon uritai Japanese flounder We are grateful to Mr Y. Kameno, Ohama Fisheries
Paralichthys olivaceus
Cooperative Association, Mr Y. Kimura, Fukui
%IRI 5–10 : 10–30: 30–60: 60–: Prefectual University and the staff of the Kagawa
Overlap index 0.4–0.5: 0.5–0.6: 0.6–0.7: Prefectural Fisheries Experimental Station for
Fig. 4 Schematic diagram of food habits of Favonigo-
assistance in sampling. We thank Dr J. Shoji, Kyoto
bius gymnauchen, Repomucenus spp., Tarphops oligole- University and Dr Y. Natsukari, Nagasaki Univer-
pis and Paralichthys olivaceus. Diet overlaps and sity, for their useful information on feeding ecology
predator-prey relationship are shown. Data on the of fish. We are also grateful to Dr Y. Hanamura,
Japanese flounder after Yamamoto et al.23 National Research Institute of Fisheries and Envi-
ronment of Inland Sea, for useful advice on
crangonind shrimp. This work was partly sup-
ported by the Fisheries Agency of Japan. Thanks
crangonid shrimp) (Fig. 4). The mean body length are also due to the referees who made constructive
of mysids, crangonid shrimp and gammarids in the comments on the manuscript.
stomach of Japanese flounder collected in this
study area was 5.0 mm (range: 1.5–10.5 mm),
8.0 mm (3.7–14.5 mm) and 3.3 mm (1.8–4.0 mm), REFERENCES
respectively (unpubl. data, M. Yamamoto 2004),
showing the size of major prey items of the floun- 1. Senta T, Amarullah MH, Yasuda M. Invitation to the study
der to be similar to that of the three dominant fish juvenile fishes occurring in surf zone. In: Go YB (ed). Pro-
ceedings of Symposium on Development of Marine Resources
species. From these results, we propose that there
and International Cooperation in the Yellow Sea and the East
is interspecific competition among the four fish.
China Sea. Mar. Res. Inst. Cheju Nat. University of Korea,
Additionally, there might be exploitative competi- Cheju. 1988; 131–146.
tion between crangonid shrimp Crangon uritai 2. Minami T, Nakabo T, Uozumi Y, Kiyono S. Benthic fish
and the four fish, because the shrimp consumed fauna off the river month of Yura, Wakasa Bay. Showa 50
mysids (mostly N. ornata) near the study field.38 A Nendo Annual Report of Kyoto Prefectural Fisheries Expri-
more detail study on daily rations and feeding mental Station. 1977; 74–100.
rhythms of these species is necessary to estimate 3. Yusa T. Life history and environment of flatfish (Japanese
the availability and carrying capacity of mysids for flounder Paralichthys olivaceus and Pleuronectidae spp.).
Japanese flounder. Aquaculture Sci. 1979; 16: 33–45.
4. Hashimoto H, Fukuura Y, Go A. Demersal fishes in the
waters of northern coast of Fukui Pref. I, Composition of
fish species, length-weight relationship and ovary matura-
Predation by the three dominant fish on larval tion on some fishes. J. Fac. Appl. Biol. Sci. Hiroshima Univ.
and juvenile Japanese flounder 1985; 24: 43–48.
5. Amarullah MH, Senta T. The R.-H push-net, a gear for study
Generally, the major cause of death among larval of juvenile flatfishes along the beach. Bull. Fac. Fish.
and juvenile Japanese flounder is predation by fish Nagasaki Univ. 1989; 65: 9–14.
and crustaceans.18 Noich et al. have reported that 6. Noichi T, Kusano M, Ueki D, Senta T. Feeding habit of fishes
newly settled flounder were preyed upon by many eating settled larval and juvenile Japanese flounder (Paral-
fish, including the three species, F. gymnauchen, ichthys olivaceus) at Yanagihama Beach, Nagasaki Prefec-
Repomucenus spp. and T. oligolepis at Yanagihama ture. Bull. Fac. Fish. Nagasaki Univ. 1993; 73: 1–6.
7. Noichi T, Kusano M, Kanbara T, Senta T. Predation by fishes
Beach, in Nagasaki Prefecture.6,7 However, far more
on larval and juvenile Japanese flounder at Yanagihama
serious predators for Japanese flounder, Trachi- Beach, Nagasaki, Japan. Nippon Suisan Gakkaishi 1993; 59:
nocephalus myops, Pseudoblennius percoides, 1851–1855.
P. cottoides and Chelidonichthy spinsus, were not 8. Mori K. Ecological study on the fishes of Yusa Bay, the Japan
abundant in this study area (Table 1).6,7 The three Sea. Bull. Natl. Res. Inst. Fish. Sci. 1995; 7: 277–388.
dominant fish preyed on small F. gymnauchen, 9. Omi H, Nabeshima Y, Kusakabe T. Seasonal change of spe-
Repomucenus spp. and T. oligolepis, but the %IRI cies of larval and juvenile fishes occurrence in the estuary in
1340 FISHERIES SCIENCE M Yamamoto and O Tominaga

the innermost area of Osaka Bay. Bull. Osaka Pref. Fish. Exp. 23. Yamamoto M, Makino H, Kobayashi J, Tominaga O. Food
Stn. 2001; 13: 61–72. organisms and feeding habits of larval and juvenile Japa-
10. Hibino M, Ohta T, Kinoshita I, Tanaka M. Fish larvae and nese flounder Paralichthys olivaceus at Ohama Beach in
juveniles occurring in the littoral zone of a tidal flat, in the Hiuchi-Nada, the central Seto Inland Sea, Japan. Fish. Sci.
bottom of Ariake Bay. Japan. J. Ichthyol. 2002; 49: 109–120. 2004; 70: 1098–1105.
11. Omi H, Kusakabe T, Nabeshima Y. Fishes occurring on arti- 24. Imada R, Nanba T. Preliminary experiments on predation of
ficial beach of southern coast in Osaka Bay. Bull. Osaka Pref. the blue-crab by using the common goby, Favonigobius
Fish. Exp. Stn. 2003; 14: 57–70. gymnauchen. Aquaculture Sci. 1981; 29: 185–189.
12. Yamamoto M, Makino H, Kagawa T, Tominaga O. Occur- 25. Kakuda S, Gushima K, Nakai K, Ogushi Y, Murakami Y. On
rence and distribution of larval and juvenile Japanese floun- the food habit of the dragonets, genus Callionymus I. Food
der Paralichthys olivaceus at sandy beaches in eastern habit of Callionymus punctatus Langsdorff. J. Fac. Appl.
Hiuchi-Nada Sea, central Seto Inland Sea, Japan. Fish. Sci. Biol. Sci. Hiroshima Uni. 1979; 18: 75–83.
2004; 70: 1089–1097. 26. Eda H. Life history of callioymid fishes in Omura Bay,
13. Imabayashi H. Feeding activities of juvenile and young bas- Nagasaki, Japan. PhD Thesis. Nagasaki University, Nagasaki.
tard halibut, Paralichthys olivaceus, in the biotic commu- 1994.
nity-I, Intraspecific relationships in the population. Nippon 27. Gonzales BJ, Taniguchi N, Okamura O, Machida Y. Diet and
Suisan Gakkaishi 1980; 46: 427–435. feeding behavior of two dragonets Calliurichthys japonicus
14. Minami T. The early life history of a flounder Paralichthys and Repomucenus hugenini in Tosa Bay, southern Japan.
olivaceus. Nippon Suisan Gakkaishi 1982; 48: 1581–1588. Fish. Sci. 1996; 62: 902–908..
15. Hirota Y, Koshiishi Y, Naganuma N. Size of mysids eaten by 28. Minami T. The early life history of a flounder Tarphops oli-
juvenile flounder Paralichthys olivaceus and diurnal change golepis. Nippon Suisan Gakkaishi 1983; 49: 533–539.
of its feeding activity. Nippon Suisan Gakkaishi 1990; 56: 29. Hashimoto H, Fukuura Y, Go A. Demersal fishes in the
201–206. waters of northern coast of Fukui Pref. II. Food habit of Sau-
16. Subiyanto, Hirata I, Senta T. Larval settlement of the Japa- rida undosquamis, Upeneus bensasi and three Het-
nese flounder on sandy beaches of the Yatsushiro Sea, erosomes. J. Fac. Appl. Biol. Sci. Hiroshima Univ. 1985; 24:
Japan. Nippon Suisan Gakkaishi 1993; 59: 1121–1128. 49–55.
17. Fujii T, Noguchi M. Feeding and growth of Japanese floun- 30. Yamamoto M, Makino H. Estimation of catch efficiency of
der (Paralichthys olivaceus) in the nursery ground. In: the two-meter beam trawl for juvenile Japanese flounder in
Watanabe Y, Yamashita Y, Oozeki Y (eds). Survival Strategies Hiuchi-nada Sea. Saibai Giken 2002; 30: 31–35.
in Early Life Stages of Marine Resources. A. A. Balkema, Rot- 31. Cortes E. A critical review of method of studying fish feeding
terdam. 1996; 141–151. based on analysis of stomach contents: application to elas-
18. Noichi T. Early life history. In: Minami T, Tanaka M (eds). mobranch fishes. Can. J. Fish. Aquat. Sci. 1997; 54: 726–738.
Biology and Stock Enhancement of Japanese Flounder. 32. Simpson EH. Measurement of diversity. Nature 1949; 163:
Kouseisha-kouseikaku, Tokyo. 1997; 25–40. 688.
19. Tanaka Y, Tominaga O, Tsusaki T, Nogami K, Fushimi H, 33. Lande R. Statistics and partitioning of species diversity, and
Tanaka M. UJNR experimental release of Japanese flounder similarity among multiple communities. Oikos 1996; 76: 5–
at Wada Bay: feeding habit of wild and released juveniles. 13.
Bull. Natl. Res. Inst. Aquacult. Suppl. 1999; 1: 125–133. 34. Chesson J. Measuring preference in selective predation.
20. Furuta S. Seasonal changes in abundance, length distribu- Ecology 1978; 59: 211–215.
tion, feeding condition and predation vulnerability of juve- 35. Pianka ER. The structure of lizard communities. Annu. Rev.
nile Japanese flounder, Paralichthys olivaceus, and prey Ecol. Syst. 1973; 4: 53–74.
mysids density in the Tottori coastal area. Nippon Suisan 36. Takahashi K, Kawaguchi K. Diel and tidal migrations of the
Gakkaishi 1999; 65: 167–174. sand-burrowing mysids, Archaeomysis kokuboi, A. japonica
21. Yamada H, Sato K, Nagahora S, Kumagai A, Yamashita Y. and Iiella oshimai, in Otsuchi Bay, northeastern Japan. Mar.
Feeding habit of the Japanese flounder Paralichthys oliva- Ecol. Prog. Ser. 1997; 148: 95–107.
ceus in Pacific coastal waters of Tohoku district, northeast- 37. Takahashi K. Distribution and migration of mysids in the
ern Japan. Nippon Suisan Gakkaishi 1998; 64: 249–258. surf-zone of a sandy beach. Gekkan Kaiyo Gougai 2001; 26:
22. Takahashi K, Hirose T, Kawaguchi K. The importance of 98–107.
intertidal sand-burrowing peracarid crustaceans as prey for 38. Hanamura Y, Matsuoka M. Feeding habit of the sand
fish in the surf-zone of a sandy beach in Otsuchi Bay, north- shrimp, Crangon uritai Hayashi & Kim, 1999, in the central
eastern Japan. Fish. Sci. 1999; 65: 856–864. Seto Inland Sea, Japan. Crustaceana 2003; 76: 1017–1023.