01_device Feeding Relationships of Dominant Fish Species 35Science Diliman (January-June 2007) 19:1, 35-46 *Corresponding author INTRODUCTION Studies on diet composition are important in community ecology because the use of resources by organisms has a major influence on population interactions within a community. Studies of species resource requirements have been used in attempts to understand factors controlling the distribution and abundance of organisms (Ross, 1986). Information about the food habits of fishes is useful in defining predator-prey relationships because predator pressure has a pervasive influence on the evolution of a population. Data on different food items consumed by fish may eventually result in identification of stable food preference and in creation of trophic models as a tool to understand complex ecosystems (Lopez-Peralta and Arcila, 2002; Bachok, 2004). Feeding Relationships of Dominant Fish Species in the Visayan Sea Annie Trixie L. Mequila* and Wilfredo L. Campos Oceanbio Laboratory, Division of Biological Sciences, College of Arts and Sciences, University of the Philippines in the Visayas 5023 Miag-ao, Iloilo E-mail: atmequila@yahoo.com Telefax: 033 3159271 ABSTRACT This study examines the diet composition of 15 fish species belonging to 11 families in the Visayan Sea. A total of 50 prey items were identified from 323 stomachs examined. Cluster analysis was used to assess similarities in the composition of prey items in individuals of similar sizes but of different species, and in individuals of the same species but of different size categories. The results showed high similarity among individuals of the same species and low similarity between different species of similar size categories. Two major clusters were formed, showing generalist feeding (high niche widths) and trophic segregation (low food niche overlap). These suggest that food resources in the Visayan Sea are not limited and that a wide range of habitats is available to the fish community. Keywords: Stomach analysis, food niche overlap, food niche width, diet composition The Visayan Sea is considered as one of the country’s most productive fishing grounds, with an average annual catch of about 200,000mt (BFAR, 2002). However, recent reports show that there has been a decline in small scale and commercial fish production in the area due to increased fishing pressure (Hermes et al., 2004). Most of the studies that have been conducted in the Visayan Sea deal with stock assessment of commercially important species. Non- commercially important species have received less attention and their biology is, to a large extent, poorly documented.This study provides empirical information on gut contents of the more abundant fish species caught by trawls in the Visayan Sea and examines the extent of feeding relationships among them. METHODS Sampling Area and Field Collection Visayan Sea is located in central Philippines between 11o and 12o N and 123o and 124o E, and covers a total area of 5,184 km2 (BFAR, 2002). It is bounded by Mequila and Campos 36 coastal areas of northeastern Panay, northern Negros Occidental, Bantayan Island, and Masbate (Fig. 1). Samples were collected during a research cruise in the Visayan Sea in July 2003 on board the RV DA-BFAR. A major activity of the cruise was a trawl survey covering 10 stations distributed within the study area. At each station, a single 1-hour tow was conducted during the daytime with the use of a two-seam trawl with a cod end mesh size of 5.08 cm. A cover net with a mesh size of 1 cm was attached to the cod end to examine gear selectivity (another study). The trawl was towed at an average speed of 4.5 knots. Specimens representing the entire size range of the more abundant fish species caught in 5 of the 10 stations were collected for stomach content analysis and preserved on board in 10% buffered seawater-formalin solution. For larger fish, an incision was made along the belly to allow the formalin to penetrate the stomach to minimize possible post mortem digestion. Laboratory Processing In the laboratory, fish were sorted and identified to species level. Prior to dissection, total length, standard length, and fork length of each individual were measured to the nearest 1mm using a ruler and total wet weight was measured to the nearest 0.01 g using an electronic balance. Stomachs were removed from the digestive tract and weighed to the nearest 0.01 g. The contents were removed and preserved in 10% buffered formalin for later analysis. After the removal of its contents, the stomach sac was weighed. Total stomach content was then calculated as the difference of stomach weight with contents minus the stomach sac. Stomach contents were examined under the dissecting microscope and identified to the lowest taxonomic level possible. Prey items were identified and grouped into major categories including: (a) microcrustaceans (amphipods, caperellids, cladocerans, cumaceans, euphausids, halocarids, isopods, mysids, Lucifer, and ostracods), (b) larval crustaceans (crab 122.5 123.0 10.5 11.0 11.5 12.0 12.5 PANAY NEG Celebes Sea Sulu Sea South China Sea Pacific Ocean N 122.5 123.0 10.5 11.0 11.5 12.0 12.5 PANAY NEG Celebes Sea Sulu Sea South China Sea South China Sea Pacific Ocean N Figure 1. Map of the Visayan Sea showing the 5 sampling stations where samples for stomach content analysis were collected during the Visayan Sea Project cruise in June 2003. Feeding Relationships of Dominant Fish Species 37 megalopa and zoea, shrimp mysis, and Cypris larvae), (c) benthic worms (annelid worms, ribbon worms, flatworms, gnathostomulids, and sipunculids), (d) echinoderms (brittle star, bipinnaria, and pentactula), and (e) urochordates (ascidians and larvaceans). Data Analysis Feeding intensity was determined by calculating the fullness index (FI) using the formula of Hureau (1969) (as given in del Norte-Campos, 1995): FI (%) = weight of stomach contents (g) x 100 body weight (g) Examination of stomach contents employed numerical analysis described by Windell and Bowen (1978), wherein food items present were identified and counted. The frequency of occurrence (FOC) of identified food items was calculated using the formula, FOC = no. of food item i x % identified total no. of food items where % identified = 100 - % unidentified. Unidentified materials were visually estimated as percentage of the total stomach content. The narrow size range of each fish species did not allow the categorization of size classes prior to the stomach content analysis. With the premise that prey differs with growth, sizing was done after prey identification using a mathematical tool. A two-step cluster analysis (using the program COMM, Piepenburg and Piatkowski, 1992) was done on FOC data, first, to determine if size categories (e.g. small, medium, and large) for each fish species could be identified on the basis of prey similarity. For example, “small” Selaroides leptolepis, Neopomacentrus filamentosus, and Leiognathus bindus were assigned with the same sizes (<100mm) although “small” N. filamentosus and L. bindus were of shorter lengths. The second step was done on the various size categories of the different species to determine similarities in gut contents with respect to size. Food niche overlap across length classes was computed using Schoener’s Index (as given in Salgado,et.al, 2004), SI xy = 1- 0.5 (∑  P xi – P yi  ) where P xi = relative frequency of food type i in the stomach of species x and P yi = relative frequency of food type i in the stomach of species y. The index ranges from 0, for entirely dissimilar diets, to 1 when the composition of the diets is identical but values >0.6 can be considered biologically significant (Wallace, 1981; Wallace and Ramey , 1983; Salgado, et al., 2004). Food niche width was also computed for each species using the Shannon-Wiener Index on data across all length classes using the formula, H’ = - ∑ (P i ) log (P i ) where P i = proportion of prey item i in the stomach of the predator. RESULTS AND DISCUSSION Fifteen species belonging to 11 families were examined in the study and their respective size groupings resulting from the initial cluster analysis are listed in Table 1. Some species did not form size classes due to their narrow length range. Out of 323 stomachs examined, only 8 were empty. A total of 50 prey items were identified excluding chyme (unidentified foods), rubble, sediment, and fish scales and bones (counted in the absence of fish in the sample). Shrimps were the most abundant prey in the stomachs of Pentapodus setosus, Upeneus asymmetricus, Plotosus lineatus, Saurida undusquamis, Parapercis alboguttata, and Lagocephalus. lunaris. Calanoids were abundant in N. filamentosus, Apogon notatus, and L. bindus; fish larvae for S. leptolepis, fish for Synodus variegatus, polychaetes for Leiognathus rivulatus, nematodes for Arothron manillensis, amphipods for Scolopsis affinis, and pelagic harpacticoids (mostly Microsetella) for Rastrelliger kanagurta. The assessment of possible dietary shifts in the different species showed that there was a general absence of variation in the pattern of prey preference with increase in size in all 15 of them. For example, L. bindus (Fig. 2a) showed overlapping size groupings while clear groupings were observed in S. variegatus (Fig. 2b). Mequila and Campos 38 Species N No. of non-empty Size Range Size Classes (TL in mm) Stomachs (TL in mm) Small Medium Large DEMERSAL F. Leiognathidae Leiognathus bindus 38 37 56 - 122 <100 101-110 >110 Leiognathus rivulatus 19 19 76 - 132 <100 >100 F. Mullidae Upeneus asymmetricus 28 27 95 - 142 <110 110-120 >120 F. Nemipteridae Pentapodus setosus 15 15 111 - 162 no size class Scolopsis affinis 16 16 128 - 213 <155 155-175 >175 F. Plotosidae Plotosus lineatus 24 21 99 - 264 <120 120-200 >200 F. Penguipeidae Parapercis alboguttata 7 7 126 - 215 no size class F. Synodontidae Saurida undosquamis 18 18 144 - 268 <200 >200 Synodus variegatus 16 16 124 - 245 <150 150-200 >200 F. Tetraodontidae Arothron manillensis 16 15 86 -119 <100 >100 Lagocephalus lunaris 18 18 79 - 227 <200 >200 PELAGIC F. Carangidae Selaroides leptolepis 34 33 78 - 153 <100 100-130 >130 F. Scombridae Rastrelliger kanagurta 26 26 142 - 208 no size class REEF ASSOCIATED F. Gobiidae Apogon notatus 15 14 71 - 85 no size class F. Pomacentridae Neopomacentrus filamentosus 33 33 66 - 139 <100 101-110 >110 Total 323 315 Table 1. List of fish species used for stomach content analysis from Visayan Sea. 98 ──┐ 102 ──┴──┐ 103 ─────┴┐ 58 ────┐ │ 65 ────┴─┴┐ 98 ─────┐ │ 56 ─────┴─┴────────────────────────────────┐ 100 ───┐ │ 95 ───┴────┐ │ 104 ─────┐ │ │ 106 ─────┴──┴┐ │ 112 ─────┐ │ │ 107 ─────┴───┴────┐ │ 73 ──────────────┴─────┐ │ 105 ──────┐ │ │ 115 ──────┴──────┐ │ │ 105 ─────────────┴──────┴──┐ │ 110 ────┐ │ │ 112 ────┴┐ │ │ 111 ─────┴───┐ │ │ 116 ─────────┴────┐ │ │ 101 ────┐ │ │ │ 107 ────┴──────┐ │ │ │ 99 ───────────┤ │ │ │ 104 ───────┐ │ │ │ │ 114 ───────┴─┐ │ │ │ │ 109 ─────────┴─┴──┴────────┴─────┐ │ 102 ─────┐ │ │ 104 ─────┴──────┐ │ │ 112 ────────────┴─┐ │ │ 106 ──────────┐ │ │ │ 100 ──────────┴───┴──────────────┴───┐ │ 122 ───────┐ │ │ 101 ───────┴──────────────────┐ │ │ 106 ──────────────────┐ │ │ │ 105 ──────────────────┴───┐ │ │ │ 58 ──────────────────┴ ──┴─────┴─ ───┘ M (101-110) M (101-110) S (<100) L (>110) 98 ──┐ 102 ──┴──┐ 103 ─────┴┐ 58 ────┐ │ 65 ────┴─┴┐ 98 ─────┐ │ 56 ─────┴─┴────────────────────────────────┐ 100 ───┐ │ 95 ───┴────┐ │ 104 ─────┐ │ │ 106 ─────┴──┴┐ │ 112 ─────┐ │ │ 107 ─────┴───┴────┐ │ 73 ──────────────┴─────┐ │ 105 ──────┐ │ │ 115 ──────┴──────┐ │ │ 105 ─────────────┴──────┴──┐ │ 110 ────┐ │ │ 112 ────┴┐ │ │ 111 ─────┴───┐ │ │ 116 ─────────┴────┐ │ │ 101 ────┐ │ │ │ 107 ────┴──────┐ │ │ │ 99 ───────────┤ │ │ │ 104 ───────┐ │ │ │ │ 114 ───────┴─┐ │ │ │ │ 109 ─────────┴─┴──┴────────┴─────┐ │ 102 ─────┐ │ │ 104 ─────┴──────┐ │ │ 112 ────────────┴─┐ │ │ 106 ──────────┐ │ │ │ 100 ──────────┴───┴──────────────┴───┐ │ 122 ───────┐ │ │ 101 ───────┴──────────────────┐ │ │ 106 ──────────────────┐ │ │ │ 105 ──────────────────┴───┐ │ │ │ 58 ──────────────────┴ ──┴─────┴─ ───┘ M (101-110) M (101-110) S (<100) L (>110) M (101-110) M (101-110) S (<100) L (>110) Figure 2a. Results of cluster analysis of L. bindus based on prey similarity by size class. S (<150) L (>200) M (150-200) M (150-200) 245 ┐ 196 ┴┐ 216 ─┴───┐ 212 ─────┴──────┐ 169 ─┐ │ 161 ─┴──────────┴──────┐ 156 ────┐ │ 149 ────┴────┐ │ 140 ─────┐ │ │ 141 ─────┴───┴─────────┴──────────┐ 181 ─────────┐ │ 124 ─────────┴──────┐ │ 162 ────────────────┴─────────────┘ S (<150) L (>200) M (150-200) M (150-200) S (<150) L (>200) M (150-200) M (150-200) 245 ┐ 196 ┴┐ 216 ─┴───┐ 212 ─────┴──────┐ 169 ─┐ │ 161 ─┴──────────┴──────┐ 156 ────┐ │ 149 ────┴────┐ │ 140 ─────┐ │ │ 141 ─────┴───┴─────────┴──────────┐ 181 ─────────┐ │ 124 ─────────┴──────┐ │ 162 ────────────────┴─────────────┘ Figure 2b. Result of cluster analysis of S. variegatus based on prey similarity by size class. Feeding Relationships of Dominant Fish Species 39 In spite of differences in the clarity of clustering, there was a general similarity in prey items across all size classes within a given species (Figs. 3a and 3b), thus showing the absence of a shift in diet. Furthermore, when prey similarities between size classes of the various species were examined, the second step of the cluster analysis showed low similarities (Fig. 4). Hence, the results show that similarity in apparent prey preference is higher among different sizes of the same species, than between similar size classes of different species. The general absence of a dietary shift is attributed to the narrow range of lengths of specimens available to the study. This may not necessarily be an artifact of size selectivity of the trawl, since the sizes of fish retained in the cover net were no different from those retained in the main net (unpublished information). Furthermore close to ¼ of the typical catch in the trawl was made up of relatively fast-swimming scombrids (Rastrelliger spp) (unpublished report). Hence, if there were larger, generally faster-swimming fish in the surveyed areas, a representative portion of them would 0 20 40 60 80 100 <100 101-110 size class (mm) F O C ( % ) chyme calanoid chaetognath Microcrustaceans others Figure 3a. Diet composition of L. bindus according to size class. 0 2 0 4 0 6 0 8 0 1 0 0 < 1 5 0 1 5 0 - 2 0 0 s iz e c l a s s ( m m ) F O C (% ) c h ym e n e m a to d e M g a s tr o p o d fis h Figure 3b. Diet composition of S. variegatus according to size class. Mequila and Campos 40 have been caught by the net. Their absence in the catches thus reflects the predominance of small and younger fish in the Visayan Sea. In most cases, the largest specimens observed from the catches were much smaller compared to their maximum reported sizes in the literature (Table 4). This is a common observation in heavily-fished waters. Fish are generally opportunistic feeders and show a relatively high degree of variability in prey preference either within or among species. The shift in feeding mode from larvae to early juveniles shows that they feed at different levels of the food chain during different stages of their life cycle. However, ontogenetic changes in feeding habits of fish as they grow do not depend on body size per se, rather these are correlated with changes in key aspects of feeding mechanisms (Wainwright and Richard, 1995). Several factors affect feeding in fish and can be categorized into intrinsic and extrinsic factors (Hourston et al., 2004). Intrinsic factors include mode of feeding, differences in swimming ability, and location in the water column, while extrinsic factors include variation in prey composition among habitat types and susceptibility of prey to predation. The assemblages formed by the cluster analysis (Fig.4) can also be related to their overall habits. The two major clusters formed clearly separate the strictly demersal species (assemblage 2) from the overlapping occurrence of pelagic, reef associated, and some demersal species (assemblage 1). Assemblage 1 can be further separated into 3 sub-groups: 1 A , 1 B , and 1 C . The first sub-group (1 A ) is formed by the pelagic carnivores S. leptolepis and R. kangurta. A reef- associated species A. notatus, was grouped together with demersal species P. lineatus and L. rivulatus in the second sub-group (1 B ) while N. filamentosus and L. bindus formed the third subcluster (1 C ). Sub-groups 1 A and 1 B had guts which typically contained large amounts of partially digested material (chyme), although planktonic prey such as fish larvae (1 A ) and calanoids (1 B ) were also common (Table 2). For sub- group 1 C, other plankton groups like larvaceans and Figure 4. Cluster analysis of size classes of 15 fish species based on prey similarity. S denotes small, M for medium, and L for large. Sl S ┐ Sl M ┴┐ Sl L ─┴───────────┐ Lb S ────────┐ │ Rk ────────┴────┴──────┐ Pl S ──┐ │ Pl L ──┴─┐ │ Pl M ──┐ │ │ An ──┴─┴───┐ │ Lr M ─────┐ │ │ Lr L ─────┴──┴───────────┴────────┐ Nf S ────┐ │ Nf M ────┴─────────┐ │ Lb M ────┐ │ │ Lb L ────┴──┐ │ │ Nf L ───────┴──────┴──────────────┴─────── Ua S ──┐ Ua M ──┴─┐ UA L ────┴─────────────┐ Sa S ─────┐ │ Sa L ─────┴────┐ │ Sa M ──────────┴───────┴──────┐ Pa ────────┐ │ Ll L ────────┴─┐ │ Ll M ──────────┴─────┐ │ Ps ────────────────┴──┐ │ Su M ─────────┐ │ │ Su L ─────────┴─────────┴─────┴───────┐ Sv M ────┐ │ Sv L ────┴─────────┐ │ Am M ──────┐ │ │ Am L ──────┴───────┴─────┐ │ Sv S -────────-───────┴─-─-──-──-─┴─-─ 1 A 1 B 2 A 1 C 2 B C lu s te r 2 C lu s te r 1 1 A 1 B 2 A 1 C 2 B C lu s te r 2 C lu s te r 1 Feeding Relationships of Dominant Fish Species 41 T a b le 2 . T w o w a y c o in c id e n c e t a b le s h o w in g v a ri a ti o n i n g u t c o n te n ts o f th e d if fe re n t fi s h s p e c ie s e x a m in e d f ro m t h e V is a y a n S e a . N o te : T a b le r o w s a re a rr a n g e d a c c o rd in g t o p re y i te m a s s e m b la g e s a n d c o lu m n s a re a rr a n g e d a c c o rd in g t o s p e c ie s -s iz e c lu s te rs 1 A 1 B 1 C p re y i te m s S I S I S I L b P I P I P I L r L r N f N f L b L b N f (S ) (M ) (L ) (S ) R k (S ) (L ) (M ) A n (M ) (L ) (S ) (M ) (M ) (L ) (L ) ** P o ly ch a e te la rv a e - - - - 0 .1 3 - - - - - - - - - - - S ili co fl a g e lla te s - - - - 0 .0 8 - - - - - - - - - - - U ro ch o rd a te s - - - 2 .2 2 - - - - - 5 .3 4 - 9 .6 3 0 .6 4 1 8 .0 3 1 5 .1 7 2 .6 1 ch a e to g n a th 0 .0 7 0 .1 4 - 5 .7 7 0 .0 2 - - - - 1 .6 7 0 .3 3 0 .0 4 - 1 5 .0 3 1 2 .3 2 7 .6 9 fis h e g g s - - - - 0 .0 2 - - - - - - - - - - 6 .4 6 A 1 P te ro p o d - - - - - - - - - - - - - - 1 .2 2 - sa lp s - - - - - - - - - - - - - 3 .4 1 3 .1 4 - C n id a ri a n s - - - - - - - - - - - - - 0 .3 9 - - co p . N a u p liu s - - - 0 .7 9 0 .0 3 - - - - - - - - - - - fi sh b o n e s - - - - - - 0 .7 8 1 .5 5 - - - - - - - - cy cl o p o id - - - 1 .7 4 5 .1 4 1 .0 4 - 0 .1 0 - 0 .1 8 0 .3 3 0 .9 8 1 .5 8 0 .2 1 0 .7 9 1 .0 3 A 2 h a rp a ct ic o id - - - 1 .2 0 1 4 .2 2 3 .4 4 - - - 0 .5 4 4 .0 0 1 .5 7 1 .1 1 0 .2 1 - 0 .4 8 D in o fl a g e lla te s 0 .0 7 0 .1 3 - 0 .4 4 8 .7 0 - - - - - - - - - 2 .5 0 - D ia to m s 0 .5 3 1 .0 6 - 8 .6 1 5 .2 2 - - - 0 .7 1 - - - - - - - a lg a e - - - 0 .4 4 - - - - - - - - - 3 .0 7 2 .5 0 - co p e p o d - - - 3 .5 4 - - - - - - - - - - 2 .5 0 - B 2 -1 B e n th ic w o rm s - - - - - - 0 .2 0 0 .2 5 - 0 .8 7 6 .5 0 - - - - - fi sh sc a le - - - 0 .2 0 - - - 4 .2 3 - - - - - - - - sq u id - - - - - - - - - - - - - - - - L a rv a c ru st a ce a n s 1 .2 8 1 .3 3 1 .2 2 1 .7 8 0 .7 9 1 .5 6 - - - - - 2 .8 0 0 .0 7 - 0 .1 9 0 .3 4 B 2 -2 p o ly ch a e te 0 .0 2 - 0 .0 5 2 .0 8 0 .1 7 - 0 .0 7 0 .6 9 - - - 0 .1 2 0 .3 1 0 .9 5 0 .3 4 0 .1 8 b iv a lv e 0 .0 9 0 .1 7 - - 0 .8 7 0 .1 8 - 0 .0 6 - - - - - 1 .1 7 - - g a st ro p o d 0 .1 6 0 .1 7 0 .1 4 - 0 .2 7 - - - - 0 .8 7 - - - 0 .3 6 - - fo ra m s - - - 0 .1 2 - - - - - - - - - - - - B 2 -3 ru b b le - - - - - - - - - - - - - - - - cr a b - - - - - - - - - - - - - - - - se d im e n t - - - - - - - - - - - - - - - - M ic ro cr u st a ce a n s 1 4 .8 0 11 .3 5 1 8 .2 5 4 .4 6 1 .3 6 - 0 .0 7 0 .2 8 3 .3 3 6 .2 4 7 .6 1 5 .0 6 3 .1 7 6 .2 2 9 .4 8 1 6 .6 3 sh ri m p 1 .0 5 0 .7 7 1 .3 2 3 .0 5 0 .0 2 2 .6 9 3 .5 8 8 .1 9 6 .3 6 - 0 .7 0 - 0 .2 1 0 .9 5 5 .3 8 - C h ym e 4 7 .5 0 4 9 .0 0 4 6 .0 0 5 3 .7 8 5 1 .9 2 8 6 .8 8 9 4 .4 0 8 2 .8 8 8 2 .1 4 7 1 .6 3 6 4 .1 3 1 8 .6 4 3 0 .5 6 2 9 .8 9 2 8 .6 7 3 2 .6 2 B 1 C a la n o id 1 4 .6 0 1 6 .4 2 1 2 .7 8 8 .6 7 1 0 .9 9 3 .3 8 0 .3 7 0 .8 8 4 .6 0 4 .0 2 2 .1 5 6 1 .0 4 6 2 .2 6 1 7 .9 7 1 5 .0 2 3 1 .9 6 F IL A 1 8 .1 0 1 7 .6 3 1 8 .5 8 - 0 .0 1 - - - - - - - - - - - E ch in o d e rm la rv a - - - 0 .8 7 - - - - - 8 .3 4 9 .7 5 - - - 0 .1 5 - F is h - - - - 0 .0 0 - - - 2 .8 6 - - - 0 .0 8 0 .4 2 - - N e m a to d e 1 .7 4 1 .8 1 1 .6 6 0 .2 4 0 .0 2 0 .8 3 0 .5 4 0 .8 9 - 0 .3 1 4 .5 1 0 .1 4 - 1 .7 0 0 .6 3 - ** d e p a u p e ra te F ee d in g R el at io n sh ip s o f D o m in an t F is h S p ec ie s Mequila and Campos 42 T a b le 2 ( C o n ti n u e d ) Tw o w a y c o in c id e n c e t a b le s h o w in g v a ri a ti o n i n g u t c o n te n ts o f th e d if fe re n t fi s h s p e c ie s e x a m in e d f ro m t h e V is a y a n S e a . N o te : T a b le r o w s a re a rr a n g e d a c c o rd in g t o p re y i te m a s s e m b la g e s a n d c o lu m n s a re a rr a n g e d a c c o rd in g t o s p e c ie s -s iz e c lu s te rs 2 A 2 B 2 C p re y i te m s U a U a U a S a S a S a L I L I S u S u S v S v A m A m S v (S ) (M ) (L ) (S ) (M ) (L ) P a (L ) (M ) P s (M ) (L ) (M ) (L ) (M ) (L ) (S ) ** P o ly ch a e te la rv a e - - - - - - - - - - - - - - - - - S ili co fl a g e lla te s - - - - - - - - - - - - - - - - - U ro ch o rd a te s - - - - - - - - - - - - - - - - - ch a e to g n a th - - - - 0 .9 1 - - - - - - 0 .7 8 - - - - - fis h e g g s - - - - - - - - - - - - - - - - - A 1 P te ro p o d - - - - - - - - - - - - - - - - - sa lp s - - - - - - - - - - - - - - - - - C n id a ri a n s - - - - - - - - - - - - - - - - - co p . N a u p liu s - - - - - - - - - - - - - - - - - fi sh b o n e s - - - - - - - - - - - - - - - - - cy cl o p o id - 1 .3 0 - - - - - - - - - - - - - - - A 2 h a rp a ct ic o id - - - - - - - - - - - - - - - - - D in o fl a g e lla te s - - - - 1 .4 6 - - - - - - 0 .3 2 - - - - - D ia to m s - - - - - - - - - - 0 .5 6 3 .6 7 - - - - - a lg a e - - - - 0 .8 4 4 .6 5 8 .3 5 2 .3 8 - 4 .3 9 2 .4 2 4 .2 3 - - - - - co p e p o d - - - 3 .0 1 1 5 .8 8 - 8 .7 5 - - 2 .2 5 2 .8 4 6 .0 6 - - 0 .9 4 - - B 2 -1 B e n th ic w o rm s - - - - 2 .0 4 - 1 .2 9 - - 2 .1 1 - - - - - - - fi sh sc a le - - - 3 .4 3 2 .4 3 1 .0 4 - - - 1 0 .0 5 - - - - - - - sq u id - - - 2 .3 6 - 1 .3 5 - 1 3 .5 7 - 1 .5 3 - - - - - - - L a rv a c ru st a ce a n s 3 .0 8 1 .1 1 6 .3 5 -. 0 .9 1 - 4 .5 2 - - - - - - - 5 .4 3 0 .1 5 - B 2 -2 p o ly ch a e te 1 .1 9 1 .6 7 3 .1 3 - - - 1 .2 9 - - - - 1 .1 7 - - 0 .6 3 - - b iv a lv e - - - - - - - - - - 1 .8 3 - - - 3 .0 4 0 .1 4 - g a st ro p o d - - - - 2 .7 2 - - 0 .4 8 - 1 .1 9 - - 3 .1 7 - 1 3 .7 3 4 .4 7 - fo ra m s - - - - - - - - - - - - - - 1 .3 5 2 .3 6 - B 2 -3 ru b b le - - - - - - - 4 .2 9 - - - - - - 0 .9 2 1 .0 4 - cr a b - 2 .5 0 - - - - 0 .3 8 5 .7 1 8 .6 4 - 0 .8 8 0 .5 5 - - - - - se d im e n t - - - - - - - 0 .7 1 3 .6 4 - - - - - - 4 .7 3 - M ic ro cr u st a ce a n s 3 4 .0 0 3 3 .7 6 3 3 .5 0 4 9 .7 0 3 1 .7 5 5 6 .2 6 - - - 8 .9 4 - 0 .1 6 - 2 .0 8 0 .8 3 0 .0 1 - sh ri m p 2 2 .1 0 2 2 .4 9 2 7 .8 4 2 0 .8 4 1 3 .8 0 1 0 .5 3 2 2 .0 2 1 8 .5 7 11 .2 1 1 6 .9 4 6 4 .6 2 3 8 .0 3 - - 0 .0 0 0 .4 5 - C h ym e 3 8 .8 9 3 1 .6 7 2 3 .3 3 4 .8 3 7 .1 7 6 .2 5 5 0 .7 1 5 2 .5 7 6 3 .0 0 3 4 .8 7 2 5 .3 8 2 9 .7 0 5 5 .0 0 4 3 .3 3 3 0 .0 0 3 6 .4 3 1 6 .6 7 B 1 C a la n o id 0 .7 4 2 .3 6 1 .6 7 - - - - - - - - 1 .2 5 - - - - - F IL A - - - - - - - - - - - - - - - - - E ch in o d e rm la rv a - - - 1 5 .8 2 2 0 .1 0 1 9 .9 2 - - - 1 0 .1 9 - - 4 .2 8 11 .0 0 - - 2 0 .1 6 F is h - 1 .1 1 - - - - 0 .5 8 - 1 2 .7 3 0 .9 2 1 .4 8 1 2 .0 5 9 .6 7 1 0 .5 0 - - 4 3 .0 2 N e m a to d e - 2 .0 4 4 .1 8 - - - 2 .1 1 1 .7 1 0 .7 9 6 .6 3 - 2 .0 2 2 7 .8 9 3 3 .0 8 4 3 .1 4 5 0 .2 2 2 0 .1 6 ** d e p a u p e ra te M eq u il a an d C am p o s Feeding Relationships of Dominant Fish Species 43 chaetognaths were also common in their guts. Sub- groups 1 A and 1 C can be considered as generally planktivorous. Observations on 1 C are consistent with overlapping feeding habits of reef-associated and demersal species in the area, particularly in stations adjacent to shallow coral reefs. For example, the pomacentrid N. filamentosus is typically found over soft bottoms of lagoons and inshore reefs around coral outcrops, rocks, and debris (Froese, and Pauly, 2006). Leiognathids in general are bottom living but small individuals feed mostly on copepods and phytoplankton, while large fish feed predominantly on benthic invertebrates. L. bindus in particular feeds on calanoid copepods, ostracods, chaetognaths, polychaete larvae, and fishes (FAO, 2001). Similarly, L. rivulatus searches for planktonic prey using a protrusible mouth or by sieving potential food through their gill rakers (Froese.and Pauly, 2006). The high incidence of chyme in subcluster 1 B , however, may not make them typical planktivores, in spite of the occurrence of planktonic prey in their guts as stated above. P. lineatus and A. notatus for instance are known to be carnivorous, feeding primarily on crustaceans, mollusks, and fishes (FAO, 1999), while A. notatus are known to feed on fish and zoobenthos (Froese and Pauly, 2006). Since trawling was done during the daytime, the large amounts of chyme suggest that P. lineatus and A. notatus fed during the night, making most of the ingested prey partially digested by the time trawling was done. The second assemblage can also be divided into three sub-groups 2 A , 2 B , and 2 C . These are largely demersal in habit because plankton (e.g., calanoids) are generally absent from their diet and they consume mostly benthic prey. Sub-group 2 A showed a high occurrence of shrimps and microcrustaceans in their guts, while the others showed common occurrences of shrimps, crabs (2 B ), and nematodes (2 c ) in their diets (Table 2). Food niche overlap was computed only for 7 demersal species (out of a total of 15) because the objective was to quantify the degree of overlap between size classes of different species. Only species with complete size classes (mostly from assemblage 2) (Fig. 4) were included in this analysis. These include S. variegatus, U. asymmetricus, L. bindus, S. affinis, P. lineatus, and A. manillensis. Generally, the values of niche overlap between size groups of various species were low (Table 3) and inconsistent with the premise that resource overlap decreases with growth. Very few species in this study showed such a trend in resource use. Examples are L. bindus and P. lineatus, and U. asymmetricus and S. affinis. Food niche overlap provides insights on the existence, nature and strength of competitive interactions (del Norte-Campos, 1995). The concept of resource overlap is usually related with the intensity of competition because of the use of common resources. Competing species will necessarily show extensive niche overlap, although high niche overlap does not always mean that there is competition. In the case of S. variegatus and A. manillensis, (Table 3), high niche overlap does not necessarily mean there is competition between them, unless food resources are limited. The presence of high prey density may result in high diet overlap between species because there is no need to partition available resources. Although no simultaneous investigation on prey availability was conducted in this study, previous reports on benthic infauna (Mequila et al., 2004) and zooplankton (Campos et al, 2002) surveys in the area show that abundances of prey for demersal and pelagic fishes are relatively high in the area. Low food niche overlap, on the other hand, may mean that food resources have become well- partitioned as a result of competitive interactions. Divergence over evolutionary time in feeding morphology and behavior of fish for example may result in utilization of different resources, thus lowering the niche overlap among previously competing species (Labropoulou and Markakis, 1998). The feeding behavior of fish examined in this study range from ram feeders (R. kanagurta and S.leptolepis) and suction feeders (L. bindus and L. rivulatus), to browsers (U. asymmetricus), and crushers (L. lunaris, A. manillensis and P.lineatus). Whether this is a result of previous competition among these species is uncertain. The high values of food niche width (H’) in Table 4, regardless of sizes, indicate that fish consumed a diverse diet, implying that they are generalists in terms of feeding. Although this seems contradictory to the low overlap values, it may mean that fish occupy a wide range of habitats, thus favoring utilization of a large number of Mequila and Campos 44 resources. According to Amundsen et al. (1996), a population with a broad niche width may be the outcome of either true generalist behavior of each individual of a population (high within-individual variation) or specialization of individuals of the population on different prey (high between-individual variation), as would be the case with ontogenetic changes in diet. The latter, however, was not observed in this study. High overall productivity in the Visayan Sea is evidenced by consistently high fisheries yield since at least the 1970s (Aprieto and Villoso, 1979), in spite of heavy fishing pressure (Hermes et al., 2004). This is the result of many factors, including its shallow depth and central location, which in turn allows not only water and nutrient exchange with various basins adjacent to it (Campos et al., 2002), but fish movement and recruitment as well. Its being surrounded by deep (>200m) water adds to the many factors that would favor high natural fish abundance in the area and the likelihood of strong competitive interactions. The results of the study, however, are not consistent with strong competition. This is attributed to an overall relative availability of food resulting from a general reduction in the abundance of consumers and a corresponding reduction on predation (consumption) on food resources (Mequila et al., 2004). Constant cropping brought about by heavy fishing, particularly of trawls which have been used in the area since the 1950s (Aprieto and Villoso, 1979) allows this. Thus, Table 4. Food niche widths (H') of fish species by size class in the Visayan Sea, July 2003 Note: Maximum total length attained as reported in the literatures are included in the table for comparison with the size observed in this study. Those indicated with SL were species with maximum length reported in standard length. Size range Niche Width (H') (TL in cm) L max Species (cm) Reference small medium large Synodus variegatus 12.4 - 24.5 30 SL Masuda, 1984 0.57 0.50 0.56 Upeneus asymmetricus 9.5 - 14.2 30 Allen, 1999 0.55 0.67 0.67 Leiognathus bindus 5.6 - 12.2 12 Allen, 1999 0.78 0.85 0.92 Scolopsis affinis 12.8 - 21.3 30 Randall et.al., 1996 0.62 0.82 0.57 Selaroides leptolepis 7.8 - 15.3 17 Masuda, 1984 0.63 0.63 0.62 Plotosus lineatus 9.9 - 26.4 32 Randall et.al., 1996 0.27 0.31 0.12 Arothron manillensis 8. - 11.9 45 SL Masuda, 1984 __ 0.64 0.51 Neopomacentrus filamentosus 6.6 - 13.9 11 Froese.and Pauly, 2006 0.53 0.42 0.70 Saurida undosquamis 14.4 - 26.8 50 SL Masuda, 1984 __ 0.45 0.73 Leiognathus rivulatus 7.6 - 13.2 12 Rau and Rau, 1980 __ 0.48 0.57 Lagocephalus lunaris 7.9 - 22.7 45 SL Masuda, 1984 __ 0.51 0.63 Rastrelliger kanagurta 14.2 - 20.8 35 Randall et.al., 1996 __ __ __ 0.68 Apogon notatus 7.1 - 8.5 10 Masuda, 1984 __ __ __ 0.32 Pentapodus setosus 11.1 - 16.2 20 Allen, 1999 __ __ __ 0.86 Parapercis alboguttata 12.6 - 21.5 22 Allen, 1999 __ __ __ 0.64 Small Medium Large Sv/Ua 0.17 0.35 0.30 Sv/Lb 0.18 0.32 0.32 Sv/Sa 0.21 0.14 0.19 Sv/Pl 0.18 0.56 0.44 Sv/Am __ 0.61 0.70 Ua/Lb 0.50 0.43 0.41 Ua/Sa 0.60 0.54 0.50 Ua/Pl 0.44 0.43 0.28 Ua/Am __ 0.35 0.28 Lb/Sa 0.16 0.16 0.24 Lb/Pl 0.64 0.34 0.33 Lb/Am __ 0.35 0.30 Sa/Pl 0.08 0.18 0.10 Sa/Am __ 0.13 0.07 Pl/Am __ 0.32 0.37 Table 3. Food niche overlaps between fish species by size class in the Visayan Sea, July 2003 Note: Sv - Synodus variegatus; Sa - Scolopsis affinis; Pl - Plotosus lineatus; Ua - Upeneus asymmetricus; Am - Arothron manillensis; Lb - Leiognathus bindus Feeding Relationships of Dominant Fish Species 45 continuous cropping together with natural high overall productivity promote generalist feeding (high niche widths) in the fish community over a wide range of habitats (low niche overlap). ACKNOWLEDGEMENTS This study was conducted as part of the Hydrological Survey of the Visayan Sea July 2003: Trawl, Oceanography, and Plankton Survey (VSea TOPS) project funded by an In-House grant from the UPV. The authors wish to thank the officers and crew of the RV DA-BFAR for the assistance during the cruise, the Visayan Sea Program for conducting the cruise, and the Conservation International for providing the student travel grant to Annie Trixie L. Mequila during the PAMS 8 symposium. REFERENCES Allen, G.R. 1999. Fishes of Southeast Asia. Periplus Editions Ltd. 292 p. Amundsen, P.A., H.M. Gabler, and F.J. Staldvik. 1996. A new approach to graphical analysis of feeding strategy from stomach contents data - modification of the Costello (1990) method. J. Fish Biol. 48: 607-614. Aprieto, V.L. and E.P.Villoso. 1979. Catch composition and relative abundance of trawl-caught fishes in the Visayan Sea. Fish. Res. J. Philpp., 4(1):9-18. Bachok, Z., M.I. Mansor, and R.M. Noordin. 2004. Diet composition and food habits of demersal and pelagic marine fishes from Terengganu waters, east coast of Peninsular Malaysia. Naga. 27:41-47. BFAR (Bureau of Fisheries and Aquatic Resources). 2002. Commercial Fishery stock assessment of the Visayan Sea in Region VI for the year 1998 to 2002. Bureau of Fisheries and Aquatic Resources Regional Office VI, Iloilo City. Unpublished technical report. Campos, W.L., R.F.C. Canto, and D.G. Estremadura. 2002. Abundance, composition and distribution of major zooplankton taxa in the interisland waters of the eastern central Philippines. UPV J. Nat. Sci. 7 (1): 66-80. Del-Norte – Campos, A.G. 1995. Ecological studies on the coexistence of the brown shrimp, Crangon crangon L. and the gobies Pomatoschistus microps Kroyer and P. minutus Pallas in shallow areas of the German Wadden Sea. University of Hamburg. Unpublished dissertation: 265 p. FAO. 2001. FAO species identification guide for fishery purposes. The living marine resources of the Western Central Pacific. Vol.5. Bony fishes part 3 (Menidae to Pomacentridae). Carpenter,K.E. and V.H. Niem. (eds) Rome. Pp.2791-3380. FAO. 1999. FAO species identification guide for fishery purposes. The living marine resources of the Western Central Pacific. Vol.5. Bony fishes part 3 (Menidae to Pomacentridae). Carpenter,K.E. and V.H. Niem. (eds) Rome. Pp.2791-3380. Froese, R.and D.Pauly. Editors. 2006. FishBase. World Wide Web electronic publication. www.fishbase.org, version (12/ 2006). Hermes, R., N.B. Armada, R.A. Aparri, E.C. Zaragoza, and U. Lohmeyer. 2004. Overexploitation in the Visayan Sea: Designing a project solution. In: In turbulent seas: The status of Philippine marine fisheries. Coastal Resources Management Project, Cebu City, Philippines. 378p. Hourston, M., M.E. Platell, F.J. Valesini, and I.C. Potter, 2004. Factors influencing the diets of four morphologically divergent fish species in nearshore marine waters. J. Mar. Biol. Ass. U.K. 84: 805-817. Labropoulou, M., and G. Markakis, 1998. Morphological- dietary relationships within two assemblages of marine demersal fishes. Env. Biol. Fishes. 51: 309-319. Lopez-Peralta, R.H. and C.A.T. Arcila. 2002. Diet composition of fish species from the southern continental shelf of Colombia. Naga. 25: 23-29. Masuda, H., K. Amaoka, C. Araga, T. Uyeno, and T. Yoshino. (eds). 1984. The fishes of the Japanese archipelago (Vol. 1:Text). Tokai University Press, Japan. 456 p. Mequila and Campos 46 Mequila, A.T., G.E. Genito, and W.L. Campos. 2004. Large scale distribution of macro- and meiofaunal assemblages in the Visayan Sea. UPV J. Nat. Sci. 9 (1): 54-69. Piepenburg, D. and U. Piatkowski. 1992. A program for computer aided analyses of ecological field data. CABIOS 8(6): 587-590. Randall, J.E., G.R. Allen, and R.C. Steene. 1996. Fishes of the Great Barrier Reef and Coral Sea. University of Hawaii Press, Honolulu. 557 p. Rau, N. and A. Rau. 1980. Commercial marine fishes of the Central Philippines. German Agency for Technical Cooperation, Germany. 623 p. Salgado, J.P., H.N. Cabral, and M.J. Costa. 2004. Feeding ecology of the gobies Pomatoschistos minutus (Pallas, 1770) and Pomatoschistus microps (Kroyer, 1838) in the upper Tagus estuary, Portugal. Sci. Mar. 68 (3): 425-434. Ross, S.T. 1986. Resource partitioning in fish assemblages: a review of field studies. Copeia. 2:352-388. Wallace, H. 1981. An assessment of diet overlap indexes. Trans. Americ. Fish. Soc., 110: 72-76. Wallace, H. and J.S. Ramsay. 1983. Reliability in measuring diet overlap. Can. J. Fish. Aquat. Sci., 40: 347-351. Wainwright, P.C. and B.A. Richard. 1995. Predicting patterns of prey use from morphology of fishes. Env. Biol. Fish. 44:97-13. Windell, J.T. and S.H. Bowen. 1978. Methods for study of fish diet based on analysis of stomach content. Pp. 219- 226. In: Methods for Assessment of Fish Production in Freshwaters. 3rd Ed. T. Bagenal (ed). Blackwell Sc. Publ., Oxford.