SPATIO TEMPORAL DISTRIBUTION OF SMALL PELAGIC FISH IN THE JAVA SEA

Time spent in seeking fish schools and potential fishing areas is the main source of fuel consumption, and it is a significant factor in raising the cost of fishing operation. Therefore, knowledge of the distribution pattern lnd aggregations of commercial fish in space and time is of considerable financial value to the fishers anAlisnlng industry because it will assist them to reduce the time and fuel consumption of the boats in locating the rich fishing grounds. To gain understanding of such knowledge, study on the spatio temporal distribution of small pelagic fish has been done in the Java Sea using commercial fishing data and hydro acoustic. lt was found that abundance of pelagic fish was seasonally variable in the Java Sea, reaching its peak in May for the inshore area and September for the offshore area. lt was mostly driven by fluctuations in the abundance of one dominant species, Sardinela spp. in the inshore and Decapterus spp. in the offshore. The high abundance occurred in the areas around the vicinity of small islands. Based on the cluster and PCA analysis, Rastrelliger spp., S. crumenophthalmus, and Sardinella spp. were considered falling into one group and found to be abundant between April and July.


INTRODUCTION
The Java Sea, which is hydrologically influenced by Pacific Ocean and the Indian Ocean, is one of important fishing areas in lndonesia.A substantial part of the growth in the catch has come from this area.Indeed, with only 7% of the total fishing area, the Java Sea contributed to amountof 32% to the total marine fish catch (DGCF, 2007\.Between 1968 and 1980, the main fishing gear used in catch growth from this area was the bottom trawlwith main target species of shrimps and demersal fish.By the end of the 1970s, the number of trawlers exceeded the license available by an estimated of 40o/o in the Java Sea area, and fishing effort was estimated to be approximately 20oh above the level required to reach the maximum sustainable yield (Potier & Sadhotomo, 1995a).
Conflicts with small scale inshore traditional fishers, particularly those operating fixed gears, had become a major political issue.Between '1981 and   1983, a ban on the use of the bottom trawl came into effect throughout the west Indonesia.Most of these trawlers, since then, were converted to pelagic fishing, particularly purse seiner, and gill neter.
Pelagic fishery in the Java Sea has become important after 1980.The fishery rapidly develops following the increase of purse seine fleet to a bigger and powerful vessel.Amongst the important species target of pelagic fishing is a community of small coastal pelagic fish, consisting scads, Decapterus spp.; sardines, Sardinel/a spp.; lndo-Pacific mackerels, Rastrelliger spp.; and trevallies, Se/ar spp.(Widodo, '1995).
The Javanese fishers, so far, rely on their traditional knowledge of habitat to search fish schooling and to target their fishing efforts.They still practice hunting in locating their fishing grounds, although they utilized fish aggregating devices and light as auxiliary devices for attracting fish schooling in their fishing strategies (Potier & Petit, 1995).Time spent in seeking fish schools and potential fishing areas is the main source of fuel consumption, and it is a significant factor in raising the cost of fishing ooeration.
It is of great importance to understand the distribution pattern and aggregations of commercial fish in space and time.A better understanding of such knowledge is of considerable financial value to the fishing industry because it will assist them to reduce the time and fuel consumption of the boats in locating the rich fishing grounds.From resource management point of view, ensuing greater efficiency of the commercialoperations is important, as itwould help to make operations more profitable.Reliable information of fish distribution and abundance and its aggregation pattern is required from the scientific community for this purpose.
Traditionally, fish distribution and abundance have been indirectly estimated through the analysis of Corresponding author: Komoleks Pelabuhan Perikanan Samudera Jl.Muara Baru Ujung-Jakatla 14440, E-Mail: wijopriono @yahoo.comhistorical data on fish catch per unit effort or through the statistical analysis of mark-recapture experiments using tagged fish.However, these indirect methods are often biased by the selectivity of fishing gears but also by artificial on the economic ground.In addition, considerable time, and substantial efforts are required in estimating fish distribution and abundance by the statisticalanalysis on the historical data (Toresen et   a/., 1998).
Following the advanced development of hydro acoustic technology, sonar, and echo sounder there have been a major impact on fisheries, especially their application in fishing system, such as searching for profitable concentration of fish.Hydro acoustic method can be considered as a suitable method for estimating fish abundance directly.In fisheries research, hydro acoustic methods have become increasingly importance over the years for studying and monitoring distribution, congregation, and migration of fish (Engas   & Ona, '1987; Misund, '1987; Thorne et al., 1987).However, there are some disadvantages on this method, 1.e., lt is not be able to directly recognize fish species under study (where this aspect is important in multrspecies fisheries), and usefulonly when the fish of interest are conveniently located; they must not be too close to the surface or the bottom, where the fish echoes are obscured by much stronger reflections from these boundaries (Maclennan & Simmonds, 1992).This paper discussed the small pelagic fish species in the Java Sea, where information derived from hydro acoustic and historical data on catch per unit of effort are used in complementary to estimate their abundance and to give comprehensive explanation on their distribution pattern, both spatially and temporally.

MATERIALSAND METHODS
Data on pelagic fish density are available from two sources, Le., from hydro acoustic survey and from records of commercial fishing.Acoustic data were collected through fishery resources survey using research vessel.Catch data were collected from big, medium, and mini purse seines, which constitute the main fishing fleet for catching pelagic fish species.

Catch Data Gollection
Catch data were collected from the records of fish catches landed by commercialfishing fleet, i.e big, medium, and mini purse seines.Catch data of the big and medium purse seines represented the offshore catch, while catch data of the mini purse seines represented the inshore catch.
-fhose two sources of data were used to describe species distribution of pelagic fish in the offshore and the inshore areas, respectively.The data were collected from their principal landing places along the north coast of Java, including Blanakan, Eretan, and Rembang for inshore 844qfr*4Z4zqfte4/t, 1 U Zr ^ 4 I -4e 4 4 4 4 V,, z 4 t W 44 4ry&4V/.LZqb, ab eZAf, * "r*' r*' rdr rdr r&.rr'r r ir.rr!. rr':r r lr re r E. catch, and Pekalongan, which is the main landing for offshore catch as well as for inshore catch (Figure 1) Data on catch weight, fishing time, fishing ground locations, and species composition of every trip of purse seiners in the pelagic fishing fleei were recorded.Monthly data on catch by species, which were available at the port administration of sampling sites were also collected.Because the exact geographical co ordinates were often not known for fishing ground positions, they were recorded in terms Geographical location of sampling sites and the main fishing areas of the fleets (purse seines). of a grid system, which divides the fishing grounds into 10' latitude x 10' longitude squares.
l-he sampling frame is defined as the list of daily landings by seines, where the sampling unit is the sample boat.The list consists of the name of the fishing boat landed; type of boat; fishing areas; and species catch.All the purse seiners landed in the selected landing places (5 landing places) were taken as samples.The next steps, one or two sub samples (a basket of fresh fish that is c.a. 30 kg) are randomly taken from a fish hold of the sampling boats.The specimens were sorted by species.Acomparison or crosscheck on the species composition of the sample and totalcatch by species from the auction was then performed.
spatio Temporal Distribution of small Pelagic Fish in the Java sea (wijopriono)   Hydro Acoustic Data Collection Acoustic data were collected from fishery resources survey in September untilOctober2002 using research vessel of RA/ Muttara lV belong to Research Institute for Marine Fisheries.Scientific echosounder Simrad EY-500, including split beam transducer system model S38-B operating at 38 KHz, was used for acquisition of the data.ltwas interconnected to GPS system, a Garmin GPS 45 personal navigator, to record positions and times.The acoustic data stored on the hard disk and dedicated software, EP-500, was used for processing echo signals.The echosounder system was callibrated, using a 60 mm copper sphere standard target sphere which has a target strength of -33.6 dB, according to the standard procedure (Foote   et a|.,1987;Knudsen, 1989; Toresen et Tracks of hydro coustic survey in september until october 20Q2 Owing to the time available and prior information that suggested the importance of central and eastern part of the Java Sea for pelagic fishery (Durand &   Widodo, 1997), the survey was then devoted to cover these areas.The survey transects covered the area from the most shallow of approximately 15 m of inshore areas to the deepest of nearly 90 m of the offshore.However, although we have covered different depth zones and habitat when the study was undertaken, surveytransects did notcoverfishing area in the middle of the Java Sea, mainly the area around Masalembo lsland, because of rough sea and limitation of time available.
Survey tracks was designed systematically in parallel transect perpendicular to the coastline in the area off south Kalimantan and triangulartransectalong the area off north coast of Java, with the same distance between the turning points (Figure 2).The distance between transects varied according to available time, but usually separated by 20 to 30 nautical mil (nmi) (Jolly & HamPton, 1990).

Data Analysis Gatch Data
Catch data of the pelagic fish landed by purse seines were analyzed, based on their catch per unit of effort in each fishing area.Catch per unit of effort is generally used as a standard measurement of relative abundance of fishery resources (Gulland, 1975).lt is usually expressed as numbers of fish caught per unit of gear per unitof time.Freon (1980) defined the time spent to search actively the fish as the best index of effort for purse seine fisheries.However, with the available data, the number of trips and time spent at sea (days at sea)were taken to estimate the effort.Number of trips, i.e., the number of trips carried out in the course of one time unit by the purse seine.Time Ind.Fish Res.J. Vol. 14 No.1 June-2008: 21-35 spent at sea (days at sea), r.e. the effective number the echo measurements into fish density estimates, of days spent at sea for each trip by the purse seine.target strength function was applied: It includes the outward and return voyage between thefishing groundsandthe harbors, thetimedevoted W =11.p,.A ....... ( 1to search for fish, catching, and hauling them onboard. ; where: tV, = number of fish in length group i Sn = average integrator value allocated to a given fish species in the region of interest p, = proportion of a given fish species in length group i A = areaof the region being surveyed (nm2) (r= wnere: o, = back scattering cross-section of fish in length group r, which is calculated from the target strength equation adopted for each species As the target strength equation for the Java Sea small pelagic species was not available, empirical equation of target strength adopted from Foote (198 /)

Tablel
Definition of species categories used in the official statistic and landing places data records Several selected important species in pelagic fishery were analyzed.Monthly catch per unit of effort values, either in the whole area or sectors, were used to describe seasonal variations of the soecies abundance.The use of these data rests on the assumption that catch per unit of effort is proportional to abundance.Interspecific species interactions, as the degree of affinity between species, was measured based on their covariations in abundance.lt was examined using pearson product moment correlation.
The correlation matrix was subsequently subjected to clustering and ordination for species grouping in which the analysis were done using hierarchical clustering, based on euclidean distance, and principal components, respectively.

HydroAcoustic
Characterizing the spatial agglomeration of pelagic fish was done based on acoustic data from fishery resources survey in September until October 2002, at the time the southeast monsoon ended or the pre northwest monsoon was started to prevail.Iocations that take spatial correlation into account.
In the analysis, the first step was io divide the transects into 5 nmi long segments and for each of these calculate an echo integration value.The geographical position of the segments was set to their mid point.From these positions and the echo integration values, linear variogram values were calculated.Different models were tried to fit these variogram values, and the best fitting model, by means of krigging, was used to interpolate echo integration values for the entire study area.The values from the best fitting model were used to draw contour maps over the distribution of pelagic fish.

RESULTS AND DISCUSSIONS
Spatial and Seasonal Distribution of the Fish Species A community of small and some large pelagic fishes, which are multispecies in nature, are exploited by pelagic fishing fleet in the Java Sea.A total of approximately 30 pelagic species was caught (Potier   & Sadhotomo, 1995).Of these, 15 species formed 90% of the totalcatch.Official statistical data records have grouped those 15 species into 6 statistical categories of fish catch as shown in Table 1.
Nevertheless, there are different names addressed for the same species or group of species between official statistic and data records at landing places.
Seasonal abundance and distributton of each species are showed in Figures 3 and 4, which were estimated based on the catch per unit of effort and percentage of its availability relative to the total soecies.
Abundance of pelagic fish species exhibited fluctuation and each showing a peculiar distribution.
Different species alternately dominated the pelagic resources both spatially and seasonally.Decapterus spp. was found to dominate the pelagic fish resources of the offshore area with the abundance much higher than the others but, in contrast, it appeared to be lesser in the inshore area.
Sardinella spp.dominated the resources of the inshore area in almost all year round with the exception of January.lt was abundant in both inshore and offshore areas in November, the time when the northwest monsoon started to prevail.The species was available more in the offshore area than the inshore area in January.
Other species composed of less than 30% of the total pelagic fish resources (Figure 4).Among them were Se/ar crumenophthalmus, Rasfre//rEter spp., Ambligaster sirm, Auxis thazard, and Ethynnus affinis, Se/arordes leptolepis, Atule mate, Alepes dieddaba, and Dussumeria acuta.In January, however, Se/ar crumenophthalmus dominated the resources in the inshore area whilst Sardinella spp.distributed more Spatio Temporal Distribution of Small Pelagic Fish in the Java Sea (Wiiopriono)   towards offshore.Another group of commercial fish, i.e.Selaroides leptolepis, Atule mate, Alepes djeddaba, and Dussumeria acuta, were found to be more abundant in the inshore than the offshore all year round.Nevertheless, their abundances have never been more than 1 0% of the total pelagic fish stock.
Overall, catch per unit of effort of pelagic fishes averaged monthly from 1998 to 2003 is presented in Figure 5.The figure shows trend of the catch per unit of effort both in the inshore and offshore areas, describing seasonal abundance of pelagic fish resources in the study area.
Seasonal abundance of pelagic fish resources in the inshore and offshore areas found to have similar pattern, each showing a single mode, with different trend.The abundance exhibited an increasing trend in both areas after falling down to the lowest in February, which was indicated by the lowest catch per unit of effort, i.e. 528.4 kg per day per vessel for the inshore area and 608.3 kg per day per vessel for the offshore area.The abundance in the inshore area has sharply increased after February, reaching to its highest in May at the time the value of the catch per unit of effort stand at 940 kg per day per vessel.At the same time, however, the abundance in the offshore area was found to continually increase and reaching a peak in September, when the value of the catch per unit of effort accounted for 1082.5 kg per day per vessel.
Figures 6, 7, and 8 show seasonal distribution and relative abundance of pelagic fish species in different fishing areas illustrated through surface plots.They are expressed as percentage of species availability (catch per unit of effort of each species) relative to the total pelagic fish stock (total catch per unit of effort).The surface plots are employed for six main pelagic fish species in six main fishing areas spreading from the central to the easternmost of the Java Sea, including the fishing areas around the waters of Karimunjawa, Bawean, Masalembo, Matasiri,   Kangean, and SambergelaP.f n general, Decapterus spp. was more abundant in the eastern fishing areas than those in the western part.High abundance of the fish in all of the fishing areas was in August until September.In March until April, stock of Decapterus spp. was concentrated around the fishing areas of the southeastern most of the Java Sea (Kangean and Sambergelap), composing of more than 80% of the total catch.The stock in these areas appeared to be sharply decreased in June, leading to the condition of low abundance in almost all of the fishing areas in the Java Sea with the exception of the area around Bawean in the west (Figure 7).Sardinella spp.made up an average of 8.4% to the total catch.lts availability was found to vary between fishing areas and seasons.However, ittends to concentrate in the fishing areas of Kangean inApril and Samber in May, composing of 45 and 34% of the total catch, respectively.Other species have greater variation in abundance and distribution pattern, both spatially, and seasonally.----I.cr\.me.ephthalmus -T* qeQapie.russpp.
--{sargtneila spp.. _ --*-S./ebtolepis.'A.Matrix correlation between six species of main pelagic fishes combined from six fishing areas is showed in Table 2. lt measured, through correlation coefficients, the relative intensity of interspecific covariation among the species.A positive correlation implies that for a given increase in abundance of one species, there is a corresponding increase in the other species.For a negative correlation, an increase in one implies a decrease in another.However.zo establishing the existence of a correlation does not imply causality.
Rastrelligerspp.exhibited a strong correlation with S. crumenphthalmus, Decapterus spp., Sardinella spp., and A. sirm (p<0.01).However, it had a strong positive covariation with the first three species and inversely to A. slm.Negative covariations were also showed between A. sirm and other species with different intensity.Auxis thazard and Ethynnus affinis showed a negative covariation with other species except for Decapterus spp.which showed a strong positive covariation indicating by significant level of correlation coefficient at p<0.01 .When this correlation matrix was subjected to ordination, it lead to species grouping as depicted in Figure 9.The species ordination was built based on principal components.

Rastrelliger spp., S. crumenophthalmus, and
Sardinellaspp. were considered belonging one group Spaflo Temporal Distribution of Smalt Petagic Fish in the Java Sea (Wiiopriono)   while the others were unique.This group was abundant between April and July where Sardinella spp.concentrated around Sambergelap and Kangean fishing areas, while Rasf relliger spp.and S. cerumenophthamus were mainly in Matasiri.The abundance tends to low in other months.Apart from this, the abundance of Decapterusspp.was high in August until October when abundance of A. sfm was low but it tends to decrease from November onward as the abundance of A.sirm decreased (see Figure 7 and 8).Abundance of A. thazard and E. affinis exhibited a similar pattern with Decapferus spp. in September but they were different in the other months.Map and surface plot showing the main fishing areas and seasonal relative abundance of pelagic fishes according to fishing area and season (month).

Spatial Agglomeration of the Fish
Bottom topography from central to eastern of the Java Sea is sloping toward the southeast.Along the area off south coastof Kalimantan is generally shallow waters with the depth ranged between 20 and 40 m.Taking longitudinal axis from northwest to southeast directions, the depth increased gradually and reaching the deep area of more than 80 m in the northern area of Madura lsland.Meanwhile, in the eastern of the study area there was a bank that called Matasiri Bank, located in the southern of Matasiri lsland with the depth of nearly g0 m.
Horizontal distribution of pelagic fish in the Java Sea in September until October is showed in Figure 10.The fish abundance is estimated from the average of every 5 nmi2 and presented in terms of relative biomass, which is expressed as the number of fish by sector (S fish nmi-2).th rh h h {a -t.-Figure 10 suggested that pelagic fish species tend to agglomerate and their abundance varied greafly between areas.Agglomerations of the fish were clearly showed in the western part of the study area, specifically in the northeastern part of Karimunjawa lsland and southern part of Kalimantan, at the pcsition between latitude of -4" and -5.8" and longitude of 110.6" and 113.3".They were also appeared in the eastern part of the study area, between Matasiri and Kangean lslands, with the geographical position between latitude of -5" and -6.2' and longitude 115.j" and 115.7'.Therewere no agglomerations of pelagic fish exhibited in the north coast of Java during this season except in the area between Bawean and Masalembo lslands, with the geographical position between latitude of -5.7" and -6.6" and longitude of 113.4" and 114.2".h   il* rao spatio Temporat Distribution of small Pelagic Fish in the Java sea (wiiopriono)   \, -l W,"]lu"'"t d*J,*^"'.'The estimation of the relative biomass revealed that the western and northern (off south coast of Kalimantan) of the study area had a greater abundance of the resources.The density of fishes ranged between 2.60x104 and 255x10a fish nmi 2 with the average of 92.4x10a fish nmi2.The highest abundance was found in the area off south of Kalimantan, between latitude of -4.66' and longitud e of 112-01'.In the eastern of the study area, between Matasiriand Kangean lslands, \ Kari^un,--" Surface plots showing the seasonal relative abundance of (A) Decapferusspp., (B) Sardinella spp , (C) Rastrettigerspp., according to fishing area and season (month)' the abundance was considered to be moderate with the density ranged from 0.79x10a to 99.61x104fish nmi-2.The average fish density within this area was 47.65x10afish nmi-2.Meanwhile, tn the area off north coast of Java the fishes were dispersed.The estimation of relative biomass revealed that the fish abundance over this area was relatively low, with the maximum density reached to only approximately 20x1}a fish nmi-2.Surface plots showing the seasonal relative abundance of (A) S. crumenophthalmus, (B) A. sfm, (C) A. thazard, and E. affinis, according to fishing area and season (month).Vertical cross-section of two meridional transects, north-south direction, the western and eastern parts of the survey area gave the illustration of fish distribution in different habitats during September-October (Figure 11).The meridional bottom topography of the Java Sea was clearly showed.As the sea bottom along the western transect (transect A) was nearly curveshaped, the deepest area was in the middle with the depth of less than 60 m.While.alonq the eastern ?n transect (transect B), itwas more southward with the depth of more than 70 m.
Abundance of pelagic fish was found to vary greatly along the water column, but in general, the abundance tends to increase with depth, specifically those from 10 m down to 50 m deep.The fishes mostly agglomerated above the bank and sloping bottom approaching the small islands vicinity.The agglomerations of high density were found to concentrate within the depth of 30 to 40 m.

Discussion
Commercial catch rates or catch per unit of effort data have been widely used in fish stock a'ssessment (e.9.Gulland,'1964; Kimura,'1981, Large,'1992).The Table 2. Spatio Temporat Distribution of Small Pelagic Fish in the Java Sea (Wijopriono)   use of these data rests on the assumption that catch per unit of effort is proportional to abundance.The effort is standardized in order to get representative index of the abundance.In this study, fishing day has been used as a standard of the effort and analysis of these time series catch per unit of effort data have heloed us to find the fisherv tendencies.
Matrix correlation of Pearson's product moment between six species of main pelagic fishes based on their monthly average catch per unit of effort (1998 to 2002)   Species ordination, based on principal component analysis upon catch weights (monthly average catch per unit of effort 1998 to 2002), of main pelagic fishes combined from six fishino areas.
Distribution of pelagic fishes based on hydro waytoobtainyearroundfiguresofspeciesdistribution acoustic analysis shows that the high abundance in the Java Sea.occurs in the areas around the vicinity of small islands and mostof theseareasarefishing groundsforpelagic Abundance of pelagic fish is seasonally variable fleets (Figure 1'1).The results of this study suggest in the Java Sea, reaching its peak in May for the that catch per unit of effort data collected from inshoreareaandSeptemberfortheoffshorearea.The commercialfishery and the resource abundances data catch is mostly driven by fluctuations in the abundance fromhydroacousticsurveyconveysimilarinformation, of one dominant species, Sardinella spp. in the in which that variations of catch per unit of effort value inshore and Decapterus spp. in the offshore.The two between fishing areas are in accordance with the species make up the average of 36 and 32ok of the variations of resource abundances derived from hydro total catch per unit of effort in the respective area.acoustic data.Although the commercialfishing areas do not cover the whole area of the populations, Thecatchperunitof effortvaluefoundtobemuch commercial fishery data can provide a cost effective lower in the Java Sea (1082.5 kg per day) than those in similar fishery existing in other parts of the world, e.9., Senegal and Chile (Saila, 1995).The scarcity of the resources has been revealed by the small number of fish shoals encountered during the hydro acoustic cruise.In most areas, the fishes are scattered in the water column and this spatial dispatch explains the use fish aggregating devices by fishermen, such as rumpon and light, in their fishing operations Pelagic species caught from the Java Sea are tropical species with a worldwide distribution (S.crumenophthatmus), Indo-pacific (R. kanagurta, A. sirm, and S. gibbosa) or a more restricted Southeast Asia (R. brachysoma, and S. fimbriata).Mostof them are typically shoaling fish which live mostly over continental shelves and make migrations across deep waters (R. kanagurta and D. macrosoma) (Fischer &  Whitehead, 1974).Their dispatch over the continental shelves is conditioned by the tolerance to salinities, some of them are euryhaltne (R. Brachysoma, S' gibossa, and S. fimbriata) some other stenohaline IR.Xanagurta, D. macrosoma, and D' russe//l) (Pelfish,1995).
Within any given community, there are a number of biotic and abiotic factors that influence the distribution, the abundance, and subsequently' the interactions of species.Depending on whether or not two species have similar or opposite respond to the same environmental factors, a certain pattern of interspecific covariation results' A positive covariation implies that the abundance of two species tend to increase and decrease together.For a negative covariation, an increase in one implies a decrease in the other.
The relative intensity of species covariation forthe main pelagic species has been measured on the basis of their correlation coefficients.lt is worth emphasizing that establishing the existence of a correlatton does not imply causality.Further, the detection of a statistically significant correlation between two species abundance patterns tell us nothing about the possible underlying reasons why this may be so' Ho*"u"t, the detection of significant interspecific covariation can be extremely helpful in generating suitable hypotheses to explain such patterns' lt may also be used as a basis for generating species grouping for which it has been done in this study' Based on the cluster and PCA analysis' Rastrettiger spp., S' crumenophthalmus, and   Sardinetlaspp.were considered falling into one group while the others were unique' This group was abundant between April and July where Sardinella spp' were concentrated around Sambergelap and Kangean fishing areas, while Rastreltiger spp' and S cruminophthamus were mainly in Matasirr' Their abundance tends to be low in other months' Two species of mackerels are caughtfrom the Java Sea, Rastrettiger kanagurta, and R' brachysoma' Potier & Sadhotomo (1995) found that most mackerels represented in the catches of large and medium purse seines were Rasf relliger kanagurta, while R brachysomawere accidentally present.On the other hand, R. brachysoma made up a bulk of the catch of inshore fleets (mini purse seines) in Rembang (Central Java) and Eretan (West Java), mainly between April and July (Hariati et at., 1995).These imply that R. kanagurta distributes more offshore in the eastern deep waters fishing areas and, conversely, R' brachysomadistribute more inshore in the centraland western shallow water of fishing areas' This phenomenon seems to confirm with ecological species related classificatron given by Longhurst &  Pauly (1987); Widodo (1995), which grouped R' kanigurta and R. brachysoma into oceanic and coastal sPecies, resPectivelY' Scads (Decapterus spp') were found to be more abundant in the eastern fishing area than the western' There are two species of scads (Decapferus spp )   caught in the Java Sea, l.e.Decapterus russe//l and D. macrosorra.Since they are grouped in one commercial category in fish landing data records, hence, impossible to give explanation with regards to the distinctive distribution pattern of the two species solely based on these data.However, referring to the investigation done during the extension of Pelfish   project (1991 to 1994), it revealed that D' macrosoma .otttyinhabited the waters of eastern part of the Java Sea and the Makassar Strait, while D' russe//lwas in the waters of Central Java and the South China Sea' Longhurst & Pauly (1987) classified D' macrosoma as Jceantc species, while D. russe//i was neritic specles.CONCLUSION 1. Variations of catch per unitof effortvalue between fishing areas derived from commerctalfishing data are in accordance with the spatial variations of resource abundances derived from hydro acoustic data.This reveals that bcth data are able to be used in complementary to explain the spatial abundance of small pelagic fish species and to give comprehensive explanation on theirdistribution oattern in the Java Sea.
2. Pelagic fishery resources in the Java Sea have undergone considerable variations in both their distributions and abundances over time' As reflected by the variations of catch per unit of effotl' the abundance reaches a peak in May in the inshore and September in the offshore fishing areas' lt is mostly driven by abundance variability of one dominant species, Sardinella spp' in the inshore and Decapterus spp. in the offshore' The two lnd.Fish Res.J. Vol.14 No.l June-2008: 21-35 species make up the average of 36 and 32o/o of the total catch per unit of effort, respectively.
3. The catch per unit of effort value in the Java Sea is 1082.5 kg per day per vessel.The small number of fish shoals encountered in the Java Sea during the hydro acoustic cruise reveals the scarcity of the resources.Results of hydro acoustic analysis show that the high abundances agglomerate in the areas around the vicinity of small islands.
However, in most areas the fishes are scattered in the whole water column and this spatial dispatch explains the use of fish aggregating devices, such as rumpon/unjam and light, in the fishing ... ... ....(3 Apr N{ay Jun Jul Aug Sept Oct n-ov Dec l\lonf h Rastrelligerspp. , A. dje(deba, O. acr.'ts * ..#... .Seasonal variation of catch per unit of effort of each pelagic fish species within the area of offshore (A) and inshore (B), averaged from 1998 to 2003.
Figure 10 from six fishing areas operations. 4. Based on theircovariation, Rastreiligerspp., Se/ar crumenophthalmus, and Sardlne//a spp.are included in one group.They are abundant between April and July where Sardinettaspp.concentrates around Sambergelap and Kangean fishing areas, while Rasfrelliger spp.and S. crumenophthamus are mainly in Matasiri.Their abundance tends to be lower in other months.