4 SURVEYS IN THE NORTHEAST INDIAN OCEAN AND SOUTH CHINA SEA

Between August 1978 and August 1983 eleven surveys were conducted in the Eastern Indian Ocean and a small part of the South China Sea. These surveys were conducted in support of various projects and often as a supplement to trawl surveys conducted by local vessels. The waters around Sri Lanka were covered three times (1978–80), those of Myanmar and Bangladesh twice (1979–80) and those of peninsula Malaysia, West Thailand and northwest Indonesia/part of Sumatra) only once in 1980. Three years later a special survey was conducted in the Maldives, that has been not considered in this report.

4.1 SRI LANKA, 1978–80

Survey objectives and effort

The surveys off Sri Lanka formed part of a general programme of development co-operation between the Governments of Norway and Sri Lanka and were sponsored and coordinated by NORAD. They were planned and conducted jointly by the Fisheries Research Station, Colombo, and the Institute of Marine Research, Bergen.

The objective was to provide information on the distribution, composition and abundance of the fishery resources on the shelf and describe their environment. The main method was to be echo integration with fishing for identification, sampling and gear trials. A special bottom trawl survey was planned for the Pedro Bank to assess the potential of this ground for a possible trawl fishery. Rough bottom on large parts of the shelf represented a limitation for a general use of the swept-area trawl method with pre-positioned fishing stations for estimating the biomass of demersal fish.

Table 4.1 Sub-divisions of the Sri Lankan shelf

The Sri Lankan shelf is shown in Figure 4.1 with the tracks from the first survey. The subdivisions used to describe the survey results are indicated in this map and described in Table 4.1. The northwest coast includes the Sri Lankan part of the Gulf of Mannar. The shelf of Sri Lanka is very narrow with a mean width of less than 15 nmi except for the Pedro Bank and the northern part. The slope is very steep and only small parts in the Gulf of Mannar and off the Pedro Bank were found to be fishable with trawl. The northern shelf, Palk Strait and Palk Bay, was too shallow for navigation with the vessel and could not be included in the investigations.

Figure 4.1 The Sri Lanka shelf with the sub-divisions used to describe the survey results and the tracks of the first survey

Table 4.2 gives some main data on the surveys. Survey I was at the end and survey II at the start of the southwest monsoon and survey III during the northeast monsoon. The pattern of survey tracks from survey I was repeated in the other surveys, but with more detailed coverage. Survey II included an additional programme of exploratory fishing in selected parts of the shelf.

Table 4.2 Sri Lanka: Review of survey information

The historical review of the state of the acoustic instruments presented in Chapter 2 showed that the conversion factors used to estimate biomass from integrator deflection corresponded to a target strength level (TS) of -35 to -36 dB/kg (for 17 cm fish). This target strength is thought to be too low and in this review an approximately 2 dB higher level has been applied which was obtained by reducing the biomass estimates presented in the first survey reports by about one third.

The results of the surveys were presented in three summary reports, Sætersdal and de Bruin, (1978), Blindheim et al., (1979) and Blindheim and Fø;yn, (1980).

Bottom topography

Based on interpretation of echosounder recordings a description of the bottom with regard to its suitability for trawling was made and presented in a chart which includes data from all coverages (Figure 4.2). Much uneven and rough bottom was found on the middle and outer shelf and especially from Colombo round the south coast beyond Trincomalee. Fishing with bottom trawl for identification and sampling had to be limited to smooth and not too uneven bottom and this may have caused bias as fish recordings off the bottom were often associated with rough bottom. Longlines were sometimes used for fishing trials on untrawlable bottom. The edge of the shelf was abrupt and the slope steep and unfishable in most parts.

Hydrography

The data showed a shift in the hydrographical environment around Sri Lanka with the monsoons. Figure 4.3 shows oceanographic conditions in two fixed profiles off the west (No. II) and southwest (No. IV) coast from each of the surveys: at the end of the southwest monsoon, at the start of the southwest monsoon and during the northeast monsoon. During the southwest monsoon the southward current brought surface water from the Gulf of Mannar and the southwest coast of India. The transitional layer lifted during the southwest monsoon and oxygen deficient water, less than 2–3 ml/l reached depths of less than 100 m on the shelf. During the northeast monsoon, in January-February the transition layer was found at more than 100 m depth, and the current was reversed with surface water moving west and northwest on the south and southwest coast.

Fig. 4.2 Observations of the character of the bottom on the shelf and upper slope

Figure 4.4 shows hydrographic sections VI and IX from the east coast and the Pedro Bank during the three surveys. The lifting of the transition layer during the southwest monsoon was pronounced with bottom water of oxygen content of 1 ml/l found at about 50 m of depth in August-September. The shoreward tilting of the isopleths indicated upwelling on the northeast shelf. The shifts in the oxygen content of the bottom water may have affected the distribution of demersal fish.

Measurements of primary productivity during the IIOE summarized in Cushing (1971a) showed a relatively high production around Sri Lanka during the northeast monsoon. Observations in this area were lacking in the southwest monsoon season, but satellite images of phytoplankton concentrations in the Indian Ocean from September-October showed a belt of high levels from the southwest coast of India continuing in the Gulf of Mannar and past the west coast of Sri Lanka, see Figure 10 Chapter 10. The general productivity on the shelf and adjacent waters around Sri Lanka should thus be expected to be fairly high.

Figure 4.3 Observed oceanographic conditions in each of the surveys in section II west coast and section IV southwest coast

Figure 4.4 Observed oceanographic conditions in each of the surveys in section VI southeast coast and section IX Pedro Bank

The acoustic surveys

Based on general experience of identifying echograms and on identification by fishing during the surveys in Sri Lanka waters, the echosounder recordings of fish were classified as follows:

Figure 4.5 shows examples of the three types of echograms. Type A was by far the most common recording. Type B was less common and often thought to have been “lost” in dense plankton recordings near the surface. Type C was stated to be quite common particularly in inshore waters in the first survey, but scarce in the two following surveys. These differences will be reflected in the biomass estimates based on the echo integration data.

The identification of the common type A recordings as semi-demersal fish was confirmed in a convincing way in a test programme during Survey II, when density estimates by acoustic integration and bottom trawling were compared (see description under “Fishing trials with the bottom trawl” below). Recordings of semi-demersal fish were encountered in all surveys in the Indian Ocean, but only in Sri Lanka were these fish found to be the main acoustic target.

Oxygen depletion in bottom water on the slope and shelf is known to affect the fish distribution and fish behaviour in many parts of the Indian Ocean and the oceanographic observations indicate that this environmental effect may be important also in Sri Lanka.

Mid-water occurrence of demersal fish is in many cases known to be more pronounced during nighttime. A repeat survey during day and night was made of a set of course tracks on the Pedro Bank and this showed about 30% higher acoustic estimates by night. A set of comparisons of mean recordings from day and night surveying of the west and southwest coast during surveys II and III did not, however, confirm this trend: in three out of four comparisons the day-levels were highest. It must be concluded that mid-water occurrence is a common behaviour of semi-demersal fish also at daytime in Sri Lankan waters. The prevailing strong currents on the south and west coasts, where the semi-demersal fish is most abundant may be another environmental factor of importance for determining their behaviour.

Figure 4.5 Examples of the three types of echo recordings:

Biomass of pelagic fish

The recordings of small and large-sized pelagic fish were not consistently distinguished in the surveys and therefore the biomass estimates are shown together in Table 4.3. On the west coast and the Pedro Bank mostly small pelagics were found, while large pelagics were found mainly along the east- and south coasts. The higher biomass from the post-monsoon survey (I) consisted mainly of small pelagics and may represent an effect of seasonal production. The pelagic trawl experiments for identification and sampling gave only low catches of small pelagic fish and the occurrence of this group in the catches from demersal hauls was also low thus confirming the generally low abundance of small pelagic in the parts of the shelf covered. Three species Sardinella sirm, Rastrelliger kanagurta and Decapterus russelli appeared in the catch records with about the same frequency.

The fish identified as large pelagics were assessed to have a biomass of some 40,000–60,000 t in the parts covered by the surveys. The source of these records was thought to be tunas and tuna-like fish and the distribution of the recordings, often on the offshore parts of the shelf indicated an oceanic continuation. The only type of these fish caught was Spanish mackerel (Scomberomorus commerson) which was probably the main representative of this group inside the shelf.

Table 4.3 Sri Lanka: Biomass estimates of pelagic fish by coastal regions and by surveys (1,000 t)

Biomass of demersal and semi-demersal fish

Table 4.4 shows the biomass estimates of semi-demersal fish, while Figure 4.6 shows their distribution in each survey. The highest densities in August-September and January-February were recorded at the Hambantota Bank and up the west coast past Colombo, while the April-May distribution indicated a shift up the west coast, which may be the result of migrations related to the northward moving current during the northeast monsoon. The densities are lower on the east coast and Pedro Bank. There is some decline in the estimates from 1978 to 1980, but this may be within the range of survey variability.

Figure 4.6 Distribution of echo intensity of semi-demersal fish.
Top: survey I, Aug-Sep 78; bottom left: survey II, Apr-Jun 79 and bottom right: survey III, Jan-Feb 80

Table 4.4 Sri Lanka: Biomass estimates of semi-demersal fish by coastal regions and by surveys (1,000 t)

Species composition of demersal and semi-demersal fish

Although possibly biased towards true bottom dwellers, the catches in the bottom trawl were the only source of information available on the composition of the semi-demersal fish. Table 4.5 shows the composition by the main families in successful hauls on the west and south coasts and Table 4.6 on the east coast and the Pedro Bank. The fauna in all of these areas was dominated by species with a preference for hard and sandy bottoms. The most common families, both by occurrence and weight of catch were snappers, emperors, groupers and sweetlips. The emperors dominated the catches on the west coast. The east coast and the Pedro Bank had a more varied fauna, surgeon fish among the more common, and with large pelagic fish, Carangidae, barracudas and Spanish mackerel. There was thus an indication of a difference in the fauna of the demersal and semi-demersal fish between the west coast and Hambantota and the east coast and the Pedro Bank, although many species were the same. The west and south regions sustained by far the largest biomass and appeared to be the most productive.

Table 4.5 Sri Lanka: Composition of semi-demersal types of fish in bottom trawl catches. Incidence (%) and % of total catch by weight. West and south coasts, all surveys pooled

Table 4.6 Composition of semi-demersal types of fish in bottom trawl. Incidence (%) and % of total catch by weight. East coast and Pedro Bank, all surveys pooled

The catches were only partly identified to species level. The following species were indicated as being the most common within the respective families listed in order of frequency of recording in the catch records:

Most of these are commercial fish of high value, and many may attain a large size. The samples of size distributions from Surveys II and III are summarized in Table 4.7. Two regions of distributions were considered for purposes of weighting: the west coast and Hambantota Bank and the east coast and the Pedro Bank. The size distributions from each of these regions were weighted with a rough estimate of the family's proportion of the “acoustic biomass” in the regions.

Table 4.7 shows an abundance of large-sized fish of these generally slow-growing species and thus gives an impression of lightly fished stocks. These data may serve as a source of reference for later studies of the state of the stocks. The original more detailed observations are available in the NAN-SIS data base.

Table 4.7 Sri Lanka: Size compositions of selected species of demersal and semi-demersal fish of surveys II and III. Fork length (%)

Fishing trials with bottom trawl

For reference purposes, a review of catch rates in the bottom trawl is presented. The objective of the fishing was usually identification and sampling of fish recorded on the echosounder. Unsuitable bottom was often a limitation, preventing fishing on dense echo recordings. A swept-area survey was made of parts of the Pedro Bank.

Table 4.8 Gulf of Mannar: Catch rates with the bottom trawl from the deep-sea ground (kg/h)

A small deep-sea trawling ground in the Gulf of Mannar around 8°45'N and 79°32'E, located in 1972 by the R/V OPTIMIST was tested in each of the surveys. The catch rates are shown in Table 4.8. The lobster was the whip lobster Puerulus sewelli which is also found off Somalia, SW India and the Tenasserim coast of Myanmar. The size was relatively small, 13– 14 cm total length for females and 11–13 cm for males. The shrimps were identified as Aristeus semidentatus, Heterocarpus gibbosus, Parapandalus spinipes and Metapenaeopsis andamanensis. Chlorophthalmus agassizi dominated the deep-sea fish.

Table 4.9 shows catch rates for fishing on the shelf of the northwest coast. Demersal and semi-demersal fish were as shown in Table 4.4, principally emperors, snappers and sweetlips. The catch rates were particularly high in the pre-monsoon survey II.

Table 4.10 shows the catch rates for the southwest coast. Emperors, represented mainly by the spangled emperor (Lethrinus nebulosus) dominated the catches of demersals as on the northwest coast with snappers and groupers following. The highest rates were from the post-monsoon season.

Table 4.9 Sri Lanka, northwest coast: Catch rates with the bottom trawl (kg/h)

Table 4.10 Sri Lanka, southwest coast: Catch rates with the bottom trawl (kg/h)

Unsuitable bottom was a severe limitation for trawl fishing on the Hambantota Bank (Table 4.11). Tests with bottom longlines gave catch rates up to 200 kg/200 hooks. Snappers, emperors and groupers represented more than 80% of the catch.

The catch rates were lower on the east coast (Table 4.12), where sweetlips and surgeonfish were added to the fauna of demersal fish.

On the Pedro Bank catch rates were highest in Survey II (Table 4.13).

Table 4.11 Sri Lanka, Hambantota Banks: Catch rates with the bottom trawl (kg/h)

Table 4.12 Sri Lanka, Batticaloa Banks: Catch rates with the bottom trawl (kg/h)

The trawl coverage of the Pedro Bank in Surveys II and III allowed estimates of the demersal fish biomass by the swept-area method. Table 4.14 shows the summary data and biomass estimates based on a catchability coefficient of 1. The acoustic estimates of the semi-demersal fish on the Pedro Bank were about 10,000 t in each of the two surveys. These may have included fish out occurring at depths above the headrope of the bottom trawl.

Table 4.13 Sri Lanka, Pedro Bank: Catch rates with the bottom trawl (kg/h)

Table 4.14 Sri Lanka, Pedro Bank: Summary of data on swept-area trawl survey inside 100 m of depth

Figure 4.7 Relationship between estimates of densities of semi-demersal fish from catch rates in bottom trawl and acoustic integration of bottom channel. +: pelagic fish; o: demersal fish. Source: Blindheim et al., 1979

A further comparison of trawl data and acoustic data was made in Survey II based on estimates of “acoustic density” from logging of integrator deflection in a 5 m bottom channel during towing and estimates of “trawl density” from catch rates and swept-areas. Figure 4.7 shows the plots from shallow hauls with a regression line estimated for the hauls in which recordings were identified as demersal fish, giving a correlation coefficient of 0.92. This relationship represents a strong confirmation of a correct identification of this type of recording.

Review of findings and later development in fisheries

The main results of the surveys as presented in the three summary reports and at a follow-up seminar in Colombo in August 1980 can be summarized as follows: the total biomass on the west, south and east shelf was 400,000–500,000 t with seasonal variation. The most important component of this was demersal and semi-demersal fish assessed at 250,000–350,000 t and consisting of emperors, snappers, groupers, sweetlips, Carangidae etc. The potential total yield from this group would be 50,000–70,000 t. Some three-quarters of these resources were located on the west and south coasts and it was recommended that they be exploited with gears other than trawls. These stocks seemed to be lightly exploited offering potential for expansion of the fisheries. The potential yield from a possible trawl fishery on the Pedro Bank was estimated at 2,000 t. The biomass of the small and large pelagics in the surveyed area was indicated at 100,000 t.

The shelf in the north - the very shallow Palk Bay and Palk Strait of about 2,800 nmi² could not be covered by the survey, but was reported to be productive with important shrimp resources and fish stocks dominated by ponyfish (Leiognatidae).

Sri Lanka's fisheries developed in the small scale sector in the years following the surveys with reported total marine landings increasing from about 150,000 t in 1979 to about 180,000 t in 1983 (FAO Yearb.Fish.Stat. Vol 50 and Vol 60). Demersal percomorphs formed part of this increase with reported landings growing from about 18,000 t to about 24,000 t. Available statistics from 1984 onwards show declining landings and changes in reporting by groups, trends which most likely must be related to political events in the country since 1983. The survey data may serve as references in future assessments of the state of the resources.

4.2 MYANMAR (BURMA), 1979–80

Survey objectives and effort

The plan of the UNDP/FAO Project BUR/77/003 “Marine Fisheries Resources Survey and Exploratory Fishing” submitted in March 1979 included approximately six months of surveys with the DR. FRIDTJOF NANSEN. The work programme for these surveys was developed jointly by representatives of The People's Pearl and Fisheries Corporation of Rangoon, FAO and IMR. The objective of the project, to which these surveys were expected to contribute, was to make an estimate of the marine fish biomass within the EEZ of Myanmar, and in particular, over its continental shelf. The estimate should serve as a basis for preliminary and tentative assessments of sustainable yields and, consequently, rational planning of further investment in the fishery industry (UNDP, 1979). The work programme as agreed in August 1979 specified the principal method as acoustic estimation of the biomass of demersal and pelagic fish, with fishing for identification and sampling purposes and for assessment of catch rates. Environmental work was to include recording of type of bottom, and hydrographical profiles from the coast to 500 m depth for temperature, salinity and oxygen.

The surveys were timed to cover two seasons, the post-southwest monsoon season in September-November 1979 and the pre-southwest monsoon season in March-April 1980. During each of these main surveys, the waters between the border with Bangladesh in the north and with Thailand in the south were covered twice. Figure 4.8 shows as an example the course tracks of the first coverage. The first and third cruises represented fixed-grid overviews, while the research efforts of the second and fourth cruises were concentrated in areas which in the preceding fixed grid survey were found to be of particular interest regarding fish abundance and distribution. The main fixed grid transects were spaced approximately 30 nmi apart and ran from a bottom depth of about 15 m seawards. During the third survey, the Coco Island region was also covered. The results were grouped by season and presented by three main areas: the Arakan coast (I), the Delta (II), and the Tenasserim coast (III). The deep-sea grounds, at 200–400 m depth off the southern coast, are also presented separately.

Table 4.15 shows the extent of the shelf by sub-areas. Relatively wide shallow inshore parts could not be covered, especially in the Delta, because of the approximately 15 m depth limit needed for the safe operation of the vessel. More than 90% of the surveys were conducted inside the shelf with the rest on the southern deep-sea grounds, but hydrographic profiles were extended off the shelf. The mean survey intensity by coverage (measured as sailed distance over the square root of the area) was high for both surveys.

The main part of the shelf is shallow, less than 100 m, and this allowed use of the 120 kHz echosounder for acoustic integration which has limited depth range. The advantage of this choice was that this higher frequency permits (as explained in Section 2) an easier distinction between recordings from fish and recordings from spurious sources, plankton, etc. The results of the surveys in Myanmar were presented in preliminary cruise reports for each survey and in two summary reports, Nakken and Sann Aung (1980) and Strø;mme et al. (1981).

Figure 4.8 Survey tracks from the first coverage. Area I: Arakan coast; Area II: the Delta; Area III: the Tenasserim coast

Table 4.15 Myanmar: Shelf areas and parts covered; survey distances and acoustic degree of coverage

The sea bottom

Based on the interpretation of echosounder recordings, a description of the bottom with special reference to its suitability for trawling was made and presented on charts which included data from all coverages (Figure 4.9).

Figure 4.9 Bottom conditions by main areas.
Legend: 1 smooth even; 2 smooth uneven; 3 rough; 4 steep

The shelf off the Arakan coast has wide areas of smooth ground in the shallower parts, but these are in places interrupted by rocks, corals and sometimes mud volcanoes. The slope is steep, rough and generally unsuitable for trawling.

The Delta area is characterized by a wide band of shallow, smooth and gently sloping bottom created by silting from the Irriwaddy and Salween rivers. At about 15° N the bottom deepens more markedly from 20–30 m to about 100 m and in this region, which extends 40–60 nmi southwards, the bottom is variable with good trawling grounds interrupted by more uneven or even rough areas.

The wide shelf off the Tenasserim coast is more varied. Beyond the 200 m depth line which lies 60–100 nmi offshore, the slope down to about 400 m is not very steep, and a deep-water ground with generally smooth bottom extends over a wide offshore area from about 13°N southwards to the border with Thailand. There are also extensive areas of trawlable bottom further inshore towards the archipelago. Between the islands and in the inlets, conditions vary greatly between even, smooth bottoms and rocky grounds.

Hydrography

The great seasonal changes in the hydrographical conditions on the shelf caused by the shifts in the monsoons and in river discharges were clearly demonstrated by some of the hydrographical observations during the two surveys. As shown in Figure 4.10, in September-October the surface water off the Delta and Arakan coasts was extensively mixed with freshwater from the river runoffs, resulting in surface salinities of 20–30° over large areas in the Delta region and northwards along the Arakan coast, and indicating a westward and northward transport of the uppermost coastal water masses. In March-April the highest salinities were found inshore, an effect of the northeast monsoon.

Figure 4.10 Salinity at 5 m depth 25 Sept-18 Oct 1979 (left) 5 March-1 April 1980 (right)

In the intermediate and deeper water masses distinct seasonal changes were evident from the distribution of temperature, salinity and oxygen in selected transects in September-October and March-April from the three parts of the coast (Figure 4.11). The main features were similar all along the coast: in September-November the transition layer between the upper homogeneous water masses and the deep water was found at depths between 70 and 150 m, while in March-April it was much shallower at 20 to 100 m depths. This lifting of the transition layer also brought poorly oxygenated water (between 1 and 2 ml/l) over large parts of the shelf. These observations confirmed previous evidence of upwelling during the northeast monsoon (Cushing, 1971b) with resulting enhanced primary production. In deeper waters, below 150–200 m, conditions were more stable. However, the difference in oxygen content between the deep water in the Bay of Bengal (Arakan) of: less than 0.2 ml/l, and that in the Andaman Sea (Delta and Tenasserim): not less than 0.8 ml/l should be noted.

Figure 4.11 Distribution of temperature, salinity and oxygen in the hydrographical sections

In addition to the effect of seasonal upwelling on enhancing productivity it is also argued that the sea off Myanmar may receive additional contributions of nutrients from river discharges.

The acoustic surveys

The recordings of fish by the acoustic systems were classified into two types based on experience in identifying echograms and on identification from sample fishing during the surveys: Type A: recordings of true larger schools or dense layers mostly in the upper water, were identified as deriving from pelagic schooling fish usually of smaller size (e.g., Clupeidae, Engraulidae, Carangidae, but also including Trichiuridae). Type B recordings showed looser aggregations or smaller schools near or at least closer to the bottom. While type A was often found at high densities this was seldom the case for type B, which was ascribed to demersal or semi-demersal fish. The identification of species in this demersal group was indicated by the relative abundance of the various types of fish occurring in bottom trawl catches (Strø;mme et al., 1981). This identification must be assumed to have been biased towards more strictly demersal fish, but indications of the more likely mid-water targets of this group are given below, in part based on a general knowledge of the behaviour of semi-demersal fish.

The historical review of the state of the acoustic instruments presented in Section 2 shows that the conversion factors used to estimate biomass from echo integrator deflection corresponded to a target strength level of -36 dB/kg (for 17 cm fish). Even if this early generation of echo integrators must be expected to have produced underestimates because of saturation phenomena in both the receiver and the integrator, a target strength of -36 dB/kg seems low and therefore an approximately 2 dB higher TS-level has been used in the present review, thus reducing the biomass estimates presented in Strø;mme et al. (1981) by about one third.

4.2.1 The Arakan coast

Pelagic fish

The distribution of the small pelagic fish differed little between the two surveys and the main part of the biomass was found on the southern inshore shelf (Figure 4.12). The biomass corrected estimates were also similar:

Appearance in the catches was the only source of information on the composition of this group although the varying catchability by size and species must be expected to have distorted these data. In these shallow waters small pelagic fish were caught both in pelagic and demersal trawls. Table 4.16 shows the occurrence of the pelagic fish by family in all trawl catches on this coast. Clupeidae and Engraulidae dominated especially by weight, followed by Trichiuridae and Carangidae.

Information on the species composition within each group was derived from their relative proportion in the weight of the catches of the three respective families in the pelagic trawl hauls (Table 4.17).

Figure 4.12 Distribution of small pelagic fish on the Arakan coast, September-November 1979 (left) and March-April 1980 (right)

Table 4.16 Arakan coast: Occurrence % in all pelagic and bottom trawl hauls of families of pelagic fish and their proportion by weight of total catch of pelagic fish (%)

Among the hairtails, Lepturacanthus savala was more common than Trichiurus lepturus. The Scombridae included Rastrelliger kanagurta as well as both Scomberomorus guttatus and S. commerson and two species of barracudas, Sphyraena barracuda and S. obtusata were caught.

Table 4.17 Arakan coast: Species composition of catches by pelagic trawl shown by proportion of total weight within families (%)

Table 4.18 Arakan coast: Occurrence (%) in all pelagic and bottom trawl hauls of families of demersal and semi-demersal fish and other groups and their proportion by weight of the total catch (%)

Semi-demersal fish

The corrected biomass estimates for this group on the Arakan coast were:

The information pertaining to the identification of this group was limited to the catch composition shown by family in all hauls in the area shown in Table 4.18, and to general information on the behaviour of species which tend to lift off the bottom.

From this it appears that the main contributors to the biomass identified as semi-demersal fish were species of the families Leiognatidae, Gerreidae, Synodontidae, Pomadasyidae, Nemipteridae and Lactariidae.

Fishing operations

An overview of catch rates recorded is presented in Table 4.19.

Table 4.19 Arakan coast: Summary of fishing operations

Part of the September-November operations related to a special shrimp survey near Sandoway; otherwise most of the fishing in this area was conducted for the purpose of identification and sampling and was of limited use for stock estimates.

A classification of the various species of fish into four economic categories was made by Strø;mme et al. (1981) based on local prices. About 80% of the catch was found to be of intermediate value, classes 2 and 3. Mean catch rates of these were fairly high, often exceeding 200 kg/h.

4.2.2 The Delta

Pelagic fish

Figure 4.13 shows the distribution of the pelagic fish recorded in the two surveys, with the main part of the biomass in a 40 nmi wide belt covering the bottom depth contours from about 15 to about 60 m. The estimates of biomass (almost doubling between the two surveys) were:

Table 4.20 shows the occurrence of the pelagic fish by family in all trawl catches in the Delta area. Trichiurids, Engraulidae and Clupeidae dominated, with Carangidae following. There is an increased proportion of Clupeidae in the second survey.

Table 4.21 shows the species composition as a percentage of total catch within each family. The effect of the shallow-water muddy delta was demonstrated by the higher number of both Clupeidae and Engraulidae, including the shallow water genera Ilisha and Thryssa. Lepturacanthus savala dominated the abundant Trichiuridae and Decapterus russelli the Carangidae. Rastrelliger kanagurta appeared in low quantities in both surveys. Scomberomorus commerson was more common than S. guttatus and Sphyraena obtusa more common than S. barracuda.

Figure 4.13 Distribution of small pelagic fish in the Delta area, September-November 1979 (upper) and March-April 1980 (lower).

Table 4.20 Delta area: Occurrence (%) in all pelagic and bottom trawl hauls of families of pelagic fish and their proportion by weight of total catch of pelagic fish (%)

Table 4.21 Delta area: Species composition of catches by pelagic trawl shown by proportion of total weight within families, (%)

Semi-demersal fish

The adjusted biomass estimates of this group in the Delta area were:

The higher biomass in the second survey could be related to the seasonal lifting of the transitional layer which brought bottom-water of less than 2 ml/l of oxygen up to a depth of less than 50 m. The main part of the semi-demersal biomass was found at shallow depths in March-April. In part of the Delta area difficulties in interpreting echo recordings were experienced because of occurrence both in mid-water and near the bottom of small shrimp of the genus Acetes.

The composition of the demersal and semi-demersal fish in this area is given in Table 4.22. The Delta fauna differed in several respects from that of the coastal regions to the north and south. Noteworthy was the presence of threadfins (Polynemidae), pike congers (Muraenesocidae) and Bombay duck (Harpadon nehereus). Sciaenidae were more common, but ponyfish less frequent than in the other areas.

Table 4.22 Delta area: Occurrence (%) in all pelagic and bottom trawl hauls of families of demersal and semi-demersal fish and other groups and their proportion by weight of the total catch (%)

Fishing operations

An overview of catch rates recorded is presented in Table 4.23.

Table 4.23 Summary of fishing operations in the Delta area

Bottom trawling during the second survey formed part of a special survey of the eastern Delta to obtain bottom trawl estimates of demersal fish, especially the Bombay duck. The maximum catch rates quoted represent the mean of the three highest catches.

An economic analysis showed that the fish of the Delta area tended to be of higher value than off the Arakan coast, with nearly half the catch in the higher price categories classes 1 and 2, and yielding catch rates of about 200 kg/h.

4.2.3 The Tenasserim coast

Pelagic fish

Figure 4.14 shows the distribution of pelagic fish on this coast. During the second survey areas of high abundance formed a belt some 20 nmi wide in the waters of the archipelago. The adjusted estimates of biomass were:

Table 4.24 shows that Engraulidae formed the main part of the biomass, with some contribution from Clupeidae in the second survey. That the Engraulidae were by far the most abundant pelagic group was further demonstrated by the proportions in the pelagic hauls only. Of an overall total catch of Clupeidae, Engraulidae and Carangidae from this gear of about 9 t in March-April, Engraulidae (anchovies) represented more than 8 t.

Table 4.25 shows the species composition of the main families. The rainbow sardine (Dussumieria acuta) was the most common clupeid, and the Indian anchovy (Stolephorus indicus) by far the most common engraulid. Rastrelliger kanagurta was caught in small quantities in nearly all hauls. Scomberomorus guttatus was more common than S. commerson and Sphyraena obtusata more common than S. forsteri and S. barracuda.

Figure 4.14 Distribution of small pelagic fish on the Tenasserim coast, September-November 1979 (left) and March-April 1980 (right)

Table 4.24 Tenasserim coast: Occurrence (%) in all pelagic and bottom trawl hauls of families of pelagic fish and their proportion by weight of total catch of pelagic fish (%)

Table 4.25 Tenasserim coast: Species composition of catches by pelagic trawl shown by proportion of total weight within families (%)

The great increase in biomass from September-November to March-April was most probably brought about by a seasonal production of small pelagic fish, mainly the short-lived Indian anchovy, a production which seems likely to relate to the upwelling during the northeast monsoon.

Semi-demersal fish

The adjusted biomass estimates of this group off the Tenasserim coast were:

The densities of this group were considerably lower than in the northern regions. As shown in Table 4.26 the most common forms in the catches were ponyfishes (Leiognathidae), grunts (Pomadasyidae), lizardfishes (Synodontidae) and false trevallies (Lactariidae). The very high proportion of ponyfish in the last survey was caused by exceptionally high catches in shallow waters. More characteristic for the area was perhaps the relatively frequent occurrence of threadfin breams (Nemipteridae) and goatfishes (Mullidae).

Table 4.26 Tenasserim coast: Occurrence (%) in all pelagic and bottom trawl hauls of families of demersal and semi-demersal fish and their proportion by weight of the total catch (%)

Fishing operations

An overview of catch rates is presented in Table 4.27.

Table 4.27 Tenasserim coast: Summary of fishing operations

The catch rates were relatively high, but since 60–80% of the catch was found to consist of species in economic class 3 the value of the bottom fish on the Tenasserim coast appears to be lower than in the northern parts.

Special trawl surveys of the deep-sea grounds

The gently falling slope in the Andaman Sea southwards from 13°N forms extensive grounds some 200 nmi long and 10–40 nmi wide within the depth range 200–500 m characterized by a predominantly even bottom suitable for trawling. Figure 4.15 shows the grounds and the distribution of fishing stations during the two surveys. Here the ambient conditions near the bottom with temperatures of 11–12° C and an oxygen content of about 0.8 ml/l were like those found on similar grounds off the southwest Coast of India and off Sri Lanka in the Gulf of Mannar, deep-sea shrimp and lobster have been found in quantities of commercial interest.

The gear used by the DR. FRIDTJOF NANSEN, an ordinary shrimp-cum-fish trawl equipped with bobbins, was likely to have had a relatively low efficiency for deep-sea shrimp and lobster. Table 4.28 shows the catch rates recorded for the 38 deep-water hauls.

Figure 4.15 The coverage of the deep-sea grounds on the slope off the Tenasserim coast

Table 4.28 Catch rates on the deep-sea ground on the Andaman Sea (kg/h)

The shrimps were Heterocarpus sp., Aristaeus semidentatus and others. The lobsters were identified as Puerulus sewelli. Commercially these catch rates were not very promising and they were considerably lower than those obtained in the Gulf of Mannar, but the total area of distribution of lobster was very extensive and further fishing trials were recommended to see whether areas or seasons of higher concentrations might be found (Strø;mme et al., 1981). Rawcliffe (1983) reported promising catches of deep-water lobster from further trials in this area.

4.2.4 Review of findings and of later research and development of fisheries

Biomass estimates by acoustic methods

The main findings were the biomass estimates summarized in Table 4.29. With totals of small pelagic and semi-demersal fish of 805,000 t in September-November 1979 and 1,600,000 t in March-April 1980. The doubling of the biomass between the two surveys, was probably related to the observed changes in the environment: enhanced production through upwelling during the northeast monsoon and the physical concentration of bottom and semi-demersal fish due to uplifting of oxygen-deficient bottom water. Obvious limitations of the surveys included incomplete coverage of extensive inshore shallow waters, and the deficiencies of the first generation echo integration system, both of which were likely to have caused underestimates.

Table 4.29 Myanmar: Estimated biomass for the shelf areas based on acoustic methods (1,000 t)

It should also be noted that the main objective of the survey was the investigation of fish with the echo integration method. The demersal fish on the shelf were thus incompletely covered since the programme did not include a general swept-area trawl survey. (A series of such surveys were carried out from November-December 1981 to March-April 1983, see below.)

Densities of fish measured as biomass per unit shelf area showed average values of 12 t/nmi² in September-November and 24 t/nmi² in March-April. These represented low to moderate levels, indicating only a moderate effect on the total production from the upwelling caused by the northeast monsoon and from the considerable river discharges. For comparison, a similar estimate of mean density of biomass measured by acoustics on the Malabar shelf (see Section 3.2.1) with its May-September upwelling during the southwest monsoon, was 67 t/nmi².

Taking the simple mean of the two biomass estimates as mean standing stock and yield fractions of 0.5 and 0.25 for small pelagic and semi-demersal fish respectively, theoretical potential annual yields of about 340,000 t and 130,000 t were obtained for the two groups respectively, a total of 470,000 t. Since the stocks are already exploited, the 1979 landings were reported to be 400,000 t, the total potential would be higher than the above estimate and can roughly be assessed at 600,000 t. A yield fraction of 0.5 used for the small pelagic fish may, however, not to be achievable for a predominantly small-scale fishery, since the highly seasonal biomass of small pelagics such as the anchovies cannot be expected to be fully utilized.

Biomass estimates from bottom trawl surveys

Table 4.30 shows a summary of the results of the fishing operations with the bottom trawl. It should be noted that these result from aimed fishing for identification and sampling. The rates may perhaps approach those of commercial fishing.

Table 4.30 Myanmar: Summary of fishing operations with the bottom trawl

A special swept-area trawl survey of a section of the eastern part of the Delta in April showed a mean density of 21 t/nmi² over approximately 6,000 nmi², assuming a catchability coefficient of 1. The total biomass of 126,000 t may be compared with the acoustic estimate of 210,000 t for semi-demersal fish in the same area. In part of this area Bombay duck was important with a density of 15.7 t/nmi², corresponding to a biomass of 42,000 t over an area of 2,700 nmi².

Shallow-water shrimp

A proper survey of the shallow water shrimp resources would have required a different type of vessel, gear and survey design. Shrimp formed, however, part of the catch in a number of hauls in shallow waters in the Delta and on the Arakan coast and the catch rate data obtained are summarized in Table 4.31. In order to limit the study to areas where shrimp was reasonably abundant only those hauls with shrimp catches exceeding 10 kg/h were included in the analyses. Nearly all of these hauls were from the 10–24 m depth zone. The small “white” shrimp, mainly Acetes spp. were excluded. Apart from giving positive indications, the information was of limited value in view of the assumed low efficiency of the gear for shrimp. Metapenaeus spp. dominated the catches with Penaeus as the second most important genus.

Table 4.31 Myanmar: Analysis of catch rates of shallow-water shrimp

Surveys with other vessels

Further important information on Myanmar's marine fish resources was obtained during an extensive programme of bottom-trawl surveys during 1981–83 also organized under the FAO/UNDP Project BUR/77/003 “Marine Fisheries Resources Survey and Exploratory Fishing”. The main objective of this programme was to supplement the acoustic surveys of pelagic and semi-demersal resources undertaken with the DR. FRIDTJOF NANSEN by determining the qualitative and quantitative distribution of the demersal resources (Rijavec and Htun Htein, 1984). More than 600 trawl hauls were made, distributed over four surveys: two in November-December of 1981 and 1982, two in April-May of 1982 and 1983.

The catch rates obtained from those trawl surveys are presented in Table 4.32 (from Rijavec and Htun Htein, 1984).

These results from the prepositioned semi-random sampling system executed by BUR/77/003 were, as expected, considerably lower than the catch rates from the aimed fishing with the DR. FRIDTJOF NANSEN, even though the fishing gear was roughly of the same size albeit of different design.

Table 4.32 Myanmar: Catch rates (kg/h) obtained in 1981–83 from a bottom trawl survey of project BUR/77/003

Rijavec and Htun Htein (1984) used their data for a swept-area assessment of the demersal fish biomass, where the catchability coefficient q was assumed to have a value of 0.5. This assumption was based on considerations of the design of the gear, cut-out lower wings, rubber bobbins groundrope, and on the results of some comparative fishing trials. The resulting estimate was 750–800,000 t, on the basis of which the authors estimated a potential yield of demersal fish in the range of 300,000–550,000 using a range of M values of 0.4–1.0, and an estimate of existing landings of 300,000 t.

Their estimate was based on catch rates from daylight hauls only, which on average were 65% higher than the night hauls. The estimate of 750–800,000 t can be compared with acoustic estimates ranging from 370,000 to 670,000 t of semi-demersal fish (day and night observations) from the DR. FRIDTJOF NANSEN surveys. The limitation of acoustic observation of fish close to the bottom has been much discussed, but is difficult to quantify. The ratio between the estimates from these two surveys was at least not unreasonable.

In a review of the available information on the marine fish resources of Burma, Pauly et al. (1984) suggested a yield estimate based on an overall M value = 0.6, a total biomass of 800,000 t demersal fish (from the 1981/83 trawl survey) and 1 million t of pelagic fish (from the DR.FRIDTJOF NANSEN surveys). With an existing catch of 300,000 t this gave a potential annual total yield of 700,000 t. However, the authors present this estimate with reservation, relating both to the validity of the method used to calculate the potential from the biomass data, and to the feasibility of Myanmar's predominantly small-scale fisheries effectively exploiting the small pelagic fish stocks. The latter reservation may be related to the lack of success of experimental fishing for small pelagic fish during special programmes of the UNDP/FAO Project BUR/77/003 “Marine Fisheries Resources and Exploratory Fishing” (Anon., 1982). However, these trials did not cover the pre-monsoon season when the high densities of small pelagics were observed in the DR. FRIDTJOF NANSEN surveys.

Both of the above reservations might mean that the 700,000 t potential was an overestimate. On the other hand Pauly et al. (1984) found that the demersal resources of the shelf of Myanmar could not have been in a state of heavy exploitation. An analysis of the composition of main groups in four trawl surveys over the period 1953–83 showed no trends of change, in contrast to the fauna of the intensively fished Andaman coast of Thailand which comprises similar stocks.

The revised estimate of 700,000 t of small pelagic fish presented in this review would, with the assumptions used by Pauly et al. (1984), give a potential total yield of 600,000 t. Unfortunately, the calculation of yield from biomass involves many uncertainties especially with the many different species of fish involved. The DR. FRIDTJOF NANSEN surveys provided, however, an extensive set of data describing the distribution, composition and abundance of the standing biomass of the main marine fish resources of Myanmar at a relatively early stage of development of light exploitation.

The reported marine fish landings of Myanmar were about 400,000 t in 1979 (FAO, 1984) and increased to nearly 600,000 t in 1990/91 (FAO, 1993). There thus seems indeed to have been a potential for increased catches. In studies of the effects of fishery development on the stocks, the results of the 1979–83 surveys may serve as important benchmarks.