3. RESULTS AND DISCUSSION

In this section the results and discussion are presented separately for shrimp farming and cage culture. In each part the results obtained using the original, pre-verification ratings are presented. These are then compared with the findings from the field verifications. Finally, the shrimp farming results are compared with a number of computer simulations that relate various siting decisions to capital and operating costs (Muir and Kapetsky, in press). Base conditions for the simulation relate to actual Johor shrimp farm conditions in the following way:

Additional comparative aspects are that both the base model and most Johor shrimp farms use aeration and supplemental feeding.

3.1 OPPORTUNITIES FOR SHRIMP FARMING

3.1.1 Suitability of Soils

Soil texture and pH were deemed the most important of the physical criteria. Soils were rated on a scale of five, from excellent to unsuitable, separately for pH and texture. The ratings are inversely related to the amount of extra time or extra cost required to make the soils usable for shrimp culture. For soils rated “poor”, special design, construction, or management would be required to overcome or minimize the limitation. For example, long-term flushing, higher pumping rates or liming could be required in the case of acid soils. For soils too sandy, lining of ponds with clay, extra dike material, or extra compaction, could be required.

The results are summarized according to the coastal areas covered by the original soil maps (Table 4): West Coast, South and Southeast Coasts and Northeast Coast.

The coastal strip of Johor within 2 km of brackishwater occupies 193 400 ha. According to the rating scheme laid out in section 2.5.2, 70 155 ha (36 percent) are soils which rate good or fair with regard to pH. No soils rated excellent for pH (Table 10).

Texturally, 74 percent of the coastal area within 2 km of brackishwater has soils which rate fair, good or excellent for pond construction according to the rating scheme laid out in Section 2.5.2 (Table 11).

Ideally, soils for shrimp farming would rate at least fair for both texture and pH. Thirty-six percent of the coastal soils, nearly 70 000 ha, are in this category (Table 12). The largest part of those, 59 700 ha, or 31 percent of the coastal soils, have excellent textures and good pH suitability according to the original rating criteria.

3.1.2 Field Verification of Soils

Eighteen shrimp farm sites were accounted for in Johor of which two were under construction and two were out of business (Table 13). Of these, 14 sites were verified by site visits including one under construction and two no longer operating. The Brackishwater Research Station at Galang Patah, engaged in experimental shrimp farming, was included in this group as a “control”.

3.1.2.1 Soil acidity

Eight farm sites occur on kranji soil. One was under construction and two were not operational. The Brackishwater Research Station at Galang Patah is on kranji and had had severe acid problems in its early years. Overcoming these problems was the focus of much applied research reported by FAO (1984). Of the operational shrimp farms on this soil, one was liming at the rate of 3.5 t/ha/cycle, another at 1 t/ha/cycle, and a third at an unknown rate. The ponds on one farm had been constructed so that there was minimal disturbance of the soil and consequently pH problems were minimal. At two other farms jarosite and iron precipitation were evident. Other farms on kranji had management problems stemming from poor design from which it was difficult to isolate pH-related difficulties.

Sihibi Mohktar (pers.comm.) verifies that kranji is a soil with acid potential. It is the only coastal soil in Johor with this characteristic. The present pH rating for the kranji (KRJ) series soil (pH score 4, “good”) is too high and should be decreased to “poor”. This change is justified because of the flushing and liming required (see Section 3.1.1). The consequences of this change in terms of surface area for shrimp farm development are summarized as follows and shown spatially in Figures 4a, 4b and 4c:

Of the two farm sites on the Harimau-Ulu Tiram association (HMU/UTR), one was out of business due to a variety of problems associated with management. Evidence of acid problems on that site was the lack of vegetation on the dikes, even though the farm had been one of the first established, and the presence of jarosite. The farm was new at the other site, but lime bags and jarosite were indicative of acid problems.

Of the two operational farms on the Rengam-Bt Lunchu association (RGMBTU) series (pH rating 2, “poor”), one had acid problems in several ponds the dikes of which had been constructed by excavating the underlying soil. Lime was being applied at the rate of 2 t/ha/cycle. The other ponds had been constructed by transporting soil from adjacent higher ground and there were no apparent acid problems. At the other farm, lime was being applied at the rate of 1 t/ha/cycle.

3.1.2.2 Economic implications of siting on acid soils

In order to establish a standard for comparision, physical and economic data collected at shrimp farm sites have been expressed in terms of a “standard pond”. The dimensions of this pond are in Figure 12, page 143 of Gedney, Shang and Cook (1984). It is Pond no. 3, part of a 14.7 ha pumpedwater pond complex. The “standard pond” includes a perimeter dike (8.26 m²) cross section on one side and pond dikes on three sides (6.02 m²) with an outer area of 10 506 m² and a water surface of 7 896 m². The dike volume is about 2 699 m³, the operating depth is 0.8 m, and the water volume about 6 317 m³.

Data on liming rates and costs suggest that liming amounts to from about 1 percent to 5 percent of total operational costs in delivered lime only. Application rates are from 0.2 t to 3.5 t/ha/cycle with the latter rate on new ponds; on older ponds the liming rate can be decreased, for example, to l t/ha for 3-year old ponds; labour costs for liming are relatively low.

According to Muir and Kapetsky (in press), the major problem of siting on acid soils is reduced yields. If it is assumed that production is reduced by 17 percent, and that liming and additional pumping are used to overcome the acidity, then the model predicts a 17 percent increase in capital costs and a 9 percent increase in operation costs.

3.1.2.3 Economic implications of siting on poor texture soils

At the farm with the “unsuitable” texture rating (the Rudua-Rusali (RDU-RSL) association, seepage losses were about 5.1 cm/d in a pond maintained at 1 m depth, equivalent to 5.1 percent by volume per day. The water management plan on that farm was to exchange 30 percent on alternate days. Thus, slightly more than one-third of the water pumped was going to make up seepage losses. Coche and Laughlin (1985) suggest that seepage losses through the pond bottom on sandy soils can amount to 10 cm.

The same seepage loss, 5.1 cm, converted to the “standard” pond amounts to 6.4 percent/day because standard pond depth is 0.8 m. Given an exchange of 20 percent/day used by Gedney, Shang and Cook (1984), the excess pumping required to maintain a 0.8 m depth to make up for seepage loss is 24 percent. The additional annual cost to make up for seepage from the standard pond is about MS 779, including fuel, maintenance and depreciation on all capital investments for pumping based on the capital and operating costs for pumping supplied by Gedney, Shang and Cook (1984).

Relatively modest increases in capital and operating costs are forecast for increasing pumping from 10 percent to 70–100 percent (Muir and Kapetsky, in press). Increases are from 2–4 percent in capital costs and 4–9 percent in operating costs. In comparison, the M$ 779 excess pumping costs for seepage is equivalent to about 6 percent of annual operating costs.

At the farm site on Marang (MRG; pH “poor”, texture “fair”), texture problems such as erosion and seepage were more serious than pH difficulties to the extent that some ponds had been lined with clay and others had been constructed with much more dike material than is usually used. Of the remaining sites on soils rating “fair” in texture (Table 13), there was some seepage and erosion of bunds at two farms. Of the other two sites, one was new and the other out of business. There was no opportunity for an interview at either.

The farm on Marang (above) used 7 m as the top-of-the-dike width and a 1:3 slope in order to reduce seepage. This is equivalent to doubling the perimeter dike width of the standard pond. The overall volume is 2.2 times greater than the standard pond. Therefore, the cost of dike construction is more than double. Further, the effective area decreases from 75 percent to 63 percent. That is, 16 percent less pond surface is available for shrimp production (63/75 x 100) for a given farm size. Thus, the gross income potential per unit pond surface area is lessened by the same amount.

With a doubling of earthworks costs, similar to the case described above, the general model predicts capital costs 29 percent above base and operating costs 13 percent above base (Muir and Kapetsky, in press). These were sufficiently high to result in an unacceptable simple rate of return.

Before doubling dike volume the same operator initially lined his ponds with clay. Based on the “standard” pond dimensions, about 1 358 m³ of clay would have to be transported of which 1 184 m³ would be required to line the bottom at 0.15 m thickness (compacted) and an additional 174 m³ would be for the lining of the inside of the dikes, also at 0.15 m thickness. The cost for transport and installation of the clay was given at M$ 5.00/m³ (in 1984). By comparison, Gedney, Shang and Cook (1984) indicate excavation costs at M$ 3.92/m³ pertaining to late 1982.

Lining with clay implies moving an additional 1 358 m³ to make room for the clay, if the lining is carried out as a remedial measure after the original construction. Retrolining a pond would be about 2.15 times as costly as the construction of the standard pond using good-texture soils. With the need for clay lining foreseen, costs are substantially reduced, amounting to about 1.34 times the cost of the standard pond.

In terms of a comparison with excess pumping costs due to seepage, costing about M$ 778/y/standard pond (above), it appears that with the need for clay foreseen in original construction, excess pumping costs could be recouped in about 4.5 years.

In summary, the field verification of soils has indicated that the pH rating for kranji soils should be decreased from “good” to “poor”. The effect of this change is to drastically reduce the area of Johor with soils rating fair or better. Ratings based on pH for other soil series or associations appear to be fairly consistent with performance although the sample sizes are small. The “fair” texture rating for the Marang series may be too high, if the one site visited is an indication of actual conditions which were “poor”; however, the sample size is too small to warrant making a change.

More information on shrimp farm performance on different soil types, along with the economic implications for capital and operating costs, is required to refine the soil acidity and texture rating systems.

3.1.2.4 Rainfall in relation to low-pH soils

Nearly 94 percent (181 221 ha) of coastal Johor soils rate poor or unsuitable for shrimp ponds because of low pH when the kranji series is included in the former category (Figures 5a, 5b, 5c; Section 3.1.2.1) This suggests that future shrimp farm development may have to be on acid soils. Where acid soils have been used for shrimp pond construction there have been problems of acidity aggravated by heavy rainfall. The purpose of this sub-investigation was to identify locations where acid soils, if developed for shrimp farming, would have the least risk of aggravated high acidity due to being located in high rainfall areas.

Nearly all of the west coast falls within 2 500–2 750 mm total rainfall range, except for the extreme southwest corner with rainfall varying from 2 000 to 2 500 mm (Figure 5a). Thus, if it is necessary to develop shrimp farms on soils with acid potential, or already acidic, then the southwest portion would provide some advantage. On the south and southeast coasts where poor-quality soils cover large areas, the lowest rainfall is in the extreme south and southwest, ranging from 2 000–2 500 mm (Figure 5b). Further north and further east rainfall increases, putting shrimp farm development on acidic soils at greater risk. On the northeast coast some of the poor-quality soils fall within the heaviest rainfall area of the state, >3 500 mm, and most of the remainder are within the 2 750–3 000 mm range, except on the southeast below Mersing where rainfall is less (Figure 5c).

Considering the entire coast of the state with regard to acidity from rainfall, risk is lowest along the southeast coast, moderate along the west coast, and relatively high along the east and northeast coasts.

3.1.2.5 Previously selected shrimp farm sites in relation to soils

Shrimp farm sites selected in studies previous to this one are found on the west coast and southeast coast soil maps (Figures 6a and 6b), but none occur on the northeast map. All of the 8 previously selected sites on the west coast are on kranji series which, as has been pointed out in Section 3.1.2.1, is a soil with acid potential. Kranji makes up 72 percent of the west coast soils within 2 km from brackishwaters.

Of the 13 previously selected sites on the south and south-east coast soil map, 5 sites are on kranji series, 5 are on kranji in combination with other series or associations which rate poor or unsuitable and 3 sites are entirely on soils which rate poor or unsuitable. Thus, from a soil suitability viewpoint, previous efforts at shrimp farm site selection appear to have been unsuccessful. However, soil characteristics on these sites, or a sub-sample of them, need to be verified before final conclusions can be reached.

3.1.3 Water Quality Time Series

Water quality time series, including pH, ammonia nitrogen and BOD were analysed from 21 locations in Johor.

Hydrogen ion concentration

Mean pH observations all fell within the “very clean” category, the best rating in suitability for aquaculture. Ranges of pH at all but 5 of the 21 stations also were within the “very clean” limits. At those 5 it was the low end of the range that exceeded the “very clean” limits. In terms of pH Johor waters appear to be well suited to aquaculture development.

Ammonia nitrogen

Ammonia nitrogen observations among 17 locations ranged into the moderately polluted catgeory, >1 ppm, at 4 locations: Power Station, Galang Patah (Sg. Pulai), Hospital and Pantai Lido (Figures 2 and 7a). At the latter station ammonia nitrogen has ranged as high as 8 ppm, well into the grossly polluted category. The frequency with which ammonia nitrogen observations fell into the slightly polluted category, >0.5 ppm, is shown in Figure 7b. Those with 10 percent or more observations in this category are Skuddai, Hospital, Endau and Segget. With the exception of Endau which is in the northeast, all of the latter locations are in the vicinity of Johor Baharu.

BOD

BOD observations ranged into the moderately polluted category, 5–10 ppm, at 15 of the 21 locations. At 8 locations BOD ranged into the grossly polluted category, >10 ppm (Figures 2 and 8a). The frequency with which BOD values fell into moderately and grossly polluted ranges is shown in Figure 8b. Locations at which moderately or grossly polluted observations occurred more than 10 percent of the time (among 16 data sets) are Pasir Gudang, Pantai Lido, Hospital, Power Station, Tebrau, Masai, Endau, Segget and West Johor Straits. With the exception of Endau, all of these locations are in the vicinity of Johor Baharu, or the western portion of Johor Straits.

Considering BOD together with ammonia nitrogen, the locations where observations ranged into the moderately polluted category were Segget, Hospital, Power Station and Galang Patah.

3.1.4 Verification of Water Quality Time Series

Data on the quality of waters at shrimp farms were not easily obtained either due to unavailability, or reluctance to allow data to be copied. However, Galang Patah BRS can be used as a water quality benchmark because it occurs in an area where conditions are less than ideal. Water quality conditions in ponds there range into the moderately and grossly polluted categories at times (ranges for 1987: BOD 4–10; pH 3.7–7.8; DO 1–8 ppm). These problems are routinely overcome by water management which includes 20–50 percent exchange per day and aeration when DO falls to 4 ppm.

Apart from Galang Patah, one shrimp farm was experiencing water quality problems caused by an adjacent farm. The short-term solution was to avoid circulating water when the other farm was discharging. On a longer term a new water inlet further removed from the offending farm was being constructed at apparent high cost.

Although water quality problems can be overcome, there are capital and operational costs associated with them, for example, electricity and upkeep of pumps, that are in excess of those incurred in areas where the water quality is better. Thus, shrimp farms located in areas where good quality water is indicated are likely to have less capital and operational expenses than those in areas where water quality is less. For example, Muir and Kapetsky (in press) show that in the general case, if production conditions are optimal, capital and production costs are reduced by 20 and 12 percent respectively.

3.1.5 Water Quality and Land Uses

The transformation of land uses affecting water quality for aquaculture from district boundaries to river basin boundaries provides the following pattern for Johor (Table 14, Figure 9).

Pulai is the most developed drainage basin in the state with regard to agriculture and urban areas, and Mersing is the least developed. According to the land-use water quality rating developed in Section 2.2.2, in addition to Pulai itself, Kesang, Batu Pahat, and Benut river drainages fall within the 81–100 percent relative development category for which a moderate impact on water quality for aquaculture is expected. Muar, with 51 percent of its area developed for agriculture and urban areas is in the little impact category. These drainages are all distributed along the west and southwest coasts of the state. According to the total land developed in the Endau and Johor Baharu drainage basins, 32 and 43 percent, respectively, only a very little impact on water quality is expected. The Sedili basin, with only 27 percent of its land developed and Mersing with only 7 percent should have good quality water for coastal aquaculture development. Figure 6 portrays water quality for aquaculture according to drainage basins.

These results are in general agreement with the water quality time series presented in the previous section. Thus, the overall interpretation of water quality for aquaculture development is: west and southwest coasts including Sg. Pulai and the West Straits - moderate impact especially around Johor Baharu, except at Muar and Batu Pahat where conditions are clean to very clean; southeast coast, including Sg. Johor and northeast coast, very little to very slight impact, except at Endau where conditions can range into polluted levels. However, it has to be kept in mind that while the effects of Singapore on water quality are implicit in the water quality time series, these effects are not taken into account in the land use approach. The results of both approaches point to the southwest coast as an area for which caution should be exercised and to the southeast coast as having relatively good water quality for development.

3.1.6 Salinity

Salinities at 4 of the 21 locations rated “good”, 15–33 ppt, for all of the observations. At 7 other locations, salinities stayed in the “fair” range, 10–15 ppt, and at the remainder, ranged into the “poor” category, <10 ppt (Figures 2 and 10a). At 4 locations (among the 16 for which individual observations were available), the frequency with which the salinity observations fell into the poor category was from 12 to 28 percent (Figure 10b). These were Kukup, Skuddai, Muar and Endau.

3.1.7 Verification of Salinity

Salinity data were available from 9 of the 11 operational shrimp farms visited. Except for 2, the salinities were within the “good” (15–33 ppt) range (Table 15). In this range, salinity did not appear to be a problem. Salinities for most of the farms were higher than the optimum for Penaeus monodon, 15–25 ppt, as described by Chen (1985), but in line with the range given for best growth of P. indicus (= merguiensis), 20–30 ppt, by Cook and Rabanal (1978). This may help to account for the fact that nearly all of the farms were raising some P. merguiensis. At the farm where salinities were lowest, 0–3 ppt for up to two weeks per year, only P. monodon, known to be tolerant of wide ranges of salinity, from 0.2 to 70 ppt (Cook and Rabanal, 1978), was being raised. At one farm, seawater was being mixed with freshwater to maintain an optimum salinity for P. monodon.

An obvious improvement in salinity criteria to better locate opportunities for shrimp farming is to define separate salinity ranges for P. monodon and P. merguiensis. Also, siting could take into account other specific differences such as the burrowing habit of P. merguiensis and thus the advantage of sandy texture soil over that of clay to facilitate this activity.

3.1.8 Access to Water and Roads

One of the fundamental assumptions to locate opportunities for shrimp farming was that farms would have to be located within 2 km from a water source. In fact, all but one of the farm sites was adjacent to, or within only a few hundred metres of a water source. The maximum distance that water was being moved to the most distant ponds was probably less than 1 km. However, at one farm planned expansion in area would mean that the most distant ponds would be up to 2.5 km from the water source.

The 2 km criterion appears to be reasonable for most farming situations when it is considered that too high elevation is likely to be encountered before too far distance to transport water horizontally becomes limiting.

The infrastructures to provide goods and services for shrimp farm development were originally considered as district centres and the primary and secondary road networks. Although infrastructure analysis was an essential part of the GIS training, the map data then available (see Figure 4) were not sufficiently recent to be used for practical development planning.

For field verification, accessibility was considered according to time and distance from farm sites to the nearest paved road. In Johor, villages and towns on paved roads are relatively closely spaced. Therefore, time and distance from a shrimp farm to a paved road are indicative of the availability of goods and services.

Of the 14 shrimp farm sites visited by road, the most isolated was 12 km and 35 minutes from a paved road (Table 16). At the other extreme, two farm sites were located adjacent to paved roads. At least one of the farms was transporting by water as well as land.

The costs of siting distant from goods and services are easily identified, but difficult to quantify. Among them are increased labour costs for time spent in transit, increased fuel costs and wear and tear on vehicles, increased delivery costs for necessities such as shrimp food, lime and post-larvae, and higher operating expenses for diesel-electric power compared with mains electricity.

Looking at both water and road access together, Muir and Kapetsky (in press) found that by extending road access from 200 m to 2 km and water access from 100 m to 500 m resulted in a 7 percent increase in capital cost and a 3 percent increase in operating cost. In the extreme, with road access 10 and 20 km and increased costs of fuel and accommodation for workers, capital costs become 31 and 81 percent higher and operating costs 20 and 39 percent greater, respectively. Extending the water supply distance to 1 km increases the capital and operating costs by 9 and 4 percent respectively.

3.2 OPPORTUNITIES FOR FISH CULTURE IN FLOATING CAGES

3.2.1 Bathymetry and Shelter

Opportunities for the further development of floating cage culture of fishes were considered for the west, south and south-east coasts. A preliminary analysis indicated that there was little shelter on the east coast, except in the lee of islands. Therefore, no analysis was made of this area.

On the west coast there are about 21 500 ha within a 2 km lee of north winds. Within this area maximum wave height should be no more than 0.6 m, given north winds of gale force. However, there are few areas along this coast which offer both shelter and sufficient depths for cage culture (Figure 11a). Current velocities, somewhat offshore from sheltered waters, are mainly 1 kt (0.51 m/sec), the upper limit for cages. Because currents are likely to be stronger over more shallow, inshore waters, it is probable that the opportunities for cage culture are few, apart from the vicinity of Kukup Island.

Sheltered areas in the extreme southwest of the south coast cover about 2 467 ha (Figure 11b); however, most of this area is too shallow for cages, except in the mouth of Sg. Pulai where, on the east shore there is rapid drop-off into 10 m well within shelter. Cages in this area could be in a fairway for navigation, however. Unfortunately, no nautical charts are available for upper Sg. Pulai and the bathymetry in those sheltered waters is unknown. Another sheltered area is off the tip of Tanjong Piai, but the drop-off is sudden and into waters in excess of 20 m.

All of the area from the west side of the Straits of Johor to the Causeway within Malaysian waters is sheltered and amounts to 2 269 ha (Figure 11c). In this stretch there are 138 ha of waters in the 5–10 m depth interval rating “good” in depth suitability for cage culture and an additional 166 ha of waters 2–5 m deep, the seaward part of which would give sufficient depth for cages. There are no current velocity data for this area.

From the Causeway to the eastern end of the Straits conditions are similar to those in the western portion, namely fully sheltered waters out to the international boundary amounting to about 2 144 ha of which 141 ha are in good depths for floating cages and an additional 110 ha are in 2–5 m depths, part of which could be used for cages (Figure 11d). There are no current data.

From the west side of Sg. Johor, along the south and southeast coasts there are 8 721 ha of sheltered waters (Figure 11e). There are no bathymetric data for Sg. Johor, but there are indications of depths of at least 10 m for the lower part of the estuary, some of which are sheltered and not in obvious navigation freeways.

Along the extreme south coast, southeast of Sg. Johor, nearly all of the waters in the 5–10 m range are sheltered, but strong currents are indicated in the vicinity well in excess of those allowable for floating cages.

On the southeast coast there are two shallow bays which offer shelter from the north. Both offer few prospects for floating cages because the deeper areas within them are marginal for cages and both are open to the winds and waves from the west and south-west. In summary, cage culture opportunities appear to be few on the west and east coasts of Johor. In the Straits of Johor there are 300 to 400 ha with well-suited to marginal depths for cages and which are also sheltered. Current velocites are unknown; however, there are cage installations on the Singpore side of the western portion of the straits which may indicate that current conditions are suitable. Undoubtedly, there are cage culture opportunites in the estuaries flowing into the straits, but lack of charts leaves the bathymetric situation unknown.

The spatial distributions of water quality, salinity and infrastructure already have been discussed in relation to shrimp culture. The same considerations apply to the location of cage culture opportunites. Briefly, water quality considerations favour floating cages in the Sg. Johor area over those in the vicinity of Johor Baharu, or Sg. Pulai. Mean salinities are favourable for Lates and Epinephelus, but the minimums experienced in some areas may be stressful, if not lethal. The Straits area and the west side of Sg. Johor are easily accessible by road, but the east side of Sg. Johor is less so.

3.2.2 Verification of Bathymetry, Shelter and Water Quality

The locations with the largest concentrations of cages were visited: Kukup (60 operators with about 1 800 cages; two interviews), and Kuala Sedili Besar (17 installations; one interview), the former in company with Mr Wang of the Johor State Fishery Office, who has been deeply involved with the development of fish farming in net cages. Cage installations were observed along the west portion of the Singapore Straits on a boat trip from Sg. Pendas to Johor Baharu (two interviews). Finally, a cage owner was interviewed at Sungai Telok Sengat.

Shelter

For the cage culture GIS, shelter was defined as the area within a 2 km band from the shoreline, within which a maximum wave height of 0.6 m would be expected, given maximum recorded wind speed.

At Kukup nearly all of the cage installations were well sheltered between the mainland and the island. At one installation further north than the others and less sheltered, the author was informed that the need for heavier materials to avoid wave damage and to accommodate greater wave motion had added about 30 percent to the cost of construction. The maximum wave heights experienced at this relatively new installation were from 0.6 to 0.9 m. At the other, well sheltered installation, waves were 0.3 to 0.6 m in the worst conditions. At Kuala Sedali Besar the cage installations experienced waves of from 0.6 to 0.9 m. These were not a problem according to one owner. Cages in the Sungai Pendas area of the Western Straits were sheltered. A maximum wave height of 0.5 m was indicated in one interview. At Sungai Telok Sengat the maximum wave height was 0.6 m. The storms that cause these waves are of short duration. Thus, there is only one indication of increased costs due to lack of “full” shelter.

Bathymetry

Cages in Johor are usually 3 m deep. At Kukup the northernmost, least sheltered installation was in 8.5 m. At one installation adjacent to the island, at low tide the landward cages had about 0.4 m and 0.6 m on the channel side. In the Western Straits, one installation was in 8 m and the other in 11 m. At Kuala Sedali Besar many installations were located adjacent to the river bank, one of which was in 8.5 m at lowest tide. At Telok Sengat the two cage installations are in relatively shallow water. At one, the lowest part of the cage is on the bottom at lowest tides. It was stated that the installation was not located in deeper waters because of too-strong currents. Thus, the depths on actual cage sites are generally in line with the criteria laid out in section 2.6.1 except that Johor cages may be located in waters less deep than believed to be suitable.

Water Quality

Water quality measurements were not available, but water quality was discussed in interviews. At Kukup, apparently there are no water quality problems. In the Western Straits area, water quality varies and is unpredictable. The operator at Telok Sengat, in business for 7 years, complained of kills caused by red tide. Although salinity varied at that location it did not affect the fish. In contrast, at Sedali Besar low salinities during the monsoon apparently make the fish susceptible to gill parasites. Another problem at that site is fish kills caused by herbicides used for land clearing.

Current Speed

Current measurements were not available at cage sites, but operators were questioned about current conditions. These were described as “strong” at the riverside site at Kuala Sedali Besar and near Sungai Pendas in the West Johor Straits, “too strong” further offshore than where cages are presently located at Telok Sengat, and “adequate” at Kukup. Current speed did not inhibit cage culture development at existing sites.

Other Criteria

In addition to water quality, current speed and bathymetry, important factors for siting cages are security and agglomeration. According to Mr Wang and to people interviewed, owners want to have their cage installations within sight of their houses and otherwise within a convenient distance. Presence of a fishing village, or of houses thus becomes a prerequisite for cage development. Agglomeration is helpful because the feed, trash fish, can be delivered more cheaply to one location than to many. Also, as most of the Lates come from Thailand, delivery of fish for stocking to a central point may also save on costs. Marketing is another aspect for which agglomeration could be an advantage. Therefore, in future, agglomeration and distance from a village must be taken into account for siting of cage culture. This could be accomplished by including shoreside villages as one of the components of the GIS. Furthermore, more attention should be paid to water quality and salinity as evidenced by the difficulty being experienced by cage culturists at Kuala Sedili Besar.