5. THE NAKED CARP FISHERY

5.1 Sources of Data

Supplementary information about the fishery has been gathered from interviews with fishermen, scientists and government officers, publications (e.g. Zhao 1975, Zhang & Chen 1980), trawler log-books and official statistics. Much of the information is incomplete, and the reliability of some sources is uncertain, so that in many cases it has been necessary to make subjective appraisals.

5.1.1 Bureau Statistics

Fishery production records of the Bureau of Aquatic Products are shown in Table 5.1 and Figure 5.1. The data for Qinghai Lake (Table 5.1) are submitted to the Bureau by the Fish Factory, and refer to weigh-bridge records rather than the logs of the trawler skippers (see sections 5.4.1, 5.9.2). The data probably are minimum estimates of the actual catch and a reflection of sales through official outlets. It is not clear whether they include records of catches by fishermen (legal and otherwise) not associated with the factory.

The Qinghai Lake fishery was the sole provincial source of fish until aquaculture became established, on a small scale, in 1981. Other provincial capture fisheries have since developed in Keluke Lake, and particularly Gyaring and Ngoring lakes on the upper Yellow River (Yang & Walker 1993b), and the discrepancies between total production figures and the sum of those for Qinghai Lake and aquaculture in 1980–85 presumably reflect these contributions.

5.1.2 Trawler Logbooks

An incomplete series of annual trawler logs dating from 1980 has been obtained from the Fish Factory. Although logs have routinely been compiled since the commencement of fishery operations in 1959, those prior to 1980 unfortunately have been destroyed. In addition, the log for 1986 is presumed lost and the records for some other years are incomplete. The records (boat number, date, length of hauls, numbers of boxes of fish, locality, wind and weather conditions) have been transferred to Lotus 123.3 (later Microsoft Excel 4.0) spreadsheets programmed to calculate the catch in kg per hour trawl haul, as an indication of catch per unit effort (CPUE). The computed data were incorporated into a database which determines CPUE for each month and each area fished.

Table 5.1
Annual fishery production (tonnes) in Qinghai Province, 1949–92. Compiled from the records of the Bureau of Aquatic Products, Xining. Figures in italics are based on the assumption that all production prior to 1981 originated from Qinghai Lake. The Fish Factory commenced operation in 19S9, and has been subject to an official annual quota of 1200 tonnes since 1987.

5.2 History of the Fishery

The earliest records of a fishery in Qinghai Lake date from the 18th century, during the reign of the Qing Dynasty Emperor Dao Guang (Wu 1987, Chen 1992). At that time Mongolian people used hook and line to catch naked carp. In summer the fish were dried, after heading and gutting, and groups of 12 fish were string together on stick and sold in lots of 10 sticks. In winter the fish were sold frozen. Pack animals were used to transport the fish to neighbouring provinces within a 150-km radius of the main distributing centre, Haiyian (Huang Yuang). It is estimated that 200,000 fish of 0.5–1 kg, or about 200 tonnes, were sold annually.

In 1909 the Mongolians moved westward and the fishery was left to local people and migrants until 1958. The numbers of people (mostly seasonal migrants from eastern Qinghai and poor rural counties) who fished the lake in this period have been estimated at between 7000 to 20,000 in each season. In 1958 the State assumed control and a salted product was introduced using salt from local lakes (Chen 1992). Fish were gutted, split from the back and dry-salted for 48 hours, then packed with salt in cat-tail bags and despatched to market by truck or train, on journeys of up to 10 days' duration. In 1959 the government established the Fish Factory on the south-east lake shore, about 135 km from Xining (some buildings were added later). Production of salted fish ceased in 1979, when the government developed alternative supplies nearer the markets. After this time the fish were frozen in blocks and consigned to various parts of China in insulated rail carriages. Ice and salt were added to the frozen blocks on long journeys. The more remote markets became redundant, however, as fishery production declined.

After the fishery was opened to commercial exploitation in 1958, the annual catch increased from 13,000 tonnes in 1959 to a maximum 28,523 tonnes in 1960. Catches declined to a minimal 2523 tonnes in 1983 before recovering to 4372 tonnes in 1986. The decline reflected intensive fishing in the early years, although social disruption associated with the cultural revolution in the 1960s may have contributed. Since 1987 annual catches by the Fish Factory officially have been limited to a quota of 1200 tonnes, apparently to allow the stock to recover. The present annual catch, including the factory quota and catches by licensed and unlicensed fishermen, is conservatively estimated at 3000 tonnes.

Recent fishery operations may be divided into eight stages:

1. Before 1958

   In the years immediately before commercial development the fishery was exploited by small lakeside commmunities, herders and other migratory workers, and there probably was intensive fishing for spawners in the Buna and Shaliu rivers in spring and early summer. The fishery was of minor commercial value, as markets were not easily accessible and the abundance of local beef and mutton would have devalued fish as a meat source. It probably provided valuable supplementary income, however, for seasonal migrants from poor areas.

2. 1958–1959

   In 1958 the fishery was officially opened to commercial exploitation, primarily as a state-owned enterprise. In 1959 a forerunner of the present factory, including a freezing plant, was established to processed the catch from two 125-hp wooden boats (19.6 m LOA) operating as pair trawlers.

3. 1960–1962

   These were boom years for the fishery. In 1960 the provincial Department of Aquatic Products was established and made responsible for commercial exploitation. An initial complement of five boats in 1960 was increased to 10 pairs in 1961–62 (three operated by the state-owned enterprise and seven by the lakeside communities). Catches attained a peak 28,523 tonnes in 1960, but declined to 18,763 tonnes by 1962.

4. 1963–1982

   Throughout most of this period 2–3 pairs of boats were active—two supplied the factory and one was based at Xalu, at the mouth of the Shaliu River, supplying the growing communities at Gangca and the nearby irrigation area. In 1979 the Bureau of Aquatic Products was established to continue and upgrade the responsibilities of the former Department Annual catches fell to about 7200 tonnes in 1963–64, and thereafter fluctuated between 2902 and 4979 tonnes.

5. 1982–1984

   The fishery was exploited by only two pairs of trawlers based at the Fish Factory. The smallest catch since the fishery commenced (2523 tonnes) was recorded in 1983.

6. 1985–1986

   Three pairs of boats were active: two based at the factory and one at the State Farm on the north shore. Catches showed a minor recovery to 4372 tonnes in 1986.

7. 1987–1988

   Catches declined further, and a quota of 1200 tonnes was imposed on factory catches to promote recovery. Since 1987 the factory ostensibly has had a monopoly on commercial operations, but some licensed fishing has been permitted.

8. 1989–1990

   The 125-hp wooden boats were phased out Two newly constructed 250-hp steel-hulled trawlers were added to the factory fleet in September 1989 and a further two were added in November 1990. Two small steel vessels operated from the north shore.

5.3 Changes in Fishing Techniques

Before 1958, 500-m seine nets were used occasionally in the rivers or near shore, and gill nets were set from the shore and from small boats. Fish were also taken with hook and line, and migrating fish in the Buha and other feeder streams were so abundant that they could be captured by beating them with a stick, or in other simple ways.

Since 1958 deep-water trawling and gill netting have been the two main methods used. The trawl wings generally have been 8–10 m long. According to Mr Ying Baicai (Bureau of Aquatic Products, Xining), the cod-end mesh size (stretched mesh) was increased from 6.5 to 7.5 cm in 1979 and the mouth of the trawl was slightly reduced in depth. The State Farm boat, however, used a mesh of 5 cm. Trawl times were increased from about 1.5 hours in the 1960s to about 2 hours in the 1970s. Other significant changes were variations in the number of trawlers and the introduction of more powerful boats. At present 80-mm stretched mesh monofilament gill nets are supplied to fishermen by the Bureau, under license, and are used in water less than 5-m deep, within 3–5 km of shore.

5.4 Changes in Trawler Catches

5.4.1 Reliability of Logbooks

The trawler logbooks have served mainly to provide annual catch records for the provincial administration, without regard for their value in monitoring long-term changes in the stock and fishing activity. Ideally, they should provide accurate records of the gear used, the length of time spent fishing, the locality fished, prevailing weather conditions and the quantity of fish caught The actual records are patchy and subject to several potential sources of error and bias. In particular:

1. Under-reporting of catches due to selling “over the side”, or fishermen's privileges in removing part of the catch for their own use. The magnitude of this effect is unknown.

2. Over-reporting of catches by including catches from illegal operators to hide their origins. This may not be important, however, in a fishery which sells only to a controlled market There is no incentive for an unauthorised fisherman to hand his catch to a legal boat as there would be no benefit to either party.

3. Omitting reports of catches from localities where restrictions apply. There is undoubtedly some fishing of the stock migrating up the rivers, although these are declared restricted zones.

4. General carelessness in estimation, especially when the catch is large or the weather is bad. This is obvious in records where the higher catches are reported in rounded figures of thousands. These deficiencies probably are of no great significance in assessing trends.

5. The trawler catch data do not necessarily correspond with weigh-bridge records because the skippers estimate the catch in terms of the numbers of crates or “boxes” (section 5.9.1). The crates used at present are 20 kg capacity, but other sizes have been used in the past and this must be allowed for in catch estimates.

5.4 2 Catch per Unit Effort in Designated Areas

The logbooks make it possible to calculate annual catch per unit effort (CPUE, kg h^-1 trawled) for different areas of Qinghai Lake. The factory management has arbitrarily divided the lake into 13 fishing areas, designated A-M (Fig. 5.2). Areas H and M are two of the areas most frequently fished, namely the lake near the Buha mouth (area H) and the South Bay, near the Fish Factory (area M). Area M is about 10% of the lake surface area.

There are insufficient data to properly evaluate the apparent changes in CPUE associated with introduction of the new steel trawlers in 1989–90, but this should be possible if complete data can be obtained for 1991–92. It is noteworthy that in 1990 83% of the catch was from area M, coinciding with the advent of the new trawlers. In comparison, during the 1980s the fleet tended to concentrate on area M at the beginning of each season, and moved to near the Buha mouth when stocks were noticeably depleted.

The total area trawled since 1959 has been only about one third of the lake, as some areas, particularly off the north shore, are too shallow and rich in algae (Cladophora) for trawling. These areas are gill-netted, however, by licensed fishermen and poachers.

A statistical summary of the CPUE data is shown in Appendix V, and annual summaries of the percentage of the estimated total catch and mean CPUE for each area are shown in Tables 5.4–5.5.

Table 5.4 shows that in the 1980s areas H and M together accounted for roughly half of the estimated total catch. The catch from area H in 1989 is low, probably because there was a major flood in the Buha in that year. As noted above, area M appears to have attained major importance since 1990.

Table 5.5 demonstrates that the mean CPUE in area M during the 1980s (except 1989) was about 400–500 kg h^-1, whereas area H tended to values of 700–1000 kg h^-1. Area M consistently has returned a smaller CPUE than several other areas, mainly at the western end of the lake.

Figure 5.3 shows a comparative plot for the two areas, relative to changes in the annual fishery production. The plot reaffirms the point that in 1980–88 area H consistently returned CPUE figures 2–3 times those of area M. The pattern differed in 1989, the year of the major Buha flood, when the trawl time in area H was comparatively short (Appendix V). The only figure available for the steel trawlers shows a significant increase in CPUE (673 kg h^-1) in area M, substantiating the observation that the trawlers now do not need to stray from the South Bay to fill their quota.

Table 5.4
Percentage of the estimated total annual catch from designated fishing areas since 1980.

See Fig. 5.2 for a map of the designated areas and Appendix V for a statistical summary

Table 5.5
Mean annual catch per unit effort (CPUE kg h^-1) in designated fishing areas since 1980.

See Fig. 5.2 for a map of the designated areas and Appendix V for a statistical summary.

5.4 3 Trends in Lakewide Catch per Unit Effort

The historical patterns of landings (Table 5.1, Fig. 5.1) illustrate the classic effects of exploitation of a newly-opened fishery. Within 3–4 years catches fell dramatically, but stabilised after the mid 1960s. Catches in the 1970s were 3558–4979 tonnes, and catches in the 1980s tended to be lower by roughly 1000 tonnes. From 1980–90 catches fluctuated between 8–15% of the recorded maximum. After 1980 factory production declined appreciably until imposition of the 1200-tonne quota in 1987.

Figure 5.4 compares the trend in factory production with the lakewide CPUE since 1980. CPUE ranged between 450–650 kg h^-1 prior to 1989, and rose to nearly 650 kg h^-1 as the new trawlers came into service.

5.4.4 Monthly Distribution of Catches

Table 5.6 shows the percentage of the estimated total annual catch in each month since 1980. Note that virtually no fishing is carried out from December to March, and only in exceptional years are significant catches taken in April or October and November. Fishing activity begins in May and increases until July or August, before declining in September.

It is remarkable that in 1990, the first year after the introduction of the new trawlers, 86% of the entire year's quota was filled in April. This is further evidence that the pattern of fishery operations has changed with the new trawlers, and that they have the potential to greatly increase the fishing effort. In this context, the lack of complete logbook data for 1991–92 is most unfortunate.

A comparable table is not produced for the monthly distributions of CPUE because fishing efficiency depends primarily upon areas fished rather than the season.

5.5 Changes in the Exploited Stock

Since 1959 the stock of large fish up to 20 years old has been heavily depleted, and the trawlers are now catching mostly smaller fish that are 5–8 years old (see below). This trend should be evident as long-term declines in the mean length and weight of fish in the commercial catch. Partial records have been kept by the Bureau of the average lengths and weights of fish sampled from the factory catch at sporadic intervals since 1964. The practice has been to take a SO kg sample of the catch, count the number of fish and so determine the average weight, and also to measure the lengths. The data, shown in Table S.7, are consistent with declines in both the length and weight of fish in the catch.

Table 5.6
Percentage of the estimated total catch since 1980 by month. See Appendix V for details.

Table 5.7
Records of the average lengths and weights of fish in 50-kg samples of the commercial catch over the history of the fishery. From records of the Bureau of Aquatic Products, Xining.

Table 5.8
Cumulative frequency distributions (percent) of random samples of the commercial catch on three occasions in 1992. Italic figures represent the percentage of the catch within approximate length and weight limits corresponding to an age of less than 7 years.

There are few historical records of the age composition of the commercial catch, but Zhang and Chen (1980) state that in 1975–77 the age of first capture was 7 years (see further section 5.8). The increase in the trawl cod-end mesh size in 1979 (section 5.3) would presumably have increased the age of first capture, although this is not borne out by recent data.

Measurements of the age composition of the commercial catch in 1989–92 are not complete, but a clear indication is obtained by using length and weight data to estimate age (cf. section 4.7.2). As a general rule, fish of length less than about 300 mm and weight less than about 350 g are likely to be less than seven years old. Table 5.8 shows cumulative frequency histograms of three random samples of the commercial catch taken in 1992. The data suggest that significantly more than half of the catch is of fish less than 7 years old. As mentioned, the effective age of first capture at present is in the range 5–8 years. Given that sexual maturity is attained at about 7 years, it is evident that large numbers of fish in the commercial catch are captured before they have had an opportunity to contribute to recruitment This is a serious imbalance, requiring action.

5.6 Current Regulations

Regulations governing the Qinghai Lake fishery are local adaptations of national regulations for management and regulation of freshwater and marine fisheries. A free English translation of the current regulations is shown in Appendix IV. The basic contents are summarised below, with some annotations on variations in practice:

• The factory is licensed to operate four 250-hp steel-hulled trawlers, working in pairs, using any techniques and exploiting any grounds other than reserved areas (see below), subject to an annual quota of 1200 tonnes. The quota was first imposed for 1987–89, and has been extended to 1992.

• The State Farm on the northern lake shore uses two small (120-hp) steel trawlers to fish without restriction for the benefit of the farm. However, it is understood that this permit has been rescinded and the trawlers are no longer active.

• About 400 licenses are issued annually for gill-net fishing by local people, using nets supplied by the Bureau. There is a minimum mesh size (70 mm), but no limit to the number of nets that may be used. Annual quotas of 1000 kg apply for full-time fishermen and 500 kg for part-time fishermen. These activities are not effectively policed (section 5.8).

• In recent years as many as 10,000 individuals have been supplied with gill-nets to fish through the ice during winter (although this is now prohibited). Apparently good catches were recovered from top set nets. It is difficult to see how this could have been controlled or assessed, given the lack of police support (section 5.8).

• Fishing is not allowed in certain reserved areas, namely the Buha and Shaliu rivers and the lake near the river mouths. These regulations are widely ignored, however, as the spawning run offers opportunities for easy capture of large fish.

• Fishing is prohibited in winter and restricted in October and November.

5.7 Present Activities

5.7.1 Authorised and Unauthorised Catches

As noted, since 1987 the Fish Factory has been limited to an official annual quota of 1200 tonnes, although it is not certain whether this limit has been applied rigorously. Trawlers no longer operate from the north shore, but there has been an officially-sanctioned increase in fishing by licensed gillnet fishermen, presently reckoned at about 800 tonnes annually. The official catch therefore is about 2000 tonnes annually.

The extent of unauthorised fishing cannot be gauged accurately. The best available estimate is from Mr Wang Fuwu, Director of the Fisheries Police General Station (advice to Team Leader, 12 September 1991). Mr Wang estimated that since 1980 the quantities of fish annually consigned from the north-shore railway stations Jianhe, Huangyu, Guan and Haergai amounted to 240 tonnes during the two-month spawning season and 300 tonnes over the remaining 10 months. Consignments from Guhe and Haiyian counties were reckoned at 360 tonnes annually, with a further unaccounted 100 tonnes. On this basis, unauthorised fishing in 1980–91 is estimated at 1000 tonnes annually.

The total annual catch at present therefore is believed to be in the vicinity of 3000 tonnes, but this probably is conservative.

5.7.2 Control and Protection

Local detachments of the Gong An (Public Security Police) are administered by the Bureau of Aquatic Products, and an additional force of Economic Police is employed directly by the Bureau. The combined “fishery police force” includes about 20 personnel. Another 20 Economic Police are deployed in each of the three counties bordering the lake (Ganghe, Gangca, Haiyian), and administered by the county offices of the provincial Bureau of Animal Husbandry & Agriculture.

The Bureau of Aquatic Products maintains Police Stations at the road crossing of the Buna River, at Giang Shigo, mid-way along the southern shore, and at Haergai, mid-way along the northern shore. Each station sells nets and fishing licenses. The police have authority to stop and search individuals and vehicles, and are equipped with three speed boats for surveillance. All landed fish must be registered at one of the three stations, and fish not registered, or not purchased from the factory, are deemed to have been illegally caught.

5.7.3 Capital Investment

Since 1989 a program of major investment by the Bureau of Aquatic Products has been underway at the Fish Factory. This has included construction and launching of four steel-hulled trawlers (each costing about US$100,000), construction of extensive housing and administration buildings and rehabilitation of the jetty and landing facilities, and will be economically justified only if returns from the fishery can be increased. The trawlers could support intensified fishing, and it is difficult to reconcile the scale of the new investments with the policy of restricted quotas.

5.8 Sustainable Yield

Zhang and Chen (1980) investigated the Qinghai Lake fishery using the classic Beverton-Holt Dynamic Pool Model. They measured the length and weight of fish from random 50-kg samples of the commercial catch, taken 10 times monthly, between July 1975 and September 1977. Their work involved a combination of original data, the results of an earlier study by Zhao (1975) and various assumptions in areas where information was lacking. The paper describes the means of estimating total mortality and natural mortality, incorporating frequency distributions of length and age derived from original data. Growth coefficients were obtained by fitting a Von Bertalanffy curve (using Walford's method), as shown in section 4.7.3. The fishable resource was estimated by the Gulland method. Their estimates of the model parameters were as follows:

Zhang and Chen considered that the size of the naked carp population was stable in the three years prior to 1975–77, although the length and weight of individual fish had declined since 1959. They used the Beverton-Holt model to explore the effects of different fishing methods and mesh sizes on the numbers and biomass of fish and the total catchable number and total catchable weight Under conditions of an age of first capture (“initial fished age”) of 10.0 years and a fishing mortality of 1.25, the estimated fishable population is 50,000 tonnes and the maximum sustainable annual yield is 4791 tonnes. In 1975–77, however, the annual yield of 4200 tonnes was obtained with age of first capture 7.0 years and fishing mortality of 0.50. The model suggested that if the age of first capture was raised to 10 years, by increasing the trawl mesh size, the weight of the catch would increase by 10%, still short of the theoretical maximum. Alternatively, an increase in fishing effort would increase the catch, but the capital investment required would offset the economic gain. Increased mesh size therefore appeared to be the best option for increased catches. In view of the apparent stability of the resource, albeit in the short-term, they argued that it was wise to maintain the mesh size and thereby maintain the annual catch at present levels.

The model developed by Zhang and Chen includes a number of assumptions that might be refined and substantiated using more recent data, but the basic conclusions probably would not change. There are considerably more sophisticated approaches than the Beverton-Holt model (e.g. Pitcher & Hart 1982), and one or more of these may best serve the future role of models in managing the fishery.

5.9 Catch Processing

This section incorporates observations by Ms Chen Sie Poh, FAO Post-Harvest Consultant, during a visit to the Fish Factory in August 1992 (Chen 1992).

5.9.1 Handling

Before 1989, the pair trawlers would venture far across the lake and remain away from base for 3–7 days. Most fish now are caught on grounds within 20 km of the factory (Area M), and a typical trip requires about 13–14 hours to land about 5 tonnes. Trawling locations generally are in offshore areas free of troublesome algal growths.

On 3 August 1992, when a pair-trawling operation was inspected by Ms Chen, the nets were cast three times, alternating between boats. At the completion of each trawl the catch was unloaded to the deck, flushed with water and scooped (using a coal scoop) into 20 kg crates. The crates were encrusted by grime, as they are not routinely washed. Part of the catch was crushed from being trodden underfoot by deckhands, or from being stacked in broken or ill-fitting crates. The loaded crates were stacked six high on the deck and remained exposed to the sun, without ice, for the remainder of the journey. The catch was not transferred to the hold evidently because this would have made unloading more difficult On return to the factory jetty the catch remained on board for 2–3 hours before being unloaded into a 10-tonne truck. The crates were returned unwashed to the deck of the trawler, ready for re-use.

5.9.2 Processing

Processing operations have been observed many times by members of the Lake Investigation Team, in the course of routine sampling. As reported by Chen (1992), the loaded truck is driven to a weigh-bridge, then to the processing room. The catch is hosed with water and the tailgate opened, unloading the fish to a part of the concrete floor cordoned by stacked freezing trays. Handlers wearing rubber boots and aprons gather the fish into 20 kg freezing trays, shuffling the top layer of fish neatly head-to-tail. Badly damaged fish and mats of algae are put aside. The packed trays are loaded onto carts, seven high and four abreast, and wheeled 50 m to be loaded into freezing trolleys. Each tray is topped with water from a hose and the trolleys are set on rails, awaiting transfer to the blast freezer. At this stage the fish are checked by an inspector to ensure that the boxes are filled. The inspector apparently has no responsibility for general standards of handling and hygeine.

The fish remain in the freezer for 12 hours before the frozen blocks are knocked out of the trays, glazed by passing through a trough of water and transferred to a cold store. Protective wrappings are not used, to minimise costs. The store may be over-loaded, making it impossible to turn over the stock on a first-in-first out basis, and is also prone to vary in temperature. The frozen fish are transferred to uninsulated, canvas-covered trucks and transported to cold stores in Xining, mainly awaiting air transport to Lanzhou (Gansu Province). Fish thawed in transit are refrozen.

5.9.3 Marketing

There are about 20 market outlets for fresh and frozen naked carp in Xining, although this may not include some of the people selling directly from street stalls. Although there are no detailed records, a large part of the frozen product is exported to Lanzhou and perhaps elsewhere. Part of the factory catch is supplied directly to restaurants or hotels in Xining and other centres.

Market prices of naked carp in Xining generally are low. The market is small partly because naked carp are regarded as an inferior product, and partly because the religious beliefs of part of the population, particularly Tibetans, discourage them from eating fish. At the time of the 1970 census Tibetans accounted for about 35% of the provincial population, although the percentage in Xining alone would be considerably smaller.

Fresh fish, supplied mainly by licensed fishermen and perhaps poachers, also command a low price of about RMB2–4 per kg for fish weighing 500–750 g, or RMB6 per kg for the less common, larger fish of 1–1.5 kg. Fish weighing less than 300 g are not easily sold. The Qinghai Lake factory markets fresh fish only in April; otherwise, most of the frozen product is sold in Lanzhou at prices below RMB2 per kg. By comparison, frozen sea fish in Xining sell for more man RMB8 per kg, live common carp and grass carp sell for about RMB10–15 per kg and fresh rainbow trout for up to RMB20 per kg.

Naked carp therefore has little market appeal. In part, this must be related to poor presentation: the thawed fish lacks texture and is then less presentable on the table. Fresh fish are regarded as only marginally more attractive. Consumers complain that the fish contain many bones, and that the gonads (the testes of stage III males: section 4.2.6) are poisonous and can cause diarrhoea unless removed before cooking. The problem may be accentuated by the factory's practice of re-freezing blocks of fish that thaw in transport, as this must increase the risk of bacterial contamination. However, the palatability of naked carp is not very different from, say common carp, and part of the marketing problem is that the products are poorly presented and have a poor image rather than any inherent qualities of the fish itself.

Improved transport links in recent years have diversified the range of market produce, bringing more competition and different choices for consumers. In principle, the government is willing to invest in improved standards of handling, processing and marketing, given some assurance that increased returns are possible. Options are considered in section 6.