Biological Conservation 1992, 63, 113-118




 THE M A N A G E M E N T OF TRADITIONAL TIDAL PONDS FOR
  A Q U A C U L T U R E A N D WILDLIFE CONSERVATION IN
    SOUTHEAST ASIA: PROBLEMS A N D PROSPECTS

                            S. Y. Lee
Department of Zoology and The Swire Marine Laboratory, University of Hong Kong, Cape D'Aguilar Road, Shek O, Hong Kong

        (Received 24 June 1991; revised version received 9 December 1991; accepted 17 December 1991)


Abstract                              the preponderance of fish and shrimp polyculture or
Despite the long history of tidal aquaculture ponds in       fish (mainly the milkfish Chanos chanos) culture in the
Southeast Asia, they still undergo a largely traditional      tambaks (Polunin, 1983). Growth of brackish water
management with little fertilization, intensification or      paddy has also been practised in some areas (e.g. the
technology input. The recent appreciation of the conser-      bheri in Bangladesh (Mahmood, 1987) and, until re-
vation values of mangrove ecosystems calls for manage-       cently, in gei wais in the Pearl River Estuary, China
ment protocols that can combine traditional exploita-        (Irving & Morton, 1988)) to make full use of the land
tional use with wildlife conservation objectives..4 case      at different seasons. Management of most tidal ponds
study on a Hong Kong tidal pond is described in which        is still largely practised in the traditional manner, i.e.
conflicts may arise when the same pond is managed si-        with little technological input, fertilization, or intensifi-
multaneously for aquaculture production and wildlife        cation (Fast, 1991).
conservation. Sedimentation rate increases as a result of       Apart from being an important source for fishery
controlled water exchange in tidal ponds, leading to        products, mangroves have traditionally been exploited
build-up in substrate level and changes in the type and       by man for forestry, including timber products for con-
amount of vegetation cover. Tidal ponds also support a       struction, charcoal, and non-timber products such as
different, and generally less diverse, fauna from the non-     tannins, dyes and medicines (Jara, 1987; Saenger, 1987).
~mpounded areas, probably a result of the larger fluctua-      More recent attention on the mangrove ecosystem has,
tion in physical conditions. Water level management for       however, focused on their conservation (e.g. Johannes
shrimp and fish culture also conflicts with waterfowl use      & Hatcher, 1986; Fortes, 1988; Gomez, 1988; Hatcher
of the ponds. Wetland reduction, due to tidal aquacul-       et al., 1989). Conflicting uses of the mangals and simi-
ture, reduces nursery areas for fish and crustaceans and      lar wetland ecosystems have created management and
makes serious inroads into mangrove swamps, which are        policy problems in many countries (Bally & Branch,
a declining world resource.                     1986; Nelson, 1986; Platt, 1987; Tompkins, 1987; Ewel,
                                  1990). With increasing understanding of the ecology of
                                  mangrove ecosystems, their importance in shoreline
INTRODUCTION
                                  stabilisation and wildlife protection has been recog-
Tidal ponds excavated from mangrove areas are a his-        nised (Johannes & Hatcher 1986; Hutchings & Saenger,
toric and yet still flourishing landscape feature through-     1987; Saenger, 1987). The impact of such a traditional
out Southeast Asia and in parts of South America          landscape feature as the tidal pond, totalling about 1.3
(Macintosh, 1983; Knox & Miyabara, 1984; Nor, 1984;         million ha in the Indo-West-Pacific in the 1980s (Mac-
Terchunian et al., 1986; Unesco/UNDP, 1987). The In-        intosh 1983), on the conservation of the coastal wet-
donesian version, the tambak, has a history of over 400       land system thus deserves special attention.
years (Schuster, 1952) and covered about 225 200 ha in         This study considers the problems of managing such
1984 (Cholik & Poernomo, t987). This area has been         tidal ponds for both aquaculture and conservation
increasing constantly and the Indonesian government is       using data on the ecology of an enclosure in Hong
planning further development (Knox & Miyabara,           Kong and published information on other Asian
1984; Naamin, 1987a; Choong et al., 1990). The Philip-       examples.
pines had 206000 ha of brackish water ponds in 1986
(Bureau of Fisheries and Aquatic Resources, 1986). In
Southeast Asia, tidal ponds have many local variations       THE ECOLOGY AND MANAGEMENT OF A
in operation and composition of the cultured species.        TRADITIONAL TIDAL SHRIMP POND IN HONG
For example, ponds in southern China are mainly used        KONG
for shrimp cultivation (Macnae, 1962), in contrast to        The study site
Biological Conservation 0006-3207/92/$05.00 © 1992 Elsevier     A detailed study of the production ecology and organic
Science Publishers Ltd, England. Printed in Great Britain      matter dynamics of a traditional tidal pond has been
                               113
114                           S. Y. Lee

carried out in the Mai Po Marshes Nature Reserve in      1990b), equivalent to 8.4 + 8.8 kg dry wt m -2 year-I.
Hong Kong (Lee, 1988 1989a,b, 1990a,b, 1991). The       This fast accretion rate strongly influences the floristics
marshes are a mangrove-dominated wetland in the        of the tidal pond.
Pearl River Estuary and have an area of about 300 ha.       Emergent macrophytes such as mangroves usually ac-
Of these, 200 ha comprise freshwater fish ponds and      celerate accretion ~(Turner, 1990). As the mangroves are
tidal shrimp ponds (gei wais) excavated from the native     progressively less frequently inundated, dispersal of
mangroves in the 1940s. The remaining 100 ha are rela-     viviparous (K. candel) and cryptoviviparous (Avicennia
tively undisturbed tidal mangroves. Both the mudflat      and Aegiceras) propagules is limited. The elevated mud
and the tidal pond mangrove communities are domi-        surface also encourages the spread of otherwise less salt-
nated by Kandelia candel (Rhizophoraceae), Avicennia      tolerant species such as reed Phragmites australis, result-
marina (Avieenniaceae) and Aegiceras corniculatum        ing in a change in the source of primary production and
(Myrsinaceae). The marshes are part of the larger Deep     also the ratio of submerged to open areas (Lee, 1990a).
Bay ecosystem (112 km2). A mudflat with maximum         Table 1 summarises the change in the amounts of vege-
water depth <6 m at high tide is the most prominent       tated and open areas in the tidal pond between dredging
component of this system. A total of over 250 bird       events, based on serial aerial photographs. The invasion
species is recorded throughout the year from the Mai      by reed resulted in a loss of open area; its coverage in-
Po Marshes (Melville & Morton, 1983). This includes       creased from <25 to >40% between 1986 and 1988. Be-
large populations of endangered species such as Saun-      cause of its high primary productivity (Lee 1990a), this
ders' gull Larus saundersi, Dalmatian pelican Pelicanus     probably raises the total primary productivity of the
crispus and, in 1990, the oriental white stork Ciconia     pond at the expense of the contribution from macroal-
boyciana (Melville & Morton 1983; Melville, 1988). The     gae (mainly Enteromorpha spp.). These represent more
marshes were declared a Nature Reserve by the Hong       easily utilisable sources of organic carbon for detriti-
Kong Government in 1976 . Although traditional use,       vores (which are in turn preyed upon by birds) than the
i.e. aquaculture, is still allowed, entry to the reserve is   vascular plants.
restricted.                            Litter from the two dominant macrophyte producers
  The local tidal ponds are mainly used for the pro-      Kandelia candel and P. australis decomposes at signifi-
duction of penaeid shrimps (Metapenaeus ensis, Pe-       cantly different rates, and supports different macroben-
naeus monodon, P. merguiensis and P. peniscillatus).      thos assemblages (Lee, 1988, 1990a,b). These two
Fish production constitutes about 20-30% of the total      macrophytes also have different degrees of spatial
annual economic return, the major species of economic      complexity. K. candel grows to about 7 m and has a
importance being the yellow-finned bream Sparus latus      more open understorey stratum whereas P. australis
(Sparidae), lady fish Elops saurus (Elopidae), striped     occurs typically as dense stands <3 m tall (shoot
mullet Mugil cephalus (Mugilidae) and Therapon jarbua      density >50 m-2) of low spatial variation. As a result,
(Theraponidae). Macintosh (1983) and Fast (1991) pro-      the two plants also support different aerial consumer
vide detailed descriptions of the general operation of     assemblages. Habitat quality therefore cannot be
tidal aquaculture ponds in Asia.                reflected merely by the maintenance of a high primary
  A study was started in 1985 on the ecology of one of     productivity.
the tidal ponds (9.1 ha) at the Mai Po Marshes. A dia-
gram showing the location of the study pond and the       Faunal composition
reserve can be found in Lee (1989a). The pond is man-      A total of 38 species of fish was recorded in the nature
aged by World Wide Fund For Nature Hong Kong          reserve for the period 1985-89, considerably lower than
(WWFHK) in the traditional manner. The objectives of      that recorded from open mangrove waters (e.g. Thong
the study were (a) to investigate important factors infl-    & Sasekumar, 1984; Pinto, 1987; Robertson & Duke,
uencing the production ecology of the tidal pond; and      1987; Thayer et al., 1987). Of all the species present,
(b) how this may relate to the planning of future man-     highest densities were recorded from the tilapias Ore-
agement strategies for the nature reserve. A model has     ochromis nilotica and O. mossambicus. Due to their
been proposed to describe the pattern, stressing the im-    high reproductive rate and adaptability in the brackish
portance of hydroperiod (draining and flooding regime)
                                Table 1. Change in the amount of vegetation cover in the
and water level management as regulators of organic      study pond with time (the last large-scale dredging took place
matter flow (Fig. 1).                                  in 1970)

Management problems                       Year        Vegetated area (ha)   % of total area
The type and amount of vegetation cover
                                1975            0.855         9.4
The Pearl River carries a high sediment load of 8.6 ×      1977            1.438         15.8
107 t year-1 (Shen, 1983). As water exchange is only      1979            2-129         23.4
possible at particularly high tides, the ponds are       1980            2.157         23-7
flooded and drained for only about five days during       1982            2-202         24-2
each tidal cycle. This results in a high rate of sediment    1983            2-493         27-4
                                1986            3.531         38-8
accretion, which was estimated in the study pond using     1988            5.879         64.6
sediment traps and found to be 1.7 cm year-l (Lee,
                      Tidal pond management in Southeast Asia                       115

                      INCURSING W A T E R

                                               HYDROPERIOD
                                    ~    E  R      LEVEL
                D E E P BAY              ~   /      ~  MANAGEMENT




                                                      EXPORT

                        A
                       STANDINC




                        BIOMASS




                   P. communis  STAND




                          MACRO-
                          ALGAE
                                                       SALE
                          P=2.~




                        PHYTO-
                       PLANKTON
                        P = 0.93


                 P - Production in pond, t dry wt.yr "1
                                           G E l WAI
                  E- Tidal flushing/water exchange
                  NU - Nutrients ]Nitrate, phosphate]
                  H - Harvest


Fig. 1. A model for particulate organic matter flow in the study pond, using the energy language of Odum (1983). Water level
       and hydroperiod management are identified as the most important elements of management.

environment, the tilapias made up about 90% of all             more economically important species such as the
individuals of the fish community and 80% of the              striped mullet and may prey upon the shrimp Metape-
harvested biomass in the study pond (Table 2). The             naeus ensis, the major economic species kept in the
high density of 2.0 individuals m -2 also resulted in            ponds.
stunted growth, with most of the individuals attaining             The highest economic return from the fish catch was
standard length < 15 cm in the first year (Fig. 2). While          from yellow-finned bream, lady fish and striped mullet.
the small size of the tilapias may benefit waders, their          Other species of commercial importance, such as Ther-
contribution to economic return is virtually nil. As            apon jarbua and Platycephalus indicus, occurred in rela-
omnivores, the tilapias also compete for food with the           tively smaller numbers.
                                        Because of its restricted water exchange and shallow
Table 2. Production by the four dominant pelagic fishes in a        depth, the tidal pond amplifies natural fluctuations in
Hong Kong tidal pond during the period September 1985 to          water quality in the estuary (Table 3). Partial loss
           January 1987
                                      of the native mangrove forests, wide fluctuations in
Species      Number of Mean wt Total % Total             physical conditions and a lower spatial complexity
          individuals (g)  wt (kg) by wt             mean that fewer species can survive in the tidal ponds.
                                      Many of the dominant animals on the tidal mudflat
Oreochromis spp.   36 406    66.99 2 429.93      77.9      (e.g. the ocypodid crabs Macrophthalmus convexum and
Mugil cephalus    1 313   164.00 221.04        7.1
Sparus latus     1 564   152.31 238.21        7.6      Uca spp., the mudskippers Periophthalmus cantonensis
Elops saurus      625   3 6 9 . 8 7 231.17     7.4      and Boleophthalmus pectinirostris) are either absent or
                                      occur only at much lower densities in the gei wai
Total         39908          3 120.35 100.0
                                      (Table 4).
 116                                         S.Y. Lee
   18   Oreochromis spp.                    N=260  Jan.8?     Table 3. Comparison of the ranges for important hydro-
                                              graphic parameters recorded from the study pond and from
                                              inner Deep Bay (Environmental Protection Department, 1988)
     J                                         Parameter         Study pond range   Inner Deep Bay




                     ,l L].
   12
                                              Surface
                                               dissolved oxygen      15.5-221.3       5.7-83.4
g                                               (% saturation)
u
                                              Surface salinity (ppt)   0-27           8.1-28.0
   6
                                              Surface temperature (°C)  10-40          19.7-30.4
                                              pH            6-8-8.0          7.2-7.9
                1                              Secchi depth (m)
                                              BOD5 (mg litre-l)
                                                           0-2-1.32
                                                            5-75
                                                                         0.2-0.7
                                                                         1.1-16-0
              F                                Surface Chl a (mg m-a)  0-8-15-9          0.2-24.0
   o ~   2   4  ['-6    s    Io  ~2  ~    ~  ~  ~o  ~
                                              POa-P (mg litre-I)    1.09-9.72         0.05-4.20
                    Standard length [ cml
Fig. 2. Size frequency distribution of tilapias Oreochromis                 such as shrimps require relatively deep water to min-
spp. from the study pond. Fish from the pond were last har-                 imise temperature fluctuations. Feeding and roosting of
vested in January 1985 so the population represents two                   most waders and other waterfowl, however, are only
years' growth. The tilapias probably have extended breeding                 possible in shallow water (<50 cm deep). Thus, confli-
periods in the pond environment, resulting in a large size-                 cts occur between the wildlife manager and the fish
         range for the individuals.
                                              farmer.
Landscape, water level and hydroperiod management                       To facilitate bird utilisation of the tidal ponds of the
Control of water level and hydroperiod are probably                    nature reserve, the water level in the ponds should be
the most important elements of traditional gei wai fish-                  kept low during high tide periods in Deep Bay, allow-
eries operation. Good timing in draining and flooding                   ing the birds to roost and forage for longer periods in
not only maximises the influx of larvae of the cultured                  the reserve. It has been found that one of the dominant
species but also enhances survival of the impounded ju-                  resident bird species at Mai Po, the Chinese pond
veniles. It is important that the general water level in                  heron Ardeola bachus on average uses the gei wais for
the tidal ponds be kept high to minimise fluctuations in                  winter feeding for about 10--15% of the time (L.
physical conditions such as temperature, in order to en-                  Young, pers. comm.). The same pattern of using the
hance survival of the cultivated species. There are,                    drained gei wais for winter feeding probably applies to
however, different water level management objectives in                   many other waders, e.g. the little (Egreta garzetta) and
different seasons. The level in the pond was generally                   great (E. alba) egrets. The landscape and hydrological
kept low (<1 m) during winter, as the farmers believe                    regime of the other gei wais under W W F H K control
that exposing the mud to moderate insolation will facil-                  have also been modified to facilitate bird use of the
itate shrimp and fish growth. Periodic exposure of the                   habitat. Water level is generally kept low to attract
mud can stimulate algal growth to enhance fish pro-                     waders, but survival and growth of the cultured fish
duction and this practice is widely adopted by tidal                    and shrimp species have been poor.
pond farmers throughout Asia (Macintosh, 1983).                        In addition to conflicting with the normal method of
Higher levels (>1 m) prevailed during the summer pe-                    gei wai water level management (to flood during high
riod to avoid heating up of the shallow water.                       tide and drain during low tide), such management is
  Wading birds and their prey have incompatible water                   also conducive to drastic fluctuations in water quality
level requirements. Fish and other commercial species                    as compared with the tidal areas. The pH of the water,
Table 4.    List of animals and plants occurring at significantly different densities in the tidal pond and on the mudflat seaward to the
                                   ponds

            Species with high densities in tidal pond                    Species with high densities on mudflat
  Animals
    Sermyla tornatella (Gastropoda)                              Boleophthalmus pectinirostris (Osteichthyes)
    Discapseudes sp. (Tanaidacea)                               Periophthalmus cantonensis (Osteichthyes)
    Oreochromis spp. (Osteichthyes)                              Macrophthalmus convexum (Brachyura)
                                                 Salinator sp. (Gastropoda)
                                                 Uca arcuata (Brachyura)
                                                 Uca acuta (Brachyura)
                                                 Littorina melanostoma (Gastropoda)

  Plants
    Phragmites australis (Gramineae)                             Avicennia marina (Avicenniaceae)
    Echinochloa crus-galli (Gramineae)                            Aegiceras corniculatum (Myrsinaceae)
    Macaranga tanarius (Euphorbiaceae)                            Acanthus ilicifolius (Acanthaceae)
                   Tidal pond management in Southeast Asia                    117

for example, is strongly affected by frequent exposure   1987), often decreasing their value as wildlife habitats.
of the acid sulphate sediment to air and this probably   The undesirable impact has been recognised by the In-
results in the low fish and shrimp yields.         donesian government and there are plans to reafforest
                              the disused tambaks (de la Cruz, 1984; Choong et al.,
                              1990).
DISCUSSION
                                Although these ponds can provide a rapid, short-
The use of tidal ponds for aquaculture has been devel-   term, economic return, a vicious circle leading to in-
oped in Southeast Asia and South America because it     creased and accelerated wetland destruction may result.
provides a relatively cheap way of tapping a rich re-    Turning mangroves into aquaculture ponds decreases
~ource. No fry or fertilizers need to be added, as stocks  nursery sites for offshore species, causing a decline in
of larvae and nutrients are carried by the water into the  capture fisheries, and therefore encouraging more
mangrove environment. Such traditional management      ponds to be built.
requires the minimum amount of manpower but has        Traditionally, attention has centred on the integra-
only limited productivity. Knox and Miyabara (1984)     tion of different types of consumptive uses of natural
suggested that tambaks in Indonesia provided about     systems (e.g. rice/fish farming in Malaysia (Ali, 1990);
112% of the total fish production by value, with good    fisheries/aquaculture (Kapetsky, 1987)). A concept of in-
employment opportunities because of their large area.    tegrated management for aquaculture and conservation
Exploitation of mangrove forests for aquaculture also    is now needed as outlined by Desaigues (1990). This re-
seems to be more profitable than forestry or capture    quires (a) more research on the impact of various tradi-
fisheries (Ong, 1982; Hatcher et al., 1989). It is there-  tional landscape and operation practices on the use of
lbre likely that more such ponds will continue to be    mangroves for wildlife conservation; (b) various man-
built and operated in the traditional manner. Tidal     agement alternatives to be explored, and negotiated so-
pond construction is now the major cause for man-      lutions achieved through a suitable decision process;
grove destruction in Latin America (Terchunian et al.,   and (c) a suitable regulatory instrument to be estab-
1986; Lahmann et al., 1987) and throughout Asia (total   lished to ensure correct implementation.
area >4.5 × 105 ha in 1980s) (Kapetsky, 1987; Naamin,
1987b). The area of tambaks in Indonesian mangroves,
fbr example has been steadily increasing, from about    ACKNOWLEDGEMENTS
1.7 × 105 ha in 1978 to 2.25 × 105 in 1984 (Polunin,
1983; Cholik & Poernomo, 1987; Naamin 1987a).       Part of this work was carried out during the tenure of a
  Nevertheless, recent studies on various methods of    John Swire Scholarship in Wetland Ecology for a PhD
brackish water pond shrimp or fish culture have indi-    degree. I am grateful for the logistic support from the
cated that such systems are not economically feasible    staff of WWFHK Mai Po Nature Reserve. Thanks are
except under semi-intensive or intensive conditions     due to Professor Brian Morton for his constructive
(Ong, 1982; Chiu et al.1987; Fast, 1991). Kapetsky     comments on the manuscript.
(1987) has therefore suggested ways of improving tidal
pond aquaculture which also prevent further destruc-
                              REFERENCES
tion of mangroves: intensification, development of non-
destructive forms of aquaculture and better manage-     Ali, A. B. (1990). Rice/fish farming in Malaysia: a resource
ment protocols. These, however, can only be of limited    optimization. Ambio, 19, 404-8.
use if excavation is still seen as an easier option.    Bally, R. & Branch, G. (1986). The Bot River estuary. Aft.
  Intensification seems to be the main direction of de-    Wildl. 40, 230-9.
                              Bureau of Fisheries and Aquatic Resources (1986). Annual
velopment in tidal pond aquaculture. However, ecolog-     Report No.36. Iloilo City, The Philippines.
ical problems such as eutrophication, introduction of    Chiu, Y. N., Posadas, B. C. & Estilo, V. J. (1987). Strategy
exotic species and of bioactive compounds (e.g. antibi-    for a cost/benefit analysis in the conversion of mangrove
otics) to the natural environment are common conse-      areas to aquaculture. In Workshop on the Conversion of
quences (GESAMP, 1991). Highly intensified establish-     Mangrove Areas to Aquaculture, UNDP/Unesco Regional
                               Project RAS/79/002, pp.86-95.
ments can be located outside wetland areas and thus     Cholik, F. & Poernomo, A. (1987). Development of aquacul-
minimise their destruction.                  ture in mangrove areas and its relationship to the man-
  The tidal pond, with its levees and sluice control of   grove ecosystem. In Papers Contributed to the Workshop on
the hydrological regime, should facilitate management     Strategies for the Management of Fisheries and Aquaculture
of wetlands as wildlife habitats. Problems, however,     in Mangrove Ecosystems, Bangkok, Thailand 23-25 June
                               1986. FAO Rep., No 370 (Suppl,)., pp. 93-104.
arise when the same pond is to be managed for conser-    Choong, E. T., Wirakusumah, R. S. & Achmadi, S. S. (1990).
vation and aquaculture production, as can be seen in     Mangrove forest resources in Indonesia. For. Ecol.
the Hong Kong example. Unlike the Hong Kong gei        Manage., 33134, 45-57.
wais, in many countries mangroves are cleared when     de la Cruz, A. A. (1984). A realistic approach to the use and
tidal ponds are built, giving a monotonous appearance     management of mangrove areas in Southeast Asia. In
                               Physiology and Management of Mangroves, ed. H. J. Teas.
as in the Indonesian tambaks (Polunin, 1983). The con-    Dr W. Junk Publisher, The Hague, pp. 65--68.
struction of these and other similar ponds therefore has  Desaigues, B. (1990). The socio-economic value of ecotones.
drastic effects on mangal landscapes (Lahmann et al.,     In The Ecology and Management of Aquatic-terrestrial
 118                              S . Y . Lee

  Ecotones ed. R. J. Naiman & H. Decamps. Parthenon Pub-         of mangrove swamps and forests in the Indo-West Pacific
  lishing, Carnforth, pp. 263-93.                    region. Adv. Mar. Biol., 6, 78-270.
Environmental Protection Department (1988). Marine Water        Mahmood, N. (1987). Effects of shrimp farming and other im-
  Quality in Hong Kong. Hong Kong Government, Hong Kong.        pacts on mangroves of Bangladesh. In Papers contributed to
Ewel, K. C. (1990). Multiple demands on wetlands. Bio-          the Workshop on Strategiesfor the Management of Fisheries and
  Science, 40, 660-6.                          Aquaculture in Mangrove Ecosystems. Bangkok. Thailand,
Fast, A. W. (1991). Marine shrimp pond grow out conditions and      23-25 June 1986. FAO Rep., No 370 (Suppl.), pp. 4~66.
  strategies: a review and prognosis. Rev. Aquat. Sci., 3, 357-99.   Melville, D. S. (1988). Hong Kong. In A Directory of Asian
Fortes, M. D. (1988). Mangrove and seagrass beds of east          Wetlands. World Conservation Union (IUCN), Gland, pp.
  Asia: habitat under stress. Ambio, 17, 207-13.             283-94.
GESAMP, (1991). Reducing environmental impacts of            Melville, D. S. & Morton, B. (1983). Mai Po Marshes. World
  coastal aquaculture. Rep. Stud. GESAMP, 47.              Wide Fund For Nature Hong Kong, Hong Kong.
Gomez, E. D. (1988). Overview of environmental problems in       Naamin, N. (1987a). Conversion of mangrove areas to aqua-
  the east Asian seas region. Ambio, 17, 166-9.             culture in Indonesia. In Workshop on the Conversion of
Hatcher, B. G., Johannes, R. E. & Robertson, A. I. (1989).        Mangrove Areas to Aquaculture. UNDP/Unesco Regional
  Review of research relevant to the conservation of shallow       Project RAS/79/002, pp. 51-65.
  tropical marine ecosystems. Oceanogr. Mar. Biol. Ann.        Naamin, N. (1987b). Impact of 'tambak' aquaculture to the
  Rev., 27, 337-414.                           mangrove ecosystem and its adjacent areas with special ref-
Hutchings, P. & Saenger, P. (1987). Ecology of Mangroves.         erence to the north coast of West Java. In Mangroves of
  University of Queensland Press, St Lucia.               Asia and the Pacific: Status and Management. Unesco/
Irving, R.T.A. & Morton, B. (1988). A Geography ofMai Po         UNDP Regional Project RAS/79/002, pp.355-65.
  Marshes. World Wide Fund For Nature Hong Kong,            Nelson, R. W. (1986). Wetland policy crisis: United States
  Hong Kong.                               and United Kingdom. Agric., Ecosys. Environ., 18, 95-121.
Jara, R. S. (1987). Traditional uses of the mangroves in the      Nor, S. M. (1984). Major threats to the mangroves of Asia
  Philippines. In Mangrove Ecosystems of Asia and the          and Oceania. In Productivity of the Mangrove Ecosystem:
  Pacific: Status. Exploitation and Management, ed. C. D.        Management Implications, ed. J. E. Ong & W. K. Gong.
  Field & A. J. Dartnall. Australian Institute of Marine Sci-      Universiti Sains Malaysia, Penang, pp. 69-78.
  ence, Townsviile, pp.114-30.                     Odum, H. T. (1983). Systems Ecology--An Introduction.
Johannes, R. E. & Hatcher, B. G. (1986). Shallow tropical         Wiley, New York.
  marine environments. In Conservation Biology: The Science      Ong, J. E. (1982). Mangroves and aquaculture in Malaysia.
  of Scarcity and Diversity, ed. M. E. Soul& Sinauer Associ-       Ambio, 11, 252-7.
  ates, Sunderland, Massachusetts, pp. 371-82.             Pinto, L. (1987). Environmental factors influencing the occur-
Kapetsky, J. M. (1987). Conversion of mangroves for pond         rence of juvenile fish in the mangroves of Pagilao, Philip-
  aquaculture: some short-term and long-term remedies. In        pines. Hyrobiologia, 150, 283-301.
  Workshop on the Conversion of Mangrove Areas to Aqua-        Platt, R. H. (1987). The advance designation approach. Envi-
  culture. UNDP/Unesco Regional Project RAS/79/002, pp.         ronment, 9, 16-43.
  96-106.                               Polunin, N. V. C. (1983). The marine resources of Indonesia.
Knox, G. A. & Miyabara, T. (1984). Coastal Zone Resource         Oceanogr. Mar. Biol. Ann. Rev., 21,455-531.
  Development and Conservation in Southeast Asia with         Robertson, A. I. & Duke, N. C. (1987). Mangroves as nursery
  Special Reference to Indonesia. Unesco & East-West           sites: comparisons of the abundance and species composition
  Centre, Jakarta and Honolulu.                     of fish and crustaceans in mangroves and other nearshore
Lahmann, E. J., Snedaker, S. C. & Brown, M. S. (1987).          habitats in tropical Australia. Mar. Biol. 96, 193-205.
 Structural comparisons of mangrove forests near shrimp        Saenger, P. (1987). Mangrove use and conservation. In Man-
 ponds in southern Ecuador. Interciencia, 12, 240-3.           grove Ecosystems of Asia and the Pacific: Status. Exploita-
Lee, S. Y. (1988). The ecology of a traditionally managed         tion and Management, ed. C. D. Field & A. J. Dar.tnall.
 tidal shrimp pond, the production and fate of macro-          Australian Institute of Marine Science, Townsville, pp.
 detritus and implications for management. PhD thesis,          97-103.
 University of Hong Kong, Hong Kong.                  Schuster, W. H. (1952). Fish-culture in brackish-water ponds
Lee, S. Y. (1989a). Litter production and turnover of the         of Java. Indo-Pacif Fish. Coun. Spec. Pubis, No 1.
 mangrove Kandelia candel (L.) Druce in a Hong Kong tidal       Shen, T. (1983). Ecological balance in the Pearl River Delta.
 shrimp pond. Estuar. Coast. ShelfSci., 29, 75-87.            Occ. Pap. Dep. Geogr. Chin. Univ. Hong Kong, No. 51.
Lee, S. Y, (1989b). The importance of Sesarminae crabs Chi-       Terchunian, A., Klemas, V., Segovia, A., Alvarez, A., Vas-
 romanthes spp. and inundation frequency on the decompo-         conez, B. & Guerrero, L. (1986). Mangrove mapping in
 sition of mangrove (Kandelia candel (L.) Druce) leaf litter       Ecuador: the impact of shrimp pond construction. Environ.
 in a Hong Kong tidal shrimp pond. J. Exp. Mar. Biol.          Manage., 10, 345-50.
 Ecol., 131, 23-43.                          Thayer, G. W., Colby, D. R. & Hettler, W. F., Jr (1987). Uti-
Lee, S. Y. (1990a). Net primary productivity, litter produc-       lization of the red mangrove prop root habitat by fishes in
 tion and decomposition of the reed Phragmites communis         south Florida. Mar. Ecol. Prog. Ser., 35, 25-38.
 (L.) Trin in a nature reserve in Hong Kong: management        Thong, K. L. & Sasekumar, A. (1984). The trophic relation-
 implications. Mar. Ecol. Prog. Ser., 66, 161-73.            ships of the fish community of the Angsa Bank, Selangor,
Lee, S. Y. (1990b). Primary productivity and particulate         Malaysia. In Proceedings of the Asian Symposium on Man-
 organic matter flow in an estuarine mangrove wetland in         qrove Environment Research and Development, ed. E.
 Hong Kong. Mar. Biol., 106, 453-63.                   Soepadmo, A. N. Rap & D. J. Macintosh. University of
Lee, S. Y. (1991). Herbivory as an ecological process in a        Malaya and Unesco, Kuala Lumpur, pp. 385-99.
 Kandelia candel (L.) Druce mangal in Hong Kong. J. Trop.       Tompkins, M. E. (1987). South Carolina's diked tidal wet-
 Ecol., 7, 337-48.                            lands: the persisting dilemmas. Coastal Manage. 15, 135-55.
Macintosh, D. J. (1983). Fisheries and aquaculture signifi-       Turner, R. E. (1990). Landscape development and coastal
 cance of mangrove swamps, with special reference to the         wetland losses in the northern Gulf of Mexico. Amer.
 Indo-west Pacific region. In Recent Advances in Aquacul-        Zool., 30, 89-105.
 ture, ed. J. F. Muir & R. J. Roberts. Croom Helm, Lon-        Unesco/UNDP (1987). Report of the Workshop on the Con-
 don, pp. 5-85.                             version of Mangrove Areas to Aquaculture. UNDP/Unesco
Macnae, W. (1962). A general account of the fauna and flora        Regional Project RAS/79/002.