Content uploaded by Matthias Halwart
Author content
All content in this area was uploaded by Matthias Halwart on Jun 10, 2014
Content may be subject to copyright.
i
CHAPTER | Topic
Culture of fi sh in rice fi elds
Edited by
Matthias Halwart
Modadugu V. Gupta
ii FAO and The WorldFish Center | Culture of Fish in Rice Fields
Culture of fi sh in rice fi elds
Edited by
Matthias Halwart
Modadugu V. Gupta
2004
Published by FAO, Viale delle Terme di Caracalla, 00100 Rome, Italy and
the WorldFish Center, PO Box 500 GPO, 10670 Penang, Malaysia
Halwart, M. and M.V. Gupta (eds.) 2004. Culture of fi sh in rice fi elds. FAO and The WorldFish Center, 83 p.
Preparation of this document:
The original manuscript was prepared by Mr. F. Yap with contributions from Mssrs. S.D. Tripathi,
G. Chapman, S. Funge-Smith, and K.M. Li. Mr. H. Guttman reviewed and condensed the manuscript.
Valuable comments and suggestions on later versions were received from Mssrs. P. Balzer,
C.H. Fernando, W. Settle, K. Gallagher, R. Labrada, and H. van der Wulp. The fi nal inputs and revisions
were provided by the editors M. Halwart and M.V. Gupta.
The designations employed and the presentation of the material in this publication do not imply the
expression of any opinion whatsoever on the part of the Food and Agriculture Organization of the United
Nations or of the WorldFish Center concerning the legal status of any country, territory, city or area or of its
authorities, or concerning the delimitation of its frontiers or boundaries.
Perpustakaan Negara Malaysia. Cataloguing-in-Publication Data
Culture of fi sh in rice fi elds / edited by Matthias
Halwart, Modadugu V. Gupta.
Includes bibliography.
ISBN 983-2346-33-9
1. Freshwater fi shes. 2. Fish-culture.
I. Halwart, Matthias. II. Gupta, Modadugu V.
639.31
Cover photos by: WorldFish photo collection
ISBN 983-2346-33-9
WorldFish Center Contribution No. 1718
© FAO and The WorldFish Center 2004
All rights reserved. Reproduction and dissemination of material in this information product for educational
and other non-commercial purposes are authorized without any prior written permission from the copyright
holders provided the source is fully acknowledged. Reproduction of material in this information product for
resale or other commercial purposes is prohibited without written permission from the copyright holders.
Applications for such permission should be addressed to the Chief, Publishing and Multimedia Service,
Information Division, FAO, Viale delle Terme di Caracalla, 00100 Rome, Italy (copyright@fao.org), or to the
Managing Editor, The WorldFish Center, PO Box 500 GPO, 10670 Penang, Malaysia
(worldfi sh-publications@cgiar.org).
Printed by Practical Printers Sdn. Bhd.
The WorldFish Center is one of the 15 international
research centers of the Consultative Group on International
Agricultural Research (CGIAR) that has initiated the public
awareness cam paign, Future Harvest.
iii
CHAPTER | Topic
Contents
Foreword ...................................................................................................................................................................................................................................
1. Introduction .................................................................................................................................................................................................................
2. History .................................................................................................................................................................................................................................
3. The Rice Field Ecosystem ............................................................................................................................................................................
3.1 Types of Riceland Ecosystem
3.2 The Wet Rice Field Ecosystem
3.2.1. Factors affecting fi sh and other aquatic organisms
3.2.2. Factors affecting plants
3.2.3. Rice fi eld fauna
3.2.4. Impact of aquatic fauna on the rice fi eld ecosystem
3.2.5. The rice fi eld as a fi sh culture system
4. Modification of Rice Fields for Fish Culture .........................................................................................................................
4.1 Increasing Dike (Bund) Height
4.2 Provisions of Weirs or Screens
4.3 Provision of Drains
4.4 Fish Refuges
4.4.1. Trenches
4.4.2. Fish pits or sumps
4.4.3. Ponds in rice fi elds
4.4.4. Rice fi elds in ponds
4.4.5. Ponds connected to rice fi elds
4.4.6. Fish pen within a rice fi eld
5. Production Systems ...........................................................................................................................................................................................
5.1 Concurrent Culture
5.1.1. Rice and fi sh
5.1.2. Rice and fi sh with livestock
5.1.3. Rice and crustaceans
5.1.4. Concurrent but compartmentalized culture
5.2 Rotational Culture
5.2.1. Fish as a second crop
5.2.2. Crustaceans as a second crop
5.3 Alternating Culture System
6. Agronomic and Aquaculture Management ....................................................................................................................
6.1 Pre-Stocking Preparation
6.2 Water Needs and Management
6.3 Fertilization
6.4 Rice Varieties
6.5 The Fish Stock
6.5.1. Species
6.5.2. Fry and fi ngerling supply
6.5.3. Stocking pattern and density
6.5.4. Fish nutrition and supplemental feeding
7. Rice-Fish Production .........................................................................................................................................................................................
7.1 Fish Yields
7.1.1. Rice-fi sh
7.1.2. Rice-fi sh-azolla
v
1
3
5
5
6
6
7
10
10
11
12
12
12
12
12
13
15
15
16
17
17
18
18
18
19
19
21
21
21
22
22
23
23
23
23
24
24
24
25
29
30
32
32
32
32
Contents
iv FAO and The WorldFish Center | Culture of Fish in Rice Fields
7.1.3. Rice and crustacean
7.1.4. Polyculture
7.2 Rice Yields
8. Pest Management .................................................................................................................................................................................................
8.1 Managing Pests with Fish Present
8.2 Management of Rice Field Weeds
8.3 Management of Invertebrates
8.3.1. Management of insect pests
8.3.2. Management of Snails
8.4 Management of Diseases
9. Impact of Rice-Fish Culture .......................................................................................................................................................................
9.1 Economics of Production
9.1.1. The “bottom line”
9.1.2. Input analysis
9.2 Benefi ts to Communities
9.2.1. Improved income status of farmers
9.2.2. Improved Nutrition
9.2.3. Public Health
9.2.4. Social Impact
9.3 Impact on the Environment
9.3.1. Biodiversity
9.3.2. Water resources
9.3.3. Sustainability
9.4 Participation of Women
9.5 Macro-Economic Impact
10. Experiences of Various Countries .....................................................................................................................................................
10.1 East Asia
10.2 SouthEast Asia
10.3 South Asia
10.4 Australia
10.5 Africa, Middle East and West Asia
10.6 Europe
10.7 The Former Soviet Union
10.8 South America and the Caribbean
10.9 The United States
11. Prospects and Program for the Future ......................................................................................................................................
11.1 Prospects
11.2 Major Issues and Constraints
11.3 Research and Development Needs
11.4 Institutional Policy and Support Services
11.4.1. Mainstreaming rice-fi sh farming
11.4.2. Popularization of the concept
11.4.3. Training and education
11.4.4. Fingerling supply
11.4.5. Financing
12. Conclusion .......................................................................................................................................................................................................................
13. Re f e r en c e s .......................................................................................................................................................................................................................
33
34
34
40
40
41
42
42
44
45
48
48
48
48
49
49
52
52
53
54
54
54
54
54
55
56
56
57
59
60
60
61
62
62
62
64
64
64
65
66
66
66
66
67
67
68
70
v
CHAPTER | TopicFOREWORD
Foreword
Rice today is grown in 113 countries in the world in a wide range of ecological conditions and water
regimes. The cultivation of most rice crops in irrigated, rainfed and deepwater systems offers a
suitable environment for fi sh and other aquatic organisms. Over 90% of the world’s rice, equivalent
to approximately 134 million hectares, is grown under these fl ooded conditions providing not only
home to a wide range of aquatic organisms, but also offering opportunities for their enhancement
and culture.
The purpose of this review is to synthesize available information and highlight the important role
that aquaculture in rice-based farming systems can play for food security and poverty alleviation.
Aquatic production, in addition to the rice crop itself, is a critically important resource for rural
livelihoods in developing countries; its local consumption and marketing are particularly important
for food security as it is the most readily available, most reliable and cheapest source of animal
protein and fatty acids both for farming households as well as for the landless.
This review describes the history of the practice and the different rice ecosystems in which fi sh
farming takes place. The various production systems, including modifi cations of the rice fi elds
necessary for integrating fi sh farming, and the agronomic and aquaculture management are
examined. Pest management in rice has evolved tremendously over the past decades, and the culture
of fi sh and other aquatic organisms can reinforce environmentally and economically sound farming
practices.
The real and potential impact of rice-fi sh farming in terms of improved income and improved
nutrition is signifi cant but generally underestimated and undervalued. Hidden benefi ts of rice-fi sh
farming such as risk reduction through diversifi cation of the farming system may have a strong
attraction to many farmers and their families. Fish can be sold directly, or may reduce the
dependence of families on other livestock which can then be traded for income. Also, fi sh from the
rice fi elds may not be sold but the production may be used to feed relatives and those who assist in
rice harvesting, a benefi t which could almost be considered essential in families with a labour
shortage.
The time for emphasizing the importance of rice-fi sh farming is particularly relevant in light of the
currently celebrated UN International Year of Rice 2004.1 Fish from rice fi elds have contributed in
the past, and continue to contribute today, towards food security and poverty alleviation of many
people in rural areas. With signifi cant changes particularly in pest management and fi sh seed
availability taking place in many rice-producing countries, there is now considerable potential for
rice-fi sh farming to further expand its contribution to improve the livelihoods and food security of
the rural families.
M. Halwart
Fisheries Department, Food and Agriculture
Organization of the United Nations
M.V. Gupta
WorldFish Center
1 The United Nations General Assembly (UNGA) declared the year 2004 the International Year of Rice (IYR) and invited the Food and Agriculture
Organization of the United Nations to act as the lead agency for the implementation of the IYR, in collaboration with partners from national,
regional, and international agencies, non-governmental organizations, and the private sector. The FAO Fisheries Department with the assistance
of Fisheries Offi cers from the Regional and Sub-Regional O ffi ces contributes to the IYR through various awareness-raising activities related to the
importance of aquatic biodiversity in rice- based ecosystems. Information is available at http://www.rice2004.com.
vi FAO and The WorldFish Center | Culture of Fish in Rice Fields
1
Introduction
“There is rice in the fi elds, fi sh in the water.” This
sentence inscribed on a stone tablet from the
Sukhothai period - a Thai kingdom that fl ourished
700 years ago - depicts a scene that must have
been as idyllic then as it continues to be now.
Having rice in the fi elds and fi sh in the water is an
epitome of abundance and suffi ciency. No other
combination would seem to be so fundamental
and nutritionally complete in the Asian context.
As such, few other plant and animal combinations
seem to be more appropriate to culture together
to improve nutrition and alleviate poverty. Fish
culture in rice fi elds provides the means for “the
contemporaneous production of grain and
animal protein on the same piece of land”
(Schuster 1955), and in this environmentally
conscious age, few other food production systems
seem more ecologically sound and effi cient.
In the strictest sense rice-fi sh farming means the
growing2 of rice and fi sh together in the same
fi eld at the same time. However, it is also taken to
include the growing of rice and fi sh serially one
after another within the same fi eld or the growing
of rice and fi sh simultaneously, side by side in
separate compartments, using the same water.
Fish by no means strictly refers to fi n-fi sh. It
means aquatic animals living in rice fi elds
including freshwater prawn, marine shrimp,
crayfi sh, crab, turtle, bivalve, frog, and even
insects.
Rice-fi sh farming is practiced in many countries
in the world, particularly in Asia. While each
country has evolved its own unique approach and
procedures, there are also similarities, common
practices and common problems.
Global recognition of, and interest in, the
potential of rice-fi sh farming in helping combat
malnutrition and poverty has been well known
for a long time. The FAO Rice Committee
recognized the importance of fi sh culture in rice
fi elds back in 1948 (FAO 1957). Subsequently it
has been the subject of discussions by the Indo-
Pacifi c Fisheries Council (IPFC), the General
Fisheries Council of the Mediterranean (GFCM),
the FAO Rice Meeting and the International Rice
Commission (IRC). IPFC and the IRC formulated
2 “Growing” is taken to mean the intentional culturing of organisms of either wild or cultured origin.
a joint program for promoting investigations to
evaluate the utility of fi sh culture in rice fi elds.
However, international interest gradually waned
over the years perhaps due to the use of chemical
pesticides and herbicides in the early attempts to
boost rice productivity.
It was not until the late 1980s when global interest
in rice-fi sh farming was renewed. Rice-fi sh farming
was identifi ed as a project of the International Rice
Research Institute’s (IRRI) Asian Rice Farming
Systems Network (ARFSN). This project, led by the
International Center for Living Aquatic Resources
Management (ICLARM), the present WorldFish
Center, was implemented as a collaborative effort
involving many institutions throughout Asia. At
the same time, the International Development and
Research Center (IDRC) of Canada co-sponsored
China’s National Rice-Fish Farming Systems
Symposium in Wuxi. The papers presented at the
symposium were translated into English and
published by IDRC (MacKay 1995). Much of the
information on China in this review was obtained
from that book.
Over the last 15 years, the spread of rice-fi sh
farming has been uneven and campaigns to
promote the practice have often been
discontinued. There are a multitude of reasons for
this including inappropriate extension campaigns,
cheap and readily available pesticides, and lack of
credit facilities.
This report seeks to review rice-fi sh farming as
practiced in different countries, explores the
similarities and differences, and identifi es
experiences that may be useful to promote rice-
fi sh culture in other parts of the world. This is not
a “how-to” manual; instead it aims to describe
how it was done or is being done in various parts
of the world.
The report is structured in four main sections and
a brief conclusion. After the introduction the fi rst
section begins with background information
including a brief history of rice-fi sh culture
(Chapter 2) and a description of the rice fi eld
ecosystem (Chapter 3). The second section then
1. Introduction
2 FAO and The WorldFish Center | Culture of Fish in Rice Fields
continues with the system itself with descriptions
of modifi cations needed for fi sh culture in rice
fi elds (Chapter 4), the various production systems
(Chapter 5), the culture techniques and
management (Chapter 6), production and yields
(Chapter 7), and pest management (Chapter 8).
The third section aims to put rice-fi sh culture in
context by discussing its importance to farmers as
well as its social and environmental impact
(Chapter 9). The fourth section reviews the
experiences and status of rice-fi sh worldwide
(Chapter 10) and concludes with the prospects
and program for the future and the lessons
learned, primarily in Asia, that can be useful in
the promotion of rice-fi sh culture in other parts of
the world (Chapters 10-11).
3
History
Both botanical and linguistic evidence point to
the early origin of cultivated rice in an arc along
continental Asia extending from eastern India
through Myanmar, Thailand, the Lao PDR,
northern Vietnam, and into southern China.
Although the oldest evidence of cultivated rice
comes from Myanmar and Thailand, wet rice
cultivation3 involving the puddling and
transplanting of rice seedlings is thought to have
been refi ned in China. In contrast to other areas,
the history of rice in river valleys and low-lying
areas in China is longer than its history as an
upland crop.
It can be assumed that once rice farming
progressed beyond shifting cultivation in forest
clearings to one involving puddled fi elds with
standing water, fi sh must have been an additional
product. Fish and other aquatic organisms would
have come in with the fl ood water, made the rice
fi eld their temporary habitat, and grew and
reproduced within the duration of the rice
farming cycle to become a welcome additional
rice fi eld product for the farmers.
It may never be known exactly when or where the
practice of deliberately stocking fi sh in rice fi elds
fi rst started. However, since it is widely
acknowledged that aquaculture started early in
China, where pond culture of common carp
(Cyprinus carpio) began at the end of the Shang
Dynasty (1401-1154 BC) (Li 1992), it is assumed
that rice-fi sh farming with stocked fi sh also started
in China. Archaeological and written records
trace rice-fi sh culture in China over 1 700 years
ago and the practice may have started when fi sh
farmers with excess fry released them in their rice
fi elds (Li 1992; Cai and Wang 1995).
Clay models of rice fi elds with fi gurines of
common carp, crucian carp (Carassius carassius),
grass carp (Ctenopharyngodon idella), and other
aquatic animals date back to the later Han
Dynasty (25-220 AD) (Bray 1986, cited in FAO
2000). The earliest written record dates from the
Wei Dynasty (220-265 AD) that mentions “a
small fi sh with yellow scales and a red tail, grown
3 Wet rice cultivation includes the IRRI rice ecosystems of rainfed lowland, fl ood-prone and irrigated rice that together make up 87% of the world’s rice area
and 96% of the rice production (IRRI 2001).
4 Note that older reports mention the term “goldfi sh” only and Ardiwinata (1957) suggests that both Cyprinus carpio as well as Carassius auratus were
included.
2. History
in the rice fi elds of Pi County northeast of
Chendu, Sichuan Province, can be used for
making sauce.” The fi sh referred to is thought to
be common carp.
Rice-fi sh culture was fi rst described by Liu Xun
(circa 889-904 AD) (Cai et al. 1995) who wrote:
“In Xin Long, and other prefectures, land on the
hillside is wasted but the fl at areas near the houses
are hoed into fi elds. When spring rains come,
water collects in the fi elds around the houses.
Grass carp fi ngerlings are then released into the
fl ooded fi elds. One or two years later, when the
fi sh are grown, the grass roots in the plots are all
eaten. This method not only fertilizes the fi elds,
but produces fi sh as well. Then, rice can be
planted without weeds. This is the best way to
farm.”
It is possible that the practice of rice-fi sh culture
developed independently in India and other parts
of the “Asian arc” of wet rice farming, but was not
documented or circulated. Apart from being
described as “an age-old practice” there are few
estimates of how long rice-fi sh farming with
deliberate stocking of fi sh has been practiced
outside China, although some authors suggest
that rice-fi sh culture was introduced to Southeast
Asia from India 1 500 years ago (Tamura 1961;
Coche 1967; Ali 1992).
Integrated rice-fi sh farming is thought to have
been practiced in Thailand more than 200 years
ago (Fedoruk and Leelapatra 1992). In Japan and
Indonesia, rice-fi sh farming was developed in the
mid-1800s (Kuronoma 1980; Ardiwinata 1957).
An early review on rice-fi sh culture showed that
by the mid-1900s it was practiced in 28 countries
on six continents: Africa, Asia, Australia, Europe,
North America and South America (FAO 1957).
Common carp was then the most popular species,
followed by the Mozambique tilapia (Oreochromis
mossambicus). In Malaysia the snakeskin gouramy
(Trichogaster pectoralis) was favored, and Nile
tilapia (Oreochromis niloticus) was used in Egypt.
Other species mentioned include buffalo fi sh
(Ictiobus cyprinellus), the Carassius4 (Carassius
4 FAO and The WorldFish Center | Culture of Fish in Rice Fields
auratus), milkfi sh (Chanos chanos), mullets (Mugil
spp.), gobies (family Gobiidae), eels, murrels or
snakeheads (Channa spp.), catfi sh (Clarias
batrachus), gouramy (Trichogaster pectoralis) as well
as penaeid shrimps (Penaeus spp.).
Coche (1967) pointed out that in most countries
rice-fi sh farming did not involve deliberate or
selective stocking of fi sh and that the species
cultured and the stocking density depended on
what came in with the fl ood waters. Thus the
species cultured usually refl ected what was living
in the waters used to fl ood or irrigate the rice
fi elds. It appears that rice-fi sh farming did not
spread out from one focal point but may have
developed independently.
5
The Rice Field Ecosystem
3. The Rice Field Ecosystem
3.1 Types of Riceland Ecosystem
Rice farming is practiced in several agro-
ecological zones (AEZs) although most of the
rice farming occurs in warm/cool humid
subtropics (AEZ 7), warm humid tropics (AEZ 3)
and in warm sub-humid tropics (AEZ 2). Cutting
across the AEZs, IRRI (1993) has categorized rice
land ecosystems into four types: irrigated rice
ecosystem, rainfed lowland rice ecosystem,
upland rice ecosystem, and fl ood-prone rice
ecosystem (Figure 1). Apart from the upland
system, the others are characterized by wet rice
cultivation. Asia accounts for over 90% of the
world’s production of rice and almost 90% of the
world’s rice land areas. In the irrigated rice
ecosystem, the rice fi elds have assured water
supply for one or more crops a year. Irrigated
lands cover over half of the world’s rice lands
and produce about 75% of the world’s rice
supply.
The rainfed lowland rice ecosystem is characterized
by its lack of control over the water and by both
fl ooding and drought problems. About one
quarter of the world’s rice lands are rainfed.
The upland rice ecosystem varies from low-lying
valleys to undulating and steep sloping lands
with high runoff and lateral water movement. The
soils vary in texture, water holding capacity and
nutrient status since these could range from the
badly leached alfi sols of West Africa to fertile
volcanic soils in Southeast Asia. Less than 13% of
the world’s rice land is upland rice.
The remaining rice lands are classifi ed as fl ood-
prone rice ecosystems (almost 8%), subject to
uncontrolled fl ooding, submerged for as long as
fi ve months at a time with water depth from 0.5
to 4.0 m or more, and even intermittent fl ooding
with brackish water caused by tidal fl uctuations.
Included here are tidal rice lands in coastal plains.
Figure 1. Rice land ecosystems (after Greenland 1997 as adapted from IRRI 1993).
Upland Rainfed lowland Irrigated Flood-prone
Level to steeply sloping fi elds;
rarely fl ooded, aerobic soil;
rice direct seeded on plowed
dry soil or dibbled in wet, non-
puddled soil
Level to slightly sloping,
bunded fi elds; non-continuous
fl ooding of variable depth and
duration; submergence not
exceeding 50 cm for more
than 10 consecutive days; rice
transplanted in puddle soil or
direct seeded on puddle or
plowed dry soil; alternating
aerobic to anaerobic soil of
variable frequency and duration
Leveled, bunded fi elds with
water control; rice transplanted
or direct seeded in puddle soil;
shallow fl ooded with anaerobic
soil during crop growth
Level to slightly sloping or
depressed fi elds; more than 10
consecutive days of medium
to very deep fl ooding (50 to
more than 300 cm) during crop
growth; rice transplanted in
puddle soil or direct seeded
on plowed dry soil; aerobic or
anaerobic soil; soil salinity or
toxicity in tidal areas
6 FAO and The WorldFish Center | Culture of Fish in Rice Fields
Flooding is not the only problem in these areas as
they may also suffer from drought as well as acid-
sulphate and/or saline soils.
Regardless of the ecosystem, fi sh can conceivably
be raised wherever wet rice cultivation is practiced.
The main determinant in the feasibility of raising
fi sh in any given rice land is the availability of
water and the water holding or dike-forming
characteristic of the soil. The volume and
seasonality of water dictate the fi sh-culture
approach for any given area. Rice lands where the
water supply is highly seasonal or constrained
have limited options for rice-fi sh farming,
whereas year-round supply of water provides
greater potential for rice-fi sh culture. Reference to
rice lands and rice fi elds in the rest of this
document refers to wet rice cultivation.
3.2 The Wet Rice Field Ecosystem
The wet rice fi eld can be described as a “temporary
aquatic environment” (Roger 1996) or “a special
type of wetland” that can be considered “a
successor of shallow marshes or swamps” (Ali
1998), which is infl uenced and maintained by
farmers’ activities. Heckman (1979) suggested
that as long as the land was farmed it would
maintain its equilibrium from year to year.
In general, the aquatic environment in rice fi elds
is characterized by shallowness, great variation in
turbidity as well as extensive fl uctuations in
temperature, pH and dissolved oxygen. Owing to
the intermittent nature of the standing water, the
aquatic fl ora and fauna, which may be rich, are
transitory in nature and must have their origins in
the irrigation canals and water reservoirs
(Fernando 1993).
This section is not meant to be exhaustive but
focuses on subjects that are relevant to the raising
of fi sh in rice fi elds. For a more comprehensive
discussion on the rice fi eld ecosystem, the reader
is directed to Heckman (1979) or Roger (1996).
The focus here is on the main aspects of the rice
fi eld ecosystem that affect the animals and plants
living in the rice fi eld as well as a brief overview of
the inhabitants themselves.
3.2.1. Factors affecting fish and other
aquatic organisms
The main factors affecting the fi sh and other
animals in the rice fi eld are the water level,
temperature, dissolved oxygen (DO), acidity
(measured as pH) and unionized ammonia
(NH3). Other factors are also important but not
to the same extent. For a more detailed discussion
on how various factors affect fi sh and other
aquatic organisms, the reader is advised to consult
Boyd (1979, 1982).
The water level in rice fi elds often varies from 2.5
to 15.0 cm depending on the availability of water
and the type of water management followed,
making it an unsuitable environment for
organisms requiring deeper waters. This is the fi rst
and often major constraint to the types of
organisms that may be able to live in the rice fi eld
environment. This is naturally not the case in
fl ood-prone rice lands.
With such shallow depth, the water is greatly
affected by weather conditions (solar radiation,
wind velocity, air temperature and rainfall). In
addition, a fl ooded rice fi eld functions like a
greenhouse, where the layer of water acts like the
glass of a greenhouse. Short-wave radiation (light)
from the sun heats up the water column and the
underlying soil, but long wave radiation (heat) is
blocked from escaping, thus raising the
temperature. Figure 2 shows the amount of heat
that can accumulate is dependent on many
factors, but usually makes the water and soil
temperature in a rice fi eld higher than the air
temperature (Roger 1996).
Maximum temperature measured at the soil/
water interface can reach 36-40°C during mid-
afternoon, sometimes exceeding 40°C during the
beginning of the crop cycle. Diurnal fl uctuations
are often about 5°C and decrease with increased
density of the rice canopy. Maximum diurnal
variations of over 16°C have been recorded in
Australia.
As all animals consume oxygen the amount of
DO is of great importance, although some
organisms are amphibious and others can use
atmospheric oxygen. The DO concentration in a
rice fi eld is the result of mechanical, biological
and chemical processes. The mechanical processes
consist of wind action and the resultant diffusion
through the air-water interface. A major source of
DO in the water column is the photosynthetic
activity of the aquatic plant biomass that can lead
to super-saturation in the mid-afternoon,
although at night the oxygen is used up by the
respiration by plants. Thus, together with
respiration by animals, bacteria and oxidation
processes, anoxic conditions result during the
7
The Rice Field Ecosystem
night and pre-dawn period (Fernando 1996).
This is more pronounced in deepwater rice fi elds,
which can become anoxic during the second half
of the rainy season (University of Durham
1987).
Respiration uses oxygen and produces carbon
dioxide (CO2) that when dissolved in water forms
carbonic acid (H2CO3), which in turn dissociates
into bicarbonates (HCO3-) and carbonates (CO3-2).
This results in the release of hydrogen ions (H+)
which increase the acidity of the water, and cause
the pH to drop. Atmospheric CO2 through natural
diffusion and agitation on the surface water and
decomposition of organic matter are other
important sources of carbon dioxide. On the
other hand, removal of CO2 from the water due to
photosynthetic activity causes the hydroxyl ions
(OH-) to increase and raises the pH of the water.
The DO level and pH of the water in a rice fi eld
are positively correlated since the DO
concentration is largely a result of photosynthetic
activity that uses up carbon and reduces the
dissolved CO2 (and thus H+ concentration),
effectively raising both pH and the DO levels.
Conversely both are lowered during the time
when respiration dominates (Figure 3).
Depending on the alkalinity (or buffering
capacity) of the water, these diurnal variations
can range from zero DO to super-saturation and
from acid to highly basic (pH>9.5) waters during
times of algal blooms (Roger 1996).
Ammonia (NH3) is an important source of
nitrogen in the rice fi eld. In its ionized form,
NH4+, ammonia is rather harmless to fi sh, while
its unionized form, NH3, is highly toxic. The
proportion of the different forms is dependent on
the pH of the water, where the NH3 concentration
increases by a factor of 10 per unit increase of pH
between pH 7 to 9 (Roger 1996). As such the
ammonia concentration in the water can cause
the death of fi sh and other organisms when the
pH of the water reaches high levels, particularly so
after applying nitrogen-rich fertilizer to the rice
fi elds.
3.2.2. Factors affecting plants
The main factors affecting the plants in the rice
fi elds are water, light, temperature, soil nutrients
(nitrogen, phosphorus, potassium and other
minerals) as well as the farming practices. The rice
fi eld fl ora consists of the rice plants as well as
many types of algae and other vascular
macrophytes. The vegetation apart from the rice
plants is often referred to as the photosynthetic
aquatic biomass (PAB). The algae alone in a rice
fi eld have been reported to develop a biomass of
several tonnes fresh weight per hectare (Roger
1996).
Figure 2. Average monthly values of maximum air temperature and of temperature in the flood water, upper (0-2 cm) and lower (2-10 cm) soil
at 1400 hr, IRRI farm, 1987 (Roger 1996).
8 FAO and The WorldFish Center | Culture of Fish in Rice Fields
A continuous fl ooding of 5.0-7.5 cm water is
considered best for optimum grain yield, nutrient
supply and weed control. When the rice starts to
ripen, the plants need very little water and usually
the rice fi elds are drained about 10 days before
harvest to make the work easier. Drying the rice
fi eld results in a drastic shift in the composition
of fl oral species as only soil algae and spore-
forming blue-green algae (cyanobacteria) can
withstand periods of dryness. The chemical make-
up of the water in rice fi elds depends initially on
its source (rainfall, fl ood water from a river, an
irrigation canal or a well). Once it becomes part
of the rice fi eld, its composition changes
drastically due to dilution by rain, dispersion of
the surface soil particles, biological activity and
most of all fertilizer application.
The amount of sunlight in a rice fi eld depends
on the season, latitude, cloud cover, as well as
the density of the plant canopy. The crop canopy
causes a rapid decrease in the sunlight reaching
the water. One month after transplanting, the
amount of light reaching the water surface may
drop by as much as 85% and after two months
by 95% (Figure 4). Shading by the rice plant can
limit the photosynthetic activities of algae in the
rice fi eld as the rice crop grows. Turbidity of the
fl ood water, density of plankton, and fl oating
macrophytes further impair light penetration.
Light availability infl uences not only the
quantity but also the species composition of
photosynthetic aquatic biomass. Many green
algae are adapted to high light conditions while
the blue-greens or cyanobacteria are regarded as
low light species. Certain species of blue-green
algae are, however, known to be resistant to or
even favored by high light intensity (Roger
1996).
Both high and low temperatures can depress
phytoplankton productivity and photosynthesis.
Similar to sunlight, the temperature may also
have a species-selective effect. Higher temperatures
favor the blue-greens while lower temperatures
stimulate the eukaryotic algae.
Soil factors also determine the composition of
algae where acid soil favors chlorophytes (green
algae) and alkaline soil fosters nitrogen-fi xing
cyanobacteria. Application of agricultural lime
(CaCO3) in acidic soil increases the available
nitrogen and promotes growth of cyanobacteria.
High amounts of phosphorus also seem to be a
decisive factor for the growth of the blue-green
algae.
Figure 3. Correlation between the Oxygen concentration of the flood water and pH in five flooded soils (P.A. Roger and P.M. Reddy, IRRI 1996
unpublished from Roger 1996).
9
The Rice Field Ecosystem
The most profound effects on the rice fi eld fl ora
may be those resulting from human intervention
or farming practices. Tillage results in the
incorporation of algae and macrophytes and their
spores into the soil and dispersion of clay particles
in the water. After being mixed with the soil, it is
likely that the motile forms of algae such as
fl agellates will be more successful at re-
colonisation since these are capable of moving to
the surface to be exposed to sunlight. The
suspension of clay particles, on the other hand,
makes the water turbid and results in reduced
amount of light available for photosynthesis.
Mineralized nitrogen is released rapidly into the
fl ood water following land preparation. This is
believed to be the reason behind algal blooms
frequently observed immediately after puddling.
The method of planting also affects algal growth.
Transplanting favors algal growth compared to
broadcasting since broadcasting results in an
earlier continuous canopy which curtails light
compared to transplanting.
Fertilization, while intended for the rice plant,
cannot but affect the growth and development of
all the aquatic organisms in the fl ood water. The
effects depend on the type of fertilizers and
micronutrients used and may vary from site to
site. Moreover, each plant and algal species also
react differently to separate applications of N, P, K
and CaCO3.
Of importance to rice-fi sh culture is the application
of nitrogen rich fertilizer such as ammonium
sulphate [(NH4)2SO4] and urea. Application results
in an increase of ammonia concentration in the
water, up to 40-50 ppm with ammonium sulphate
and less than half of that with urea. Phosphorus
application, which is often done at monthly
intervals, stimulates algal growth and thus
productivity. Otherwise it has no effect on the
animals in the rice fi eld.
Surface application of NPK frequently results in
profuse algal growth with the planktonic forms
developing fi rst followed by the fi lamentous forms
that persist longer. Nitrogen-rich fertilizer favors
growth of eukaryotic algae while inhibiting the
growth of blue greens. In phosphorus-defi cient
soils, the addition of phosphorus fertilizers or
phosphorus-rich manure enhances the growth of
algae. Calcium is rarely a limiting factor to algal
growth in rice fi elds, but liming stimulates the
growth of blue-greens by raising the pH. The use of
organic manure may temporarily reduce algal
growth during the active decomposition stage, but
may later favor the growth of blue-green algae.
The composition of aquatic plants in a rice fi eld
may also be determined by the organisms in the
fi eld, which may be pathogens, antagonists or
grazers. Certain bacteria, fungi and viruses are
pathogenic and infl uence succession. Some algae
are antagonistic by releasing substances that
Figure 4. Relation between plant height and incident light intensity transmitted under the new canopy (Kurasawa 1956 from Roger 1996).
10 FAO and The WorldFish Center | Culture of Fish in Rice Fields
inhibit growth. Finally, there are the animal
grazers - organisms that rely on the aquatic plants
as food, such as cladocerans, copepods, ostracods,
mosquito larvae, snails and other invertebrates.
In the experimental rice plots of IRRI in the
Philippines, primary productivity has been
measured to range from 1.0 to 2.0 g C·m-2·day-1,
but in most cases would range from 0.2 to 1.0 g
C·m-2·day-1. These values are similar to the
productivity values reported in eutrophic lakes.
3.2.3. Rice field fauna
The rice fi eld has a surprisingly great biodiversity,
perhaps the greatest of any tropical rainfed system,
where Heckman (1979) recorded a total of 589
species of organisms in a rice fi eld in Thailand. Of
these, as many as 233 were invertebrates (excluding
protozoans) representing six phyla of which over
half were arthropod species. In addition, there
were 18 fi sh species and 10 species of reptiles and
amphibians. A similar number of fi sh, snails, crabs
and larger insects are reported in Cambodia
(Gregory and Guttman 1996).
Rice fi elds also serve as the habitat for birds and
wildlife for part or all of their life cycle. Ali (1998)
lists at least 13 bird species and 6 small mammals
that may be found in rice fi elds.
The rice fi eld biodiversity is under threat not only
due to changing farming practices with widespread
mechanization and use of chemical inputs, but
also environmental degradation leading to the
disappearance of permanent reservoirs (or refuges)
for organisms within the vicinity of the rice fi elds
(Fernando et al. 1979). Rice fi elds used to be, and
remain, a rich source of edible organisms in many
areas. Heckman (1979) found that one vegetable
and 16 animal species were collected in a single
rice fi eld in Thailand. Similar fi gures are found in
other areas of Southeast Asia (Gregory 1996;
Gregory and Guttman 1996). Balzer et al. (2002)
reported about 90 aquatic species (excluding
plants) that are collected by Cambodian farmers in
their rice fi elds and used daily by rural households.
Such diversity of food from a rice fi eld, while still
common in many areas, is reported to be
decreasing (Halwart 2003b).
3.2.4. Impact of aquatic fauna on the
rice field ecosystem
The aquatic fauna plays an important role in
nutrient recycling. Whether as primary or
secondary consumers, animals excrete inorganic
and organic forms of nitrogen and phosphorus
and are a major factor in the exchange of
nutrients between soil and water. Among the
organisms, the benthic oligochaetes (family
Tubifi cidae) have received special attention
because they can move between the reduced soil
(which lies beneath the shallow oxidized layer)
and the fl ood water. Together with ostracods and
dipteran larvae, oligochaetes respond positively
to nitrogen fertilizer if applied by broadcasting,
but not when applied by deep placement.
Indigenous snail populations on the other hand
are strongly affected negatively by broadcast
application of N fertilizer (Simpson 1994).
Fish plays an important role in the nutrient cycle
of the rice fi eld ecosystem. Cagauan (1995) lists
four ways how fi sh may infl uence the nutrient
composition of the fl ood water and the oxidized
surface soil as well as the growth of the rice plant.
First, by contributing more nutrients to the rice
fi eld through faeces excretion as well as through
decomposition of dead fi sh. Second, by the
release of fi xed nutrients from soil to water when
the fi sh swims about and disperses soil particles
when disturbing the soil-water interface. Third, by
making the soil more porous when fi sh disturb
the soil-water interface, fi sh increase the nutrient
uptake by rice. Finally, fi sh assist in the recycling
of nutrients when they graze on the photosynthetic
biomass and other components of the
ecosystem.
More specifi cally, fi sh affect the nitrogen cycle in
a rice fi eld. Cagauan et al. (1993) found that a
rice fi eld with fi sh has a higher capacity to
produce and capture nitrogen than one without
fi sh (Table 1). At the same time, fi sh may help
conserve nitrogen by reducing photosynthetic
activity (by grazing on the photosynthetic aquatic
biomass and by increasing turbidity) and thus
keeping the pH lower and reducing volatilization
of ammonia. This may be important as nitrogen
losses through ammonia volatilization have been
estimated to be from 2 to 60% of the nitrogen
applied (Fillery et al. 1984).
Fish also affect the phosphorus cycle. Phosphorus
is often a limiting nutrient for primary production
as it often becomes fi xed in the soil and is
unavailable to plants in the rice fi eld. Fish, by
disturbing the soil, increase soil porosity and
promote phosphorus transfer to the soil. On the
other hand, by grazing on the oligochaete
population, fi sh may have exactly the opposite
11
The Rice Field Ecosystem
effect as oligochaetes also increase soil porosity.
Plots without fi sh were found to have higher soil
porosity because of the presence of undisturbed
oligochaetes. Fish have been found capable of
reducing oligochaete population in a rice fi eld by
80% (Cagauan et al. 1993).
3.2.5. The rice field as a fish culture
system
In principle, as long as there is enough water in a
rice fi eld, it can serve as a fi sh culturing system.
However, a rice fi eld is by design intended for
rice and therefore conditions are not always
optimum for fi sh. At the most basic level is the
fact that rice does not necessarily need standing
water at all times to survive. Rice can be
successfully grown in saturated soils with no
Unit Rice Rice-Fish
Total production (kg N/crop) 465.60 476.80
Total fl ow to detritus (kg N/ha/crop) 447.10 456.80
Total throughput (kg N/ha/crop) 1 122.22 1 183.60
Throughput cycled (kg N/ha/crop)
(includes detritus) 334.40 346.30
Cycling index (%) 59.60 58.50
Mean path length 11.45 12.11
Table 1. Summary statistics of N models of lowland irrigated rice fields with and without fish.
Source: Cagauan et al. (1993)
PARAMETER NORMAL RANGE
RICE FISH
1. Depth of Water Minimum: saturated soils with no fl ooding;
Ideal: Continuous fl ooding starting at 3 cm
depth gradually increasing to max of 15 cm by
60th day. Complete draining 1 – 2 weeks before
harvest (Singh et al. 1980).
0.4-1.5 m for nursery and 0.8-3.0 m for grow-out
(Pillay 1990)
2. Temperature Water and soil temperature of up 40°C and
fl uctuations of up to 10°C in one day
apparently with no deleterious effect.
25°-35°C for warmwater species. Stable
temperature preferable. Feeding may slow down
at temperatures below or above normal range.
Metabolic rate doubles with every 10°C rise.
3. pH of water Neutral to alkaline. 6.5-9.0 (Boyd 1979).
4. Oxygen Important during seedling stage for
development of radicles. Preferably at near-saturation or saturation level
(5.0-7.5 ppm depending on temperature).
5. Ammonia High levels of ammonia common immediately
after fertilization. Un-ionized ammonia highly toxic. Ionized form
generally safe.
6. Transparency or Turbidity Immaterial. Important for growth of natural food. Very high
level of suspended soil particles may impair
respiration.
7. Culture Period 90-120 days for HYV; up to 160 days for
traditional varieties. 120-240 days depending upon species and
market requirement.
Table 2. Comparison between environmental requirements of fish and rice.
standing water (Singh et al. 1980), and recent
evidence on the system of rice intensifi cation
suggests that intermittent irrigation may increase
rice yields. However, even with a continuously
standing column of water, a fl ooded rice fi eld is
not necessarily an ideal place for growing fi sh.
The water temperature can reach very high levels.
Also, rice requires fertilizer which increases the
total ammonia level in the water and can thus
increase the highly toxic (to fi sh) un-ionized
ammonia level in the water. Rice does not require
oxygen in the water - an element essential for
most fi sh. Finally, rice farming requires other
human interventions which may be detrimental
to the survival and/or growth of fi sh, such as
mechanical weeding or herbicide application.
Some of the contrasting requirements of rice and
fi sh are summarized in Table 2.
12 FAO and The WorldFish Center | Culture of Fish in Rice Fields
Several physical modifi cations have been devised
over the years in order to make the rice fi eld
better suited for fi sh culture. Most are common
to many countries and may have been developed
independently from each other as a result of a
“common sense” approach that characterizes
many traditional practices.
All modifi cations have the basic goals of
providing deeper areas for the fi sh to grow
without inundating the rice plants and of limiting
escape from and access to the rice fi eld. This is
achieved either by making portions of the rice
fi eld deeper than the ground level for the fi sh,
or conversely, by creating areas higher than the
ground level for the rice or other crops. There are
four physical improvements that are commonly
made to prepare rice fi elds for fi sh culture. The
fi rst is to increase the height of the dike or bund
to allow deeper water inside the fi eld and/or to
minimize the risk of it being fl ooded. The second
is the provision of weirs or screens to prevent the
fi sh from escaping as well as keeping predatory
fi sh from coming in with the irrigation water.
The third, which is not always practiced but often
recommended, is provision of proper drains and
fi nally, provision of deeper areas as a refuge for
the fi sh. Details of the various modifi cations
have been described by various authors (e.g.
FAO et al. 2001) and this section will provide a
complementary overview.
4.1 Increasing Dike (Bund) Height
Rice fi eld embankments are typically low and
narrow since the usual rice varieties do not require
deep water. To make the rice fi eld more suitable for
fi sh, the height of the embankment needs in most
cases to be increased. Reports on rice-fi sh culture
from various countries show embankments with
a height of 40-50 cm (measured from ground
level to crown). Since the water level for rice does
not normally exceed 20 cm, such embankments
will already have a freeboard of 20-30 cm. This is
suffi cient to prevent most fi sh from jumping over.
The height of the embankments cannot of course
be increased without a corresponding increase in
the width. There are no hard and fast rules as to
the fi nal width, but generally it is within the range
of 40-50 cm.
4. Modifi cation of Rice Fields for
Fish Culture
4.2 Provisions of Weirs or Screens
Once the fi sh are inside the rice fi eld, efforts are
made to prevent them from escaping with the
water, regardless of whether it is fl owing in or out.
To prevent loss of the fi sh stock, farmers install
screens or weirs across the path of the water fl ow.
The screens used depend on the local materials
available. FAO et al. (2001) list three types of
screens: bamboo slats, a basket, and a piece of
fi sh net material (even a well-perforated piece of
sheet metal).
4.3 Provision of Drains
In general rice fi elds are not equipped with gates
for management of water levels. The common
practice is to temporarily breach a portion of the
embankment to let the water in or out at whatever
point is most convenient. This is understandable
since typically dikes are no more than 25-30 cm
high with an almost equal width. Using a shovel,
a hoe or bare hands, water can be made to fl ow
in or out. Repairing the dike afterwards is just as
easy.
The larger dike required for rice-fi sh culture
makes it more diffi cult to breach, and it will also
take more effort to repair. It is therefore advisable
to provide a more permanent way of conveying
water in or out just like in a regular fi shpond,
although this may incur an extra cost. Generally
reports do not contain enough detail on the type
of water outlets installed, but among these are
bamboo tubes, hollowed out logs, metal pipes
or bamboo chutes (FAO et al. 2001; IIRR et al.
2001).
4.4 Fish Refuges
A fi sh refuge is a deeper area provided for the
fi sh within a rice fi eld. This can be in the form
of a trench or several trenches, a pond or even
just a sump or a pit. The purpose of the refuge
is to provide a place for the fi sh in case water in
the fi eld dries up or is not deep enough. It also
serves to facilitate fi sh harvest at the end of the
rice season, or to contain fi sh for further culture
whilst the rice is harvested (Halwart 1998). In
conjunction with the refuge, provisions are often
13
Modifi cation of Rice Fields for Fish Culture
made to provide the fi sh with better access to the
rice fi eld for feeding.
There are various forms of refuges ranging from
depressions in a part of the rice fi eld, to trenches
to a pond adjacent to the fi eld connected with
a canal. A multitude of systems have been
reported, but they all follow the same principles.
This section will provide a brief overview of the
various types of refuges that are practiced in rice-
fi sh culture, divided into trenches, ponds and
pits or sumps. It should be noted that it is not
uncommon to combine trenches with ponds or
pits, and also that these designations are rather
imprecise as it is a gradual change from a trench
to a lateral pond and likewise from a pit to a pond
and a rather academic issue, of limited practical
value, to determine when a trench becomes a
lateral pond and vice versa.
4.4.1 Trenches
Before describing the various ways trenches have
been used in rice-fi sh culture, it is worthwhile to
note that trenches can have three functions: as
a refuge should water levels drop, a passageway
providing fi sh with better access for feeding in the
rice fi eld and as a catch basin during harvest (De
la Cruz 1980).
There are several ways the trenches could be
dug. The simplest way involves just digging a
central trench longitudinally in the fi eld. Figure 5
illustrates the great variations on this rather simple
theme (Koesoemadinata and Costa-Pierce 1992).
Xu (1995a) reported on the practice to dig
trenches in the shape of a cross and even a
“double-cross”, a pair of parallel trenches
intersecting with another pair, in larger rice fi elds
(from 700 up to 3 000 m2).
The trenches are just wide enough and deep
enough to safely accommodate all fi sh during
drying and weeding and usually require only
the removal of two rows of rice seedlings. In this
manner, the trenches do not signifi cantly affect the
production of the rice crop. Reported widths are
approximately 40-50 cm (Koesoemadinata and
Costa-Pierce 1992) and a suggested minimum
depth is 50 cm, measured from the crown of the
bund to the bottom of the trench resulting in the
bottom of the trench being 25-30 cm to below
the fi eld level (Ardiwinata 1953). Sevilleja et al.
(1992) reported a design with a 1 m wide central
trench with water from a screened inlet fl owing
directly into it a narrow peripheral trench.
Another experimental design in the Philippines
used an ”L-trench” involving two sides of the rice
fi eld, with a width of 3.5 m occupying 30% of the
rice fi eld area.
For fi ngerling production, the ditches are dug
together with 50-70 cm deep 1 m2 pits or sumps
at the water inlet and outlets. Rice seedlings are
planted along both sides of each ditch and three
sides of each pit to serve as “a fence” (Wan et al.
1995).
A variation, reported from China, is a “wide
ditch”5 measuring 1 m wide and 1 m deep, placed
laterally along the water inlet side of the rice fi eld
with a ridge rising about 25 cm above the fi eld
level. It is constructed along the side of the ditch
that is away from the embankment. To allow
the fi sh to forage among the rice plants, 24 cm-
wide openings are made along the ridge at 3-5 m
intervals. These ditches occupy around 5-10% of
the rice fi eld area.
Having a small number of trenches limits the
area for raising fi sh. To provide more area for
them, farmers sometimes dig shallow trenches
(also referred to as furrows or ditches) using
the excavated soil to form ridges where rice is
transplanted. In this manner trenches and ridges
alternate with one another throughout the whole
rice fi eld (Figure 6; Li 1992). The dimensions
of the ridges and ditches are not hard and fast,
varying from one place to another. Ridges range
from 60 to 110 cm to accommodate 2 to 5 rice
seedlings across (Li 1992; Ni and Wang 1995;
Xu 1995a). Ditches range from 35 cm wide by
30 cm deep to 50 cm wide and 67 cm deep (Li
1992; Xu 1995a; Xu 1995b; Ni and Wang 1995).
One or two ditches may be dug across all the
ridges to connect them and improve the water
fl ow. During transplanting water is only in the
trenches. Afterwards the fi elds are fi lled up to
the top of the ridge. Although this method can
improve low-yielding rice fi elds since it makes
multiple use of available resources (Ni and Wang
1995), Wan and Zhang (1995) noted the limited
adaptability of this approach since the method
requires a lot of work that must be repeated each
year. Extension efforts in Jiangxi Province, China,
5 The words “trench” and “ditch”’ are synonymous here since the two words are used interchangeably in the literature on rice-fi eld fi sh culture.
14 FAO and The WorldFish Center | Culture of Fish in Rice Fields
have been successful in establishing this model in
0.5% of the rice-fi sh farming area.
By utilizing the dikes of the rice fi elds to cultivate
dryland crops the fi eld can be described as a
multi-level system. One such system is the surjan
system (Figure 7) found in coastal areas with
poor drainage in West Java, Indonesia. The dikes
are raised to function as beds for dryland crops.
The trenches, the rice area and the dikes form
three levels for the fi sh, rice and dryland crops
(Koesoemadinata and Costa-Pierce 1992).
Xu (1995a) described a development resulting
in a seven-layer rice-fi sh production system
practiced in Chongqing City, China. The seven
“layers” were: sugarcane on the ridges, rice in the
fi elds, wild rice between the rows of rice, water
chestnuts or water hyacinth on the water surface,
silver carp in the upper layer of the water column,
grass carp in the middle layer, and common carp
or crucian carp at the bottom. In order to utilize
rice fi elds comprehensively for better economic,
ecological and social benefi ts, many experiments
on multi level systems have been set up such as
Figure 5. Design and construction of fish trenches in Indonesian rice+fish farms or minapadi (Koesoemadinata and Costa-Pierce 1992).
1– peripheral trench; 2 – diagonal trench; 3 – crossed trenches; 4 – Y-shaped trench; 5 – peripheral with one central longitudinal trench; 6 –
peripheral with two equidistant transverse trenches; 7 – latticed trenches.
15
Modifi cation of Rice Fields for Fish Culture
rice-crab-shrimp-fi sh in Jiangsu Province, rice-
fi sh-mushroom in Helongjiang Province, rice-
fi sh-animal husbandry-melon-fruit-vegetables in
Guizhou Province, and rice-lotus-button crab in
Beijing (Li Kangmin, pers. comm.).
Figure 6. Rice ridge and fish ditch farming system in China (Li Kangmin 1992).
4.4.2 Fish pits or sumps
In some countries sumps are provided as the
only refuge without any trench, for example
when traditional beliefs do not allow major
modifi cations of rice fi elds as in the rice terraces
of the Philippines (Halwart 1998). Coche (1967)
found that farmers in Madagascar dig one sump
for every 100 m2, each measuring 1 m in diameter
and around 60 cm deep. A “stalling pond” was
also provided to hold fi ngerlings.
Sumps can serve as a catch basin during harvest
in addition to providing refuge for the fi sh.
Figure 8 illustrates sumps of 1-2 m width and
depth dug in the center of the rice fi eld for this
purpose (Ramsey 1983). Sumps may just be
simple excavations but modifi cations exist such
as sumps lined with wooden boards to prevent
erosion or a secondary dike built around them
(Ramsey 1983). In Bangladesh, farmers excavate
a sump occupying 1-5% of rice fi eld area with a
depth of 0.5-0.8 m (Gupta et al. 1998).
4.4.3 Ponds in rice fields
Another approach to provide a relatively deeper
refuge for fi sh in a rice fi eld is the provision of
a pond at one side of the rice fi eld. There is no
clear-cut boundary as to when a “trench” becomes
wide enough to be considered a pond.
In Indonesia the payaman or lateral pond (Figure
9) is used in rice fi elds that are located right
beside a river. The pond is constructed here so
that water from the river has to pass through the
pond to get into the rice planting area. A dike
separates the pond from the rice planting area.
Openings are made along the dike to enable the
water to fl ow freely to the rice and allow the fi sh
to forage within the rice fi eld. When the rice fi eld
is drained, the pond serves as a refuge for the fi sh,
making it possible to catch them after the rice
harvest. According to Koesoemadinata and Costa-
Figure 7. Design of Indonesian rice-fish-vegetable farm or surjan
(Koesoemadinata and Costa-Pierce 1992).
B
B
B. Raised beds planted with dryland crops
A. Sunken beds planted with rice
C. Fish trenches
BB
BB
AA
AA
C
CCCC
CCC
A. Sunken beds planted with rice
B. Raised beds planted with dryland crops
C. Fish trenches
Figure 8. Cross-section of Ifugao rice terraces in the Philippines
showing traditional fish harvesting pits (Ramsey 1983).
B
C
A
A
1- 2 m
1- 2 m
16 FAO and The WorldFish Center | Culture of Fish in Rice Fields
Pierce (1992) it is a way of making “better use of
an unproductive part of a rice fi eld.”
A Philippine rice-fi sh model involves the
provision of a minimum of 500 m2 fi shpond
in any one-hectare rice fi eld. In India, instead
of providing a pond only at one end of the rice
fi eld, the West Bengal State Fisheries Department
introduced a design involving two ponds, one at
each end of the rice fi eld (Figure 10). The ponds
have a top width of 18 m and pond bottom width
of 1.5 m. They are 1.5 m deep measured from
the fi eld level. The rice fi eld has a total length of
125 m (inclusive of 3 m dikes). Thus the ponds
actually cover 28% of the gross rice fi eld area and
the dikes about 4.8%. Even with such a large area
devoted to fi sh, farmers in the area who used the
deepwater pond system reportedly were still able
to realize an annual harvest of 5.1-6.4 mt of rice
per ha (Ghosh 1992).
The lateral pond design is the most popular form
of rice fi eld modifi cation in Jiangxi Province,
China (Wan et al. 1995). A small pond is dug at
one end of the fi eld, or shallow pond(s) between
the rice fi elds can be made. The ponds are 1 m
deep and occupy only 6-8% of the total fi eld area.
The ponds are supplemented by 30-50 cm deep
ditches that cover about one-third of the total
pond area.
With the lateral pond, farmers have the option of
making temporary breaches along the partition
dike separating the pond from the rice fi eld to
interconnect the fi shpond with the rice fi eld,
therefore allowing the fi sh to graze among the
rice plants. Water for irrigating the rice has to
pass through the fi shpond. By draining the rice
fi eld and repairing the breach, the fi sh are made
to congregate in the pond compartment and their
culture continues independent of the agronomic
cycle of the rice. Thus the fi sh, if still under-sized,
can be cultured through the succeeding rice crop
if necessary. This model makes it possible to take
advantage of the mutualism between rice and fi sh
while desynchronizing the fi sh culture cycle from
that of rice.
Another option is to maintain a deepwater
fi shpond centrally located in the rice fi eld as is
reported from hilly areas in Southern China. In
Sichuan province, where per caput fi sh production
is low and rice-fi sh farming is perceived as a
promising way of increasing it, circular ponds
made of bricks and cement are placed in the
middle of the rice fi elds (Halwart, pers. comm.).
Ghosh (1992) reported on a 1.5 m deep pond in
India that measured 58 x 58 m in the center of a 1
ha rice fi eld (Figure 11). Note that in the fi gure the
fi shpond deceptively looks much larger than the
rice area when in fact it occupies exactly one-third
of the total area.
4.4.4 Rice fields in ponds
The sawah-tambak rice fi eld - fi sh pond
combination (Figure 12) – in Indonesia is
unique to the low-lying (1-2 m above sea-level)
coastal areas of East Java. These areas are fl ooded
throughout the wet season but lack water during
the dry season. Farmers construct 1.4-2.0 m
high dikes around their land with a 3 m wide
peripheral trench parallel to the dike. A second
dike is built around the rice fi eld that is low
enough to be fl ooded over (Koesoemadinata
and Costa-Pierce 1992).
Figure 9. Design and construction of Indonesian rice+fish farm with
lateral pond or payaman (Koesoemadinata and Costa-Pierce 1992).
B
F
C
G
F
D
H
E
E
A
25 m
4 m
25 m
A. Village
B. River
C. Pond
D. Rice field
E. Inlet with screen (river °% pond)
F. Inlet (pond °% rice field)
G. Oil lamp to attract insects
H. Outlet with screen (rice field °% rice field)
40 cm
100 cm
50 cm
400 cm
2500 cm
40 cm
40 cm
Figure 10. Typical rice-fish pond in West Bengal State Fisheries
Department, India (Ghosh 1992).
C
3 m 18 m
125 m
83 m 18 m 3 m
3 m
74 m
80 m
3 m
Perimeter dike
Lateral pond
Rice plot
Lateral pond
Lateral pond
Perimeter
dike G.L.
Lateral
plot
Rice plot
Section thru C
1 m
18 m 83 m
3 m
1.5 m
1.5 m
Lateral pond
17
Modifi cation of Rice Fields for Fish Culture
Milkfi sh (Chanos chanos) and silver barb (Barbodes
gonionotus) are the main species raised in the
polyculture system, although the common carp
and the giant freshwater prawn (Macrobrachium
rosenbergii) may be grown together with both
species. These are adaptable for either concurrent
or rotational systems of rice-fi sh culture.
4.4.5 Ponds connected to rice fields
In the most important rice-fi sh farming area
in peninsular Malaysia (northwestern Perak),
the practice is to dig a small pond at the lowest
portion of the land, separate from the rice fi eld,
which is connected with the rice fi eld through the
inlet/outlet gate (Ali 1992). The pond is typically
no more than 6-8 m in length and width and has
a depth of 2 m. Fish can graze in the rice fi eld
and still seek refuge in the sump pond when the
water in the rice fi eld is low or too hot. When the
rice is harvested, the pond is drained and the fi sh
harvested as well. Small fi sh are left behind to
provide stock for the next season.
This type of system was also reported from China
(Ni and Wang 1995) with a 1.5 m deep pond that
was used for fry production. Fish are concentrated
in the pond only during harvest time. Once the
subsequent rice crop is planted and established,
the fi sh are allowed to graze freely again.
A similar system was promoted in Cambodia
(Guttman 1999) by connecting small ponds dug
for households under a “food for work” scheme
with the adjacent rice fi elds. The fi sh were often
kept in ponds until the Khmer New Year (mid-
April) as the fi sh prices were at a peak then.
4.4.6 Fish pen within a rice field
Farmers in Thailand set enclosures within the
natural depressions of a rice fi eld to grow fry
to 7 cm fi ngerlings for direct stocking into the
rice fi elds. The enclosures are made of plastic
screens or - less prevalently - bamboo fencing.
Fish are stocked in these enclosures after the
fi rst rains when the water has reached 30-50
cm. Owing to the turbidity during this period,
plankton productivity is low and the fi sh have
to be fed. Farmers try to reduce the turbidity by
surrounding these depressions with a low dike.
For added protection from predators, the net
pen material is embedded in the dike (Sollows
et al. 1986; Chapman 1992; Fedoruk et al. 1992;
Thongpan 1992; Tokrishna 1995; Little et al.
1996).
A net pen can be a useful option in deepwater
rice fi elds where fl ood waters over 50 cm might
persist for four months or longer. This has been
tried in Bangladesh using a 4 m high enclosure
(Gupta 1998). However, investment costs of
the net enclosures to contain the fi sh have
often made the operation uneconomical and
unsustainable.
Figure 11. Central fishpond within a rice field in India (Ghosh 1992).
A
Perimeter dike
Rice plot
Central pond
100 m
58 m
58 m
100 m
58 m
55 m
Central pond
E
Figure 12. Design and construction of Indonesian coastal rice+fish
farm or sawah-tambak (Koesoemadinata and Costa-Pierce 1992).
A. Main dike
B. Fish trench
C. Second dike
D. Area planted with rice
B
B
A
A
C
C
80 m
120 m
D
D
3 m 2 m
2 m
3 m 4 m 3 m
1m
0.5m
18 FAO and The WorldFish Center | Culture of Fish in Rice Fields
5. Production Systems
In categorizing the production systems, it is not
possible to completely divorce the purely physical
design aspects from the cropping practices. This is
because a particular cropping practice may require
some specifi c physical modifi cations although the
converse may not be true. A particular
modifi cation does not necessarily limit the
cropping practice to be employed. Farmers can
always sell their fi sh as fi ngerlings if they fi nd it
fi nancially advantageous to do so, or conversely
grow them to larger size as table fi sh. Farmers in
many areas routinely switch between, or cycle
through, rotational and concurrent practices
using the same rice fi eld.
This section will describe the two main production
systems, concurrent culture – growing the fi sh
together with the rice in the same area - and
rotational culture – where the rice and fi sh are
grown at different times. The fi nal part will
mention an alternating system that is really a type
of rotational culture, but distinct enough to
warrant a separate section.
5.1 Concurrent Culture
The growing of fi sh simultaneously with rice is
what comes to mind for most people when rice-
fi sh culture is mentioned. This is often referred in
short as “rice+fi sh” (Yunus et al. 1992; Roger
1996). As mentioned earlier, physical modifi cations
are required to make a rice fi eld “fi sh-friendly”. The
timing in stocking fi ngerlings is crucial since if
stocked too soon after the rice is planted, some fi sh
species are likely to damage the newly planted
seedlings (Singh et al. 1980), and if too late there
may be a multitude of predator species in the
fi elds.
It should be mentioned that the earliest and
still most widely practiced system involves the
uncontrolled entry of fi sh and other aquatic
organisms into the rice fi eld. Coche (1967)
called this method the “captural system of rice-
fi sh culture.” This can only be considered a rice-
fi sh culture system if the fi sh are prevented
from leaving once they have entered the rice
fi eld. In this system, the organisms often
depend wholly on what feed is available
naturally in the fi eld, although it is not
uncommon for farmers to provide some type of
supplementary feeds.
This system is often practiced in rainfed areas and
plays an important role in many rice-producing
countries, for example in Thailand where rainfed
areas constitute 86% of the country’s rice area
(Halwart 1998), as well as in the Lao PDR (Funge-
Smith 1999) and Cambodia (Guttman 1999;
Balzer et al. 2002). The transition from a pure
capture system and a capture-based culture system
is gradual and has been described as a continuum
(Halwart 2003b).
5.1.1 Rice and fish
The stocking and growing of fi sh in a rice fi eld is
basically an extensive aquaculture system that
mainly relies on the natural food in the fi eld. On-
farm resources and cheap, readily available
feedstuff are often provided as supplementary
feeds, particularly during the early part of the
growing cycle. For the management of the rice
crop, compromises are made with respect to the
application of fertilizer, which is done judiciously.
The use of pesticides is minimized and when
applied the water level may be lowered to allow
the fi sh to concentrate in the refuge.
One constraint of the concurrent system is that
the growing period of the fi sh is limited to that of
rice, which is usually 100 to 150 days.
Consequently the harvested fi sh are small,
especially if early-yielding rice varieties are used.
This can be partly remedied by the use of larger
fi ngerlings, but there is a limit to this since large
fi sh may be able to dislodge the rice seedlings.
Another solution is to limit the production to
that of large-size fi ngerlings for sale to farms
growing table fi sh. The increased demand for
fi ngerlings for growout in cages during the late
1970s in Indonesia was one of the catalysts that
helped popularize rice-fi sh farming.
This system is practiced widely although there are
many variations of the basic theme. For example,
in the minapadi - literally “fi sh-rice” system - of
Indonesia, the rearing of fi sh is not one
continuous process. It consists of three distinct
rearing periods that are synchronized with the
rice cultivation. Two different explanations have
been given for such a procedure: not to subject
the fi sh to very turbid conditions (Ardiwinata
1957) and not to adversely affect rice yields
(Koesoemadinata and Costa-Pierce 1992). The
19
Production Systems
fi rst period takes place from 21 to 28 days between
rice transplanting and fi rst weeding; the second
period during the 40 to 45 days between the fi rst
and second weeding; and the third, during the 50
days between the second weeding and the
fl owering of the rice plants.
The fi rst and second rearing periods may be
considered the nursery periods for growing the fry
to fi ngerling size. The rice fi eld is stocked at the
rate of 60 000 fry·ha-1. During the fi rst weeding,
the fi sh stock is confi ned to the trenches. Before
the second weeding, the fi ngerlings are harvested
and sold. In the third growing period, 8-10 cm
fi ngerlings are stocked at the rate of 1 000 to
2 000 fi sh·ha-1 for the production of food fi sh.
To have more food available for the fi sh, the
Chinese have introduced the growing of azolla
together with the fi sh and rice. Aside from serving
as food for the fi sh, azolla is also a good nitrogen
source for the rice because of its nitrogen-fi xing
capability (Liu 1995). This system works well in
either fi elds with pits or with rice on the ridges:
azolla on the surface of the water and fi sh within
the water column (Yang et al. 1995). The fi eld
must have suffi cient water and good irrigation
and drainage. The proportion of pits and ditch as
to the total area depends on the desired yields of
rice and fi sh.
Yang et al. (1995) found that both fi sh and rice
yields varied according to the ridge width or ditch
width. Fish yields also vary according to the
species cultured and the stage at which they are
harvested (Wang et al. 1995). The output of fi sh
was highest using “food fi sh” followed by carp fry,
catfi sh fry (Clarias gariepinus), and the lowest
yield with grass carp. Chen et al. (1995) reported
a 70% increase in fi sh yield with azolla over
culture without azolla.6
5.1.2 Rice and fish with livestock
Carrying the concept of integration one step
further, livestock rearing may also be integrated
with rice-fi sh systems. This has been tried in many
areas but is not as common as the integration of
livestock with pond culture.
The most common form of integration is
probably the rice-fi sh-duck farming. The
6 The system used “fi ne feed” to feed pigs that produce manure for the rice fi elds and “beer left-overs” as supplementary feed.
7 The term “prawn” is used for freshwater species and “shrimp” for marine and brackishwater organisms.
integration of one hundred laying ducks with a
one ha rice-fi sh farm resulted in the production of
17 031 eggs/year in addition to the rice and fi sh
(Syamsiah et al. 1992). It should be noted that
ducks are also known to feed on snails, and this
combination of biological control agents has
been suggested for controlling the various life
stages of golden apple snails in rice fi elds (Halwart
1994a; FAO 1998).
5.1.3 Rice and crustaceans
Crustaceans raised in rice fi elds range from crabs
and crawfi sh to prawns and shrimp.7 This is
being practiced in many coastal areas relying
either on natural recruitment or in stocked
fi elds.
In the southern United States, crawfi sh
(Procambarus clarkii) are stocked in their adult
stage to serve as broodstock unlike most other
aquaculture systems where juveniles are stocked.
Reproduction occurs in the rice fi eld and it is the
offspring that are harvested. The broodstock are
released in the month of June after the rice has
reached 10-25 cm and the rice fi eld is already
fl ooded. While the rice is growing, the crawfi sh
reproduce and grow. By August the rice is ready
for harvesting. Two weeks before harvesting, the
rice fi eld is drained to make harvesting easier. By
this time all the crawfi sh are expected to have
completed their burrowing (NAS 1976).
The rice stubble left after harvesting re-grows as a
ratoon crop when the fi eld is re-fl ooded and the
new growth is foraged directly by the crawfi sh
(Chien 1978). Loose plant material decomposes
and serves as food for zooplankton, insects,
worms and molluscs, that make up a large part of
the crawfi sh diet. Although any type of vegetation
can serve as forage for crawfi sh, rice appears to be
more widely used. When the fi eld is re-fl ooded
after the rice harvest, the young crawfi sh are
fl ushed out of their burrows and partial/selective
harvesting can start as early as December and
proceed through April/May to June/July
depending upon the desired cropping pattern.
Crawfi sh are harvested at 15-60 g size by using
traps made of plastic or wire screens with ¾ inch
mesh and baited with gizzard shad or carp. Lanes
between the stands of rice are provided to allow
the harvesting boats to move freely.
20 FAO and The WorldFish Center | Culture of Fish in Rice Fields
Although the river crab or mitten-handed crab
(Eriocheir sinensis) has been cultured with rice in
China for less than 12 years, there are now almost
100 000 ha devoted to its culture8 (Wang and
Song 1999). The rice fi eld is used either as a
nursery for the production of crab juveniles (or
“button-crab”); growout for the production of
marketable-sized crabs (125 g); or as a fattening
area for rearing undersized crabs (50-100 g).
The rice fi eld is modifi ed with a peripheral trench
(2-4 m wide, 1 m deep), a cross trench (0.8-1.0 m
wide, 0.5-0.8 m deep) and a sump (20-60 m2, 1
m deep) as a nursery-rearing-harvesting “pond”.
In total 15% to 20% of the total area is modifi ed.
To prevent crabs from escaping, a wall of smooth
material (plastic or corrugated sheet) is installed
(Li 1998).
While saltwater is needed for egg hatching and
rearing the larvae at the initial stage, at later
stages the larvae can develop into crabs in a
freshwater or near-freshwater environment. Li
(1998) identifi ed the stage stocked in rice fi elds
as zoea that in four months attain “stage V zoea”
at 40 to 200 individuals per kg. Wang and Song
(1999) found megalopa9 stocked needed to be
slowly acclimated (six to seven days) to near
freshwater condition (below 3 ppt) for better
survival when stocked in freshwater. It is at this
stage that they are either reared into button-crabs
or reared directly into adults. For the production
of button-crabs, the rice fi elds are stocked at the
rate of 4.5-7.6 kg·ha-1. For growing into
marketable crabs, the stocking rate is 75-150
kg·ha-1. These are harvested upon reaching the
size of 125 g.
Supplementary feed is given consisting of a mix
of trash fi sh, snail, clam or viscera of animals
(40%); vegetables, sweet potatoes, pumpkin, rice
or wheat bran, leguminous cakes (25%); and
terrestrial grass or duck weeds (35%). The trash
fi sh and other animal protein source are steamed
and minced fi nely during the early stage of
growth. The vegetable materials are stewed and
are given during the middle stage. At the late
stage animal feeds are again given in order to
fatten the crabs and develop the gonads that
make the crabs even more prized. Pellet feeds are
also used in some places.
Good water management is essential and about
20 cm of the water is changed every three days or
one-third of the water of the entire fi eld every 10
to 15 days. The dissolved oxygen level is
maintained at a level above 4 ppm throughout the
culture period. Basal manuring and top-dressing
with urea are applied two to three times a year.
The rice crop is harvested at “frost’s descent” and
the crabs by October and November when the
gonads are ripe. The time of harvest may be
advanced if the temperature should abruptly drop
since the crabs have a tendency to burrow when
the temperature is low. The crabs are concentrated
in trenches by irrigating and draining prior to the
rice harvest. The crabs are caught when they crawl
out of the trenches at night by using bottom trap
nets or by draining the water.
The giant freshwater prawn (Macrobrachium
rosenbergii), as well as another prawn species
(M. nipponensis), grow together with rice in China.
The physical preparations are the same as for river
crabs in terms of providing trenches, sumps and
screens; so are pre-stocking preparations up to the
liming stage (Li 1988). Thereafter, submerged
aquatic plants are planted in the trenches to cover
one-half to one-third of the water surface.
For M. rosenbergii, the stocking rate is 3 pieces·per m2
of 1.5 cm sized juveniles.10 The M. nipponensis on the
other hand may be stocked as 4-6 cm size brooders
at 3.0-3.8 kg·ha-1 and allowed to breed, or as
juveniles at 23-30 pieces·per m2. The feed consists
of soybean milk and fi sh gruel for the early stages
(seven to eight days after stocking the fry) and
pelleted feeds or a mixed diet of wheat bran or rice
bran and some animal protein source thereafter.
The M. rosenbergii is fed a higher protein diet.
M. rosenbergii is harvested before the temperature
drops too low. Harvest for M. nipponensis can start
on a selective basis by late November or early
December. The undersized animals are left to
grow for the total harvest by May or June before
the rice planting season.
In coastal rice fi elds encroached by saltwater, it is
common for saltwater shrimps to enter the rice
fi elds with the fl oodwater and grow among the
rice plants. In the Mekong Delta area in Vietnam
8 This includes pens and cages set in lakes, ponds, and rice fi elds.
9 Megalopa is the last larval stage of crabs before they metamorphose into fully-formed juvenile crabs. It is the most likely the more accurate designation
of the crab larvae when stocked in the rice fi elds.
10 This rather low stocking rate is due to the aggressive behavior of the prawn.
21
Production Systems
some farmers have been successful in growing
shrimp together with a traditional tall rice crop in
a brackishwater environment. Supplementary
feeding results in higher yields even when the feed
consists of nothing more than “rice bran, broken
rice and rotten animals” (Mai et al. 1992).
5.1.4 Concurrent but
compartmentalized culture
Rice culture and fi sh culture both require water
and in some circumstances the rice and fi sh are
cultured side by side sharing the water. One
advantage of this set-up is that fi sh rearing
becomes independent from rice, making it
possible to optimize the conditions for both rice
and fi sh. However, the synergistic effect of rice and
fi sh on each other is no longer present. Generally
there is only a one-way infl uence from fi sh to rice
in the form of nutrient-enriched water.
In the rice culture zone of Senegal, environmental
changes have forced the rice farmers to diversify
and integrate fi sh culture in their farming
operations (Diallo 1998). Owing to two decades
of drought, the foreshore mangrove areas have
expanded resulting in the salinization of surface
and ground water. To protect their rice fi elds
against the infl ow of saltwater, farmers built
fi shponds along the foreshore area to produce
fi sh. The fi shponds range from 500 to 5 000 m2
(30 cm deep with 1 m deep peripheral canal).
During the fi rst rain, the gates of the rice fi elds
and fi shponds are opened to allow the rainwater
to wash away any salt that may have accumulated.
Then the gates are closed and the rainwater and
surface runoff are collected for both the rice
planting and fi sh growing operations. After the
rice has been planted from mid-August to mid-
September, the seaward gates are opened during
the spring tides. Coastal fi sh attracted by the fl ow
of freshwater come into the ponds and are
trapped. No attempt is made to control the
species and the number of fi sh that enter. The rice
fi elds and fi shponds are fertilized with cattle and
pig manure and ash. The fi sh are fed rice bran,
millet bran and sometimes termites.
The fi sh are harvested either when the rice is
about to mature or just after the rice has been
harvested from December to January, when the
fi sh have been growing from 120 to 150 days.
Harvesting is done during low tide by draining
the pond with a basket locally known in Senegal
as etolum placed at the end of the drainpipe.
5.2 Rotational Culture
5.2.1 Fish as a second crop
In Hubei and Fujian provinces, China, raising fi sh
during the fallow period or as a winter crop is
practiced to make use of the rice fi eld when it
otherwise would not be used (Ni and Wang 1995).
Elsewhere in China it does not seem to be as
widely practiced as concurrent culture. In
Indonesia, particularly West Java, the art of rotating
fi sh with rice has been developed to a greater
degree and can be traced back to 1862 or earlier.
The Indonesians call raising fi sh as a second crop
palawija or “fallow-season crop.” Instead of growing
another rice crop or soybeans or maize after one
rice crop, some Indonesian farmers grow fi sh. The
only physical modifi cation required is the raising of
the dike to hold water. Without the rice, the entire
rice fi eld can be operated and managed just like a
regular fi shpond from three to six months a year. It
can be used for growing table fi sh or producing
fi ngerlings. The production of two or three crops of
fi ngerlings instead of one crop of table fi sh is done
by some farmers in Indonesia to avoid problems of
poaching or fi sh mortality due to infestation by
predators such as snakes, birds and water insects
(Koesomadinata and Costa-Pierce 1992).
Raising fi sh, in this case common carp as palawija,
was described in detail by Ardiniwata (1957). The
rice fi eld is fl ooded with the rice stubble, either
trodden down or cut off and stacked together with
loose rice-straw, before or after the fi rst fl ooding.
Within two or three days the water becomes putrid
due to the decomposition of plant materials and is
released and replaced with new water. Water depth
is maintained between 30-80 cm.
Carp fi ngerlings are stocked at a density that is
based on the magnitude of the rice harvest and the
size of the fi ngerlings. The rule of thumb is to
stock from 500 to 700 fi ngerlings (5-8 cm long)
for one tonne of padi (unhusked rice) harvested.
Sometimes large fi ngerlings (100 g) are also
stocked at the rate of 10% of the main stock. These
larger fi sh keep the soil surface loose by their
activities. Alternatively, 10-day old carp fry may be
stocked at the rate of 100 000 fry·ha-1 for growing
into fi ngerlings. This practice often results in high
mortality but is apparently resorted to only if no
other area is available as a nursery.
Marketable fi sh are harvested in 40 to 60 days,
fi ngerlings after only 4 weeks. There is enough time
22 FAO and The WorldFish Center | Culture of Fish in Rice Fields
for a second, third or even fourth crop of fi sh prior
to the next rice planting season, depending on the
availability of water. The stocking density is
increased by 25% during the second fi sh cycle but
then reduced since there is a risk of running out of
water before the fi sh have reached marketable size.
In Indonesia, a short growing period is possible
since the local preference is for small fi sh averaging
125-200 g (Costa-Pierce 1992). Table fi sh are
harvested by draining the fi eld, forcing the fi sh into
trenches where they are picked by hand. The fi eld
is left to dry for two days, repairs made and rice
straw turned over and the fi eld is ready once again
for another crop of fi sh. To harvest fi ngerlings, a
temporary drainpipe covered with a fi ne meshed
screen is installed and then the water level is
carefully lowered until it is only in the trenches.
Fingerlings left in puddles on the trench fl oor are
gathered fi rst, and when only a little water is left,
the fi ngerlings concentrated at the screened outlet
are carefully scooped out and placed in holding
vessels for distribution.
Another Indonesian system is called penyelang or
“intermediate crop” where farmers who double-
crop rice with an adequate water supply year-round
fi nd it possible to raise fi sh in between the two rice
crops. Since the seedbeds occupy only a very small
portion of the rice fi eld, the farmers can use the rest
of the rice fi elds for growing fi sh during a period of
1-1½ months suffi cient to produce fi ngerlings.
Some farmers let fi sh breeders use their rice fi elds
during this period (Koesoemadinata and Costa-
Pierce 1992). The whole rice fi eld can be operated
as a fi shpond and with the widespread use of the
high-yielding varities (HYVs) that make possible
four to fi ve crops of rice in two years, the penyelang
is reported to be more popular than the palawija
described earlier.
The fi elds are stocked after they have been tilled
and made ready for the next rice crop and are
already clean and free from rice stubble
(Ardiwinata 1957). This makes them suitable for
rearing carp fry and are sought after by fi sh
breeders. The same stocking density is used as in
palawija (100 000 fry·ha-1). Fingerlings are
harvested after only one month. If used for
growing marketable fi sh, the stocking is 1 000
fi sh·ha-1 (8-11 cm). As long as trenches are
provided, whether peripheral or otherwise, the
fi sh may remain during the plowing and
harrowing process.
5.2.2 Crustaceans as a second crop
Along the western coast of India the low-lying
coastal rice lands are left fallow after one crop of
salt-tolerant rice (Pillay 1990). The dikes are
raised after the rice is harvested (in September)
and tidal water is allowed to inundate the fi eld
carrying with it shrimp larvae and fry. This natural
stocking process continues for two to three
months with every spring tide. Lamps are installed
over the inlet to attract the shrimp larvae and
conical bag nets installed at the sluice gates to
prevent the trapped shrimps from getting out.
Selective harvesting may start as early as December
allowing of the earliest shrimps to enter. Regular
harvesting thins the stock resulting in better
growth rate for the remaining stock. With such
uncontrolled stocking, several species are
harvested but mainly of Penaeus indicus,
Macrobrachium rude and Palaemon styliferus.
This system of shrimp culture is an old practice in
India, but lately due to the high value of shrimps
farmers are devoting greater attention to managing
the shrimp stock through better water management
and fertilization. Many farmers now no longer
leave the stocking to chance preferring instead to
stock at a controlled density using hatchery-
produced postlarvae, particularly of P. monodon.
5.3 Alternating Culture System
Another alternative is an alternating system since
rice takes from 105 to 125 days to mature
depending on the variety, but fi sh can be
marketable as fi ngerlings in as short as 30 to 45
days. Fish therefore can also be a good “time-
fi ller” crop. By alternating between rice-fi sh and
fi sh-only farming, rice fi elds can be productive
throughout the year and higher incomes can be
realized. A farmer may practice two rice crops and
then a fi sh-only crop, or two rice-fi sh crops
followed by a fi sh-only crop, with the latter
becoming more popular in parts of Indonesia
(Koesoemadinata and Costa-Pierce 1992).
Ironically enough, even if rice is the main crop,
fi sh are raised year-round in the rice fi eld rather
than rice. In a survey in West Java, farmers who
practiced two rice-fi sh crops followed by a fi sh-
only system had a net return to input of 173% per
year as against 127% for those practicing a rice-
rice-fi sh system and 115% for those practicing
rice-rice-fallow system (Yunus et al. 1992).
23
Agronomic and Aquaculture Management
6. Agronomic and Aquaculture
Management
As mentioned earlier rice and fi sh sometimes have
confl icting requirements. Growing fi sh in the rice
fi eld does require some modifi cations to the
management to ensure that the fi sh get their
necessary requirements and to facilitate fi sh
survival and growth during certain critical periods.
This section focuses on the additional or modifi ed
management interventions that are needed for
rice-fi sh culture.
6.1 Pre-Stocking Preparation
Whether the modifi cation is in the form of
trenches, lateral ponds or higher and wider dikes,
nothing suggests that one form of modifi cation
can be considered superior to others. The type of
modifi cation used is based on a combination of
different factors: the terrain, soil quality, water
supply, traditions, exposure to other methods,
past experiences, relative importance given to
either rice and fi sh, whether fi ngerling or food fi sh
is desired and the fi nancial resources available.
Although generally rice is the main crop in any
rice-fi sh farming activity, there are exceptions
where rice is planted or ratooned for the purpose
of providing forage for the culture organism.
6.2 Water Needs and Management
Water is the most important single factor in any
agricultural production. Merely supplying
adequate water to enable a previously non-
irrigated area to produce a dry season crop more
than doubles the total annual production as rice
production is often higher during the dry season
than during the wet season. It is estimated that
rice requires a minimum of 1 000 mm of water
per crop, which is inclusive of both
evapotranspiration and seepage and percolation
(Singh et al. 1980). This is equal to 10 000 m3 per
hectare per crop.
Wet rice cultivation uses water either for a
continuous submergence or intermittent
irrigation. The latter has advantages, besides
saving on water, but it may not be the best option
for rice-fi sh culture since it requires concentrating
the fi sh in trenches or sumps every time the rice
fi eld is dry. For rice-fi sh culture it is preferable to
adopt continuous submergence where the rice
fi eld is kept fl ooded from the transplanting time
to about two weeks before harvest.
Continuous fl ooding up to the maximum
tolerated by rice without affecting its rice
production is recommended. In most literature
this is a standing water depth of from 15 to 20 cm
(Singh et al. 1980; Rosario 1984; Koesoemadinata
and Costa-Pierce 1992). At that depth, and with
the fi sh refuge of whatever form having a depth of
50 cm below fi eld level, the effective water depth
of 65-70 cm is available to the fi sh in the refuge.
This is suffi cient to provide the fi sh with a cooler
area when shallow water over the rice fi eld warms
up to as high as 40°C. The increased water depth
means a greater volume of water for rice-fi sh
farming. Despite the fact that seepage and
percolation may be higher with deeper standing
water, fi sh, unlike rice, do not consume water.
Thus a farm with a rice-fi sh system operates
similar to an extensive aquaculture system.
6.3 Fertilization
Application of fertilizers, organic or inorganic,
benefi ts both rice and fi sh. The presence of
adequate nutrients increases the growth of
phytoplankton, which may be consumed directly
by the fi sh or indirectly through supporting
zooplankton production.
Early speculations indicated that rice-fi sh farming
might use from 50% to 100% more fertilizers
than rice farming without fi sh (Chen 1954)
where the additional fertilizer was deemed
necessary to support phytoplankton production
as the base of the fi sh culture food chain. Recent
reports indicate that the presence of fi sh in the
rice fi eld may actually boost rice fi eld fertility and
lower fertilizer needs.
Experiments in China indicate that the organic
nitrogen, alkaline nitrogen and total nitrogen in
the soil are consistently higher in fi elds with fi sh
than in the control fi elds without fi sh (Wu 1995).
Wu attributed this to the fact that fi sh in the rice
fi eld consume weeds and are able to assimilate
30% of the weed biomass. The rest is excreted that
helps maintain soil fertility since nutrients,
otherwise locked up in weeds, are released.
24 FAO and The WorldFish Center | Culture of Fish in Rice Fields
Further experiments showed that rice-fi sh plots
require less fertilizer than rice-only plots. On
average the control plots used 23% more fertilizer
than the rice-fi sh plots (Li et al. 1995b). In
summary, the Chinese experiments indicate that
less, not more, fertilizer is required in rice-fi sh
farming.
Fertilizer applied on rice-fi sh farms by
incorporating the nitrogen fertilizer thoroughly
in the soil during land preparation results in
higher rice yields than when broadcast on the
surface (Singh et al. 1980). Subsequent
fertilization by applying urea as mudballs or as
briquettes is a technique found to increase
fertilizer effi ciency by slowing down the release of
the fertilizer. This avoids the problem of high
ammonia concentration in the water, which may
adversely affect fi sh growth. If the fertilizer is
broadcast on the surface, the rice fi eld should be
drained to expose the planted area and confi ne
the fi sh to the refuge trench or pond. Initial
fertilization ought to be at the same level as in a
rice-only farm since at this stage the fi sh are still
small and cannot be expected to contribute
signifi cantly to the soil fertility. Less fertilizer
should be needed in subsequent applications.
No difference has been found between applying
the phosphorus fertilizer on the surface or
incorporating it in the soil. However, surface
application is believed to be better for promoting
plankton growth in the water. Split applications
of phosphorus may be better for sustained
plankton production without hampering rice
production as long as they are made before
tillering. If applied at a later time, this should be
on top of the normal requirements for rice. An
application rate of 30-50 kg P2O5·ha-1 is often
reported as optimum for algal growth (Singh et
al. 1980).
Organic fertilizers benefi t both rice and fi sh. In
addition to nutrients, the particles can also act as
substrates for the growth of epiphytic fi sh food
organisms. Animal manure should be considered
an input to benefi t the fi sh in addition to
inorganic fertilizers applied primarily for the rice
(Sevilleja et al. 1992). Manure should be applied
several weeks before transplanting and the fi elds
kept fl ooded for complete decomposition and to
avoid any toxic effects (Singh et al. 1980).
Fertilization is a complex issue and varies greatly
depending on the particular location. Providing
general statements runs the risk of over-
simplifying the issue, but there is evidence that
nutrients are more effi ciently utilized in rice-fi sh
systems compared to rice-only systems, this effect
being more enhanced particularly on poorer and
unfertilized soils where the effect of fi sh may be
greatest (Halwart 1998).
6.4 Rice Varieties
With the development of HYVs of rice, several
issues affecting rice-fi sh culture have emerged.
Among these are concerns about the unsuitability
of short-stemmed varieties because of the deeper
standing water required in rice-fi sh farming. This
may be unfounded. Rosario (1984) listed varieties
that have been successfully used for rice-fi sh
farming that included one variety that has a tiller
height of only 85 cm, and this concern may only
apply to areas of moderate to deep fl ooding (≥ 50
cm).
The reduced growing period may be of greater
concern, as many new varieties mature within
approximately 100 days or less. With such a short
culture period for fi sh there is a need to either
stock large fi ngerlings, with the associated
problems in fi sh dislodging and eating rice plants,
or to harvest the fi sh early for further on-growing.
The result is that this may make rice-fi sh farming
a less attractive option in areas where large size
fi sh are preferred. It should be noted here that in
Southeast Asia small-sized fi sh are highly
acceptable, particularly so in the Philippines and
Indonesia.
6.5 The Fish Stock
6.5.1 Species
The fi sh to be stocked in rice fi elds should be
capable of tolerating a harsh environment
characterized by: shallow water, high (up to 40°C )
and variable temperatures (range of 10°C in one
day), low oxygen levels and high turbidity (Hora
and Pillay 1962; Khoo and Tan 1980). Fast growth
is also mentioned as a desirable characteristic so
that the fi sh could attain marketable size when
the rice is ready for harvest.
With such adverse environmental conditions that
a fi sh could tolerate, it would seem that very few
of the commercially valuable species are hardy
enough to qualify. This, however, is not the case.
A review of rice-fi sh farming practices around the
world reveals that practically all the major
freshwater species now being farmed, including a
25
Agronomic and Aquaculture Management
salmonid and even a few brackishwater species,
have been successfully raised in a rice fi eld
ecosystem as well as several crustacean species
(Table 3).
The species farmed in rice fi elds include 37 fi nfi sh
species (from 16 families) and seven crustaceans
(from four families). Molluscs, primarily snails
and some clams are often harvested from rice
fi elds, but there is little information that these are
deliberately stocked.11 The same is true with frogs
and freshwater turtles.
Two groups of fi sh stand out in rice-fi sh farming:
cy prinids a nd tilapias. The c ypri nids, par ticular ly
the common carp and the Carassius have the
longest documented history, having been
described by early Chinese writers. The common
carp has fi gured prominently since ancient times
up to the present and is raised in rice fi elds in
more countries than the other species. The grass
carp and silver carp fi gure prominently,
particularly in China, and the silver barb
(Barbodes gonionotus) in Bangladesh, Indonesia,
and Thailand, and the Indian major carps such
as catla (Catla catla), mrigal (Cirrhinus cirrosus)
and rohu (Labeo rohita) in Bangladesh and
India.
The Mozambique tilapia (O. mossambicus) used to
fi gure prominently in early literature, but is
increasingly replaced by the Nile tilapia
(O. niloticus) in many places. The Nile tilapia is
now as widely used as the common carp in rice-
fi sh farming.
Although rice-fi sh farming of the gouramis,
specially Trichogaster spp., and climbing perch
(Anabas testudineus) initially relied on natural
stock, it is now cultured in Thailand using
hatchery produced fry.
The crayfi sh (Procambarus clarkii) can also be
considered a major species in rice fi eld aquaculture
since these are being raised in hundreds of
thousands of hectares of rice fi elds in the
American south. The practice is not widespread,
mostly in the United States and to a limited extent
in Spain (Halwart 1998).
Among the many species available for raising in
rice fi elds, the choice is based on availability,
marketability or desirability as food. In the
Philippines, tilapia is the species of choice since
carp does not have a wide market outside some
small regional pockets. In Indonesia, people
prefer common carp and silver barb over tilapia
and these are therefore the species of choice for
raising in rice fi elds. In China, people are more
familiar with the various species of carp. With
their long history of aquaculture, Chinese farmers
are aware of the advantages of polyculture over
monoculture so that polyculture of various
species of carps seems to be the rule.
6.5.2 Fry and fingerling supply
The availability of seed12 to stock the rice fi elds is
in many areas a determining factor for the choice
of culture species. It is also a critical part of any
type of aquaculture development and is subject to
the same factors as seed production targeted for
pond and cage culture.
Hatchery and nursery technologies for most, if
not all, of the freshwater fi sh species that are
currently being cultured in rice-fi sh systems are
well established. However, getting the required
number of fi ngerlings of the desired species at a
particular time remains a problem in many
areas. This is especially acute in countries where
mass production and distribution are still
centralized in a government agency rather than
in the hands of private producers. The issue of
what is a suitable policy for the promotion of fi sh
seed for aquaculture development is wide
ranging and a thorough discussion is not possible
in this report. Suffi ce it to say that general
guidelines for the development of fi sh seed
supply for aquaculture in general also hold true
for rice-fi sh culture.
Some common problems associated with seed
production and distribution are seed quality,
genetics (broodstock quality), hatchery
management and administration, transportation
and stocking. It is best to involve as many people
as possible in decentralized production and
distribution of fi sh seed. Decentralization
11 Rice-clam (Hyriopsis cumingii) culture is practiced in Jiangsu Province, P.R. China. Farmers use rice fi elds as nursery for small pearl clams and
then the small freshwater clams are hanged in ponds, pools, reservoirs or lakes. A rice-fi sh-frog model was tested in Jiangxi Province in early 1984.
The experiment was conducted to control rice pests and diseases by frogs as well as fi sh. The farmed frogs included the black spotted frog Rana
nigromaculata, Rana plancyli, Rana tigrina rugulosa, Rana limnocharis, Microhyla butleri, and the toad Bufo bufo gargarizans stocked at rates of 4950/ha
and 9900/ha (Li Kangmin, pers. comm.)
12 This term includes fi nfi sh fry and fi ngerlings as well as crustacean equivalents, such as post-larvae (PL), zooea or megalop.
26 FAO and The WorldFish Center | Culture of Fish in Rice Fields
Table 3. List of fi sh and crustacean species recorded as having been or being farmed in rice fi elds.
Scientifi c Name Common Name(s) Countries Where Cultured
A. FINFISH
Family Anabantidae Anabas testudineus Climbing perch Malaysia, Thailand, Indonesia
Family Cichlidae Etroplus maculatus Orange chromide India
Etroplus suratensis Pearl spot/Green chromide India
Oreochromis mossambicus India, China, Taiwan,
Zimbabwe, Sri Lanka, Malaysia,
Thailand, Indonesia, Philippines
Oreochromis niloticus Nile tilapia Egypt, Korea, Philippines,
China, Bangladesh, Thailand,
Cote d’ Ivoire, Gabon, Tanzania
Paratilapia polleni Madagascar
S. hornorum x S. niloticus hybrid tilapia Brazil
Tilapia macrochir Cote d’ Ivoire
Tilapia melanopleura Pakistan
Tilapia rendalli Malawi
Tilapia zillii Egypt, Philippines
Family Cyprinidae Amblypharyngodon mola India
Aristichthys nobilis Bighead carp China, Thailand, Taiwan
Carassius auratus Goldfi sh China, Japan, Madagascar,
Vietnam, Indonesia, Italy
Catla catla Catla India, Bangladesh, Indonesia
Cirrhina mrigala Mrigal India, Bangladesh, Indonesia
Cirrhinus reba Reba carp Bangladesh
Ctenopharyngodon idella Grass carp China, Bangladesh
Cyprinus carpio Common carp China, India, Korea, Philippines,
Indonesia, United States,
Japan, Thailand, Vietnam,
Madagascar, Brazil, Italy,
Bangladesh, Hong Kong, Spain,
Taiwan, Hungary, Pakistan
Hypophthalmichthys molitrix Silver carp China, India, Korea, Philippines,
Indonesia, Bangladesh
Labeo bata Bangladesh
Labeo collaris Vietnam
Labeo rohita Rohu India, Bangladesh, Indonesia
Mylopharyngodon piceus Black carp China
Osteochilus hasseltii Indonesia
Puntius gonionotus Minnow/Tawes Vietnam, Thailand, Bangladesh,
India
Puntius javanicus (=Barbodes
gonionotus) Java carp/Silver barb Indonesia, China
Puntius pulchelus Minnow India
Puntius sophore Pool barb India
Puntius ticto Ticto barb India
Rasbora danoconius Slender rasbora India
Tinca tinca Tench Italy
Family Osphronemidae Osphronemus gouramy
Trichogaster pectoralis Snakeskin gourami Malaysia, Pakistan, Indonesia
Trichogaster sp. Thailand
Trichogaster trichopterus Malaysia
Family Helostomatidae Helostoma temmincki Indonesia, Malaysia continue >
27
Agronomic and Aquaculture Management
< continued
Scientifi c Name Common Name(s) Countries Where Cultured
Family Anguillidae Anguilla japonica Japan, Taiwan, India
Family Channidae Channa striata
(=Ophiocephalus striatus) Carnivorous snakehead Malaysia, Thailand, India,
Bangladesh
Channa gachua India
Channa punctatus India
Chanos chanos Philippines, Indonesia, India
Ophicephalus maculatus Vietnam, Taiwan
Ophicephalus striatus Snakehead India, Malaysia, Indonesia,
Philippines, Vietnam
Family Cobitidae Misgurnus anguillicaudatus Loach Japan, Korea, Philippines
Family Centropomidae Lates calcarifer Seabass, baramundi Australia, Thailand, Singapore,
Philippines, Malaysia,
Bangladesh, India, Myanmar,
India, Vietnam, Kampuchea,
Taiwan, China
Family Mugilidae L. parsia Gold-spot mullet India
L. tade Tade mullet India
Liza sp. India
Mugil cephalus Grey mullet India
Mugil corsula Mullet Bangladesh, India
Mugil dussumieri India
Mugil parsia India
Mugil tarde India
Rhinomugil corsula Corsula India
Family Clariidae Clarias batrachus India, Thailand, Indonesia,
Malaysia
Clarias gariepinus China
Clarias macrocephalus Omnivorous catfi sh Malaysia
Family Pangasiidae Pangasius hypophthalmus Sutchi catfi sh Cambodia
Family Ictaluridae Ictalarus lacustris Channel catfi sh United States
Ictalarus punctatus Channel catfi sh United States
Family Siluridae Parasilurus asotus Amur catfi sh Korea, Vietnam
Family Atherinidae Atherina bonariensis Kingfi sh Argentina
Family Curimatidae Prochilodus argentes Curimatá pacu Brazil
Leporinus elongatus Brazil
Prochilodus cearanesis Brazil
Family Pimelodidae
Other species:
Family Heterpneustidae Heteropneustes fossilis Stinging catfi sh India, Bangladesh
Family Pomacentridae C. dimidiatus Chocolatedip chromis India
C. ternatensis Ternate chromis India
Chromis caeruleus Green chromis India
Family Mastacemblidae Macrognathus aculeatus India
Mastacembelus armatus Tiretrack eel India
Mastacembelus panealus Barred spiny eel India
Family Aplocheilidae Aplocheilus panchax Blue panchax India
Family Nandidae Nandus nandus Gangetic leaffi sh India
Family Notopteridae Notopterus notopterus Bronze featherback India
continue >
28 FAO and The WorldFish Center | Culture of Fish in Rice Fields
< continued
Scientifi c Name Common Name(s) Countries Where Cultured
Family Ambassidae Ambassis nama Elongata glass-perchlet India
Ambassis ranga Indian glassy fi sh India
Family Gobiidae Glossogobius giuris Tank goby India
Pseudapocryptes lanceolatus Vietnam
Family Catostomidae Ictiobus cyprinellus Bigmouth buffalo United States
Family Centrarchidae Micropterus salmoides United States
Family Atherinidae O dontesthes bonariensis Silverside/Pejerrey Argentina
Family Polynemidae Polydactylus sexfi lis Sixfi nger threadfi n Bangladesh
Family Bagridae Mystus gulio Tengra/Long whiskers catfi sh India
Mystus sp. Bangladesh
Family Centrarchidae Lepomis sp. United States
Family Osphronemidae O sphronemus goramy Giant gourami Malaysia
Family Plecoglossidae Plecoglossus altivelis Ayu fi sh Japan
Other species: Beterotris niloticus Cote d’ Ivoire
B. CRUSTACEANS
Family Natantia Macrobrachium dayanum India
Macrobrachium lamarrei India
Macrobrachium mirabile India
Macrobrachium niponensis China
Macrobrachium rosenbergii Vietnam, Bangladesh, Brazil,
India, Indonesia, China
Macrobrachium rude India
Family Penaeidae Penaeus indicus India, Vietnam
Penaeus merguiensis India
Penaeus monodon India, Bangladesh
Penaeus semisulcatus India
Penaeus stylifera India
Family Metapenaeidae Matepenaeus brevicornis India
Metapenaeus ensis Vietnam
Metapenaeus lysianassa Vietnam
Metapenaeus tenuipes Vietnam
Metapenaues dobsonii India
Metapenaues monoceros India
Family Astacura Procambarus clarkii United States, Japan
Procambarus zonangulus United States
Family Brachyura Eriocheir sinensis River crab China
Other species: Palaemon styliferus India
Parapenaeopsis sculptilis India
Acetes sp. India
Note: Scientific names are listed as originally cited.
29
Agronomic and Aquaculture Management
overcomes many problems of distribution and
spreads the benefi ts of development more evenly.
Special consideration should be given to the
participation of women and disadvantaged
groups such as landless families.
A fi sh seed network is a group of people producing
and distributing fi sh seed in an informal but co-
ordinated manner. As seed production and
distribution develops, people involved in the
network adopt more specialized roles. These
networks are also important for information
exchange. Most government hatcheries experience
problems with seed distribution because they
operate outside these informal networks. To
maximize the opportunities for the poor, the
following are recommended: promote small
rather than large hatcheries; train people in the
skills required for a range of network activities
such as fry nursing, fi ngerling transportation, and
hapa manufacturing; and organize micro-credit
schemes to support people in fi sh seed networks.
6.5.3 Stocking pattern and density
Much like aquaculture using fi shponds, rice-fi sh
culture may involve the stocking of young fry for
the production of fi ngerlings (nursery operation)
or the growing of fi ngerlings into marketable fi sh
(growout operation). Rice-fi sh farming may
either be the culture of only one species
(monoculture) or a combination of two or more
species of fi sh and crustaceans (polyculture).
Thus the stocking density varies depending on
the type of culture as well as the number of
species used. A fi nal factor determining the
stocking is the type of modifi cations to rice fi elds
that has been made and what is considered the
fi sh culture area. The variation is so great that it is
diffi cult to provide even generalized guidelines,
but Table 4 gives some information from several
countries.
The stocking rate negatively affects the survival
rate of fi ngerlings (for example, grass carp) and
average body weight (ABW). At a density of
15 000 fi ngerlings·ha-1, the survival rate was 3%
higher than at 30 000 fi ngerlings·ha-1, while the
ABW was 11.4 g heavier than at 22 500
fi ngerlings·ha-1 and 20.6 g heavier than at 30 000
fi ngerlings·ha-1 (Yang et al. 1995).
Polyculture or stocking a combination of species
makes it possible to take advantage of all the
available food niches in the rice fi eld ecosystem,
aside from being able to manage a wider variety
of pests. For example, a combination of common
carp and grass carp has been found effective in
controlling insects, snails and weeds because of
the different feeding habits of the two species.
Table 4. Stocking densities for rearing fish in rice fields (Gupta et al. 1998; Li and Pan 1992; Sevilleja 1992;Quyen et al. 1992; Costa-Pierce 1992).
Stocking Density (fi sh/ha)
Concurrent Rotational
Monoculture
Oreochromis niloticus 3 156 to 5 000 10 000
Cyprinus carpio 3 000 to 3 400
Barbodes gonionotus 3 017
Polyculture
O. niloticus + C. carpio 3 000 + 2 000
3 070 total (6 000 to 10 000) + (4 500 to
5 000)
C. carpio + B. gonionotus 4 667 total
Multispecies (carp+barb+ tilapia) 9 323 total
C. carpio + C. auratus + C. idella (1 500 to 2 250) + (750 to 1 200) + (300 to 450)
O. niloticus + C. carpio + C. idella (6-10 cm: 6 000 to 9 000 or 3 cm: 12 000 to 18 000) +
(300 to 600) + (150 to 300)
B. gonionotus + M. rosenbergii 26 000 + (5 000 to 20 000)
Fingerling production
1-3 cm C. carpio (30 days) 70 000 – 100 000
3-5 cm C. carpio (50 days) 10 000 – 15 000
5-8 cm C. carpio (50 days0 6 000 – 10 000
5- 8 cm C. carpio (50-90 days) 1 500 –3 000
8-11 cm C. carpio (30 days) 1 000 – 2 000
30 FAO and The WorldFish Center | Culture of Fish in Rice Fields
Research indicates that although yield increases
with higher stocking density (positive
correlation), this should be compared with the
increased mortality and associated increase in
costs of stocking. A positive correlation has been
found between fi sh production and stocking
density (Gupta et al. 1998). At a mean stocking
density level of 3 825 fi ngerlings per ha during
the dry season and 2 948 per ha during the wet
season in Bangladesh, the average production
was 233 kg·ha-1 and 118 kg·ha-1 respectively. At
stocking densities of more than 6 000·ha-1
during the wet season the mean production
reached 571 kg·ha-1. On the other hand, a
negative correlation was found between the
stocking density and recover y rate such that a 1%
increase in the stocking density the survival rate
decreased by 0.14% with an insignifi cant
decrease in harvest size.
6.5.4 Fish nutrition and supplemental
feeding
Fish graze and feed on a wide range of plant and
animal organisms; preferences however vary
between species as well as with the stage of
development within species. For example, among
cyprinids the common carp has the widest range
in food and can feed on a variety of plant and
animal matter. Another important factor is the
presence and abundance of food organisms, for
example it has been shown that juveniles of the
rice-consuming aquatic snails P. canaliculata may
become a major food item of common carp in
rice fi elds (Halwart et al. 1998). Table 5 provides
an overview of the diets of different species of
tilapias (Bowen 1982).
The capability of O. mossambicus and T. zillii to
consume weeds even in a pond or rice fi eld
situation has also been reported (Hauser and
Lehman 1976), with T. zillii regarded as more
superior as a natural “weedicide”. Although listed
as a phytoplankton, feeder studies indicate that
the Nile tilapia may prefer certain categories of
algae such as fi lamentous cyanobacteria over
diatoms and green algae (Micha et al. 1996). The
species is not considered macrophytic but in a
culture situation the Nile tilapia is known to feed
on chopped terrestrial plants such as Napier grass
and aquatic plants including water spinach
Ipomoea aquatica as well as brans, cassava or
termites.
The rice fi eld ecosystem is rich in phytoplankton,
zooplankton, macrophyton, benthos, detritus
and bacteria. If the different types of natural food
organisms available in a rice fi eld ecosystem are
fully exploited by stocking a proper combination
of fi sh species, Li and Pan (1992) estimated that it
can support up to a maximum of over 500 kg·ha-1
of fi sh as shown in Table 6. This estimate of the
natural carrying capacity of a rice fi eld as an
aquaculture system is by no means a constant
fi gure, as it will vary from place to place and from
season to season. However, to produce more than
the natural carrying capacity or to ensure that
adequate nutriments are available at all times, it
may be necessary to apply supplemental feeds.
Farmers use fertilizers to increase the naturally
occurring food organisms in the rice fi eld and
supplements to feed the fi sh directly. The use of
supplemental feeds is necessary if a certain degree
of intensifi cation is desired since the natural food
in a rice fi eld is not suffi cient to support a higher
biomass of fi sh. Supplemental feeding functions
in much the same way in rice fi elds as it does in
fi shponds.
Diana et al. (1996) found that starting
supplemental feeding late had little effect on the
fi nal harvest and since fi sh culture in rice fi elds is
often limited in duration by the rice growing cycle
(120 days), this has two implications. First, if the
rice fi eld is used as a nursery for the growing of fry
to fi ngerlings, feeding may not be necessary for as
long as the fi eld is adequately fertilized. Second, if
older fi ngerlings are used to grow food fi sh,
feeding is essential from the very start.
Species Diet Reference
T. rendalli Macrophytes, attached periphyton Caulton (1976, 1977); Denny et al. (1978)
S. mossambicus Macrophytes, benthic algae,
phytoplankton, periphyton, zooplankton,
fi sh larvae, fi sh eggs, detritus
Bowen (1979, 1980); Man and Hodgkiss
(1977); Munro (1967); Naik (1973);
Weatherley and Cogger (1977)
S. aureus Phytoplankton, zooplankton Fish (1955); Spataru and Zorn (1976, 1978)
S. niloticus Phytoplankton Moriarty and Moriarty (1973)
T. zillii Macrophytes, benthic invertebrates Abdel-Malek (1972); Buddington (1979)
Table 5. Diets reported for adult tilapias in natural habitats (Bowen 1982).
31
Agronomic and Aquaculture Management
Supplemental feeds often consist of what is
available in the locality. Consequently rice bran is
a common supplemental feed in practically all
rice producing countries. In Bangladesh, wheat
bran and oil cake are used as well (Gupta et al.
1999) and in the Philippines, where coconut is an
important product, copra meal (Darvin 1992) is
employed. In China, feed may consist of wheat
bran, wheat fl our, oilseed cakes (rapeseed,
peanuts, soybeans, for instance), grasses and
green fodder (Wang and Zhang 1995; Li et al.
1995; Chen et al. 1998; He et al. 1998); and in
Malawi, maize bran and napier grass (Chikafumwa
1996), to name a few examples. Wang and Zhang
(1995) showed that the use of supplemental
feeding results in higher survival rate of 67% as
against 56.1% without supplemental feeding and
with a corresponding increase in unit yield of
337.5 kg·ha-1 and only 249 kg·ha-1, respectively.
Carp species Type of Food Potential Fish
Production (kg·ha-1) Utilization Rate (%) Food Conversion
Factor Potential Fish Production
Ave. Max.
Grass Aquatic Weeds 30 000-53 000 65 120 78 195
Silver Phytoplankton 9.3 70 40 30 59
Bighead Zooplankton 15 25 10 7.5 16
Common Benthos 4 25 45 118.2
Total 160.5 388.2
Add:
Detritus
and
bacteriaa48.2 117.2
Grand total 208.7 504.2
a Approximately 30% of total fish production
Table 6. Estimates of fish production from natural food in rice fields (Li and Pan 1992).
Formulated diets in mash, crumble or pellet form
are now increasingly used because of their greater
availability. Although more expensive than farm
by-products, they have the advantage of being
available at the volume required if needed and are
more convenient to store, handle and apply.
For more details on the types of supplemental
feed, the reader is directed to the extensive
literature on supplemental feeding in semi-
intensive pond aquaculture. In all cases of
supplemental feeding it should be remembered
that most feeds either incur a direct cost by having
to purchase the feed, or an opportunity cost in
that the input could be put to other uses (for
example fed to livestock) or sold. In addition,
when employing supplemental feeds, the water
quality may become an important issue as it can
deteriorate rather quickly if the fi eld is “overfed”.
32 FAO and The WorldFish Center | Culture of Fish in Rice Fields
7.1 Fish Yields
Similar to most aquaculture operations, the
amount of fi sh that can be harvested in rice-fi sh
farms varies greatly. The harvest of aquatic
animals from any rice fi eld is a function of several
factors such as: water depth and water supply,
presence of predators, species, stocking density,
whether monoculture or polyculture is practiced,
size of fi sh at stocking, and the rearing period.
Seasonal variations in natural productivity, and
whether fertilization and/or supplemental
feeding have been applied also affect fi sh
production.
Table 7 attempts to combine yields for several
systems in different countries, but these fi gures are
only indicative and great variations exist between
identical systems even within the same country.
The total production fi gures are only one aspect of
the issue. The production costs as well as the value
of the product are other important aspects.
7.1.1 Rice-fish
Fish production varies with stocking density, size
at stocking and whether or not supplementary
feeds were used. Without feeding the production
per crop can range from 100 to 750 kg·ha-1·yr-1
(Zhang 1995), with feeding the result might be
1 812 kg·ha-1·yr-1.
In the Indonesian minapadi system, the yield
varies from 75 to 100 kg·ha-1 and the fi sh weight
between 50-70 g. Where O. mossambicus is stocked
instead of carp, the fi rst stocking is made with
1 000 to 10 000 fry together with a few hundred
adults per hectare. Six weeks later the largest are
harvested for consumption and the rest restocked
for further growing (Khoo and Tan 1980).
In Basse Casamance, Senegal, rice–fi sh alternating
with fi sh only culture results in fi sh yields ranging
between 963-1 676 kg·ha-1 in ponds fertilized
with animal manure and fed farm by-products,
and 590 kg·ha-1 from the rice fi eld. A typical
harvest would consist of Sarotherodon melanotheron
(50%), O. guineensis (40%), Hemichromis fasciatus
(2%), Mugil (5%) and Penaeus notialis (3%). In
addition, fry and fi ngerlings may also be present
and may constitute from 5-8% of the harvest
(Diallo 1998).
7. Rice-Fish Production
Stocking of large fi ngerlings directly into rice
fi elds in Thailand yielded from 146-363 kg·ha-1,
while growing fry in a nursery pond before
transferring to rice fi elds ranged from 88-263
kg·ha-1. Rice yields were noted to have increased
on subsequent studies (Deomampo 1998).
In Iran, production averaged 1 580 kg·ha-1 with
feeding and 695 kg.ha-1 of fi sh without (172 days
culture period) and a rice yield of 7 014 kg·ha-1
(personal communication, Mr Ibrahim Maygoli ,
Shilat Aquaculture Division Head, Tehran, Iran,
30 August 1999).
7.1.2 Rice-fish-azolla
Fish yields using azolla vary widely. Liu (1995)
reported fi sh yield of 1 000 kg·ha-1 by stocking a
species-mix consisting of 100 H. molitrix and 300
C. carpio with 100 C. idellus and 7 500 O. niloticus.
This was attributed to the different species
complementing each other according to their
feeding habits and effi ciency. Both fi sh and rice
yield were found by Yang et al. (1995) to vary
with ridge width or ditch width. At constant ditch
width, fi sh production varied from 841 kg to
736 kg to 676 kg·ha-1 at 53 cm, 80 cm, and
106 cm ridge width respectively while rice yields
varied from 13-14 t. At constant ridge width of
53 cm, fi sh yields were 613 kg, 702 kg and 784 kg
respectively for ditch widths of 40 cm, 46 cm and
106 cm respectively while rice yields varied from
9.4 to 10.1 and 10.4 t.
Wang et al. (1995) reported that fi sh yields also
vary according to the species cultured and the
stage at which they were harvested. Output of fi sh
was highest in the rice-azolla-food fi sh at
536 kg·ha-1 followed by rice-azolla-C. carpio fry at
419 kg and rice-azolla-catfi sh (C. gariepinus) fry at
324 kg. The lowest fi sh yield was obtained with
C. gariepinus fry at 280 kg·ha-1. Wang also
obtained the highest yield with African catfi sh fry
grown in a rice fi eld without azolla at 717 kg·ha-1.
The highest fi sh yield was reported by Chen et al.
(1995) using a polyculture of H. molitrix, C. carpio
and crucian carp with 7 038 kg.ha-1 for rice-azolla-
fi sh as against only 4 119 kg·ha-1 for rice-fi sh
combination. The high yields were obtained by
using “fi ne feed” to feed pigs which produced
manure for the rice fi elds and “beer left-overs” as
supplementary feed.
33
Rice-Fish Production
7.1.3 Rice and crustacean
Crawfi sh yields from rice fi elds range from 1 120-
2 800 kg·ha-1 depending upon the length of the
harvest period (Dela Bretonne and Romaire
1990). River or mitten handed crabs yield 227-
303 kg·ha-1 button-crabs. The yield of marketable
crabs ranges from 303-454 kg·ha-1 at a stocking
rate of 75-150 kg·ha-1. Penaeid shrimp yield in
India ranges from 3 kg·ha-1 in deepwater rice
Fish Yield (kg·ha-1)
Bangladesh China India Indonesia Philippines Thailand Vietnam
Concurrent
Monoculture
High Range 188-239a223-263n
Low Range 125-156a2 000-3 100d143k43.7-59.7o48-79t
Polyculture
High Range 187-605b750-1 500e500-2 000h2 000-3 500l606-636p468-1 472r677u
187 prawn
+21 fi shv
Low Range 116 –396b150-300f500-700h78-303o87.7-363.3s
Rotational
Monoculture
Range 80-367m406-527q
Polyculture
Maximum >1500f
Range 300-450f815-2 135i
Concurrent-Deepwater
Polyculture
Range 1 320-3 211c300g3-1 100 j
Table 7. Unit production of fish in rice fields, various countries.
a) Gupta et al. (1998), ditch or sump, using C. carpio, B. gonionotus or O. niloticus. High range - boro (dry) season; low range - aman (wet) season.
b) Gupta et al. (1999), ditch or sump, using two (minimum figure) or more species (maximum figure). High range - boro (dry) season; low range -
aman (wet) season.
c) Gupta et al. (1999), excavated ponds with average depth of 0.5 m during dry season and minimum retention of 0.9 m for 7.93 months. Minimum figure is
that of adopters; maximum, that of research farmers raising fish up to 9 months.
d) Chen (1995), based on ridge-ditch system with Clarias leather, feed applied.
e) Xu (1995), based on ridge-ditch system with C. idella, C. carpio and H. molitrix.
f) Zhang (1995), unspecified species but can be assumed to be polyculture of different cyprinids as is the usual practice in China.
g) Wan et al. (1995), based on one experimental run only using C. carpio+C. carassius+Oreochromis sp.
h) Dehadrai (1992), high range - Khazan system (brackishwater) in Goa with shrimps+perches; low range – irrigated/rainfed with murrels+ catfish+carp.
i) Dehadrai (1992), brackishwater system with P. monodon+mullets.
j) Ghosh (1992) lower value represents production of natural stock of unspecified species and higher value on polyculture of Indian major carps+
Chinese carps+catfish.
k) Koesomadinata and Costa-Pierce (1992) minapadi system with C. carpio.
l) Koesomadinata and Costa-Pierce (1992), based on annual yield for sawah-tambak with stock of C. chanos+C. carpio+P. javanicus+M. rosenbergii or
P. monodon.
m) Yunus et al. (1992), the lower value represents penyelang crop and the higher value, palawija both using C. carpio.
n) Saturno (1994), wet season using pond refuge with O. niloticus for lower value; Israel et al. (1994) dry season using pond refuge with O. niloticus for
higher value.
o) Fermin et al. (1992), wet season crop with trench refuge, using C. carpio+O. niloticus.
p) Torres et al. (1992) dry season crop with trench refuge using O. niloticus.
q) Sevilleja (1992) based on single trial using fallow ricefield to raise C. carpio+O. niloticus.
r) Fedoruk and Leelapatra (1992) based on Thailand Dept. of Fisheries 1983 figures.
s) Thongpan et al. (1992) based on on-farm rice-fish farming research in Ubon, Northeast Thailand.
t) Mai et al. (1992). M. rosenbergii production in ricefield canals in Mekong Delta.
u) Cantho Univ. College of Agric. (1997), pond or canals connected to ricefield using three cyprinid species.
v) Mai et al. (1992), polyculture of M. rosenbergii and P. gonionotus.
34 FAO and The WorldFish Center | Culture of Fish in Rice Fields
plots relying on natural stock of mixed species to
over 2 135 kg·ha-1 in shallow brackishwater rice
fi elds stocked with P. monodon (Ghosh 1992).
7.1.4 Polyculture
Stocking multiple species or polyculture generally
results in higher yields than monoculture. The
high fi gures from the sawah-tambak of Indonesia
and the deepwater rice in Bangladesh are all
based on polyculture: C. chanos + C. carpio +
B. gonionotus + M. rosenbergii or P. monodon in the
case of Indonesia and six species of Indian and
Chinese carps in the case of Bangladesh. Higher
yields with polyculture of O. niloticus and/or
B. gonionotus with other carps than monoculture
of either species have also been reported by Gupta
and Rab (1994) in Bangladesh.
Gupta et al. (1998) found the combination of any
two species among C. carpio, B. gonionotus, and
O. niloticus resulted in lower yields than only one
of the species. When farmers added different carp
species such as H. molitrix, L. rohita, C. catla,
C. cirrhosus and C. idella, the production surpassed
monoculture (Table 8). The apparent difference
in the average production for all species is not
signifi cantly different. During the dry season 66%
of the farmers preferred C. carpio while during wet
season 54% preferred B. gonionotus.
In summary, it is diffi cult to either predict what
the yield will be in any particular area or advise
(without local trials) what stocking practice is the
best. Overall, there are indications that polyculture
gives better yields, but not any polyculture.
Likewise, although increased stocking density and
feed inputs increase yields (within certain limits),
this has to be compared with the associated
increase in costs. Usually local trials are needed to
assess which would be the best mix to provide the
farmer with the highest net profi t and least risk.
While the magnitude of fi sh harvest in a
concurrent rice-fi sh farming system may be
unspectacular compared to the harvest in an
intensive or even a semi-intensive pond
aquaculture, this is perhaps not the main point.
Rice is, after all, the main crop. What is more
important is that with some additional expense
and effort and without having to acquire more
land, a rice farmer can actually produce fi sh and
thus diversify the household’s options in terms of
food security as well as income generation. The
fact that the presence of the fi sh may actually help
boost rice production and may reduce, if not
completely eliminate, the need to use pesticides
and fertilizers can be seen as an added bonus.
7.2 Rice Yields
Much has been said about the mutualism of fi sh
and rice. Mutualism implies benefi cial effects on
each other. Rice acts as a nitrogen sink and helps
reduce the ammonia that may be released by the
fi sh and in so doing helps make the water cleaner
for the fi sh. Figure 13 shows the interrelationship
Table 8. Production, harvest size and recovery rate of fish at various stocking densities during boro (dry) and aman (wet) seasons in
Bangladesh 1992-94. Standard deviations are in parenthesis (Gupta et al. 1998).
Species No. of cases Stocking
Density per ha Average weight
at harvest (g) Recovery (%) Fish Production
(kg·ha-1)
Boro seasons (1993 & 1994)
C. carpio 96 3 400 (1 107) 115 (56) 53.8 (24.5) 204 (133)
B. gonionotus 13 3 017 (319) 95 (72) 65.0 (22.3) 188 (154)
O. niloticus 8 3 156 (442) 108 (25) 69.5 (12.1) 239 (75)
C. carpio + B. gonionotus 13 3 070 (324) 107 (42) 59.3 (15.4) 187 (64)
C. carpio + O. niloticus 1 4 667 86 39.6 158
B. gonionotus 2 3 643 (909) 25 (4) 50.5 (35.4) 47 (37)
Multispecies 12 9 323 (7 503) 241 (255) 49.1 (24.4) 605 (385)
All 145 3 825 (2 814) 121 (96) 55.6 (23.4) 233 (197)
Aman seasons (1992-1994)
C. carpio 4 4 090 (2314) 54 (19) 76.8 (13.4) 156 (77)
B. gonionotus 53 3 130 (603) 58 (29) 66.4 (15.6) 125 (90)
C. carpio + B. gonionotus 20 3 771 (1611) 53 (38) 61.7 (22.0) 116 (85)
Multispecies 21 6 778 (2834) 214 (146) 34.1 (20.7) 396 (256)
All 98 4 082 (2148) 90 (97) 59.0 (22.3) 184 (179)
35
Rice-Fish Production
between rice, fi sh and the environment in a rice-
fi eld ecosystem (Ni and Wang 1995). To a large
extent mutualism does exist. However, this does
not mean that the presence of rice necessarily
makes it possible to produce more fi sh. To the
contrary, the presence of rice hinders fi sh
production since the biological needs of the rice
and the fi sh are rather disparate. An example of
this was found by Rothuis et al. (1998b) in
Vietnam where the rice-seeding rate negatively
affected the fi sh yield. Dense stands of rice
suppressed phytoplankton growth as nutrient
availability was reduced, shading increased and
the access of fi sh into the rice fi eld restricted.
Without rice, the rice fi eld can be managed like a
fi shpond and higher fi sh yields may be expected.
It would seem a simple matter to obtain a
defi nitive answer to what happens to the rice
when fi sh are stocked considering the growing
body of literature on rice-fi sh farming.
Unfortunately it is not so simple. While many of
the papers available have specifi c fi gures on rice
yields of rice-fi sh farms, only a few have any
information on what the rice yields would have
been without the fi sh under the same
circumstances or what may be considered control
fi gures. Often the assertions are anecdotal. As
Lightfoot et al. (1992) pointed out, “many
authors have quoted farmers (or quoted other
authors who quoted farmers) to elevate to the
status of conventional wisdom the increase in rice
yield when fi sh are stocked.”
From the nearly 200 documents consulted, only
18 had control fi gures based on fi rst-hand data
that could serve as a basis for obtaining a clearer
picture on the effects of fi sh on rice yield. The 18
documents include two graduate school theses
and one annual report in addition to some
scientifi c papers presented in symposia,
Figure 13. Flow of energy in a rice field ecosystem (Ni and Wang 1995).
36 FAO and The WorldFish Center | Culture of Fish in Rice Fields
workshops or conferences covering fi ve countries.
The selection of paired data where both the rice-
fi sh culture and rice-only culture were done by
the same farmer is important in order to remove
the “skill factor”. As Waibel (1992) has pointed
out, it is possible that farmers who adopted rice-
fi sh farming are just better farmers.
It is well to start with the Philippines that has the
earliest comparative rice yield fi gures. In trials
using O. niloticus throughout the Philippines, on
average the rice yield was not signifi cantly lower
in rice-fi sh plots (NFAC 1980). More recent
studies have consistently shown higher rice yields
in rice-fi sh fi elds than in rice-only fi elds, between
14-48% (Table 9a). The same pattern of increased
rice yields in fi elds with fi sh has emerged from
Bangladesh (Gupta et al. 1998).
Studies in China follow the trends in the
Philippines and Bangladesh with some exceptions
(Table 9b). All provinces, apart from Jiangsu,
showed higher yields with fi sh than without them.
In West Bengal, India, fi eld trials in deepwater rice
testing the effect of supplementary feeding on the
fi sh stocked resulted in 4-11% higher rice
production in the rice-fi sh plots in both with and
without feeding. However, rice yields were slightly
lower (by 2-5%) in rice-fi sh using cow-dung (poor
in nitrogen and phosphorus) as fertilizer, but
higher (8-43%) using chicken manure rich in
nitrogen and phosphorus (Mukhopadhyay et al.
1992).During the dry seasons of 1993 and 1994,
an average of 82.4% of 34 farms practicing rice-
fi sh farming reported higher yields in fi elds with
fi sh. During the wet seasons of 1992 to 1994,
56.2% of 25 farms reported higher yields. Rice
yields in the fi elds with fi sh were, on average,
higher by 6.4% during the dry season in 1994 and
19.5% in 1993, and during the wet season, 12.7%
in 1992 and 9.8% in 1993 (Gupta et al. 1998) as
shown in Table 10.
In Indonesia, side-by-side trials consistently
showed higher rice yield (22-32%) in the rice-fi sh
plots compared to control plots without fi sh (Fagi
et al. 1992), regardless of season and whether the
plots were weeded or not, or whether herbicides
were used or not. Purba (1998) concluded in his
study in North Sumatra that although the rice-
fi sh system decreases the effective area for growing
rice, its impact on the total rice production of the
country is minor and can be ignored. In Thailand,
under all topographic conditions rice yields were,
on average, higher in rice fi elds stocked with fi sh
(Thongpan et al. 1992). In Vietnam, the yield was
lower, but statistically signifi cant. The rice yield
was observed in rice fi elds with B. gonionotus
(Rothuis et al. 1998c), but there was no control
without fi sh.
In order to obtain an overall perspective of the
situation, the frequency distribution of the
percentage increase in rice production was
determined when fi sh were present. Data from
the trials were averaged considering only one
variable, with or without fi sh. However, for trials
with treatments, for example use of different
fertilizers, the result of each treatment was entered
separately. Although this approach may not be
rigorous enough for the result to be considered as
defi nitive by some purists, by pooling the results
of the various workers from fi ve different countries
in Asia an overall picture of the impact of rice-fi sh
farming on rice is possible (Figure 14).
The analysis demonstrates that, although higher
rice yields were not always obtained with the
introduction of fi sh, the majority (80%) resulted
in higher yields of 2.5% or more. The results seem
convincing enough: growing fi sh in rice fi elds
does generally result in higher yields than growing
rice without fi sh.
These results indicate that although the area for
rice cultivation is decreased in rice-fi sh culture,
the mutualism with fi sh possibly together
increases inputs and/or better management and
more than compensates for this loss in area
through greater yield. The increase in yield in turn
seems to be due to the increased number of grains
per panicle13 (Table 11) and possibly in
combination with a decrease in the incidence of
whiteheads14 (Magulama 1990).
In summary, rice fi elds where fi sh are stocked will
likely have a higher yield because the rice fi eld
will have less weeds and less stemborers. Less
weeds to compete with the rice for soil nutrients
and less pests cannot but contribute to the
production of more and bigger grains, and a
reduced occurrence of unfi lled grains. In short,
rice fi elds with fi sh have healthier rice plants than
those without fi sh.
13 A panicle is defi ned as the terminal shoot of a rice plant that produces grains.
14 Whiteheads are empty panicles and are so called because of the appearance of the affl icted rice plants. They result mainly from stemborer attacks that
cause the lower portion of the rice stems to be cut. Drought and desiccation may also cause whiteheads.
37
Rice-Fish Production
Table 9a. Effect of fish on rice yield, paired results from various places 1977-94.
System/Location/Year Rice Yield (kg·ha-1)References
With fi sh W/out fi sh More (Less)
BANGLADESH
S/Da, Mymensingh/Jamalpur, dry
1993-94 4 980 4 555 425 Gupta et al.1998
S/Da, Mymensingh/Jamalpur, wet
1992-94 3 811 3 496 315 -ditto-
INDIA
Sumpb/no feed, Chinsura 1987 1 729 1 574 155 Mukhopadhyay et
al 1992
Sump/fed, Chinsura 1987 1 741 -ditto- 167 -ditto-
Sump/no feed, Gosaba 1987 2 122b2 039 83 -ditto-
Sump/fed, Gosaba 1987 2 130b-ditto 91 -ditto-
Sump/cdd, Sabang 1987 1 602 1 677 (75) -ditto-
Sump.cmd, Sabang 1987 2 399 -ditto- 722 -ditto-
Sump/cd, Girirchalk 1987 2 850 2 920 (70) -ditto-
Sump/cm, Girirchalk 1987 3 160 -ditto- 240 -ditto-
INDONESIA
Tr/0-we, Sukamandi, wet 1988-89 6 620 5 430 1 190 Fagi et al 1992
Tr/1-we, Sukamandi, wet 1988-89 7 130 6 700 430 -ditto-
Tr/2-we, Sukamandi, wet 1988-89 7 380 7 300 80 -ditto-
Tr/wcidee, Sukamandi, wet 1988-89 7 280 6 970 310 -ditto-
Tr/0-w, Sukamandi, dry 1989 4 220 3 430 790 -ditto-
Tr/1-w, Sukamandi, dry 1989 4 690 4 170 520 -ditto-
Tr/2-w, Sukamandi, dry 1989 5 570 5 280 290 -ditto-
Tr/w-cide, Sukamandi, dry 1989 4 970 4 560 410 -ditto-
Trench/TSPf, Sukamandi, dry 1989 7 994 6 060 1 934 -ditto-
PHILIPPINES
Trench, 11 regionsg 1977-78 5 739 5 939 (200) NFAC 1980
Trench, Cavite 1986-87 7 100h4 750 2 350 Fermin 1992
Trench, 20 x 20i, Laguna 1988 2 392 2 348 380 Magulama 1990
Trench, 40 x 10i, Laguna 1988 2 693 2 199 494 -ditto-
Trench, 30 x 10i, Laguna 1988 3 142 2 381 761 -ditto-
Trench, 20 x 15i, Laguna 1988 2 431 2 431 0 -ditto-
Trench, Nueva Ecija 1989 6 300 6 200 100 Torres et al. 1992
Pondj, Nueva Ecjia 1989 6 100 “ (100) -ditto-
Pondj, Nueva Ecija, wet season 1990k4 929 4 177 752 Israel et al. 1994
Pondj, Nueva Ecija, dry season 1991k6 098 4 294 1 804 Israel et al. 1994
THAILAND
ns, Dom Noi, wet 1985 1 890l1 790 100 Thongpan et al. 1992
ns, Khoo Khad, wet 1985 1 630l1 510 120 -ditto-
ns, Amnart Charoen 1987 2 537l2 014 523 -ditto-
ns, Kheuang Nai 1987 2574m2 372 202 -ditto-
ns, Det Udom 1987 2 651m2 427 224 -ditto-
Legend: TSP - triplesuperphosphate
a) Sump or ditch, involved 107 farms during 3 rainy seasons (aman) in 1992-94 and 149 farms for 2 dry seasons (boro) in 1993-93.
b) Central sump provided, deep water rice used.
c) Average of two plots.
d) Composted cow dung (cd) and dried chicken manure (cm) tested as fertilizers.
e) Trench, 0-w, 1-w, 2 w (0, 1 & 2 weeding respectively); w-cide (herbicide used).
38 FAO and The WorldFish Center | Culture of Fish in Rice Fields
f) 7 levels of TSP against 1 control, w/fish rice-yield figure is average of 7 levels .
g) Nationwide field testing in 13 pilot provinces, figures represent average of 15 field-test results.
h) Average of 1986 and 1987 runs.
i) Refers to the four rice-planting patterns tested.
j) Pond refuge within ricefield.
k) Average harvests from 15 farmers using pond refuge.
l) Average harvest of 12 farmer cooperators in Khoo Khad and 13 in Amnart Charoen.
m) Average of tests using 5 different rice varieties in Kehung Nai and 3 in Det Udom.
LEGEND: Tr-trench; RAF- Rice-azolla-fish; Rdg-ridge; WRdg-wide ridge; R/D –ridge/ditch; ns – not specified.
a) Average of two treatments.
b) X-trench 0.33 m wide x 0.4 m deep w/ sump (2.5x1x1m) at intersection.
System/Location/ Year Rice Yield (kg/ha) Reference
With Fish W/out Fish More (Less)
Tr, Hunan, early 1980-83 3 272 2 734 538 Nie et al. 1992
Tr, Hunan, median 1980-83 5 596 5 138 458 -ditto
Tr, Hunan, late 1980-83 8 595 6 218 2 377 -ditto
ns, Hubei 1983 7 774a6 375 1 398 Wu 1995
ns, Hubei 1984 7 569a6 573 996 -ditto
RAF, ns. 1985-86 7 096 6 493 603 Wang et al. 1995
ns, ns 1985-86 6 905f-ditto- 411 -ditto-
Tr w/sumpb, Jiangsu 1985 8 667 9 054 (387) Li et al. 1995
Tr w/ sump Jiangsu 1986 7 884 7 929 (45) -ditto-
Tr w/sump, Jiangsu 1987 7 998 7 996 (2) -ditto-
Rdg, Anhui 1987 7 125 6 150 975 Yan et al. 1995
WRdg, Anhui 1987 6 870 -ditto- 720 -ditto-
Bed, Anhui 1987 6 990 -ditto- 840 -ditto-
Conventional, Anhui 1987 6 795 -ditto- 645 -ditto-
R/D, Guilin, early 1987 7 632 6 135 1 497 Cai et al. 1995
R/D, Guilin, late 1987 6 750 6 225 525 -ditto-
R/D, Wuzhou, early 1987 11 654 11 037 617 -ditto-
R/D, Wushou, late 1987 6 606 6 206 400 -ditto-
R/D, Qinzhou 1987 5 537 4 857 680 -ditto-
Tr, Yunnan 1986 6 500 5 800 700 Chen 1995
Tr, Yunnan 1987 7 100 6 400 700 -ditto-
Tr, Yunnan 1988 7 000 6 500 500 -ditto-
ns, Hubei 1988 8 250 7 650 600 Lin et al. 1995
Table 9b. Effect of fish on rice yield, results from China, 1980-87.
Season Year No. of
Cases Rice yield (kg.ha-1)
Control plot Integrated plot Cases with higher yields
from integrated plots (%) Mean difference in yield
from control (%)
Boro
(Irrigated) 1993 10 3 957 4 651 70.0 +19 (-13.3 to +57.6)
1994 24 4 804 5 117 87.5 + 6.4 (-30.0 to +19.0)
All 34 4 555 4 980 82.4 +10.25 (-13.3 to 57.6)
Aman
(Rainfed) 1992 15 3 749 4 108 67.0 +12.7 (-21.3 to +55)
1993 10 3 121 3 364 40.0 + 9.9 (-30.6 to –66.7)
All 25 3 498 3 811 56.2 11.6 (-21.3 to 66.7)
Table 10. Rice yields from integrated and rice-fish plots and mono-cropped rice plots. Ranges are in parentheses (Gupta et al. 1998).
39
Rice-Fish Production
No.Grains per Panicle % Empty Grains 1000-Grains wt (g)
Rice-Fish Control Rice-Fish Control Rice-Fish Control
WU 1995
Early-1 94.0 87.0 8.4% 13.0% 24.8 24.8
Early-2 107.0 7.8% 24.8
Late –1 104.0 111.6 19.7% 21.4% 28.5 28.6
Late- 2 116.8 19.0% 28.7
YAN ET AL. 1995a
Ridge 107.9 105.0 18.6% 21.6% 30.2 29.0
Wide Ridge 115.6 19.7% 28.6
Bed 112.2 23.2% 30.0
Conventional 114.0 25.6% 29.1
LI ET AL. 1995
1985 153.3 152.2 10.9% 8.6% 29.1 29.8
1986 138.3 142.6 12.4% 12.1% 28.6 28.2
1987 152.5 152.7 17.4% 16.4% 28.8 28.9
CAI ET AL. 1995a
Guilin, early 126.0 117.0 13.6% 17.7% 28.3 27.8
Guilin, late 118.0 105.0 17.9% 21.6% 27.0 26.9
Wuzhou, early 124.3 118.4 11.0% 12.5% 25.6 24.8
Wuzhou, late 127.7 109.6 19.8% 21.2% 25.3 24.8
Qinzhou 125.4 121.1 17.0% 27.8% 26.6 25.3
No. Grains per m2
MAGULAMA 1990
20 x 20a30 535 26 121 26.1% 32.0% 25.4 25.0
40 x 10 37 954 28 352 23.5% 33.4% 25.1 24.5
30 x 10 44 175 31 642 23.2% 33.5% 25.1 24.9
20 x 15 37 107 34 546 24.8% 32.0% 24.8 24.8
a) Treatment consisted of planting patterns, numbers refers to rice plants.
SUMMARY:
Total number of data rows: 20
No. of grains/panicle:
Total instances higher in rice+fish plot: 17
Average percentage higher in rice +fish plot: 9.9%
% empty grains
Total instances lower in rice+fish plots: 15
Average percentage lower in rice+fish plots: 13%
1000-grain weight
Total instances higher in rice+fish plot: 13
Average percentage higher in rice+fish plots: 1.1%
Table 11. Comparative characteristics of rice grown with and without fish, the Philippines and China (data sources as indicated in table).
Figure 14. Frequency distribution of percentage increase in rice
yield as a result of raising fish in a rice field based on published data
from China, India, Indonesia, the Philippines and Thailand, 1977 to
1992 as summarized in Tables 9a and 9b.
40 FAO and The WorldFish Center | Culture of Fish in Rice Fields
8. Pest Management
8.1 Managing Pests with Fish
Present
Pest management includes many options falling
into four major categories: mechanical, chemical,
cultural and biological. The fi rst is the most widely
used and the one with the longest tradition,
together with natural control that is considered
part of biological control. Weeding is perhaps the
best example of this, but also includes cultural
techniques such as water level control. Chemical
pest management is relatively new and widespread,
particularly popular for its perceived effectiveness
and for the fact that it is not labor intensive.
Unfortunately, insecticide applications in rice
have been proven to become a major problem
because they destabilize the ecosystem and trigger
pest resurgence thus creating an even more critical
situation than without their use. Biological control
of pests has a range of applications from favoring
certain organisms that are predators of certain pests,
to use of disease resistant rice varieties. Particularly
when pesticide-related health impairments are
included, natural control is the most profi table
option for farmers (Rola and Pingali 1993). An
integrated approach using various management
options termed Integrated Pest Management or
IPM is the preferred choice for plant protection in
rice,15 and in fact has been adopted as the national
plant protection strategy by most rice-producing
countries.
Integrated pest management16 encompasses all
four management options outlined above and
attempts to optimize their use. The following
sections will examine the available options and
their established or potential impact on fi sh in
the rice fi eld. The main pest organisms to manage
are weeds, pathogens and invertebrates (mainly
snails and insects); although rats and crabs may
also cause a lot of damage in some areas.
One reason why farmers can no longer catch fi sh
in their rice fi elds like they used to, especially
if irrigation comes from river water, is the
increased use of pesticides. The use of chemicals
15 Except in organic farming practices.
16 IPM means “the careful consideration of all available pest control techniques and subsequent integration of appropriate measures that discourage the
development of pest populations and keep pesticides and other interventions to levels that are economically justifi ed and reduce or minimize risks to
human health and the environment. IPM emphasizes the growth of a healthy crop with the least possible disruption to agro-ecosystems and encourages
natural pest control mechanisms.” - FAO International Code of Conduct on the Distribution and Use of Pesticides.
is often cited as one of the major constraints
in the popularization of rice-fi sh farming
(Koesomadinata 1980; Cagauan and Arce 1992).
Yet stocking fi sh in rice fi elds actually reduces
pest infestation, and thus also reduces if not
eliminates the need for application of herbicides
and insecticides and particularly molluscicides
where snail predatory fi sh are cultivated (Waibel
1992; Cagauan 1995; Halwart 2001a, b, 2004a).
The practical and economic advantages of using
fi sh instead of chemicals are often obvious.
The effectiveness of fi sh as a bio-control agent
depends on how well they are distributed within
a rice fi eld. If fi sh stay mostly in the pond refuge
then they cannot be effective in controlling rice
pests. Halwart et al. (1996) found that in rice
fi elds provided with a 10% pond refuge, and
stocked with either C. carpio or O. niloticus, more
fi sh were present among the rice plants than in
the pond. Since feeding is a major impulse for
the diurnal activity of the fi sh, the distribution
pattern supports the hypothesis that fi sh are
potentially important in controlling pests.
Although farmers stocking fi sh tolerate a higher
level of pest infestation before spraying is
economically justifi ed (Waibel 1992), a high level
of pest infestation is always a possibility. In such
a situation, the use of pesticides as well as other
control methods should be considered based on
the potential costs and losses in terms of rice yield
and fi sh harvest. The important characteristics to
be considered in the selection of any pesticide to
be applied in a rice-fi sh farm can be summed up
as follows:
• relative safety to fi sh - should be tolerated
by fi sh at the recommended dosage effective
against the target insect species;
• rate of bio-accumulation - should not
accumulate or persist in rice and should be
metabolized into non-toxic compounds and
excreted by fi sh; and
• rate of degradation and persistence - should
either volatilize, bio-degrade or chemically
41
Pest Management
degrade shortly after its application, preferably
within a matter of days.
There are of course other factors such as safety
for humans and livestock and relative effi cacy are
also important considerations, which, at any rate,
apply whether or not fi sh are cultured with rice.
There are four major groups of pesticides used in
rice fi elds: herbicides, insecticides, fungicides and
molluscicides. Herbicides are considered the least
toxic and insecticides generally the most toxic to
humans. Current changes in rice culture including
high labour costs and increasing nitrogen
fertlilization appear to be resulting in increased
herbicide and fungicide use, respectively. Several
herbicides and fungicides are known to have high
non-target toxicities and therefore need to be
critically examined.
Rice-fi sh farmers tend to avoid pesticides,
mainly because the risk of killing the fi sh is high
particularly when pesticides with high fi sh toxicity
are applied. The use of non-toxic or low-toxic
compounds is viewed cautiously as well since even
though the consumption of contaminated fi sh is
not likely to cause immediate death or illness it
may result in residues and bio-accumulation of
these so-called “safe” pesticides.
In the aggregate, most countries today favor IPM
practices and particularly when fi sh are stocked
in rice fi elds the natural control option has been
shown to be the most profi table choice for farmers.
In cases where the use of pesticides may be the
only option, precautionary measures should be in
place to safeguard the fi sh17 and other non-target
organisms as well as the consumers’ health.
8.2 Management of Rice Field
Weeds
There are several practical options in controlling
weeds in rice fi elds: land preparation, water depth
variation, mechanical weeding, herbicide use and
stocking of herbivorous fi sh.
At a water depth of 15 cm or more, weed species
such as Echinochloa crusgalli stop growing and
most plants die (Arai 1963). Manna et al. (1969)
also reported how water depth negatively affected
the incidence of grass weeds and sedges in rice
17 In order to ensure the safety of the fi sh, most writers recommend that the fi sh be concentrated in the trenches, sumps or ponds prior to spraying and a
temporary embankment built to prevent the water from the rice fi eld getting into the fi sh refuge. Only when the toxicity of the pesticide has dissipated,
are the fi sh allowed to return to the rice fi eld.
fi elds. The fact that a rice fi eld stocked with fi sh
needs a certain water depth generally makes the
control of weeds easier.
Mechanical weeding is perhaps the most
frequently used way of controlling weeds, and
although stirring up the water and causing
turbidity may affect fi sh growth negatively, the
frequency is unlikely to signifi cantly impact on
the fi sh production. It is, however, a very labor
intensive way of controlling weeds and as such
often carries a high opportunity cost (particularly
in areas integrated in a cash economy).
Herbicides are used extensively, but are not
considered a serious problem in rice-fi sh farming.
If a herbicide is applied, it is normally done
immediately after transplanting. Fish are stocked
10 to 14 days after application (Torres et al. 1992).
Further, it is also possible to select a herbicide
which can be tolerated by fi sh even at relatively
high levels. Cagauan and Arce (1992) together
with Xiao (1992) listed nine types of herbicides
being used in rice culture in Asia.
Tests showed that C. carpio, M. rosenbergii, and
a freshwater clam (Corbicula manilensis) have
very high tolerance limits for 2,4-D or MCPA
(Chlorophenoxyacetic acids) (Cagauan and Arce
1992; Xiao 1992). 2,4-D’s toxicity to aquatic
organisms depends on the species of organism,
the formulation of the chemical, and the surface
water system parameters such as pH, temperature,
and water chemistry. 2,4-D is readily excreted in
the urine of animals and does not bio-accumulate.
However, some authors (for instance Beaumont
and Yost 1999) maintain that the 2,4-D type
of herbicides have been associated with cancer,
citing several writers to support their contention
that these types of chemicals are tumor promoters.
2,4-D is currently in a re-registration process with
the US EPA.
Introducing fi sh to the rice fi eld can reduce
the amount of weeds in several ways. To the
herbivorous species of fi sh, weeds are part of
their diet. To bottom feeding species, weeds just
happen to be in the way. In the process of looking
for food, the muddy bottom of a rice fi eld is tilled
giving little chance for the submerged weeds to
anchor their roots in the soil thus affecting their
growth and proliferation. In rice fi elds stocked
42 FAO and The WorldFish Center | Culture of Fish in Rice Fields
with B. gonionotus and C. carpio in Bangladesh,
farmers have observed that weeds are eaten
directly by the fi sh or are uprooted and die off
when the soil is disturbed by the browsing fi sh
- resulting in reduced weed infestation (Gupta et
al. 1998).
In China, fi sh have been found to be more
effective in weed control than either manual
weeding or use of herbicides. C. idellus was
the most effective species for this purpose and
especially effective for controlling 21 different
species of weeds, such as Echinochloa crusgalli,
Eleocharis yokoscensis, Cyperus difformis, Rotala
indica, Sagittaria pygmaea, Monochoria vaginalis,
and Marsilea quadrifolia. The introduction of fi sh
reduced the amount of weeds in one rice fi eld
from 101 kg to only 20 kg after fi ve weeks, while
in an adjacent rice fi eld with no stocked fi sh the
weed biomass increased from 44 kg to 273 kg
during the same period (Wu 1995).
C. carpio eat young roots, buds and underground
stems of weeds in the rice fi eld, although
ingestion may be incidental rather than deliberate
as they forage on benthic organisms. Only weeds
with their roots anchored to the soil (such as
Cyperaceae and Poaceae families) are foraged but
not free fl oating weeds (Satari 1962).
O. mossambicus and the Redbelly tilapia (T. zillii)
can also be used to control weeds. T. zillii is
especially effective (Hauser and Lehman 1976).
O. niloticus is not regarded as a weed feeder
and is more effective in consuming blue-green
algae (Anon. 1971 as cited by Moody 1992),
although Magulama (1990) found that it can also
contribute to the reduction of weeds. Two other
species found to be effective in weed control are
B. gonionotus and Trichogaster pectoralis (Khoo and
Tan 1980).
8.3 Management of Invertebrates
Insects and other invertebrate pests, primarily
snails and, in certain areas, crabs may cause
damage to the rice crop during particular growth
stages. The following section deals primarily with
the management of insect and snail pests.
The application of pesticides to reduce insects
and other invertebrates has several consequences
that are of importance to rice-fi sh culture, since
some of the pesticides directly affect the fi sh and
in other cases reduce the food organisms for the
cultured species.
8.3.1 Management of insect pests
Insect pests may be classifi ed into two general
types: those that affect rice production and those
that do not but are nevertheless considered
as pests because of public health reasons, for
instance mosquitoes. The effectiveness of fi sh
in controlling insect pests is infl uenced by
hydrological, biological and agricultural factors.
Fish have been shown to play a signifi cant role in
reducing some insect species populations in rice
fi elds. Their interaction with benefi cial organisms
is less clear. It should be noted that insect pest-
predator dynamics are usually well balanced in a
rice ecosystem that is not disrupted by the use of
insecticides. Halwart (1994a) concludes that the
presence of fi sh in fl ooded rice further reinforces
the stability and balance of pest-predator
interactions in the ecosystem.
In Bangladesh, the population of useful insects
such as lady beetle, spider and damsel fl y, was
5-48% higher in rice-fi sh farms compared to
rice-only farms 10-12 weeks after transplantation,
but later on the converse was observed. However,
pest infestation was 40-167% higher in rice-only
farms during all stages of rice growth (Gupta et
al. 1998).
Mosquitoes and midges pass part of their life-cycle
in the water and while not considered harmful
to rice plants, they are still considered as pests.
Some early work on stocking fi sh in rice fi eld was
mainly aimed at controlling mosquitoes rather
than producing food fi sh with the exception
of China where combined raising of Gambusia
and common carp resulted in the reduction of
anopheline and culicine larval populations by 90
and 70%, respectively (WHO 1980 in Pao 1981).
The rice planthoppers and leafhoppers usually
rest on the middle or lower parts of the rice plants
to suck plant juices during the day and climb to
the upper part of the rice plant to feed at night
or in the early morning. C. carpio and C. idellus
over 6.6 cm in length were found to be effective
in reducing planthoppers and leafhoppers,
respectively (Xiao 1992). C. idellus are the most
effective fi sh against the hoppers followed by
C. carpio and O. niloticus (Figure 15). Yu et al.
(1995) suggest that C. idellus are effective because
of consuming the outer leaves of the rice plants
where the planthoppers oviposit their eggs. In
addition, the fi sh also consume planthoppers
that fall down in the water. So as not to depend
purely on chance, Xiao (1995) recommends that
“a rope be pulled over the rice plants” in order to
43
Pest Management
drive the planthoppers down to the water surface
where they are accessible to the fi sh. In Vietnam,
a rice-fi sh farm recorded 3 800 hoppers·m-2 as
against hundreds of thousands of hoppers·m-2 in
surrounding infested areas (Tuan 1994).
Yu et al. (1995) report that observations in China
indicate 47-51% less stemborers in rice-fi sh fi elds
compared to rice-only fi elds. They also found a
reduction of between 28-44% in the attack rate
compared to rice-only fi elds. Magulama (1990)
observed that whitehead incidence, a clear sign
of stemborer infestation, in experimental plots
in the Philippines was 11% in rice-fi sh fi elds
and 18% in rice-only plots (Figure 16). Halwart
(1994a) observed low stemborer infestation
levels in both rice-only and rice-fi sh treatments in
three consecutive seasons. In the fourth season,
however, he noted a statistically signifi cant 3%
reduction in yellow stemborer (Scirpophaga
incertulas) infestation as whiteheads in rice fi elds
with O. niloticus and 5% lower with C. carpio
compared to control fi elds without any fi sh where
an 18% infestation was prevalent. The control
mechanism is likely to be predation by fi sh on the
neonate stemborer larvae which, after hatching,
often suspend themselves from the rice leaves
with a silken thread to disperse to other hills.
Conversely, the number of leaffolders
(Cnaphalocrocia medinalis), sometimes also called
leaf rollers, was actually higher in rice-fi sh fi elds
than in rice-only fi elds in China. Rice-fi sh fi elds
had 90 to 234 leaffolders per 100 hills as against
12 to 149 in rice-only fi elds. Fish apparently do
not eat the leaffolder larvae while the presence
of fi shwaste and deep water may have favored
oviposition, hatching and feeding of the insect
larvae. However, Hendarsih et al. (1994) noted
Figure 15. Effect of different species of fish on rice planthopper nymphs in rice+fish farms. NW -- normal water depth, DW—Water kept at 10
cm, None – No fish, GCarp – Grass Carp, CCarp – Common Carp, NileT- Nile Tilapia, Mixed – All 3 Species. Shangyu County, Zhejiang Province,
China (data source: Yu et al. 1995).
Figure 16. Incidence of whiteheads on rice plants in fields stocked
with Nile tilapia and in fields without fish (data source: Magulama
1990).
44 FAO and The WorldFish Center | Culture of Fish in Rice Fields
that damage to rice due to leaffolders was 50%
lower for Indonesian rice-fi sh farmers, although
this was not found to be statistically signifi cant.
Chemical insecticides are generally more toxic
than herbicides and may have to be applied even
while the fi sh is still growing in the rice fi eld.
Xiao (1992) maintains that pesticides are not
incompatible with rice-fi sh culture and that these
can be applied safely provided the following
points are followed:
• a suitable type is selected;
• a safe dosage is used;
• proper delivery methods are used;
• application period is properly timed; and
• pre-application preparations are undertaken
to protect the fi sh.
There has been no systematic evaluation of
the different insecticides as to their toxicity to
different species of fi sh as well as to their rate of
bio-accumulation in fi sh. What is available are a
number of tests on the more prevalent insecticides
in various places as reviewed by Cagauan and Arce
(1992) and Xiao (1992) (Tables 12, 13).
It is important to note here that besides the
statistical signifi cance also the economic
signifi cance of the data should be considered and
that, with or without the presence of fi sh, “there
are no good data to support any use of insecticides
in tropical irrigated rice” (Settle, pers. comm.).
8.3.2 Management of snails
One of the latest pests to hit the rice fi eld in
Southeast Asia is the golden apple snail, Pomacea
canaliculata. This snail, which is of Latin American
origin, has invaded most of the rice production
areas in Asia (Halwart 1994b). Two species
were imported from Florida, USA, in 1980 as a
potential food and export crop in the Philippines
with a second batch imported from Taiwan in
1984 by two separate private groups (Edra 1991).
Seemingly harmless when fi rst introduced, they
are now known to be capable of completely
devastating rice fi elds with newly emerging rice
plants.
The use of fi sh as a biological control for snails
has been recognized for some time. The review
of Coche (1967) lists work done in Uganda,
Mozambique and the Congo as early as 1952
to 1957. Then the concern was to control snails
that serve as intermediate hosts to Schistosoma
spp., a trematode that causes schistosomiasis - a
debilitating disease in humans that is also known
as bilharzia.
To control apple snails, most farmers and
government agricultural agencies used chemical
molluscicides, mainly organo-tin compounds.
Increasing awareness of the hazards posed by
organo-tins on humans and livestock led to
banning of these in some countries. In the
Philippines, the agricultural chemical companies
have shifted to metaldehydes after their approval
by relevant authorities. Farmers do not fi nd the
metaldehydes to be as effective since they are
applied in bait form and have to be ingested by
the target snails to cause any damage.
Fish are a far better, biological control option.
In the Philippines, a three-year program started
in 1990 as part of the strategic research in the
Asian Rice Farming Systems Network (ARFSN)
specifi cally evaluating the potential of O. niloticus
and C. carpio under laboratory and fi eld (both on-
station and on-farm) conditions (Halwart 1994a).
Experiments on the feeding response and size-
specifi c predation in a controlled environment
suggested that common carp is the preferred
biocontrol agent capable of daily consumption
rates of up to 1 000 juvenile snails, also feeding on
larger snails (Figure 17, Halwart et al. 1998). These
results in combination with new data on the snail
population ecology resulted in fi eld experiments
testing combinations of different snail and fi sh
densities (Figure 18, Halwart 1994a). Results were
then further tested for their long-term probability
and robustness by developing a snail population
dynamics model that identifi ed fi sh in rice as one
of the key determining snail mortality factors
(Heidenreich and Halwart 1997; Heidenreich et
Figure 17. Number of juvenile Pomacea canaliculata snails (less than
5 days old) consumed per 24 hours by single fish (Cyprinus carpio and
Oreochromis niloticus) as the initial snail density is varied (Source:
Halwart 1994a).
45
Pest Management
al. 1997). In Indonesia, a preliminary screening
pointed at four species with potential for snail
control: C. carpio, O. niloticus, B. gonionotus,
and O. mossambicus (Hendarsih et al. 1994).
Among these, C. carpio was identifi ed as the best
candidate and found to be capable of consuming
up to 40 young snails in one day, with the other
three species consuming only 84-87% of that
number within four days. The fi ndings have
been applied in Vietnam where IPM has been
identifi ed by FAO as the most suitable approach
for snail control with carp being the preferred fi sh
species for biological control (FAO 1998).
8.4 Management of Diseases
The role of fi sh in a rice fi eld is not limited to
controlling the proliferation of weeds, snails, and
some insect pests. In China, the Taoyuan County
Agricultural Bureau in Hunan province has found
that raising C. idellus in rice fi elds controlled rice
sheath blight disease (Xiao 1992). The disease
Pesticide group/
common name
48-hour LC50 (ppm of formulated product)
and toxicity ranka48-hour LC50
(ppm of formulated product)
O. niloticus O. mossambicus C. carassius O. niloticus O. mossambicus C. ca rassius
INSECTICIDES
Carbamate
BMPC 5.6 –
6.7 ht - 28.3 mt 5.4-6.12 - 25.1
Carbaryl 3.10 ht - - - 2.93 - -
Carbofuran 2.27 ht 2.4 ht - - 1.97 1.72 -
MTMC 68.0 mt 52.0 mt - - 50.0 46.9 -
MTMC + Phenthoate 9.56 et - - - - 0.47 - -
PMC 6.05 6.0b- 34.75 - - - -
PMP 59.0 mt - - 3.8 mt 47.1 - 19.6
Organophosphate
Azinphos ethyl 0.028b0.023b0.009 - - 0.002
Chlorpyrites 2.0 ht 1.34 - ht 1.3 1.19 -
Diazinon 45.0 mt - 40.7 2.2 - 15.2
Methyl parathion 25.7 mt - 13.4 19.0 - 11.0
Monocrotophos 1.2 ht 47.6 0.31 ht - 33.10 -
Triazophos 5.6 ht - - - -
Organochlorine
Endosulfan 5.8 ht - 1.3 1.3 - 1.6
Synthetic pyrethroid
Permethrin 0.75 et 1.3 ht - -0.75 - -
Cypermethrin 0.63 et - - 0.63 - -
HERBICIDES
2-4-D
Agroxone (MCPA)
Rilof (piperophos) 27.5 mt
Machete (Butachlor) 1,4 ht 1.3
Modown (bifenox) 149.0 lt 102.0 lt 127.0 102.0
EPTAM D (EPTC) 71.5 mt 49.5 mt 54.4 49.5
Trefl an (trifuralin) 308 lt 170.0 lt 225.0 170.0
a Ranking of pesticides from Koesomadinata and Djadjaredja (1976) for 48-hour LC50:< 1 = extremely toxic (et); 1 – 10 = highly toxic (ht); 10 to 100 =
moderately toxic (mt); and >100 = low toxic
b 24-hour LC50
Table 12. Toxicity of different insecticides and herbicides expressed as 48- and 96-hour LC50 to O. niloticus, O. mossambicus, and C. carassius
tested at the Freshwater Aquaculture Center – Central Luzon State University, Philippines (abridged from Cagauan and Arce 1992).
46 FAO and The WorldFish Center | Culture of Fish in Rice Fields
Table 13. Median tolerance limits ( TLM) of common carp (Cyprinus carpio) to various pesticides (abridged from Xiao 1992).
Formulated Product TLM (ppm) 48-hours Toxicity grade
INSECTICIDE Trichlorfon 6.2 medium
Dichlorvos 4.0 medium
Fenitrothion 4.4 medium
Malathion 9.0 medium
Rogor <40.0 low
Methyl Parathion 5.0 medium
Phosmet 5.3 medium
Phenthoate 2.0 medium
Baytex 2.0 medium
Tsumacide 15.3 low
Landrin 38.1 low
Bassa 12.6 low
Etrofolan 4.2 medium
Chlordimeform 15.2 low
Rotenone 0.032 high
Bramaxymil octamoate 0.0 high
BACTERICIDE EBP 5.0 medium
IBP 5.1 medium
Edinphensop 1.3 medium
Oryzon 6.7 medium
Plictran 14.6 low
Thiophanate methyl 11.0 low
Blasticidin >40.0 low
Kasugamycin 100.0 low
CAMA 10.0 medium
Phenazine >10.0 low
Triram 4.0 medium
HERBICIDE 2,4-D >40 low
DMNP 14.0 low
Propanil 0.4 high
Nitrofen 2.1 medium
Benthiocarb 3.6 medium
Amine methanearsonates 3.7 medium
GS 13633 0.86 high
Hedazhuang 34.0 low
Oradiazon 3.2 medium
Prometryne 23.5 low
Glyphosate 119.0 low
Pentachlorophenol 0.35 high
OTHERS Zinc Phosphide 80.0 low
Propargit 1.0 medium
Lime 140.0 low
47
Pest Management
incidence index in rice+fi sh plots ranged from
8.5-34.2 in early rice and 2.4-26.4 in late rice as
against 24.1-55.0 and 4.7-41.7 in the controls,
respectively (Figure 19). Similar results were
observed in Shangyu County, Zhejiang Province
(Yu et al. 1995) where disease incidence was
lower by 9.9-19.6% in normal depth rice+fi sh
plots.
Yu et al. (1995) offered three mechanisms that
enable fi sh to mitigate the effects of fungal
infection. First, the fi sh stripped the diseased
leaves near the bottom of the rice plants that
therefore diminished the sources of re-infection
in the fi eld. Second, after the bottom leaves of
the plants were stripped, improved ventilation
and light penetration made the microclimate
unfavorable to the fungus. Third, long-term,
deepwater conditions prevented any germination
of spores and re-infection.
Xiao (1992) reports that C. idellus feed directly on
the sclerotia (compact masses of fungal hyphae
with or without host tissue) of the sheath blight
and digest them after 24 hours. Secretions from
the fi sh also appear to slow down the germination
of hyphae and reduce infection. However, the
fi sh are effective only when the infection occurs
at the water surface. Once the infection spreads
upward, away from the water surface, the fi sh are
ineffective.
Figure 18. Abundance of live Pomacea canaliculata snails collected
two days after rice harvest in 50 m² plots with pond during the wet
season (A) and 200m² plots with pond during the dry season (B) at
low (0.18 snails·m²), medium (0.48 snails·m²and high (1.32 snails/
m²) initial snail infestation levels, Muñoz, Nueva Ecija, Philippines.
CC = Cyprinus carpio, ON = Oreochromis niloticus, low = 5000
fi sh·ha-1, high = 10 000 fi sh·ha-1. Bars are means of 3 replications.
Means within the same snail infestation (low, medium, high) with
a common letter are not signifi cantly different at the 5% level by
DMRT (Source: Halwart et al. 1998).
Figure 19. Incidence index of rice sheath blight disease in rice grown
with fi sh and without fi sh, Tau Yuan Agricultural Bureau, Tao Yuan,
China (data Source: Xiao 1992).
48 FAO and The WorldFish Center | Culture of Fish in Rice Fields
9. Impact of Rice-Fish Culture
It is the impact of rice-fi sh that ultimately should
determine whether this is a worthwhile endeavor
for rice farmers. The impact of rice-fi sh culture can
be measured in many ways, but this section will
focus on the direct economic impact followed by
its impact on household nutrition, public health
and its role in poverty alleviation. Environmental
issues then follow.
9.1 Economics of Production
9.1.2 The “bottom line”
In order to assess whether raising fi sh in the rice
fi eld is really worth the extra effort, available
comparative cost and returns fi gures for rice-fi sh
and rice-only farming were examined. Specifi cally
considered are only those cases where both
fi gures were obtained within the same locality
during the same period of time. Many of the
papers available do have some cost and returns
fi gures for the rice-fi sh operation, but usually lack
the fi gures for rice only. These are not included in
this analysis. As can be seen in Tables 14 and 15,
the percentage differences in the net returns vary
widely from one country to another and from one
year and one place to another within the same
country (Yu et al. 1995). However, by and large,
the presence of fi sh had the effect of increasing
the net returns.
In Bangladesh the net returns from rice-fi sh
was over 50% greater than that from rice
monoculture. The higher net returns were
probably due to the lower mean costs of rice
cultivation and higher rice yields in addition to
the fi sh yield from integrated farms (Gupta et al.
1998). In China, the increase varied from 45 to
270%. Growing fi sh was almost three times more
profi table than rice alone (Yan et al. 1995a). Lin
et al. (1995) related the economic benefi ts of
rice-fi sh farming to an increase in rice yields and
savings in labor and material inputs. Rice yields
in rice–fi sh culture were 8% higher, labor input
19% lower, and material costs were 7% lower
(savings in the cost of controlling diseases and
pests). Additionally, fi sh production increased
the net income.
18 Thongpan et al. (1992) noted that during the dry season of 1985, rice-fi sh culture had higher returns than rice monoculture, which unfortunately was
not presented in detail in the paper. Subsequently, two other farms showed higher profi tability in the rice-fi sh culture during the rainy season of 1985.
Indonesian fi gures show that having two crops
of rice-fi sh and using the rice fi eld for a short
intermediate crop or penyelang of fi sh has a
116% higher return than having two crops of
rice and leaving the rice fi eld fallow for two
months or so. Purba (1998) concluded that
the rice-fi sh system is a profi table technology
and that its adoption is likely to increase farm
household income, labor absorption, and better
liquidity.
In the Philippines, rice-fi sh farms yielded a 27%
higher net return with fi sh compared to a single
crop of rice (Sevilleja 1992). In addition, it has
been demonstrated that it is possible to achieve a
three-fold increase in profi tability of rice farming
by culturing fi sh as well as rice (Fermin 1992;
Israel et al. 1994).
Thailand, in contrast to previously mentioned
countries, showed lower net returns in the rice-
fi sh fi elds than in the rice-only fi elds. The Thai
fi gures indicate that profi tability in the rice-fi sh
fi elds was only 80% that of rice monoculture.
Thongpan et al. (1992) attributed this to the
high initial investment in rice-fi sh culture.18
A survey of 76 farms in the Mekong Delta of
Vietnam (Rothuis et al. 1998a) showed a 16%
lower rice yield and a 20% lower overall net
return in farms that allocated part of their area to
rice-fi sh culture. Mai et al. (1992) reported that
from three farms in the Mekong Delta, the net
returns from the rice fi elds with unfed shrimps
was 52% higher than that of rice monoculture
and 176% higher in the rice fi elds where shrimps
were fed with rice bran and decomposing
animals.
9.1.2 Input analysis
An analysis of what inputs are needed is of
importance considering that high input costs
will exclude the poorer sections of rural areas.
Detailed cost and returns of rice monoculture
with the rice-fi sh system are available for
Bangladesh, Indonesia, the Philippines and
Vietnam.
49
Impact of Rice-Fish Culture
Except for Indonesia, all the other cases
consistently showed an increase in the overall
labor requirement when fi sh are raised in the rice
fi eld, with the amount of increase varying from
only 10% to as high as 234%. This was mainly
due to the need to prepare the rice fi eld for fi sh
stocking as well as for fi sh harvesting. However,
in some specifi c activities connected with the rice
crop such as fertilizing, weeding and pesticide
applications, the presence of fi sh actually lessened
the labor required. Again the amount varies from
activity to activity and from one area to another as
shown in Table 16.
In terms of fertilizer expense Bangladesh,
Indonesia and the Philippines showed from 4%
to 14% lower fertilizer costs in rice-fi sh fi elds,
while Vietnamese fi gures indicate a 96% increase.
The same countries showed signifi cantly lower
costs of chemical pesticides in rice-fi sh farms (44-
86%). However, in Vietnam pesticide applications
were higher in homesteads practicing rice-fi sh
farm.
9.2 Benefits to Communities
9.2.1 Improved income status of
farmers
The immediate benefi ciaries of the production
of fi sh and often improved rice yield in rice-
fi sh farming are the farmers who adopt the
technology. Although it seems obvious, Ruddle
and Prein (1998) have pointedly stated, “the
existence of such a relationship has not been
demonstrated unequivocally.” However, the fact
that many farmers in different countries continue
to practice it year after year, even without any
government program, would seem to be proof
enough of the benefi ts derived from this type of
rice farming.
Models developed using linear programming
techniques on a 2.3 ha farm in Guimba, Nueva
Ecija, Philippines, show that the adoption of
rice-fi sh farming technology can generate an
additional 23% more farm income by raising
Rice+Fish System, Year, Period, (Source) Rice+Fish Rice Only % More or (Less)
Amount Total Amount Total
BANGLADESH
Ditch/Sump, boro (dry) 1994, (Gupta et al. 1998)a
Rice Income
Fish Income
Rice Expenses
Fish Expenses
749
195
(302)
(72)
690
(326)
8.5%
(7.4%)
Net Returns 570 364 56.6%
Ditch/Sump, aman (wet) 1993, (Gupta et al. 1998)a
Rice Income
Fish Income
Rice Expenses
Fish Expenses
464
183
(121)
(31)
444
(137)
4.5%
(11.6%)
Net Returns 495 307 61.2%
CHINA
WRDG Grow-out 1987, one crop (Yan et al. 1995)b
Rice Income
Fish Income
Rice Expenses
Fish Expenses
559
864
(131)
(202)
562
(158)
(0.9%)
(17.1%)
Net Returns 1 090 404 169.8%
Unsp. Grow-out 1988, one crop, (Lin et al. 1995)
Net Returnsb588 405 45.2%
Table 14. Summary of cost and returns from rice+fish and rice-only culture, Bangladesh and China. All figures in USD·ha-1·crop-1 or USD·ha-1·yr-1
as indicated and are rounded to the nearest unit. The last column compares rice+fish against rice only farming in terms of income from rice
only, expenses incurred for rice and the net returns.
LEGEND: WRDG –Wide Ridge
a) Original figures in Bangladesh Taka (BDT), converted to USD at the 1994 rate of USD1.00=BDT39.00. Gross rice income not given but was derived using net
benefit from rice and rice expenses.
b) Original figures in Chinese Yuan (CNY), converted to USD at the 1987-88 rate of USD1.00=CNY3.72.
50 FAO and The WorldFish Center | Culture of Fish in Rice Fields
LEGEND: BW/DWR –Brackishwater Deep Water Rice
a) Extrapolated to 1 ha from weighted average of 6 farms of 0.35-1.0 ha for rice-rice-fallow and 0.5 -1.5 ha for minapadi-minapadi-fi sh.
b) Original fi gures in Philippine Peso (PHP), converted to USD at 1991 rate of USD1.00= PHP27.48.
c) Original fi gures in Vietnam Dong (VND), converted to USD1.00=VND11 000 as given by authors .
d) Even farmers not adopting rice-fi sh farming maintained a small fi shpond accounting for the fi sh.
Rice+Fish System, Year, Period, (Source) Rice+Fish Rice Only % More or (Less)
Amount Total Amount Total
INDONESIA
Minapadi-Minapadi-Fish vs Rice-Rice-Fallow 1988,
one year, (Yunus et al. 1992)a
Rice Income
Fish Income
Rice Expenses
Fish Expenses
1 518
490
(621)
(122)
1663
770
(8.7%)
(19.4%)
Net Returns 1244 576 116.0%
PHILIPPINES
Trench 1986, one crop, (Sevilleja 1992)
Rice Income
Fish Income
Total Expenses
674
126
(506)
700
-
(469)
(3.7%)
7.9%
Net Returns 294 231 27.3%
Trench 1986, one crop, (Sevilleja 1992)
Rice Income
Fish Income (incl. own consumption)
Rice Expenses
Fish Expenses
1098
607
(322)
(242)
757
(390)
45.0%
(17.4%)
Net Returns 1141 367 210.9%
Pond Refuge 1991-92, one year, (Israel et al. 1994)b
Rice Income
Fish Income (incl. own consumption)
Total Expenses
2077
1126
(1860)
1579
(1143)
31.5%
62.7%
Net Returns 1343 436 208.0%
THAILAND
Unspec. 1984-85, one year, (Thongpan et al. 1992)
Net Returns 121 160 (24.4%)
VIETNAM
BW/DWR. 1988, one year, (Mai et al. 1992)
Net Returns from Rice Monoculture
Net Returns from Rice and Shrimps: fed
Net Returns from Rice and Shrimps: not fed 105
58
38 176.3%
(34.9%)
Ricefi eld w/homestead, pond and dike (Rothuis et
al. 1998)c
Rice Income
Fish Income
Income from homestead and dike
Rice Variable Expenses
Fish Variable Expenses
Homestead/dike variable expenses
Total farm fi xed cost
888
89
175
(544)
(66)
(98)
(176)
1060
6d
119
(600)
(3)
(91)
(157)
(16.2%)
1383.3%
47.1%
(9.3%)
2100.0%
7.7%
12.1%
Net Returns 268 334 (19.8)
Table 15. Summary of cost and returns from rice+fish and rice-only culture, selected Southeast Asian countries. All figures in USD·ha-1·crop-1
or USD·ha-1·yr-1 as indicated and are rounded to the nearest unit.
51
Impact of Rice-Fish Culture
Bangladesh 1994a
(Gupta et al. 1998) Indonesia 1988
(Yunus et al. 1992) Philippines,1991-92,
(based on Israel et al 1994)cVietnam 1994-95e
(Rothuis et al. 1998)
Rice +
Fish Rice
Only
%
more
(less)
Rice +
Fish Rice
Only
%
more
(less)
Rice +
Fish Rice
Only
%
more
(less)
Rice +
Fish Rice
Only
%
more
(less)
GROSS RETURNS 943.56 689.77 36.8% 2 087.54 1 663.02 25.5% 3 202.70 1 579.37 102.8% 1 152.55 1 186.00 (2.8%)
Rice 748.59 689.77 8.5% 1 518.24 1 663.02 (8.7%) 2 077.03 1 579.37 31.5% 888.45 1 060.18 (16.2%)
Fish 194.97b 569.30 1 125.67 89.00 6.45 1,28.9%
Others 175 119 46.7%
COSTS 374.4 325.7 15.0%
743.55 770.21 (3.5%) 1 701.17 1 095.20 55.3%
Labor 158.28 153.34 3.2% 449.11 528.72 (15.1%) 720.93 404.57 94.3% 299.80 261.28 14.7%
Dikes, Refuge
& Repairs 13.92 43.87 7.79 463.5%
Land
Preperation 35.90 35.44 1.3% 54.18 90.65 (40.2%) 93.28 93.28 0.0%
Seeding
(Pulling/
Handling) 7.01 9.53 (26.4%) 27.97 27.08 3.3%
Transplanting 32.13 32.49 (1.1%) 31.92 40.79 (21.8%) 77.98d 54.20d 43.9%
Fertilizing 5.78 11.20 (48.4%) 14.71d 13.64d 7.8%
Pest
eradication 10.31 20.30 (49.2%) - -
Weeding 23.00 32.54 (29.3%) 12.88 18.75 (31.3%) - -
Rice
Harvesting 53.33 52.87 0.9% 303.37 337.49 (10.1%) 251.68 208.58 20.7%
Stocking 1.48 3.74
Feeding, other
fi sh tanks 16.27 34.45
Fish
Harvesting 5.93 173.24
Irrigation & Water
Management 6.85 158.36 48.02 229.8% 63.17 36.00 75.5%
Inputs 218.48 156.89 39.3% 607.76 421.20 44.3%
Rice Seed 17.05 19.23 (11.3%) 18.76 17.57 6.8% 93.19 95.61 (2.5%) 72.97 66.63 9.5%
Fertilizer 60.31 70.38 (14.3%) 86.53 90.22 (4.1%) 149.32 164.87 (9.4%) 197.02 100.34 96.4%
Chemicals 0.97 7.10 (86.3%) 27.19 49.11 (44.6%) 15.11 53.45 (71.7%) 33.09 14.44 129.1%
Fingerlings 44.08 - 78.47
120.09
45.66 -
Feeds 7.21 - 7.53 56.73
23.87 -
Fuel 173.32 107.27 61.6%
Fixed Costs 79.62 75.62 5.3% 75.95 84.60 372.49 269.43
NET RETURNS 569.21 364.10 56.3% 1 343.99 892.81 50.5% 1 343.16 436.14 208.0%
a) Dry season (boro) crop. Original currency in Bangladesh taka (BDT), converted at USD1.00 = BDT 39.
b) Fish yield does not include wild fi sh.
c) Constructed using farm by farm data from Israel et al (1994), original currency in Philippine Peso (PHP) converted at the 1991 rate of USD1.00 = PHP27.48
d) Transplanting includes labor for weeding and fertilizing includes labor for pesticide application.
e) One-year operation of one-hectare farm w/ rice fi eld, homestead, dike and pond based on double rice crop and one fi sh crop. The data entered in this table
is not complete and do not add up as they do for the other countries since the manner of presentation in the original paper did not lend itself to reformatting.
Original fi gures were in Vietnamese dong (VND) and were converted at the rate of USD1=VND11 000. The difference between the gross returns is reported to be
not statistically signifi cant.
Table 16. Relative cost of labor and material inputs in rice+fish culture and rice only culture.
52 FAO and The WorldFish Center | Culture of Fish in Rice Fields
fi sh as well in 0.5 ha. This increases to 91% if
the entire 2.3 ha area is stocked with fi sh, even
if rice production remains constant and farm
requirements for cash and labor increased by
22% and 17%, respectively (Ahmed et al. 1992).
One indication that fi sh farming in rice fi elds
must be satisfactory (economically or otherwise)
from the farmers’ perspective is that in many
cases farmers on their own continue or even
expand the extent of their rice-fi sh farms after
having tried the technology. For example,
Zambian farmers wanted to continue with rice-
fi sh farming although researchers had found it to
be uneconomical (Nilsson and Blariaux 1994).
In Northeast Thailand, the total rice fi eld area
stocked with fi sh increased each year from 1985
to 1987 in spite of a dismal showing the fi rst
year (Thongpan et al. 1992). It has been pointed
out that nutritional benefi ts and lowered risk of
production may provide strong motivation for
rice farmers to diversify and that rice-fi sh farming
can be “profi table” in many ways including from
social, environmental, or ecological point of
views (Halwart 1999).
9.2.2 Improved nutrition
One benefi t that is often assumed, but never
supported by solid evidence, is that farmers who
culture fi sh in their rice fi elds have improved
nutrition. Villadolid and Acosta (1954) and
Coche (1967) and other writers postulated
that fi sh could prevent protein defi ciency and
contribute to the improved socioeconomic
welfare of populations. Yet in the case of rice-
fi sh farming there are no fi gures available as to
how much the caloric and protein intake or the
per caput fi sh consumption of farmer families
have been increased by the availability of fi sh
once these are grown in their own rice fi elds. For
example, it is estimated that home consumption
accounts for 35% of the production in Northeast
Thailand, but no absolute fi gure was given
(Mackay 1992). To complicate the matter, direct
consumption of the animals cultivated depends a
great deal on the market value of the product and
the economic status of the farmer.
In the Philippines, and most likely elsewhere,
farmers may be less inclined to have the
“additional burden” of raising fi sh if its main
purpose is to improve their own nutrition.
Farmers will likely culture fi sh if they believe they
can earn extra cash out of it beyond what they are
already earning from rice. Horstkotte-Wesseler
(1999) found no reduction in food expenses
in households practicing rice-fi sh culture as
all fi sh of marketable size produced were sold
and none consumed in the household. Income
augmentation was the most frequent reason
provided for engaging in rice-fi sh, additional
food only ranked third (Saturno 1994). In
Bangladesh, it was pointed out that extra income
was the most appreciated benefi t from growing
fi sh (70%) followed by “increased food for the
family” (59%) (Gupta et al. 1998).
Improvements of a farming household’s nutrition
as a result of culturing fi sh in the rice fi elds may
just be an incidental and perhaps even indirect
effect, such as being able to buy meat or chicken
as a result of the extra cash earned from fi sh. The
main benefi t of rice-fi sh farming is often seen as
providing an opportunity to earn cash.
Improvement in the local community’s nutrition
has been cited as one of the benefi ts of rice-fi sh
farming. With greater availability of fi sh, the local
population of a rice farming community will have
easy access to fi sh at affordable prices. However, in
a free market the farmer may opt to sell the fi sh to
a trader at a higher price than what the neighbors
can afford. The trader in turn may opt to bring
the fi sh to the nearest urban center where prices
are higher. This is a common situation in most
fi shing communities in the Philippines where fi sh
can be diffi cult to fi nd in the local market having
been siphoned off to the cities.
Nevertheless, particularly in more remote areas
and where the mixed forms of capture and culture
are prevalent, it is estimated that fi sh and other
aquatic organisms from rice fi elds provide a very
important component of the daily diet, hence
also the term “rice-fi sh societies” (Demaine and
Halwart 2001). The nutritional contribution
extends from micronutrients and proteins to
essential fatty acids that are needed for visual
and brain development. Recognizing this, the 20th
Session of the International Rice Commission
recommended its member countries to pay
increased attention to the nutritional value of fi sh
and other aquatic organisms from rice fi elds (FAO
2002; Halwart 2003a). A recent FAO/IUCN study
in Lao PDR confi rms the urgent need for further
focus on this issue (Meusch et al. 2003).
9.2.3 Public health
There are two public health vectors against which
fi sh have been employed: mosquitoes and snails.
53
Impact of Rice-Fish Culture
Mosquitoes are known carriers of malaria and
dengue fever. Certain species of freshwater snails
serve as hosts to trematodes (Schistosoma spp.)
that cause schistosomiasis should it enter the
human bloodstream. A third aspect is that rice-
fi sh culture may reduce the use of agricultural
chemicals that pose a health hazard to humans.
In some areas, where there is a tradition of using
nightsoil and/or there is a lack of latrines, human
infections with fi sh borne trematodes may be an
issue when fi sh from rice fi elds are eaten raw or
semi-preserved.
Field surveys in China indicate that mosquito
larvae densities in rice fi elds with fi sh were only
12 000·ha-1 as against 36 000·ha-1 in rice fi elds
without fi sh (Wang and Ni 1995). In other
studies mosquito larvae were observed in only
one of nine rice fi elds stocked with fi sh, being
completely absent in the other eight, whereas
in other rice fi elds not stocked with fi sh, the
density of mosquito larvae ranged from 32 000 to
128 000·ha-1. In Indonesia, fi sh were found to be
even more effective in controlling mosquitoes than
DDT. After fi ve years of fi sh culture in rice fi elds,
malaria cases decreased from 16.5% to 0.2% in a
highly endemic area for malaria (Nalim 1994). In
a control area using DDT the malaria prevalence
remained steady at 3.4% during the same period.
The effect of fi sh on the schistosoma-carrying
snails is less clear. As reviewed by Coche (1967)
fi sh were tested in the past for that purpose
in many parts of Africa where schistosoma
was endemic. At an experimental level, good
results were obtained when the Louisiana red
swamp crayfi sh was introduced into small rain-
fi lled quarry pits to control the schistosome-
transmitting Biomphalaria and Bulinus snails
in Kenya. Later work on fi sh as snail predators
has focused more on the golden apple snail as
was discussed in the section on rice pests, and
for which purpose it has been found effective
(Halwart 1994a; Halwart et al. 1998; Hendarsih
et al. 1994; FAO 1998). In countries such as
China, black carp (Mylopharyngodon piceus) is
used to control snails that are intermediate
hosts in parasite transmission. In Katanga, the
majority of snails in rice fi elds were controlled
by Haplochromis mellandi and Tilapia melanopleura
stocked at 200 fi sh·ha-1 and 300 fi sh·ha-1,
respectively. Halwart (2001) concludes that well-
maintained aquaculture operations contribute,
often signifi cantly, to the control of insects and
snails of agricultural and medical importance, and
that integrated management programes should be
pursued to keep vectors and pests at levels where
they do not cause signifi cant problems.
Often overlooked is the fact that fi sh in the
rice fi elds can reduce the use of chemical
pesticides. Despite the fact that some pesticides
are considered safe to use in rice-fi sh farming
due to their low toxicity, low tendency to bio-
accumulation, and short half-life, pesticides are
still poisons and may be carcinogenic or harmful
in other ways. Their use and misuse is a serious
public health issue that may become more serious
than mosquitoes and snails. Fish are potentially a
good herbicide and insecticide and stocking can
greatly reduce, if not completely eliminate, the
need for using chemical pesticides. The presence of
fi sh discourages farmers from applying pesticides
(Saturno 1994). The reduction or elimination of
the need to apply chemicals cannot but result in
an environment that is safer and healthier for the
people.
9.2.4 Social impact
It seems far-fetched that stocking fi sh in rice fi elds
can have a signifi cant impact on the society as
a whole, particularly so with isolated cases of
technology adoption by one or a few farmers
widely dispersed. However, when there is a large-
scale adoption involving an entire community
the social impact can be quite profound.
The use of fallow rice lands for fi sh culture by
landless farmers in Indonesia as described by
Ardiwinata (1957) is one such case. The situation
prevailing in Indonesia in the past was that
landless tenants were allowed to use the rice fi elds
for fi sh culture during the fallow season, giving
birth to the palawija system. Nowadays, the use
of the rice fi elds for fi sh production during the
fallow season is not limited to landless tenants,
but involves fi sh breeders requiring a larger area
for raising fi ngerlings (Koesoemadinata and
Costa-Pierce 1992; Fagi et al. 1992). In real-estate
development jargon such a scheme is called time-
sharing, an effi cient use of a resource giving a
chance for the landless to have access to land,
however temporary.
Although the Indonesian example may be
unique, in general adoption of rice-fi sh
farming should result in job creation. Physical
modifi cations of rice fi elds to accommodate and
harvest fi sh require extra labor. In the Philippines
ancillary activities connected to tilapia fi ngerling
production are:
54 FAO and The WorldFish Center | Culture of Fish in Rice Fields
• diking and excavation;
• making hapa-nets, harvesting seines and other
fi sh culture accessories;
• renting out water pumps, harvesting nets,
oxygen tanks, etc.;
• repair of pumps and making steel hoops for
scoopnets, etc.;
• harvesting, sorting and packing of fi ngerlings;
and
• transport of fi ngerlings.
Each type of activity is done by a different
person. This makes it possible to operate a tilapia
hatchery without incurring a large capital cost or
having a wide range of equipment or maintaining
more personnel than necessary. As none of these
aspects have been quantifi ed and documented,
there is little good information available on the
amount of labor generated.
9.3 Impact on the Environment
The impact of rice farming on the environment,
including its contribution to the greenhouse
effect, should be a matter of concern to everyone.
There is no doubt that the development of rice
lands has resulted in the loss of natural wetlands
and marshlands, although this made a difference
between widespread famine and food suffi ciency
in many parts of the world. This section, however,
will only examine what impact the introduction
of fi sh may have on the ecosystem of an existing
rice fi eld.
9.3.1 Biodiversity
A rice fi eld is known to be the habitat of a diverse
assemblage of species (Heckman 1979; Balzer et
al. 2002). Intensifi cation of rice cultivation with
an associated increase in chemical pesticide use
is reducing this diversity (Fernando et al. 1979).
Since rice-fi sh farming often reduces the need
to use chemicals for pest control, this assists in
preserving a diverse rice fi eld biota. Utilizing
the existing - native - species for rice-fi sh culture
serves to actively preserve the biodiversity.
9.3.2 Water resources
With fi sh in the rice fi eld, a greater water depth
has to be maintained and more water may be
required, an issue raised half a century ago by
Schuster (1955). Even without fi sh, rice farming
consumes large volumes of water. For rice culture
in general, Singh et al. (1980) and Sevilleja et al.
(1992) estimated that a crop needs a minimum
of 1 000 to 1 500 mm of water, respectively. If a
hectare of rice fi eld produces 10 mt of rice, it still
takes from 1 to 1.5 m3 of water to produce 1 kg
of paddy.
Fish are a non-consumptive user of water, and
while they can degrade the water they do not use
it up. If cleaned, the same water can be returned
and reused by the fi sh. The increased water use is
due to percolation and seepage (P&S) and leakage
(L), which increase with rice-fi sh culture due to the
deeper water maintained, a purely physical process
that takes place with or without the fi sh. Sevilleja
et al. (1992) estimated that the water requirement
for rice culture was 1 662 mm while rice-fi sh
culture required up to 2 100 mm, or 26% more
than rice monoculture. The main water losses are
attributable to P&S (67%), followed by L (21%).
Thorough puddling during land preparation, good
maintenance of the dikes and proper sealing of
inlets and outlets may reduce the losses.
9.3.3 Sustainability
Wet rice cultivation has been practiced for at least
4 000 years, and its long history indicates that
traditional rice farming is basically sustainable.
What is less certain is whether the dramatic
increases of rice production made possible by the
“green revolution” are sustainable (Greenland
1997). Global warming, sea level rise, increased
ultraviolet radiation and even availability of water
are all expected to have an adverse impact on
rice production. However, such scenarios are far
beyond the level and scope of this report, and for
the foreseeable future it can be assumed that rice
farming will continue. Further, it seems likely that
the culture of fi sh in rice fi elds can enhance the
sustainability of rice farming, since indications
are that the presence of fi sh makes the rice fi eld
ecosystem more balanced and stable. With fi sh
removing the weeds and reducing the insect pest
population to tolerable levels, the poisoning of
the water and soil may be curtailed.
9.4 Participation of Women
In most of the rice-producing countries of
Asia, women are already an integral part of
the farm labor force. The integration of fi sh
culture into the rice farming activity will likely
expand women’s participation further. There
are no socioeconomic data quantifying possible
involvement of women in rice-fi sh farming
activities but as Dehadrai (1992) has amply
stated, any “projected new opportunities for
55
Impact of Rice-Fish Culture
women in rice-fi sh farming emanate largely from
the known and well documented involvement of
women in the management of rice in Asia.” A
benefi cial aspect may be that the presence of fi sh
in the rice fi elds could save precious time that
women and children otherwise spend fi shing
in other areas, although this effect is somewhat
counterbalanced by the extra work needed for
the rice-fi sh management.
9.5 Macro-Economic Impact
There are three macro-economic issues on
which the widespread adoption of rice-fi sh
farming technology could impact: food security,
employment generation, and national income.
However, such discussions will be in the realm
of speculation since most countries do not have
separate statistics on rice-fi sh farming areas nor
rice and fi sh yields in such areas.
Speculations, however, indicate that the potential
impact is tremendous. If 5% of the irrigated rice
lands in the Philippines were stocked with fi sh,
the production would increase by 29 000 t worth
US$ 35 million and provide 5 900 t of protein
(Ahmed et al. 1992). Cai et al. (1995a) estimated
that if 10% of the rice fi elds south of the Huai He
River, China, were used, the commercial fi sh yield
would be 346 000 t at a yield of 300 kg/ha, and 5
billion full-size fi ngerlings. With such production
potential the ecological and economic benefi ts
would be considerable.
Coche (1967) summed it up very well by saying
that fi sh culture in rice fi elds is technically an
almost ideal method of land use, combining
the production of both vegetal and animal
proteins. Its further development is important,
as it may contribute to a guarantee of the world
food supply. Widespread adoption of rice-fi sh
farming as a strategy to substantially narrow the
gap between the protein supply and demand is
a potential option for any major rice-producing
country. All it requires is the political will to push
through with it.
56 FAO and The WorldFish Center | Culture of Fish in Rice Fields
10. Experiences of Various Countries
As far as can be ascertained from the available
literature, rice-fi sh farming is still practiced in
quite a few countries as shown in Figure 20.
There are no hard statistics on the total extent
of rice-fi sh farming globally but estimates for
the major countries are available (Table 17). The
world’s rice-fi sh farms are concentrated within
South Asia, East Asia and Southeast Asia but
there are also some notable developments in
Africa. This chapter mainly provides a historical
perspective and reports on the current status in
major regions.
10.1 East Asia
China
China, with 27.4 million ha of rice land, is
second only to India in terms of hectarage but
is fi rst in terms of rice production with about
166 million t.19 It is the world’s largest aquaculture
producer with an inland production of
28 million t,20 and rice-fi sh culture has always
been given a strong emphasis in China. It also
19 FAOSTAT data (2003).
20 FAO FISHSTAT data (2002), excluding aquatic plants.
has the oldest archaeological and documentary
evidence for rice-fi sh farming.
However, it was not until after the founding
of the People’s Republic of China in 1949
that rice-fi sh culture developed quickly in the
whole country. In 1954 it was proposed that
development of rice-fi sh culture should be spread
across the country (Cai et al. 1995a), and by
1959, the rice-fi sh culture area had expanded to
666 000 ha. From the early 1960s to the mid-
1970s there was a temporary decline in rice-fi sh
farming. This was attributed to two developments:
fi rst, the intensifi cation of rice production
that brought with it the large-scale application
of chemical inputs; and second, the ten-year
Cultural Revolution (1965-75) during which time
the raising fi sh was considered a bourgeois way of
making money and was offi cially discouraged.
Improved rice varieties, use of less toxic chemicals
and political changes (production-contract or
production responsibility system) reversed the
earlier trends of the 1960s and 1970s. The new
Figure 20. Map of the world showing areas where rice-fish and/or rice-crustacean farming is practiced.
57
Experiences of Various Countries
system allowed individual families, rather than
the commune, to become the main production
units. In addition, the rapid development of
aquaculture required a large supply of fry and
fi ngerlings. This demand was partly met by
fi ngerling production in rice fi elds.
In 1983, the Ministry of Agriculture, Animal
Husbandry and Fisheries (now the Ministry of
Agriculture) organized the First National Rice
Fish Culture Workshop. The workshop resulted
in the establishment of a large coordination
group for Eastern China to popularize rice-fi sh
farming techniques. Also various other provinces,
autonomous regions and municipalities
undertook such measures in line with local
conditions. As a result, by 1996 China had 1.2
million ha of rice-fi sh farms producing 377 000 t
of fi sh (Halwart 1999).
Thus it can be seen that in China rice-fi sh farming
is promoted actively as a viable option for rice
production. It is part of the program not only of
fi shery institutions, but also of agencies involved
in rice production. In addition, it receives
considerable support at the ministerial level of
government.
Japan
Rice-fi sh farming appears to be of minor
importance in Japan and there is not much
literature on the subject. After reaching a peak
production of 3 400 t in 1943 due to war-time
food production subsidies, carp production in
rice fi elds decreased to only 1 000 t during the
1950s. In 1954 only 1% of Japan’s 3 million ha
of rice land was used for carp culture (Kuronoma
1980) and it is no longer practiced on a signifi cant
scale, if at all (Pillay 1990).
Korea
In Korea, rice-fi sh farming started only in the
1950s and never spread widely because the fi sh
supply from inland waters was suffi cient to meet
the limited demand for freshwater fi sh (Kim et
al. 1992). Inland production accounted for only
1.7% of the total fi sh production of 3.3 million
t in 1987. As of 1989 only 95 ha of rice fi elds
were being used for fi sh culture, and only for
the growing the most popular species of loach
(Misgurnus anguillicaudatus).
10.2 Southeast Asia
Indonesia
Rice-fi sh farming is believed to have been practiced
in the Ciamis area of West Java, Indonesia,
even before 1860 although its popularization
apparently started only in the 1870s. Ardiwinata
(1957) attributed the expansion of fi sh culture
in rice fi elds to the profound changes in the
governing system during the Preanger regency in
West Java in 1872, during which the possession
Country
Rice Rice-fi sh
Total Irrigated Rainfed Lowland Floodprone Upland
(‘000 ha) (‘000 ha)
Bangladesh 10 245 22 47 23 8 ?
Cambodia 1 910 8 48 42 2 ?
P.R. China 33 019 93 5 - 2 1204.9
Egypt 462 100 - - - 172.8
India 42 308 45 33 7 15 ?
Indonesia 10 282 72 7 10 11 138.3
Korea, Rep. 1208 91 8 - 1 0.1
Lao PDR 557 2 61 - 37 ?
Madagascar 1 140 10 74 2 14 13.4 (highlands)
Malaysia 691 66 21 1 12 ?
Philippines 3 425 61 35 2 2 ?
Sri Lanka 791 37 53 3 7 ?
Thailand 9 271 7 86 7 1 25.5 (culture)
2966.7 (capture)
Vietnam 6 303 53 28 11 8 40.0 (Mekong delta)
Table 17. Distribution of rice and rice-fish area, by environment (Halwart 1999).
58 FAO and The WorldFish Center | Culture of Fish in Rice Fields
of rice fi elds was made hereditary. The pressure
on the arable land by the growing population
caused the rental rates to go up. Tenants started
to utilize their fi elds by stocking fi sh, generally
common carp, or by raising other crops. Fish
culture was popular because the capital required
was minimal, and the landowners did not expect
a share of the fi sh. This practice is what is called
palawija or fallow-season crop.
The spread of palawija outside its point of origin
in Java is attributed to the Dutch administrators
who promoted the concept. By the 1950s some
50 000 ha of rice land were already producing
fi sh. The development of irrigation systems also
contributed to the expansion of the area used
for rice-fi sh farming. The average area of rice-
fi sh farming increased steadily after Indonesia
became independent in 1947 and rice-fi sh farms
covered 72 650 ha in 1974, but declined to less
than 49 000 in 1977. The decline was attributed,
ironically, to the government’s rice intensifi cation
program (Koesoemadinata and Costa-Pierce
1992). However, the surging demand for carp
fi ngerlings brought about by the proliferation
of fi sh cages in dams and reservoirs stimulated
expansion once again. The area utilized reached
an all time high of 138 000 ha in 1982, but
declined to 94 000 ha in 1985.
Recent reports indicate that rice-fi sh farming
is on the upswing. The 1995 fi gures from the
Directorate General of Fisheries indicate a
total area of over 138 000 ha. The resurgence
has been attributed to a drastic change in rice
production practices in 1986 when integrated
pest management (IPM) was declared the offi cial
national pest control strategy. At present rice-fi sh
farming is practiced in 17 out of 27 provinces
in Indonesia. In summary, the development of
rice-fi sh farming can be attributed to landless
tenants who wanted an extra income during
the fallow season for rice. The government’s rice
intensifi cation program, promoting heavy use
of chemical pesticides, was the major reason for
its decline in the early to mid-1970s. Its growth
at present has been attributed to the increased
demand for fi ngerlings to stock fi sh cages, which
makes it a purely market-led development.
Thailand
Integrated rice-fi sh farming is believed to have
been practiced for more than 200 years in
Thailand, particularly in the Northeast where
it was dependent upon capturing wild fi sh for
stocking the rice fi elds. It was later promoted by
the Department of Fisheries (DOF) and expanded
into the Central Plains. The provision of seed
fi sh and technology helped in popularizing the
concept. Rice yields in rice-fi sh farms in the 1950s
increased by 25-30% and the fi sh yields ranged
from 137 to 304 kg·ha-1·crop-1 (Pongsuwana
1962). As a measure of the importance given to
rice-fi sh farming, the DOF established a Center
for Rice-Fish Farming Research in Chainat in
the Central Plains in 1968. However, during the
1970s, Thailand, like the rest of Asia, introduced
the HYVs of rice and with it the increased use
of chemical pesticides. This resulted in the near
collapse of rice-fi sh farming in the Central Plains
as farmers either separated their rice and fi sh
operations or stopped growing fi sh altogether. In
1974 the research center in Chainat was closed.
However, rice-fi sh farming did not completely
vanish and in recent years it has recovered,
particularly in the Central Plains, North and
Northeast Regions. In 1983 rice fi eld culture
fi sheries was practiced on 2 820 ha mainly in
the Central, North, and Northeast Provinces.
This grew to 23 900 ha in 1988 and was further
expanded to 25 500 ha in 1992. Such a steep
increase resulted from a general decrease in
the availability of wild fi sh made worse by the
occurrence of the ulcerative disease syndrome in
wild fi sh stock . Fedoruk and Leelapatra (1992)
attributed the recovery to more discriminate use
of HYV; the emergence of pesticides that when
properly applied are not toxic to fi sh; the growing
perception of the economic benefi ts of rice-fi sh
farming, and its promotion in special projects
assisting disadvantaged farmers, among other
factors.
Little et al. (1996) concluded that the
development of rice-fi sh systems is unlikely to be
homogeneous in the Northeast Region. The high
expectations of farming communities is thought
to be a major constraint to the wider adoption of
rice-fi sh systems where off-farm employment was
the norm as the major means of livelihood until
the economic crisis in mid-1997. The increasing
frequency of directly broadcasting rice seeds and
using machines for fi eld preparation are signs of
the growing labor shortage. The shortage may
favor the development of more easily managed
pond culture rather than the more laborious
rice-fi sh system. On the other hand, adoption of
rice-fi sh systems in the Northeast Region may be
biased towards those who are betteroff and have
access to labor and other resources.
59
Experiences of Various Countries
Malaysia
In Malaysia, from where reports on the practice
of rice-fi sh farming appeared as early as 1928, the
rice fi elds have always been an important source
of freshwater fi sh. Before the 1970s when farms
still practiced single-cropping, integrated rice-fi sh
farms were the major suppliers of freshwater fi sh,
especially for snakeskin gouramy (T. pectoralis),
catfi sh (Clarias macrocephalus), and snakehead
(Channa striata). Fish production from rice
fi eld started to decline with the introduction
of the double-cropping system and with it the
widespread use of pesticides and herbicides (Ali
1990).
Vietnam
Vietnam has a strong tradition of integrating
aquaculture with agriculture. The Vietnamese
system involves the production of livestock,
vegetables, and fi sh in a family farm and does
not necessarily involve rice. While fi sh, shrimps
and other aquatic organisms were traditionally
caught in the rice fi elds, these were reported to
have become scarce ever since chemical pesticides
started to be used (Mai et al. 1992). Le (1999)
reports fi ve common rice-fi sh culture systems
being practiced in Vietnam, but gives no fi gures
on the area involved. The fi ve systems are fi sh-
cum-rice for nursery and growout, fi sh-cum-rice
for growout only, shrimp-cum-rice, fi sh/rice
rotation and shrimp/rice rotation.
The Philippines
In the Philippines, fi sh are traditionally allowed
to enter the rice fi elds with the irrigation water
and are later harvested with the rice. The earliest
mention of stocking fi sh in a rice fi eld in the
Philippines was made in 1954 (Villadolid and
Acosta 1954), but it was not until 1974 when rice-
fi sh farming became part of a research program of
Central Luzon State University (CLSU). In spite of
the lower rice yields (on average 3.8%), in 1979
the government proceeded to promote rice-fi sh
farming nationwide. The decision was based on
the results of the economic analysis that even
with a reduced rice production, the farmer would
still be economically ahead due to the additional
income from the fi sh. After a peak of 1 397 ha
involving 2 284 farms in 1982 the program was
discontinued in 1986. At that time it covered only
185 ha (Sevilleja 1992) despite the fact that the
average production of rice from rice-fi sh farms
was above the national average.
Sevilleja (1992) did not offer any explanation for
the sudden drop in the participation by 1983;
however records show that 1983 was one of the
worst El Niño years in recent history and the
drought badly affected agriculture (Yap 1998).
The year 1983 also marked the start of political
turmoil and relative politico-economic stability
did not return until 1990. The failure of the rice-
fi sh promotion in the Philippines should also be
viewed against the political milieu. In 1999, a
more modest rice-fi sh program was launched.
10.3 South Asia
Rice-fi sh farming is known to have been practiced
in India, Bangladesh and Sri Lanka and much
of the history, current practice and potential is
highlighted by Fernando and Halwart (2001) in
their paper on fi sh farming in irrigation systems
with special reference to Sri Lanka.
India
Having the world’s largest area devoted to rice
cultivation at 42 million ha as of 1994, India
produces a considerable amount of fi sh from
its rice fi elds. A report on the status of rice-fi sh
farming in India (Ghosh 1992) indicates that
India has rice-fi sh farms covering 2 million ha,
which is the largest reported area for rice-fi sh
culture for a single country. Rice-fi sh farming is
considered an age-old tradition in the states of
West Bengal and Kerala, but it is limited to capture
systems in the Ganges and Brahmaputra plains.
The practice cuts across different ecosystems, from
the terraced rice fi elds in the hilly terrain in the
north to coastal pokhali plots and deepwater rice
fi elds. In between are the mountain valley plots of
northeastern India and rainfed or irrigated low-
land rice fi elds scattered all over India. The species
involved are just as diverse with over 30 species of
fi nfi sh and some 16 species of shrimps listed as
being cultured in Indian rice fi elds. Most of the
non-carp species and penaeid shrimp species are
from natural stocks entering the rice fi eld with the
fl ood waters. Production rates are varied, ranging
from 3 kg·ha-1·year-1 in the deepwater rice plots
relying on natural stock of mixed species to over
2 t·ha-1·year-1 of Tiger shrimps (P. monodon) in
shallow brackish water rice fi elds (Ghosh 1992).
Bangladesh
Farmers in Bangladesh have been harvesting
fi sh from their rice fi elds for a very long time.
60 FAO and The WorldFish Center | Culture of Fish in Rice Fields
The description of the traditional practice in
Bangladesh that follows came from Dewan
(1992). Farmers construct ponds of different
sizes in low-lying areas of the fi eld and when
the ponds and rice fi elds are full of water during
the monsoon, carp fry are released, following no
specifi c stocking density. The small ponds may be
provided with brush shelters, but no fertilizers
or feed are applied. The fi sh are harvested over
a period extending from the time the rice is
harvested in November-December up to March.
In the coastal areas, marine shrimps such as the
various penaeids including P. monodon may also
be cultured. The traditional bheri system is used
wherein the rice fi elds are enclosed by small
embankments complete with inlet channels
and sluice gates. Fields vary in size from 3 to 50
ha. Both rotational and concurrent systems are
practiced. Occasionally, the freshwater prawn
(M. rosenbergii) may also be cultured. Prawn fry
gathered from nearby rivers are stocked after the
monsoon rains have washed out the salinity from
the rice fi elds.
Intensive studies and surveys undertaken
from 1992 to 1995 in Bangladesh showed
improvement in income and food availability for
most of the respondents to the extent that 89% of
the farmers involved planned to continue with the
practice (Gupta et al. 1998). CARE-Bangladesh
promoted rice-fi sh culture in all its projects as an
integral part of its IPM strategy (Nandeesha and
Chapman 1999).
Bangladesh is one of the few countries actively
promoting rice-fi sh farming and pursuing a
vigorous research and development program.
NGOs in Bangladesh are likewise showing
increasing interest in rice-fi sh farming. Among the
more successful NGO efforts was the Noakhali
Rural Development Program in 1989 which used
the rotational system to produce from 223 to
700 kg·ha-1 of mixed species of fi sh in 50 fi elds
planted with local rice varieties (Haroon et al.
1992). More recently, CARE has become the most
active NGO involved in rice-fi sh farming.
Thousands of farmers in Bangladesh have
experimented with rice-fi sh culture and have
developed practices to suit their own farming
systems. Both table fi sh and fi ngerlings are being
produced with farmers generally concentrating
on fi sh seed during the dry season, which is an
irrigated crop. The adoption rate among the
project participants has been in the range of
10-40% depending on the area and sex of the
participant. Initially the adoption rate was lower
among females, but the activity is reported to be
gaining popularity among both male and female
groups. Increased income and fi sh consumption
have been noted among families adopting rice-
fi sh culture in Bangladesh.
10.4 Australia
A large commercial rice grower in Newcastle,
New South Wales is stocking common carp in
rice fi elds on a trial basis. The intention is to
eventually stock 5 000 ha with common carp on
a concurrent basis with rice. The fi sh produced
will be used as raw materials for pet food
(personal communication, Mr. Jonathan Nacario,
Consultant, 12 October 1999).
10.5 Africa, Middle East and West
Asia
Apart from Egypt, Africa has 10 rice producing
countries with a total rice land area of 6.8
million ha. Nigeria has the largest rice area with
1.7 million ha, followed by Madagascar and
Guinea with 1.2 million ha and 1.1 million ha,
respectively. In terms of rice production Nigeria
is fi rst with 3.8 million t, followed by Madagascar
with 2.36 million t.
Madagascar
The earliest report on rice-fi sh culture in Africa
comes from Madagascar. As early as 1928.
Legendre (cited in FAO 1957) reported on
the practice in Madagascar on the culture of
Paratilapia polleni, Carassius auratus and Cyprinus
carpio in rice fi elds. This was followed by another
report in 1938 on poultry-raising and fi sh culture
in rice fi elds. Based on the report of Coche (1967),
the level of technology in Madagascar at that
time appears to have approximated that of Asia,
although stocking was lighter. Both concurrent
and rotational systems relying on entry of natural
fi sh stock were practiced. In 1952 the government
initiated a program to promote fi sh culture in
fi shponds and rice fi elds. Local capacity in the
mass production of fi ngerlings was developed in
1972. Only in 1979 was suffi cient progress made
for the government to promote rice-fi sh culture.
Fingerling supply remained a major constraint
until 1985 when the government promoted private
sector participation in fi ngerling production. By
the end of the 1980s it was realized that without
continued external assistance the government
would be unable to sustain the operation (Van
61
Experiences of Various Countries
den Berg 1996). An average yield of 80 kg·ha-1
indicates that culture techniques at the farm level
still need to be improved (Randriamiarana et al.
1995).
A country with almost 900 000 ha of rice fi elds
does have a great potential for rice-fi sh farming,
as about 150 000 ha could be suitable for rice-
fi sh farming. A potential annual production of
300 000 t of edible fi sh has been projected from
the said areas. Rice-fi sh culture in Madagascar was
signifi cant enough to be mentioned in a country
study done by the US Library of Congress (Metz
1994).
Malawi
Farmers in Malawi are just beginning to grow rice
and fi sh together as well as fi sh and vegetables.
Although not specifi cally mentioned, the fi sh
involved are apparently tilapia, where O. shiranus
and/or T. rendalii are reportedly the principal
species in the country.
Zambia
Rice-fi sh culture trials have been reported for
Zambia by Coche (1967) but failed to take off.
In 1992-93, FAO again introduced the concept
during the implementation of the Aquaculture
for Local Community Development Programme
(ALCOM). Although the project was discontinued
when economic analysis showed that income
from the fi sh and the additional rice harvested
failed to compensate for the additional cost of
culturing fi sh, many farmers continued with
the practice on their own (Nilsson and Blariaux
1994).
Senegal
In Senegal, low-land farmers have resorted to
integrating fi sh culture with rice farming due to
environmental changes that endangered their
rice farms (Diallo 1998). Seawater encroaching
on their rain-fed coastal rice fi elds forced them
to build fi shponds to prevent tidal waters from
inundating their rice fi elds. In the process they
also produce fi sh.
Other African Countries
Congo-Katanga (now known as Shaba province
of the Republic of Zaire) and Rhodesia (now
Zimbabwe), Ivory Coast, Gabon, Liberia and
Mali and Benin are reported to have conducted
rice-fi sh culture trials (Coche 1967; Nzamujo
1995; Vincke 1995). More recent activities for
West Africa have been documented by Moehl
et al. (2001). Integrated aquaculture trials have
been limited to fi sh with only livestock in both
Cameroon (Breine et al. 1995) and Rwanda
(Verheust et al. 1995).
Egypt
Egypt, which is the biggest rice producer in both
the Middle East and the African continent, started
with a capture-type of rice-fi sh farming based
totally on occasional fi sh stock coming in with
the irrigation water. Limited experiments using
carps in the early 1970s were conducted with
encouraging results (Essawi and Ishak 1975).
The rice-fi sh farming area expanded considerably
using reclaimed salt-affected lands and in 1989
reached a peak of 225 000 ha. As rice prices
increased, however, HYVs were adopted and
reclaimed lands were used for rice monoculture.
This resulted in a drop in the rice-fi sh area to
172 800 ha by 1995. Nonetheless the 1995 fi sh
production from rice fi elds accounted for 32% of
the total aquaculture production in the country
(Shehadeh and Feidi 1996). Since then 58 000 ha
of farmland have been added producing 7 000 t
of C.carpio in 1997 (Wassef 2000).
Iran
Iran begun rice-fi sh culture trials in 1997
(personal communication, Mr Ibrahim Maygoli,
Shilat Aquaculture Division Head, Tehran,
Iran, 30 August 1999). With good results
obtained, 18 farms with a total area of 12 ha
adopted the technology. Chinese major carps
are used concurrently with rice, sometimes
with supplementary feeding. Productions over
1.5 t of fi sh per ha together with 7 t of rice have
been achieved with a high survival rate (96%),
despite an average water temperature of only
23°C during the culture period. In addition, 70
farms have adopted a rotational rice-fi sh farming
system where the rice fi eld is stocked with trout
during the winter months when the average
water temperature is 12°C, yielding 640 kg·ha-1.
Concurrent culture of M. rosenbergii with rice is
also being tried.
10.6 Europe
Rice is not a major crop in Europe and is relatively
important only in Italy (216 000 ha of rice land)
producing 59% of the European Union’s (EU)
62 FAO and The WorldFish Center | Culture of Fish in Rice Fields
rice production. Spain with 86 000 ha comes
a distant second, contributing only 25% of the
EU production. The other European countries
producing rice are Albania, Bulgaria, France,
Greece, Hungary, Macedonia, Romania, and
Yugoslavia.
Italy
Rice-fi sh culture was introduced to Italy at the
end of the 19th century and was to progressively
become important during the subsequent 40 years.
The main species were C. carpio, C. auratus and
Tinca tinca. The rice fi elds were used to produce
fi ngerlings that had a ready market among pond
owners and angling society. The practice gradually
declined and by 1967 it was no longer considered
an important activity. The cause of its decline was
traced to economic, social and technical factors.
As rice farmers abandoned traditional practices
to increase rice production, the production of
fi sh became less and less compatible with these
new practices (Coche 1967). There is a renewed
interest in investigating fi sheries management
in rice fi elds including ecological and economic
aspects under modern methods of cultivation at
the University of Bologna.
Hungary
In Hungary where irrigated rice land once covered
45 000 ha, C. carpio was cultured in the fl ooded
fi elds by the cooperative and state farms to reduce
production costs. In the absence of marine fi sh,
freshwater fi sh commanded a good price thus
boosting the farmers’ income. It was also reported
that fi sh helped keep the fi elds clean. With the
total rice hectarage down to only 5 000 ha as of
1992, there is no published information as to
whether any of the rice fi elds are still cultivating
fi sh.
10.7 The Former Soviet Union
Although wheat is the most important grain for
most of the former Soviet Union countries, rice
is grown in some of the Central Asian republics
and many have tried or practiced rice-fi sh
culture.
Fernando’s et al. (1979) listing of publications
dealing with the aquatic fauna of the world’s
rice fi elds had 55 entries from the former Soviet
Union, of which 12 dealt specifi cally with rice-
fi sh culture. This is a large number considering
that the bibliography had a total of 931 entries
from 61 different countries and territories. By way
of comparison the US had a total of 70 papers
listed, 89 for India, and 54 for Japan.
The most authoritative historical review for this
region is by Meien (1940).
10.8 South America and the
Caribbean
Although rice is produced in nine countries
in South America and eight countries in the
Caribbean, the culture of fi sh in rice fi elds is not
widespread. As early as the 1940s, experiments
were being conducted in Argentina on the culture
of kingfi sh (Atherina bonariensis) in rice fi elds as
a food fi sh and for the control of mosquitoes
(Macdonagh 1946 as extracted from FAO 1957).
Attempts were also made to introduce the
concept in the British West Indies and the British
Guiana in the early 1950s (Chacko and Ganapati
1952 as extracted from FAO 1957).
Experiments on integrating fi sh culture with
rice production are, or were, being conducted in
Brazil, Haiti, Panama and Peru, but only Brazil
appears to have had some degree of commercial
success. Extensive rice-fi sh culture had its
beginnings in the valley of Rio São Francisco
(northeast) and in the rice fi elds in the south.
In the northeast, farmers became interested
in semi-intensive rice-fi sh culture using native
fi sh species caught in lakes along the river such
as curimatá pacu (Prochilodus argentes), piau
verdadeiro (Leporimus elongatus), and mandi
arnarelo (Pimelodus clarias). Experiments on
intensive rice-fi sh culture were also conducted in
the Paraíba basin using the C. carpio and Congo
tilapia (T. rendalli) (Guillen 1990). The outlook
for rice-fi sh culture is thought to be favorable
for the region because of its suitable climate
and irrigated areas. Recent FAO-facilitated
community work focuses on the promotion of
aquaculture and other integrated production
methods in rice-based systems in Guyana and
Suriname.
10.9 The United States
Rice-fi sh farming used to be considered
important in the United States. After the rice had
been harvested, the rice lands were fl ooded and
stocked mainly with C. carpio, bigmouth buffalo
(Ictiobus cyprinellus), and channel catfi sh (Ictalurus
punctatus ). In 1954, some 4 000 ha of woodlands
in Arkansas were diked, fl ooded, and stocked
63
Experiences of Various Countries
with fi sh. In 1956 this increased to 30 000 ha and
reportedly produced 3 200 t of fi sh. Demand for
fi ngerlings shot up and new hatcheries had to be
put into operation.
The growing importance of rice-fi sh farming and
the need to improve existing practices led the
US Congress to enact the Fish Rice Rotation Act
of 1958 for the Secretary of the Interior (who
then had jurisdiction over the Fish and Wildlife
Service) to implement. Its objective was “to
establish a program for the purpose of carrying
on certain research and experimentation to
develop methods for the commercial production
of fi sh on fl ooded rice acreage in rotation with
rice fi eld crops, and for other purposes.” To carry
out the studies on rice-fi sh rotation a research
station, which was to become the Stuttgart
National Aquaculture Research Center (SNARC),
was established in Stuttgart, Arkansas.
By 1960, a survey of 53 selected farmers in the
states of Arkansas, Louisiana and Mississippi
showed that 20.4% of the total water surface area
was used for fi sh culture. At that time there were
1.25 million ha of irrigated rice lands in the US
and the potential for fi sh culture was considered
great. Coche (1967) thought the industry had
bright prospects, saying, “There is little doubt
that a new area of intensive development can be
forecast for fi sh culture in the vast complex of US
rice fi elds.”
As technology evolved and because of new
economic realities, interest in rice-fi sh farming
appears to have waned sometime after the 1960s.
This can be surmised from the shift in the research
direction of SNARC.
Nonetheless, the concept of fi sh-rice rotation on
a commercial scale is far from dead in the US.
However, instead of fi nfi sh, crawfi sh are now
being rotated with rice. Two crawfi sh species
are popular because of their hardiness and
adaptability, the red swamp crawfi sh (Procambarus
clarkii) and to a certain extent the white river
crawfi sh (P. zonangulus). The life-cycle of crawfi sh
and environmental requirements lend very well
to being rotated with rice and even with rice and
soybeans. Most of the crawfi sh produced in the
US now come from the rice fi elds of the southern
states (De La Bretonne and Remaire 1990).
64 FAO and The WorldFish Center | Culture of Fish in Rice Fields
11.1 Prospects
It is now an opportune time to promote rice-fi sh
farming. Integrated rice-fi sh farming has been
practiced for some time but has failed to become
so common as to become second nature to rice
farmers. Interest in rice-fi sh farming over the years
has waxed and waned among policy-makers,
scientists, extension workers and farmers in
different countries. This is understandable given
the circumstances during particular periods. Now
is a good time to rekindle the interest among all
sectors since policy-makers, researchers, extension
workers and farmers might be more receptive due
to the convergence of four factors.
First, capture fi sheries has in many areas reached
its limit. Increasing aquaculture production is
one obvious solution to meet growing demands,
and the world’s rice fi elds represent millions
of hectares of fi sh growing areas. The 1996
World Food Summit agreed “to promote the
development of environmentally sound and
sustainable aquaculture well integrated into rural,
agricultural and coastal development.”
Second, there is a growing recognition of the
need to “work with” rather than “against”
nature. Integrated pest management (IPM) is
being promoted in the place of extensive use of
pesticides, and fi sh have been found to be an
effective pest control agent. Chemical pesticides
are a double-edged sword that can be as injurious
to human health and the environment as to its
targeted pests.
Third, fresh water is a limited resource and the
integration of fi sh with rice is one way of using
water more effi ciently by producing both aquatic
animals and rice. In addition, new land suitable
for aquaculture is limited and the culture of fi sh
together with rice is an effective way of utilizing
scarce land resources.
Fourth, rice is not a purely economic commodity;
in many countries it is a political commodity as
well. The farm gate price of rice is not always
based on providing a just economic return to the
farmers, but often has political implications such
as national food security and export potential.
The market, however, usually determines the
price of fi sh. While growing fi sh in a rice fi eld
11. Prospects and Program for the Future
entails minimal incremental costs, it is one way of
augmenting the farmers’ income.
These developments serve as an impetus for
promoting rice-fi sh farming. Together, these
trends cover various concerns of all sectors
involved in rice farming.
11.2 Major Issues and Constraints
Several concerns over rice-fi sh culture have
been identifi ed (in a working paper prepared
for the 16th Session of the International Rice
Commission, 1985).
• The greater water depth required in rice-fi sh
farming than in traditional rice cultivation
may be a limiting factor if the water supply
is inadequate. As discussed earlier, increased
leaks, seepage and percolation due to
maintaining deeper standing water in rice-fi sh
culture can increase water needs signifi cantly.
• Fish cause damage to rice plants which
they uproot and eat them. Destructiveness
of fi sh on the rice crop has been observed,
particularly when bottom-dwelling C. carpio
are stocked too early after crop establishment
and the transplanted rice seedlings have
not developed a good root system, or when
herbivorous fi sh such as C. idellus are stocked
at larger sizes capable of consuming whole
plants. These problems can easily be avoided
by good management practices including
species selection, stocking size and timing of
stocking.
• More fertilizers are needed to increase the
primary productivity of the water and feed the
fi sh. Increased fertilization is assumed since
both the rice and the phytoplankton require
nutrients. The increased fertilization was fi rst
estimated by Chen (1954) to range from 50
to 100%. However, experience has shown
that in most cases the fertilizer requirement
decreased with the introduction of fi sh (Gupta
et al. 1998; Israel et al. 1994; Yunus et al.
1992). Cagauan (1995) found that a rice fi eld
with fi sh has a higher capacity to produce and
capture nitrogen (N) than one without fi sh.
• A small percentage of the cultivable area is lost
through the construction of drains and shelter
holes resulting in reduction of the paddy
yield. Again, experience has shown that the
65
Prospects and Program for the Future
rice yield often increases in rice-fi sh culture
and thus the excavation of a small part of the
rice fi eld (normally no more than 10%) often
results in no net loss but rather a net gain in
rice production.
• The use of short-stemmed, high yielding rice
varieties is limited by the deeper standing
water required for rice-fi sh farming. Even
IR36, which has a tiller height of 85 cm, has
been successfully used for rice-fi sh farming.
Costa-Pierce and De la Cruz (1992) found that
widespread use of HYV was not considered a
major constraint in rice-fi sh culture in most
countries,21 neither was pesticide usage. In
fact, as was pointed out at the 19th Session of
the International Rice Commission, the case of
the P.R. China with 1.2 million ha under rice-
fi sh farming in a rice area almost exclusively
planted with modern varieties shows that the
use of these varieties does not appear to be a
constraint for rice-fi sh farming (Halwart 1999,
Table 17).
• The use of pesticides will be limited. It is
argued here that reduced use of pesticides is
an advantage to farmers, the communities
and the environment in general. Studies
undertaken in Bangladesh have revealed that
rice-fi sh farmers use less than 50% pesticides
than that used by rice-only farmers (Gupta
et al. 1998). Saturno (1994) observed that
farmers are less likely to use pesticides when
fi sh are stocked in their rice fi elds and still
enjoyed high yields. Kenmore and Halwart
(1998) have pointed out that elimination of
nearly all pesticides in rice fi elds of farmers
who have undergone IPM training results in
a higher biodiversity of frogs, snails, aquatic
insects and others which frequently is used by
farmers in a sustainable manner.
• The farmer has to make a greater initial
investment for installations in the rice fi eld
(higher bunds, drains, shelter holes). The
initial investment is a factor that retards a
widespread adoption of rice-fi sh culture. It is a
disadvantage in increasing a farmer’s fi nancial
exposure, but the potential returns can be very
rewarding and the risks are often low.
• The practice of multiple cropping (several
annual rotations) will be limited because the
fi elds are fl ooded for a shorter period - four
months compared with six to eight months, in
the case of the annual crop. On the contrary,
continuous fl ooding from six to eight months
is advantageous to rice-fi sh farming since it
makes it possible to grow the fi sh to larger
size.
Many constraints that are not inherent to rice-
fi sh farming, but apply to aquaculture and
agriculture in general, such as lack of seeds and
credit facilities, have been identifi ed (Costa-Pierce
and De la Cruz 1992). Some are site-specifi c, for
example the natural fl ooding cycle (Bangladesh,
Cambodia and Vietnam) and poor soils
(Indonesia and Thailand). However, it is argued
that the major constraint to adoption by more
farmers is the fact that rice-fi sh farming is not part
of the mainstream agronomic practice.
11.3 Research and Development
Needs
There is a need to refi ne rice-fi sh farming, where
the thrust is on improving fi sh production
without affecting rice production. De la Cruz et
al. (1992) identifi ed possible areas and topics for
research for various countries. Topics common
to several countries where rice-fi sh farming is
practiced or has high potential are:
• Ecological studies specifi cally on food webs
and nutrient cycle in a rice fi eld ecosystem;
• Determination of the carrying capacity and
optimum stocking densities;
• Development of rice fi eld hatchery and/or
nursery system;
• Development of rice-fi sh farming models
specifi c to different agroclimatic zones;
• Optimum fertilization rates and fertilization
methods;
• Evaluation of new fi sh species for rice fi eld
culture;
• Evaluation of different fi sh species in the
control of rice pests and diseases;
• Development of fi sh aggregating and fi sh
harvesting techniques for rice fi elds; and
• Optimal rice planting patterns for rice-fi sh
farming.
Other topics identifi ed are not necessarily specifi c
to rice-fi sh farming and may be covered by regular
aquaculture research, such as fi sh nutrition and
feed development, or in agronomy, for example
weed ecology and management. Long-term,
“wish list” research includes the development of
new rice varieties for different rice-fi sh systems.
21 With the exception of the Philippines.
66 FAO and The WorldFish Center | Culture of Fish in Rice Fields
Fernando and Halwart (2000) argue that a
systematic approach to fi sh farming development
is needed at irrigation system level which will
alleviate most of the constraints that are met
when trying to promote fi sh farming in rice
fi elds only. One important task is to classify rice-
producing areas for their suitability for rice-fi sh
farming, considering the capacity of the irrigation
infrastructure, general soil characteristics, physical
requirements as well as the socio-economic
situation. The result could serve as a guide as to
where to concentrate greater effort in promoting
rice-fi sh culture. The availability of materials from
China may be useful to fi eld-test some systems for
possible adoption in other countries.
It will be useful if socioeconomic studies are
conducted before and after the introduction/
promotion of rice-fi sh culture. Baseline data
on income status and diet will be important in
assessing the full impact of rice-fi sh technology.
Deepwater rice systems warrant more studies as
such areas could be natural places for fi sh culture.
Low yields of such systems could potentially be
compensated by fi sh yields as Dehadrai (1992)
reported yields of 1 100 kg·ha-1·crop-1 in India
and 650 kg·ha-1 in four months in Bangladesh
(Ali et al. 1993), although the system was not
found fi nancially viable due to the cost of the 4 m
high net enclosure.
The rising sea level may necessitate research into
brackishwater rice-fi sh farming. Penaeid shrimps
grown concurrently with rice in brackish water
as demonstrated in Vietnam (Mai et al. 1992),
and in India, the pokhali and Khazan systems,with
salt-resistant rice are reported to produce
885-2 135 kg·ha-1·crop-1 of giant tiger shrimps
and mullets and 500-2 000 kg·ha-1·crop-1 of
shrimps and perches, respectively. The sawah-
tambak (Indonesia) may be appropriate for
low-lying coastal areas suffering from saltwater
intrusion as it produces 2 000-3 500 kg·ha-1·year-1
of brackishwater species (such as penaeid shrimps,
milkfi sh and seabass). It may also be possible to use
abandoned shrimp farms for rice-shrimp farming,
as many such farms were originally rice fi elds.
11.4 Institutional Policy and
Support Services
11.4.1 Mainstreaming rice-fish farming
People involved in rice production often regard
rice-fi sh farming as a novelty, and standard
literature on plant protection in rice production
(e.g. Heinrichs 1994; Reissig et al. 1986) does not
mention fi sh as a possible bio-control agent or
rice-fi sh culture. To address this, rice-fi sh farming
should be made part of the agriculture curriculum
in universities and colleges, and recognized as a
viable farming system.
If possible the agriculture ministry, or its
equivalent, in rice producing countries should
make integrated rice-fi sh farming part of the
standard agronomic practice so it becomes a
logical and viable option for farmers.
Since IPM is now an accepted approach to pest
control this is a logical entry point for raising fi sh
in rice fi elds. However, suitable curricula for the
Farmer Field Schools still need to be developed.
11.4.2 Popularization of the concept
Many farmers are aware that fi sh can be cultured
with rice, but few realize the advantages. A major
concern is likely to be how to deal with insect
infestations when growing fi sh in the fi elds.
Since governments are often promoting IPM
for rice cultivation, the culture of fi sh should
be considered as part of IPM methods as fi sh
cultivation can be effective in strengthening
other non-chemical IPM strategies (Kamp and
Gregory 1994) and better utilization of resources.
Increased income and a healthy crop of rice
reinforce farmers’ acceptance of non-chemical
IPM and rejection of pesticides (Kenmore and
Halwart 1998).
Rice-fi sh farming should become part of public
awareness so the culture of fi sh in rice fi elds
becomes as integral to rice growing as fertilizer
application. In fact not too long ago, before
the promotion of chemical pesticides, fi sh and
other aquatic organisms were the most natural
thing to have in the fl ood water of rice fi elds.
This continues to be the case for example in
parts of Cambodia, the Lao PDR and other
parts of Southeast Asia where pesticide use is
negligible.
11.4.3 Training and education
Generating public awareness alone is not suffi cient
however. It may lead to frustration if suitable
technologies cannot be delivered. Farmers should
know where to turn for assistance. To do this it
is necessary to train and re-orient agricultural
extension offi cers. Agriculturists rather than
fi sheries offi cers should be targeted for such
67
Prospects and Program for the Future
training since they are the persons who are most
often in contact with the rice farmers.
Beyond short-term training for agricultural
extension offi cers, agricultural school curricula
should include rice-fi sh culture as a viable farming
system, and the role of fi sh in pest management
should also be taught. Textbooks on rice farming
should include sections on rice-fi sh farming. All
those involved in rice production should be made
aware that the advantages of rice-fi sh farming go
beyond the production of fi sh.
11.4.4 Fingerling supply
A vital input in rice-fi sh farming is fi sh seed for
stocking. In countries where aquaculture is not
an important industry, fi ngerlings are scarce and
expensive. There are many issues pertaining to how
to successfully promote fi ngerling production,
but this is common for all aquaculture and
not specifi c to rice-fi sh culture. Any effort to
promote a wider adoption of rice-fi sh farming
needs to be accompanied by developing local
capability in fi ngerling production. This could
be done through the rice farmers themselves as
has been successfully done in Madagascar where
a network of private fi ngerling producers was
set up gradually. As a private producer became
operational, fi ngerling distribution by the
government in that area was discontinued. In the
next step, extension services for rice-fi sh farmers
in the area were included in the marketing
strategy for fi sh seed producers, ranging from
demonstrating their own rice-fi sh operations to
organizing meetings. To achieve this, fi ngerling
producers were trained in marketing methods,
teaching skills, and extension methods. Activities
were supported by a small but highly qualifi ed
group of government extension agents (Van den
Berg 1996).
11.4.5 Financing
Financing may be required since the raising
of dikes and excavation of ponds or trench
refuges may incur extra expenses beyond what
is normally required for rice farming. Often the
amounts involved (US$ 500 or less) are small
enough to fall within the scope of micro-credit.
Even if hundreds of farmers are to be fi nanced
in each locality the total amount involved will
certainly be within the capability of rural banking
facilities to service. The more critical issue is often
to get the fi nancing body to accept this farming
practice as a viable venture, as aquaculture has
had diffi culties in being seen as a low risk farming
option.
68 FAO and The WorldFish Center | Culture of Fish in Rice Fields
Rice-fi sh farming offers tremendous potential
for food security and poverty alleviation in rural
areas. It is an effi cient way of using the same
land resource to produce both carbohydrate and
animal protein concurrently or serially. Water is
similarly used to simultaneously produce the two
basic foodstuffs.
Fish in the rice fi eld has been shown to be capable
of eradicating weeds by eating or uprooting them.
It also devours some insect pests not the least of
which are stemborers. Experience has shown that
the need for chemical pesticides is greatly reduced
and in many instances even eliminated. Fish also
add to the rice fi eld’s fertility and can reduce
fertilizer requirements. Integrating aquaculture
with agriculture results in an effi cient nutrient
use through product recycling since many of the
agricultural by-products can serve as fertilizer
and feed inputs to aquaculture (Willmann et al.
1998). This in turn leads to more fi sh for the
household and can put more cash in the pocket.
An important side effect is a cleaner and healthier
rural environment.
Other economic impacts can be expected. Rice
fi eld modifi cations may need extra labor beyond
what is available within the family, leading to
rural employment. Increased fi ngerling demand
may spur the growth of the hatchery and fi ngerling
production business and all other ancillary
activities, such as making of hapa nets, harvesting
seines, fabrication of hand tools, installation and
repair of pumps, among others. Fish need to be
marketed and perhaps even processed before
marketing. Thus there is a potential to generate
additional employment.
The reality is, however, that the adoption rate
of rice-fi sh farming is very low. China with 1.2
million ha used for rice-fi sh farming is clearly the
world leader, but this fi gure represents only 3.92%
of its irrigated area. Surprisingly, it is outside Asia
where the rice-fi sh farms are extensive relative to
the irrigated rice fi elds. In Egypt, the rice-fi sh farm
area represents 37.4% of the irrigated area and
in Madagascar, 11.75%. Within Southeast Asia,
Thailand is reported to have 2.966 million ha
devoted to rice-fi sh farming and another 25 500
ha related to stocking and managing the fi sheries.
In all the rest of Asia, the adoption rate is only a
little over 1% or there are no statistics available
12. Conclusion
on the extent of rice-fi sh farming. Should the
adoption rate increase to an average 10% of the
irrigated rice fi elds (68.07 million ha), even an
annual yield of only 150 kg·ha-1 would mean
more than 1 million t of fi sh annually. This fi gure
does not include rainfed areas that also have a
potential for fi sh production.
In order to realize this potential, there is a need
for a fundamental shift in attitude towards
rice-fi sh farming in all sectors involved in rice
production, from policy-makers to extension
offi cers and farmers. At present rice-fi sh farming at
best is considered a novelty and at worst a fringe
activity that does not merit serious consideration
in the formulation of national rice production
strategies, and is often relegated to a limited set
of projects. Further, fi shery technologists and
scientists are not the appropriate people to best
reach out to rice farmers, or to whom rice farmers
would listen. The message must be carried by the
rice people.
To integrate fi sheries and agriculture, Willman et
al. (1998) recommend multi-sectoral integration
between various government agencies involved in
river basin and coastal development and various
government agencies that may be involved in
fi sheries and agriculture. However, the authors
also acknowledged the diffi culty involved in such
integration. While ideal, the case of promoting
a more widespread adoption need not involve
too many agencies; in fact it should involve only
those involved in agriculture.
The various sub-sectors in agriculture need to
recognize rice-fi sh farming as a distinct and viable
farming system that farmers can choose to adopt
wherever the physical conditions are appropriate.
If rice-fi sh farming is seen as a viable agronomic
practice, many of the expenses that go into
raising fi sh in rice fi elds will be part of legitimate
expenses where supervised credit is involved.
Fisheries agencies have an important role to
play, in seeing that good quality fi ngerlings are
available at the time required by farmers.
Proper guidelines should also be in place to
safeguard that the fi sh culture component not
be overdone to the detriment of rice production.
With good fi sh production and high prices
farmers tended to enlarge the refuge areas in Viet
69
Conclusion
Nam (Halwart 1998). Purba (1998) concluded
in Indonesia that an increase in fi sh demand and
price would decrease rice production, as the ratio
of the refuge to the rice planting area becomes
excessive. It should be clear that the objective of
raising fi sh with rice is to increase fi sh production
without lowering rice yields.
With such a shift at the top level, agricultural
extension agents can be properly trained to
promote and demonstrate the “new” technology.
In this manner, the popularization of rice-
fi sh integration will not be limited to a few
farmers under a special project, although it
may be initiated in such a manner. Widespread
introduction of rice-fi sh concepts to communities,
coupled with demonstrations in farmers’ own
fi elds, and linking of the rice-fi sh approach with
the IPM Farmer Field Schools (Kenmore and
Halwart 1998) is likely to result in sustained
adoption. The farmers themselves are the most
effective agents of change. For improved contact
with adopters, person-to-person channels are the
best mechanisms to obtain information about
new technologies. These channels include direct
contact with other farmers, extension workers
and technical specialists. In India, about 85%
of the farmers mentioned other farmers as their
sources of information (Librero 1992).
In summary, in order to popularize rice-fi sh
culture, the concept should become part of the
agricultural system rather than the fi sheries
system. The fi sheries agencies will need to put
further efforts in the establishment of viable
national fi sh seed production and distribution
system operated by the private sector so that
fi ngerlings of the desired species are readily
available to the farmers. Only then can more fi sh
be found in the rice fi elds.
70 FAO and The WorldFish Center | Culture of Fish in Rice Fields
Abdel-Malek, S.A. 1972. Food and feeding habits of some
Egyptian fi shes in Lake Qurun. I. Tilapia zillii (Gerv.)
B. according to different length groups. Bull. Inst.
Oceanogr. Fish. Cairo. 2:204-123.
Ahmed, M., M.A.P. Bimbao and R.C. Sevilleja. 1992. The
economics of rice-fi sh in Asian mixed farming system
- a case study of the Philippines, p. 207-216. In C.R. De
la Cruz, C. Lightfoot, B.A. Costa-Pierce, V.R. Carangal and
M.P. Bimbao (eds.) Rice-fi sh research and development
in Asia. ICLARM Conf. Proc. 24, 457 p.
Ali, A.B. 1990. Rice-fi sh farming in Malaysia: a resource
optimization. Ambio 19(8): 404-408.
Ali, A.B. 1992. Rice-fi sh farming in Malaysia: past, present
and future, p. 69-76. In C.R. De la Cruz, C. Lightfoot, B.A.
Costa-Pierce, V.R. Carangal and M.P. Bimbao (eds.) Rice-
fi sh research and development in Asia. ICLARM Conf.
Proc. 24, 457 p.
Ali, A.B. 1998. Rice agroecosystem and the maintenance
of biodiversity, p. 25-36. In B.M. Nashriyak, N.K. Ho, B.S.
Ismael, A.B. Ali and K.Y. Lun (eds.) Rice agroecosystem
of the Muda Irrigation Scheme, Malaysia. Mintmada
Malaysia, 250 p.
Ali, M. H., N.I. Miah and N.U. Ahmed. 1993. Experiences in
deepwater rice-fi sh culture. Rice Farming Systems
Division, Bangladesh Rice Research Institute, Gazipur,
Bangladesh. Mimeo, 28 p.
Anonymous. 1971. Algicide for direct-seeded rice. IRRI
Reporter 2/71:2-3.
Arai, M. 1963. Theory of rice crop protection. I. Weed
control, p. 228-251. In M. Matsubayashi, R. Ito, T.
Nomoto, T. Takase and N.Yamada (eds.) Theory and
practice of growing rice. Fuji Publs. Co., Ltd., Tokyo,
Japan.
Ardiwinata, R.O. 1953. Common carp culture in rice fi elds.
Vorhink Van Hoeve Vol. V. Bandung’s Gravenhage (in
Indonesian with English summary).
Ardiwinata, R.O. 1957. Fish culture in the rice fi elds in
Indonesia. Proc. Indo-Pacif. Fish. Coun. 7:119-154.
Balzer, T., P. Balzer and S. Pon. 2002. Traditional use and
availability of aquatic biodiversity in rice-based
ecosystems – I. Kampong Thom Province, Kingdom
of Cambodia. In M. Halwart and D. Bartley (eds.) FAO
Inland Water Resources and Aquaculture Service. CD
ROM, ISBN 92-5-104820-7. FAO, Rome. Available at
ftp://ftp.fao.org/fi/CDrom/AwarnessAgrBiodiv/
default.htm.
Beaumont, J.J. and K.J. Yost. 1999. Phenoxy herbicides and
cancer. In Agricultural chemicals and cancer website.
Univ. of Calif. Center for Occup. and Env. Health, revised
2 March 1999.
Bowen, S.H. 1979. A nutritional constraint in detritivory
by fi shes: the stunted population of Sarotherodon
13. References
mossambicus in Lake Sibaya, South Africa. Ecol.
Monogr. 49(1):17-31.
Bowen, S.H. 1982. Feeding, digestion and growth -
qualitative considerations, p. 141-156. In R.S.V. Pullin
and R.H. Lowe-McConnell (eds.) The biology and
culture of tilapias. Proc. Int. Conf. on the Biol. and Cult.
of Tilapias, ICLARM Conf. Proc. 7, 432 p.
Bowen, S.H. 1980. Detrital nonprotein amino acids are
the key to rapid growth of tilapia in Lake Valencia,
Venezuela. Science. 207:1 216-1 218.
Boyd, C.E. 1982. Water quality management for pond fi sh
culture. Elsevier, Amsterdam, 318 p.
Boyd, C.E. 1979. Water quality in warmwater fi sh ponds.
Auburn University Agricultural Experiment Station,
359 p.
Breine, J.J., G.G. Teugels and F. Ollevier. 1995. Preliminary
results of integrated fi sh farming at the fi sh breeding
research station, Foumban, Cameroon, p. 381-386. In
J.J. Symoens and J.C. Micha (eds.) Proceedings of the
Seminar on “The Management of Integrated Agro-
Piscicultural Ecosystems in Tropical Areas”, 16-19 May
1994, Brussels. Tech. Centre for Agric. and Rural Co-op.
(CTA) Royal Academy of Overseas Sciences, Brussels,
587 p.
Buddington, R.K. 1979. Digestion of an aquatic macrophyte
by Tilapia zillii. J. Fish. Biol. 15(3):449-456.
Cagauan, A.G. 1995. Overview of the potential roles of
pisciculture on pest and disease control and nutrient
management in rice fi elds, p. 203-244. In J.J. Symoens
and J.C. Micha (eds.) Proceedings of the Seminar on
“The Management of Integrated Agro-Piscicultural
Ecosystems in Tropical Areas”, 16-19 May 1994, Brussels.
Tech. Centre for Agric. and Rural Co-op. (CTA) Royal
Academy of Overseas Sciences, Brussels, 587 p.
Cagauan, A.G., C.R. De la Cruz and C. Lightfoot. 1993.
Nitrogen models of lowland irrigated ecosystems with
and without fi sh using ECOPATH. Paper presented
during the 3rd Asian Regional Workshop on Integrated
Rice-Fish Research and Development, 6-11 June 1993,
Sukamandi Research Institute for Food Crops, West
Java, Indonesia.
Cagauan, A.G. and R.G. Arce. 1992. Overview of pesticide
use in rice-fi sh farming in Southeast Asia, p. 217-234.
In C.R. De la Cruz, C. Lightfoot, B.A. Costa-Pierce, V.R.
Carangal and M.P. Bimbao (eds.) Rice-fi sh research and
development in Asia. ICLARM Conf. Proc. 24, 457 p.
Cai, R., D. Ni and J. Wang. 1995a. Rice-Fish culture in China:
the past, present and future, p. 3-14. In K.T. MacKay (ed.)
Rice-fi sh culture in China. International Development
Research Centre (IDRC) Ottawa, Canada, 276 p.
Caulton, M.S. 1976. The importance of pre-digestive food
preparation to Tilapia rendalli Boulenger when feed
71
References
on aquatic macrophytes. Trans. Rhod. Sci. Assoc. 57:
22-28.
Caulton, M.S. 1977. A quantitative assessment of the
daily ingestion of Panicm repens L. by Tilapia rendalli
Boulenger (Cichlidae) in Lake Kariba. Trans. Rhod. Sci.
Assoc. 58(6):38-42.
Chacko, P.I. and S.V. Ganapati. 1952. Fish culture in paddy
fi elds. Ind. Comm. J. Madras No. 8:3.
Chapman, G. 1992. Fry nursery techniques in the rice-fi sh
systems of Northeast Thailand, p. 130-144. In C.R. De la
Cruz, C. Lightfoot, B.A. Costa-Pierce, V.R. Carangal and
M.P. Bimbao (eds.) Rice-fi sh research and development
in Asia. ICLARM Conf. Proc. 24, 457 p.
Chen, D., H. Ying and M. Shui. 1995. Rice-azolla-fi sh in rice
fi elds, p. 169-176. In K.T. MacKay (ed.) Rice-fi sh culture
in China. International Development Research Centre
(IDRC) Ottawa, Canada, 276 p.
Chen, H.L., A.T. Charles and B.T. Hu. 1998. Chinese integrated
fi sh farming, p. 97-109. In J.A. Mathias, A.T. Charles and
B. T. Hu (eds.) Integrated Fish Farming. Proc. Workshop
on Integrated Fish Farming, 11-15 Oct. 1994, Wuxi,
Jiangsu Prov., P.R. China. CRC Press, 420 p.
Chen, T.P. 1954. The culture of tilapia in rice paddies in
Taiwan. Jt. Comm. Rural Recontr. China Fish. Ser. 2. 29 p.
Chien, Y.H. 1978. Double cropping rice Oryza sativa and
crawfi sh Procambarus clarkii (Girard). Louisiana State
University and Agricultural and Mechanical College,
Baton Rouge LA, USA. 85 p. M.S. thesis.
Chikafumbwa, F.J.K. 1996. Use of terrestrial plants in
aquaculture in Malawi, p. 175-182. In R.S.V. Pullin, J.
Lazard, M. Legendre, J.B. Amon Kothias and D. Pauly
(eds.) The Third International Symposium on Tilapia in
Aquaculture. ICLARM Conf. Proc. 41, 575 p.
Coche, A.G. 1967. Fish culture in rice fi elds. A worldwide
synthesis. Hydrobiologia 30:11-44.
Costa-Pierce, B. and C.R. De la Cruz. 1992. Working group
reports on on-station research methodology, p. 415-
424. In C.R. De la Cruz, C. Lightfoot, B.A. Costa-Pierce, V.R.
Carangal and M.P. Bimbao (eds.) Rice-fi sh research and
development in Asia. ICLARM Conf. Proc. 24, 457 p.
Costa-Pierce, B.A. 1992. Rice-Fish systems as intensive
nurseries, p. 117-130. In C.R. De la Cruz, C. Lightfoot,
B.A. Costa-Pierce, V.R. Carangal and M.P. Bimbao (eds.)
Rice-fi sh research and development in Asia. ICLARM
Conf. Proc. 24, 457 p.
Darvin, L.C. 1992. Status of rice fi elds as hatcheries/
nurseries for tilapia in the Philippines, p. 145-150.
In C.R. De la Cruz, C. Lightfoot, B.A. Costa-Pierce, V.R.
Carangal and M.P. Bimbao (eds.) Rice-fi sh research and
development in Asia. ICLARM Conf. Proc. 24, 457 p.
Dehadrai, P.V. 1992. Opportunities for women in rice-fi sh
culture. In C.R. De la Cruz, C. Lightfoot, B.A. Costa-Pierce,
V.R. Carangal and M.P. Bimbao (eds.) Rice-fi sh research
and development in Asia. ICLARM Conf. Proc. 24, 367 p.
De la Bretonne, L.W. Jr. and R.P. Romaire. 1990. Crawfi sh
culture: selection, pond construction and water quality.
SRAC Publication No. 240, January 1990. Southern
Regional Aquaculture Center, Stoneville, MS, USA.
De la Cruz, C.R. 1980. Integrated agriculture-aquaculture
systems in the Philippines with two cases on
simultaneous and rotational rice-fi sh culture, p. 209-
223. In R.S.V. Pullin and Z.H. Shehadeh (eds.) Proc. of
the ICLARM-SEARCA Conf. on Integrated Agriculture-
Aquaculture Farming Systems, 6-9 August 1979,
Manila, Philippines.
De la Cruz, C.R, C. Lightfoot, B.A. Costa-Pierce, V.R. Carangal,
and M.P. Bimbao (eds.) 1992. Rice-fi sh research and
development in Asia. ICLARM, Manila, Philippines,
457 p.
De la Cruz, C.R. 1979. Integrated agriculture-aquaculture
farming systems in the Philippines. ICLARM/SEARCA
Conf. on Integration of Agriculture-Aquaculture
Farming Systems, 6-9 August 1979, Manila, Philippines,
29 p.
Demaine, H. and M. Halwart 2001. An overview of rice-
based small-scale aquaculture, p. 189-197. In IIRR,
IDRC, FAO, NACA and ICLARM. Utilizing different
aquatic resources for livelihoods in Asia: a resource
book. International Institute of Rural Reconstruction,
International Development Research Centre, Food
and Agriculture Organization of the United Nations,
Network of Aquaculture Centers in Asia-Pacifi c, and
International Center for Living Aquatic Resources
Management, 416 p.
Denny, P., D.W. Bowker and R.G. Bailey. 1978. The importance
of littoral epiphyton as food for commercial fi sh in the
recent African man-made lake, Nyumba ya Mungu
reservoir, Tanzania. Biol. J. Linn. Soc. 10(1):139-150.
Deomampo, N.R. 1998. Farming systems, marketing and
trade for sustainable aquaculture, p. 201-217. Annex
IV-5 of ADB/NACA, 1998. Aquaculture sustainability
and the environment. Report on a regional study
and workshop on aquaculture sustainability and the
environment. Bangkok, Thailand. Asian Development
Bank and Network of Aquaculture Centers in Asia-
Pacifi c, 492 p.
Dewan, S. 1992. Rice-fi sh farming in Bangladesh: past,
present and future, p. 11-16. In C.R. De la Cruz, C.
Lightfoot, B.A. Costa-Pierce, V.R. Carangal and M.P.
Bimbao (eds.) Rice-fi sh research and development in
Asia. ICLARM Conf. Proc. 24, 457 p.
Diallo, A. 1998. Integrated farming: a new approach in
Basse Casamanche, Senegal, p. 257-264. In J.A. Mathias,
A.T. Charles and B. T. Hu (eds). 1998. Integrated Fish
Farming. Proc. Workshop on Integrated Fish Farming,
11-15 Oct. 1994, Wuxi, Jiangsu Prov., P.R. China. CRC
Press, 420 p.
Diana, J.S., C.K. Lin and Y. Yi. 1996. Timing of supplemental
feeding for tilapia production. Journal of the World
Aquaculture Society 27 (4):410-419.
Edra, R. 1991. Introduction of the golden snail and
escalation of its infestation in Philippine ricelands
72 FAO and The WorldFish Center | Culture of Fish in Rice Fields
(Abstract), p. 11. In B.O. Acosta and R.S.V. Pullin
(eds.) 1991. Environmental impact of the golden
snail (Pomacea sp.) on rice farming systems in the
Philippines. ICLARM Conf. Proc. 28, 34 p.
Essawi, A.K. and M.M. Ishak. 1975. A preliminary study on
raising carp in rice fi elds of Sakha Research Station,
Arab Republic of Egypt. CIFA/75/SE1. Paper presented
at the FAO/CIFA Symposium on Aquaculture in Africa,
30 Sept.-6 Oct. 1975, Accra, Ghana.
Fagi, A.M., S. Suriapermana and I. Syamsiah. 1992. Rice-fi sh
farming research in lowland areas: The West Java case,
p. 273-286. In C.R. De la Cruz, C. Lightfoot, B.A. Costa-
Pierce, V.R. Carangal and M.P. Bimbao (eds.) Rice-fi sh
research and development in Asia. ICLARM Conf. Proc.
24, 457 p.
FAO. 1957. Fish culture in rice fi elds, p. 193-206. In A
preliminary review and annotated bibliography from
Section II and III of Proc. Indo-Pacif. Fish. Coun. 7. FAO,
Rome.
FAO. 1998. Fish-farming in Vietnamese rice fi elds fi ghts
golden apple snail pest. News & Highlights, FAO Web
Page http://www.fao.org/NEWS/1998/980410-e.htm
FAO. 2000. Small ponds make a big difference – integrating
fi sh with crop and livestock farming. FAO, Rome, 30 p.
FAO. 2002. Report of the International Rice Commission,
Twentieth Session. Bangkok, Thailand, 23-26 July 2002.
FAO, Rome, 46 p.
FAO/ICLARM/IIRR. 2001. Integrated agriculture-
aquaculture. A primer. FAO Fisheries Technical Paper
407. FAO, Rome, Italy, 149 p.
Fedoruk, K. and W. Leelapatra. 1992. Rice fi eld fi sheries in
Thailand, p. 91-104. In C.R. De la Cruz, C. Lightfoot, B.A.
Costa-Pierce, V.R. Carangal and M.P. Bimbao (eds.) Rice-
fi sh research and development in Asia. ICLARM Conf.
Proc. 24, 457 p.
Fermin, F.V. 1992. The adaptation of rice-fi sh farming
technology: the case of Mang Isko in Cavite,
Philippines, p. 333-338. In C.R. De la Cruz, C. Lightfoot,
B.A. Costa-Pierce, V.R. Carangal and M.P. Bimbao (eds.)
Rice-fi sh research and development in Asia. ICLARM
Conf. Proc. 24, 457 p.
Fernando, C.H. 1993. Rice fi eld ecology and fi sh culture
– an overview. Hydrobiologia 259:91-113.
Fernando, C.H., J.I. Furtado and R.P. Lim. 1979. The aquatic
fauna of the world’s rice fi elds: a general introduction
and bibliography of references to the aquatic fauna of
rice fi elds with an introduction to the ecology of the
rice fi eld fauna. Wallacean Suppl. 2 March, 105 p.
Fernando, C.H. and M. Halwart. 2000. Possibilities for the
integration of fi sh farming into irrigation systems.
Fisheries Management and Ecology 7:45-54.
Fernando, C.H. and M. Halwart. 2001. Fish farming in
irrigation systems: Sri Lanka and global view. Sri Lanka
J. Aquat. Sci. 6:1-74.
Fillery, I.R.P., J.R. Simpson and S.K. de Datta. 1984. Infl uence
of fi eld environment and fertilizer management on
ammonia loss from a fl ooded soil. Soil Science Soc.
Am. J. 48(4):914-920.
Fish, G.R. 1955. The food of tilapia in East Africa. Uganda J.
19:85-89.
Funge-Smith, S.J. 1999. Small-scale rural aquaculture in the
Lao PDR. FAO Aquaculture Newsletter, Nos. 22 and 23.
Ghosh, A. 1992. Rice-fi sh farming development in India:
past, present and future, p. 27-43. In C.R. De la Cruz,
C. Lightfoot, B.A. Costa-Pierce, V.R. Carangal and M.P.
Bimbao (eds.) Rice-fi sh research and development in
Asia. ICLARM Conf. Proc. 24, 457 p.
Greenland, D.J. 1997. The sustainability of rice farming. CAB
International, UK, 273 p.
Gregory, R. 1996. Rice fi eld fi sheries handbook. Cambodia-
IRRI-Australia Project. Phnom Penh, 37 p.
Gregory R. and H. Guttman, H. 1996. Capture or culture?
ILEIA Newsl. 12 (2):20-21, July 1996.
Guillén, F.A. 1990. Current status of integrated rice-fi sh
culture in Latin America and the Caribbean. Int. Rice.
Comm. Newsl. 39:196-207.
Gupta, M.V. 1998. Social and policy issues involved in
adoption of integrated agriculture-aquaculture-
livestock production systems in Bangladesh, p. 229-
244. In J.A. Mathias, A.T. Charles and H. Baotong (eds.)
Integrated fi sh farming. CRC Press, Boca Raton, New
York.
Gupta, M.V. and M.A. Rab. 1994. Adoption and economics
of silver barb (Puntius gonionotus) culture in seasonal
waters in Bangladesh. ICLARM Tech. Rep. 41, 39 p.
Gupta, M.V., J.D. Sollows, M. Abdul Mazid, A. Rahman, M.G.
Hussain and M.M. Dey. 1998. Integrating aquaculture
with rice farming in Bangladesh: feasibility and
economic viability, its adoption and impact. ICLARM
Tech. Rep. 55, 90 p.
Gupta, M.V., M. Abdul Mazid, Md. S. Islam, M. Rahman, and
M.G. Hussain. 1999. Integration of aquaculture into
the farming systems of the fl ood-prone ecosystems
of Bangladesh: an evaluation of adoption and impact.
ICLARM Tech. Rep. 56, 32 p.
Guttman, H. 1999. Rice fi eld fi sheries – a resource for
Cambodia. NAGA, The ICLRAM Q. 22(2):11-15.
Halwart, M. 1994a. Fish as biocontrol agents in rice. Margraf
Verlag, Weikersheim, Germany, 169 p.
Halwart, M. 1994b. The golden apple snail Pomacea
canaliculata in Asian rice-farming systems: present
impact and future threat. International Journal of Pest
Management 40(2):199-206.
Halwart, M. 1998. Trends in rice-fi sh farming. FAN, the FAO
Aquaculture Newsletter, April 1998. No. 18.
Halwart, M. 1999. Fish in rice-based farming systems
–trends and prospects, p. 130 - 141. In Dat van Tran
(ed.) International Rice Commission - assessment and
orientation towards the 21st century. Proceedings of
the 19th Session of the International Rice Commission,
7-9 September 1998, Cairo, Egypt, 260 p.
Halwart, M. 2001a. Fish as a component of integrated
73
References
pest management in rice production, p. 115-117. In
FAO/ICLARM/IIRR Integrated agriculture-aquaculture
– a primer. FAO Fish. Tech. Pap. 407, 149 p.
Halwart, M. 2001b. Fish as biocontrol agents of vectors and
pests of medical and agricultural importance, p. 70-75.
In IIRR, IDRC, FAO, NACA and ICLARM. Utilizing different
aquatic resources for livelihoods in Asia: a resource
book. International Institute of Rural Construction,
International Development Research Center, Food
and Agriculture Organisation of the United Nations,
Network of Aquaculture Centers in Asia-Pacifi c and
International Center for Living Aquatic Resources
Management, 416 p.
Halwart, M. 2003a. Recent initiatives on the availability
and use of aquatic organisms in rice-based farming,
p. 195-206. In Proceedings of the 20th Session of the
International Rice Commission. Bangkok, Thailand, 23-
26 July 2002. FAO, Rome.
Halwart, M. 2003b. Ricefi eld fi sheries and rice-based
aquaculture – underestimated and undervalued
resources, p. 36-47. In FAO Inland Water Resources
and Aquaculture Service. Review of the State of World
Fishery Resources: Inland Fisheries. FAO Fisheries
Circular No. 942, Rev. 1. Rome, FAO, 60 p.
Halwart, M., M. Borlinghaus and G. Kaule 1996. Activity
pattern of fi sh in rice fi elds. Aquaculture 145:159-170.
Halwart, M., M.C. Viray and G. Kaule 1998. Cyprinus carpio
and Oreochromis niloticus as biological control agents
of the Golden Apple Snail Pomacea canaliculata
– effects of predator size, prey size and prey density.
Asian Fisheries Science 11:31-42.
Halwart, M., S. Funge-Smith and J. Moehl. 2002. The role
of aquaculture in rural development, p. 71-83. In FAO
Inland Water Resources and Aquaculture Service.
Review of the state of world aquaculture. FAO Fisheries
Circular No. 886 (Rev 2). FAO, Rome, Italy.
Hauser, W.J. and W.F. Lehman 1976. Raising fi sh for weed
control subject of study. The Rice Journal 79(5):10-11,
22, 24.
He, Q.H., H.Q. Sheng and G.H. Xu. 1998. The use of
supplementary feeds for semi-intensive fi sh culture
practices in reservoirs in China, p. 307-324. In J.A.
Mathias, A.T. Charles and B.T. Hu (eds.) Integrated Fish
Farming. Proc. Workshop on Integrated Fish Farming,
11-15 Oct. 1994, Wuxi, Jiangsu Prov., P.R. China, CRC
Press, 420 p.
Heckman, C.W. 1979. Rice fi eld ecology in Northeastern
Thailand. Monographs Biologicae 34:1-228. Dr. W. Junk
Publishers, The Hague.
Heidenreich, A. and M. Halwart. 1997. Fish farming and
other Integrated Pest Management strategies against
Pomacea snails in rice fi elds: a computer simulation, p.
287-290. In Y.Q. Zhou, H.Q. Zhou, C.Q. Yao, Y. Lu, F.Y. Hu,
H. Cui and F.H. Din (eds.) The Fourth Asian Fisheries
Forum. Asian Fisheries Society, Manila, Philippines.
Heidenreich, A., Poethke, H.J., Halwart, M. and Seitz,
A. 1997. Simulation der Populationsdynamik von
Pomacea canaliculata (Prosobranchia) zur Bewertung
von Managementmassnahmen. In J. Pfadenhauer
(ed.) Verhandlungen der Gesellschaft für Ökologie.
Proceedings of the Meeting of the German Society for
Ecology 27: 441-446.
Heinrichs, E.A. (ed.) 1994. Biology and management of rice
insects. IRRI/Wiley Eastern Limited, New Delhi, India,
779 p.
Hendarsih, S., S. Suriapermana, A. Fagi and I. Manwan. 1994.
Potential of fi sh in rice-fi sh culture as a biological
control agent of rice pests (Abstract), p. 32. In C.R. De
la Cruz (ed.) Role of fi sh in enhancing rice fi eld ecology
and in integrated pest management. Summary report
of the 3rd Asian Regional Rice-Fish Farming Research
and Development Workshop, 6-11 June 1993,
Sukamandi Research Institute for Food Crops, West
Java, Indonesia. ICLARM Conf. Proc. 43, 50 p.
Horskotte-Wesseler, G. 1999. Socio-economics of rice-
aquaculture and IPM in the Philippines: synergies,
potentials and problems. ICLARM Tech. Rep. 57, 225 p.
IIRR, IDRC, FAO, NACA and ICLARM. 2001. Utilizing
different aquatic resources for livelihoods in Asia: a
resource book. IIRR, Cavite, Philippines. International
Development Research Center, Food and Agriculture
Organization of the United Nations, Network of
Aquaculture Centers in Asia-Pacifi c and International
Center for Living Aquatic Resources Management,
416 p.
IRRI. 1993. Rice research in a time of change. IRRI, Los
Baños, Laguna, Philippines.
Israel, D.C., R.C. Sevilleja, A.V. Circa and R.D. Cocio. 1994.
Rice-fi sh culture: an output risk programming analysis.
Freshwater Aquaculture Center, Central Luzon State
University, Muñoz, Nueva Ecija, Philippines.
Kamp, K. and R. Gregory. 1994. Fish cultivation as a means
to increase profi tability from rice fi elds: implications
for integrated pest management, p. 29. In C.R. De la
Cruz (ed.) Role of fi sh in enhancing rice fi eld ecology
and in integrated pest management. Summary
report of the Third Asian Regional Rice-Fish Farming
Research and Development Workshop, 6-11 June
1993, Sukamandi Research Institute for Food Crops,
West Java, Indonesia, 50 p.
Kenmore, P. and M. Halwart 1998. Functional
agrobiodiversity, Integrated Pest Management, and
aquatic life management in rice. In Proceedings of the
FAO/CBD International Workshop on Opportunities,
Incentives, and Approaches for the Conservation
and Sustainable Use of Biodiversity in Agricultural
Ecosystems and Production Systems. FAO, Rome, Italy.
Khoo, K.H. and E.S.P. Tan. 1980. Review of rice-fi sh culture
in Southeast Asia, p. 1-14. In R.S.V. Pullin and Z.H.
Shehadeh (eds.) Proc. of the ICLARM-SEARCA Conf. on
Integrated Agriculture-Aquaculture Farming Systems,
6-9 August 1979, Manila, Philippines, 258 p.
74 FAO and The WorldFish Center | Culture of Fish in Rice Fields
Kim, B.H., H.D. Kim and Y.H. Kim. 1992. Rice-fi sh farming
systems and future prospects in Korea, p. 63-68. In C.R.
De la Cruz, C. Lightfoot, B.A. Costa-Pierce, V.R. Carangal
and M.P. Bimbao (eds.) Rice-fi sh research and
development in Asia. ICLARM Conf. Proc. 24, 457 p.
Koesoemadinata, S. 1980. Pesticides as a major constraint
to integrated agriculture-aquaculture farming
systems, p. 45-51. In R.S.V. Pullin and Z.H. Shehadeh
(eds.) Proc. of the ICLARM-SEARCA Conf. on Integrated
Agriculture-Aquaculture Farming Systems, 6-9 August
1979, Manila, Philippines, 258 p.
Koesoemadinata, S. and B.A. Costa-Pierce. 1992.
Development of rice-fi sh farming in Indonesia: past,
present and future, p. 45-62. In C.R. De la Cruz, C.
Lightfoot, B.A. Costa-Pierce, V.R. Carangal and M.P.
Bimbao (eds.) Rice-fi sh research and development in
Asia. ICLARM Conf. Proc. 24, 457 p.
Koesoemadinata, S. and Djadjaredja. 1976. Some aspects
on the regulation of agricultural use of pesticides in
Indonesia with reference to their effects on inland
fi sheries. Inland Fisheries Research Institute Cont. 3, 14 p.
Kuronoma, K. 1980. Carp culture in Japanese rice fi elds.
(Edited and condensed by R.S.V. Pullin from the
author’s original work published in 1954), p. 167-174.
In R.S.V. Pullin and Z.H. Shehadeh (eds.) 1980. Proc. of
the ICLARM-SEARCA Conf. on Integrated Agriculture-
Aquaculture Farming Systems, Manila, Philippines, 6-9
August 1979, 258 p.
Le, T.L. 1999. Small-scale aquaculture in the context of
rural livelihood development in Vietnam. FAO/NACA
Expert Consultation on “Sustainable Aquaculture for
Rural Development”, 25-31 March 1999, Chiang Mai,
Thailand.
Li, D., N. Wu and T. Zhou. 1995. Effect of fi sh on growth and
development of rice, p. 209-212. In K.T. MacKay (ed.)
Rice-fi sh culture in China. International Development
Research Centre (IDRC), Ottawa, Canada, 276 p.
Li, K. 1988. Rice-fi sh culture in China: A review. Aquaculture
71:173-186.
Li, K. 1992. Rice-fi sh farming in China: past, present and
future, p. 17-26. In C.R. de la Cruz, C. Lightfoot, B.A.
Costa-Pierce, V.R. Carangal and M.P. Bimbao (eds.) Rice-
fi sh research and development in Asia. ICLARM Conf.
Proc. 24, 457 p.
Li, K. 1998. Rice aquaculture systems in China: a case of rice-
fi sh farming from protein crops to cash crops. In E.L.
Foo and T. Della Senta (eds.) Integrated bio-systems
in zero-emissions applications. Proc. of the Internet
Conf. on Integr. Biosystems. Microbial Resources
Center (MIRCEN), Unesco. http://www.ias.unu.edu/
proceedings/icibs. 14 p.
Li, K. and Y. Pan. 1992. Ricefi elds as fi sh nurseries and
growout systems in China, p. 151-164. In C.R. De la
Cruz, C. Lightfoot, B.A. Costa-Pierce, V.R. Carangal and
M.P. Bimbao (eds.) Rice-fi sh research and development
in Asia. ICLARM Conf. Proc. 24, 457 p.
Li, X., W. Huaixun and Z. Yontai. 1995b. Economic and
ecological benefi ts of rice-fi sh culture, p. 129-138.
In K.T. MacKay (ed.) Rice-fi sh culture in China.
International Development Research Centre (IDRC),
Ottawa, Canada, 276 p.
Librero, A.R. 1992. Technology transfer in rice, corn
and wheat in Asia and the Pacific, p. 19-67 In
Technology transfer for small farmers in Asia.
Report of the APO Study Meeting on Transfer of
Technology to Small Farmers held in India, 10-19
Dec. 1990. Asian Productivity Organization, Tokyo,
Japan, 307 p.
Lightfoot, C., P. Pingali and L. Harrington. 1993. Beyond
romance and rhetoric: sustainable agriculture and
farming systems research, NAGA, ICLARM Q. 16(1):
17-18.
Lightfoot, C., A.A. van Dam and B.A. Costa-Pierce. 1992.
What’s happening to the rice yields in rice-fi sh
systems?, p. 177-184. In C.R. De la Cruz, C. Lightfoot,
B.A. Costa-Pierce, V.R. Carangal and M.P. Bimbao (eds.)
Rice-fi sh research and development in Asia. ICLARM
Conf. Proc. 24, 457 p.
Lin, X., L. Zhang and G. He. 1995. Economic analysis of rice-
fi sh culture, p. 247-252. In K.T. MacKay (ed.) Rice-fi sh
culture in China. International Development Research
Centre (IDRC), Ottawa, Canada, 276 p.
Little, D.C., P. Surintaraseree and N. Innes-Taylor, 1996. Fish
culture in rainfed rice fi elds of northeast Thailand.
Aquaculture 140 (1996):295-321.
Liu, C.C. 1995. Rice-azolla-fi sh cropping system, p. 210-
208. In K.T. MacKay (ed.) Rice-fi sh culture in China.
International Development Research Centre (IDRC)
Ottawa, Canada, 276 p.
Macdonagh, E.J. 1946. Piscicultura del Pejerrey en el arrozal
de la Faculdad de Agronomia de la Plata. Rev. Fac.
Agron. La Plata 26:33-50.
MacKay, K.T. 1992. Why do rice-fi sh research on farms?,
p. 393-397. In C.R. De la Cruz, C. Lightfoot, B.A. Costa-
Pierce, V.R. Carangal and M.P. Bimbao (eds.) Rice-fi sh
research and development in Asia. ICLARM Conf. Proc.
24, 457 p.
MacKay, K.T. (ed.) 1995. Rice-fi sh culture in China.
International Development Research Centre (IDRC),
Ottawa, Canada, 276 p.
Magulama, E.E. 1990. Effect of planting pattern in rice-fi sh
production system. University of the Philippines at Los
Baños. 94 p. M.S. thesis.
Mai, T.D., T.D. Le, K.S. Dang, N.M. Pham and N.D. Nguyen.
1992. Rice fi eld aquaculture systems in the Mekong
Delta, Vietnam: potential and reality. In C.R. De la Cruz,
C. Lightfoot, B.A. Costa-Pierce, V.R. Carangal and M.P.
Bimbao (eds.) Rice-fi sh research and development in
Asia. ICLARM Conf. Proc. 24, 457 p.
Man, H.S.H. and I.J. Hodgkiss. 1977. Studies on the ichthyo-
fauna in Plover Cove Reservoir, Hong Kong: feeding
and food relations. J. Fish. Biol.11:1-13.
75
References
Manna, G. B., M.S. Chaudhury and A.R. Bedekar. 1969.
Control of weeds in rice fi elds. Oryza 6(2):90-94.
Meien, V.A. 1940. Fish farming in rice fi elds. (Russian). Food
Ind. Leningrad and Moscow, 96 p.
Metz, H.C. (ed.) 1994. Country Study - Madagascar. Federal
Research Division. Library of Congress, Washington,
DC, USA. (http//lcweb2.loc.gov/frd/cs/mgtoc.html)
Meusch, E., J. Yhoung-Aree, R. Friend, and S. Funge-
Smith. 2003. The role and nutritional value of
aquatic resources in the livelihoods of rural people
– a participatory assessment in Attapeu Province, Lao
PDR. FAO and IUCN, Bangkok, Thailand, 34 p.
Micha, J.C., J.P. Descy and F. Laviolette. 1996. Consumption
of phytoplankton by Oreochromis niloticus in Lake
Muhazi (Rwanda). (Abstract), p. 346 In R.S.V. Pullin, J.
Lazard, M. Legendre, J.B. Amon Kothias and D. Pauly
(eds.) The Third International Symposium on Tilapia in
Aquaculture. ICLARM Conf. Proc. 41, 575 p.
Moehl, J.F., I. Beernaerts, A.G. Coche, M. Halwart and V.O.
Sagua. 2001. Proposal for an African network on
integrated irrigation ands aquaculture. Proceedings
of a workshop held in Accra, Ghana, 20-21 September
1999. Rome, FAO, 75 p.
Moody, K. 1992. Fish-crustacean-weed interactions, p. 185-
192. In C.R. De la Cruz, C. Lightfoot, B.A. Costa-Pierce, V.R.
Carangal and M.P. Bimbao (eds.) Rice-fi sh research and
development in Asia. ICLARM Conf. Proc. 24, 457 p.
Moriarty. C.M. and D.J.W. Moriarty. 1973. Quantitative
estimation of the daily ingestion of phytoplankton
by Tilapia nilotica and Haplochromis nigripinnis in Lake
George, Uganda. J. Zool. 171:15-24.
Mukhopadhyay, P.K., D.N, Das and B. Roy. 1992. On-farm
research in deepwater rice-fi sh culture in West Bengal,
India, p. 255-272. In C.R. De la Cruz, C. Lightfoot, B.A.
Costa-Pierce, V.R. Carangal and M.P. Bimbao (eds.) Rice-
fi sh research and development in Asia. ICLARM Conf.
Proc. 24, 457 p.
Munro, J.L. 1967. The food of a community of East African
freshwater fi shes. J. Zool. 151(3):389-415.
Naik, I.U. 1973. Studies on Tilapia mossambica Peters in
Pakistan. Agric. Pak. 24(1):47-76.
Nalim, S. 1994. The impact of fi sh in enhancing rice
fi eld ecosystem. In C.R. De la Cruz (ed.) Role of fi sh
in enhancing rice fi eld ecology and in integrated
pest management. Summary report of the 3rd
Asian Regional Rice-Fish Farming Research and
Development Workshop, 6-11 June 1993, Sukamandi
Research Institute for Food Crops, West Java, Indonesia.
ICLARM Conf. Proc. 43, 50 p.
Nandeesha, M.C. and G. Chapman. 1999. Aquaculture
development in Bangladesh through capacity
building: experiences of CARE. Paper presented at the
Joint FAO/NACA Consultation for the Development of
a Regional Collaborative Programme on Sustainable
Aquaculture for Rural Development, 29-31 March
1999, Chiang Rai, Thailand, 11 p.
NAS. 1976. Making aquatic weeds useful: some
perspectives for developing countries. Report of an Ad
Hoc Panel of the Advisory Committee on Technology
Innovation, Board of Science and Technology
for International Development, Commission on
International Relations. National Academy of Sciences,
Washington D.C., 175 p.
NFAC. 1980. 1979 Annual Report. National Food and
Agriculture Council. Quezon City, Philippines.
Ni, D. and J. Wang. 1995. Material cycles and economic
returns in a rice-fi sh ecosystem, p. 177-182. In K.T.
MacKay (ed.) Rice-fi sh culture in China. International
Development Research Centre (IDRC), Ottawa, Canada,
276 p.
Nilsson, H. and D. Blariaux. 1994. Rice-cum-fi sh trials in
Luapula Province, Zambia. ALCOM Field Document No.
25. Aquaculture for Local Community Development
Programme, GCP/INT/555/SWE; GCP/RAF/277/BEL.
FAO. Harare, Zimbabwe, Jan. 1994, 23 p.
Nzamujo, F. 1995. Aquaculture in Benin, p. 413-421. In J.J.
Symoens and J.C. Micha (eds.) Proceedings of the
Seminar on “The Management of Integrated Agro-
Piscicultural Ecosystems in Tropical Areas”, 16-19 May
1994, Brussels. Tech. Centre for Agric. and Rural Co-op.
(CTA) Royal Academy of Overseas Sciences, Brussels,
587 p.
Pillay, T.V.R. 1990. Aquaculture principles and practices.
Fishing News Books, Oxford, England U.K., 575 p.
Pao, L.L. 1981. The present status of biocontrol of mosquitoes
in China, p. 54-77. In M. Laird (ed.) Biocontrol of medical
and veterinary pests. Praeger Inc.
Pongsuwana, V. 1962. Progress of rice fi eld fi sh culture in
Thailand. p. 157-163. Proceedings of the 10th Indo-
Pacifi c Fisheries Council (IPFC) Session II:157-163.
Purba, S. 1998. The economics of rice fi sh production
systems in North Sumatra, Indonesia: an empirical
and model analysis. Farming Systems and Resource
Economics in the Tropics, Vol. 31. Wissenschafverlag
Vauk Kiel KG.
Quyen, M.V., L.T. Duong, D.K. Son, P.N. Minh and N.D. Nghia.
1992. Rice fi eld aquaculture systems in the Mekong
Delta, Vietnam: potential and reality, p. 105-115. In
C.R. de la Cruz, C. Lightfoot, B.A. Costa-Pierce, V.R.
Carangal and M.P. Bimbao (eds.) Rice-fi sh research and
development in Asia. ICLARM Conf. Proc. 24, 457 p.
Ramsey, P. 1983. Rice-fi sh practices in Ifugao Province,
Philippines. ICLARM Newsl. 6(3):8.
Randriamiarana, H., A. Rabelahatra and J. Janssen. 1995.
Rice/fi sh farming in Madagascar: the present situation
and future prospects and constraints, p. 353-371. In
J.J. Symoens and J.C. Micha (eds.) Proceedings of the
Seminar on “The Management of Integrated Agro-
Piscicultural Ecosystems in Tropical Areas”, 16-19 May
1994, Brussels. Tech. Centre for Agric. and Rural Co-op.
(CTA) Royal Academy of Overseas Sciences, Brussels,
587 p.
76 FAO and The WorldFish Center | Culture of Fish in Rice Fields
Reissig, W.H., E.A. Heinrichs, J. Litsinger, K. Moody, L. Fiedler,
T.W. Mew and A.T. Barrion, 1986. Illustrated guide
to integrated pest management in rice in tropical
Asia. International Rice Research Institute, Los Banos,
Laguna, Philippines.
Roger, P.A. 1996. Biology and management of the fl oodwater
ecosystem in rice fi elds. International Rice Research
Institute, Los Banos, Laguna, Philippines, 250 p.
Rola, A.C and P.L. Pingali. 1993. Pesticides, rice productivity
and farmers health: an economic assessment.
International Rice Research Institute, Los Banos,
Laguna, Philippines, 100 p.
Rosario, W.R. 1984. Rice-fi sh culture, p. 189- 206. In
Philippines (BFAR) Freshwater Aquaculture Extension
Training Manual, Vol. III Technology. USAID-Bureau
of Fisheries and Aquatic Resources, Quezon City,
Philippines, 248 p.
Rothuis, A.J., D.K. Nhan, C.J.J. Richter and F. Ollevier. 1998a.
Rice with fi sh culture in the semi-deep waters of the
Mekong Delta, Vietnam: a socio-economical survey.
Aquaculture Research 29:47-57.
Rothuis, A.J., D.K. Nhan, C.J.J. Richter and F. Ollevier. 1998b.
Rice with fi sh culture in the semi-deep waters of
the Mekong Delta, Vietnam: Interaction of rice and
fi sh husbandry management on fi sh production.
Aquaculture Research 29:59-66.
Rothuis, A.J., L.T. Duong, C.J.J. Richter and F. Ollevier. 1998c.
Polyculture of silver barb, Puntius gonionotus, Nile
tilapia, Oreochromis niloticus (L.), and common carp,
Cyprinus carpio, in Vietnamese rice fi elds: feeding
ecology and impact on rice and rice fi eld environment.
Aquaculture Research 29:649-660.
Ruddle, K. and M. Prein. 1998. Assessing potential
nutritional and household economic benefi ts of
developing integrated farming systems. In J.A.
Mathias, A.T. Charles and H. Baotong (eds.) Integrated
fi sh farming. CRC Press, Boca Raton, New York.
Satari, G. 1962. Wet rice cultivation with fi sh culture. A
study on some agronomical aspects. Univ. Indonesia,
Bogor, Indonesia. 126 p. Ph.D. thesis (in Indonesian,
English summary).
Saturno, J.O. 1994. Comparative analysis of rice-fi sh
farming and rice monoculture in Nueva Ecija. 117 p.
Master’s thesis.
Schuster, W.H. 1955. Fish culture in conjunction with rice
cultivation. World Crops 7:11-14; 67-70.
Sevilleja, R.C. 1992. Rice-fi sh farming in the Philippines:
past, present and future, p. 77-90. In C.R. De la Cruz,
C. Lightfoot, B.A. Costa-Pierce, V.R. Carangal and M.P.
Bimbao (eds.) Rice-fi sh research and development in
Asia. ICLARM Conf. Proc. 24, 457 p.
Sevilleja, R.C., A.G. Cagauan, E.A. Lopez, C.R. de la Cruz
and A.A. van Dam. 1992. Component technology
research in rice-fi sh systems in the Philippines, p.
373-384. In C.R. De la Cruz, C. Lightfoot, B.A. Costa-
Pierce, V.R. Carangal and M.P. Bimbao (eds.) Rice-fi sh
research and development in Asia. ICLARM Conf.
Proc. 24, 457 p.
Shehadeh, Z.H. and M. Pedini 1997. Aquaculture
development research priorities in Africa and West
Asia, p. 43-46. In J.H. Annala (ed.) Fisheries and
aquaculture research planning needs for Africa and
West Asia, ICLARM Conf. Prof. 50, 80 p.
Singh, V.P., A.C. Early and T.H. Wickham 1980. Rice
agronomy in relation to fi sh culture, p. 15-34. In R.S.V.
Pullin and Z.H. Shehadeh (eds.) Proc. of the ICLARM-
SEARCA Conf. on Integrated Agriculture-Aquaculture
Farming Systems, 6-9 August 1979, Manila, Philippines,
258 p.
Spataru, P. and M. Zorn. 1976. Some aspects of natural
feed and feeding habits of Tilapia galilaea (Artedi)
and Tilapia aurea (Steindachner) in Lake Kinneret.
Bamidgeh 28:12-17.
Spataru, P. and M. Zorn. 1978. Food and feeding habits
of Tilapia aurea (Steindachner) (Cichlidae) in Lake
Kinneret (Israel). Aquaculture 13(1):67-79.
Syamsiah, I., S. Suriapermana and A.M. Fagi. 1992. Research
on rice-fi sh culture: past experiences and future
research program, p. 287-293. In C.R. De la Cruz, C.
Lightfoot, B.A. Costa-Pierce, V.R. Carangal and M.P.
Bimbao (eds.) Rice-fi sh research and development in
Asia. ICLARM Conf. Proc. 24, 457 p.
Tamura. 1961. Carp cultivation in Japan, p. 103-120. In G.
Borgstrom (ed.) Fish as Food. Vol. 1. Academic Press,
New York.
Thongpan, N., M. Singreuang, C. Thaila, S. Mankheng, S.
Kaeowsawat and J.D. Sollows. 1992. On-farm rice-
fi sh farming research in Ubon Province, Northeast
Thailand, p. 301-314. In C.R. De la Cruz, C. Lightfoot,
B.A. Costa-Pierce, V.R. Carangal and M.P. Bimbao (eds.)
Rice-fi sh research and development in Asia. ICLARM
Conf. Proc. 24, 457 p.
Tokrishna, R. 1995. Integration of agriculture, livestock and
fi sh farming in Thailand, p. 245-263. In J.J. Symoens
and J.C. Micha (eds.) Proceedings of the Seminar on
“The Management of Integrated Agro-Piscicultural
Ecosystems in Tropical Areas”, 16-19 May 1994,
Brussels. Tech. Centre for Agric. and Rural Co-op. (CTA)
Royal Academy of Overseas Sciences, Brussels, 587 p.
Torres, J., N.A. Macabale and J.R. Mercado. 1992. On-farm
rice-fi sh farming systems research in Guimba, Nueva
Ecija, Philippines, p. 295-300. In C.R. De la Cruz, C.
Lightfoot, B.A. Costa-Pierce, V.R. Carangal and M.P.
Bimbao (eds.) Rice-fi sh research and development in
Asia. ICLARM Conf. Proc. 24, 457 p.
Tuan, N.A. 1994. Culturing fi sh in rice fi elds as a means
of preventing insect infestations: preliminary
observations (Abstract), p. 31. In C.R. De la Cruz (ed.)
Role of fi sh in enhancing rice fi eld ecology and in
integrated pest management. Summary report of
the 3rd Asian Regional Rice-Fish Farming Research
and Development Workshop, 6 -11 June 1993,
77
References
Sukamandi Research Institute for Food Crops, West
Java, Indonesia. ICLARM Conf. Proc. 43, 50 p.
Tuong T.P., S.P. Kam, L. Wade, S. Pandey, B. Bouman and B.
Hardy (eds.) 2001. Characterising and understanding
rainfed environments. International Rice Research
Institute, Los Banos, Philippines.
Univ. of Durham. 1987. Hydrology of deepwater rice
fi elds in Bangladesh and Thailand: general report on
contracts for U.S. Overseas Dev. Adm. and Comm. of
the European Community, Univ. of Durham Dept. of
Botany, Durham, England, 57 p.
Van den Berg, F. 1996. Lessons from Madagascar. FAO
Aquaculture Newsletter12 (4):14-16.
Verheust L., E. Rurangwa and F. Ollevier. 1995. Tilapia-
Clarias polyculture in Rwanda with special emphasis
on the integration of pisciculture-agriculture
animal husbandry, p. 401-412. In J.J. Symoens and
J.C. Micha (eds.) Proceedings of the Seminar on
“The Management of Integrated Agro-Piscicultural
Ecosystems in Tropical Areas”, 16-19 May 1994,
Brussels. Tech. Centre for Agric. and Rural Co-op. (CTA)
Royal Academy of Overseas Sciences, Brussels, 587 p.
Villadolid, D.V. and P.A. Acosta. 1954. Raising tilapia for
food and profi t. Pamphlet Department of Agriculture
and Natural Resources, Bureau of Fisheries, Manila,
Philippines, 7 p.
Vincke, M.M.J. 1995. The present state of development
in continental aquaculture in Africa, p. 27-61. In
J.J.Symoens and J.-C. Micha (eds.) Proceedings of the
Seminar on “The Management of Integrated Agro-
Piscicultural Ecosystems in Tropical Areas”, 16-19 May
1994. Tech. Center for Agric. and Rural Co-op. (CTA)
Royal Academy of Overseas Sciences, Brussels, 587 p.
Waibel, H. 1992. Comparative economics of pesticide use
in rice and rice-fi sh farming, p. 245-254. In C.R. De la
Cruz, C. Lightfoot, B.A. Costa-Pierce, V.R. Carangal and
M.P. Bimbao (eds.) Rice-fi sh research and development
in Asia. ICLARM Conf. Proc. 24, 457 p.
Wan, B. and Q. Zhang. 1995. New techniques for raising
fi sh in fl ooded rice fi elds, p. 85-90. In K.T. MacKay (ed.)
Rice-fi sh culture in China. International Development
Research Centre (IDRC), Ottawa, Canada, 276 p.
Wan, Q., K. Li, P. Li, H. Gu and X. Zhou. 1995. Developing rice-
fi sh culture in shallow waters in lakes, p. 67-76. In K.T.
MacKay (ed.) Rice-fi sh culture in China. International
Development Research Centre (IDRC), Ottawa,
Canada, 276 p.
Wang, J. and D. Ni, 1995. A comparative study of the
ability of fi sh to catch mosquito larva, p. 217-222.
In K.T. MacKay (ed.), 1995. Rice-fi sh culture in China.
International Development Research Centre (IDRC)
Ottawa, Canada, 276 p.
Wang, B. and Q. Zhang. 1995. Cultivating different breeds
of fi sh in rice fi elds, p. 139-146. In K.T. MacKay (ed.)
Rice-fi sh culture in China. International Development
Research Centre (IDRC), Ottawa, Canada, 276 p.
Wang, G. and Song Q. 1999. Culture of Chinese mitten-
handed crab. Aquaculture Asia 6(2):18-21.
Wassef, E.A. 2000. Status of aquaculture in Egypt. World
Aquaculture 31 (1):29- 32, 60-61.
Weatherley, A.H. and B.M.G. Cogger. 1977. Fish culture:
problems and prospects. Science 197(4302):427-430.
Willmann, R., M. Halwart and U. Barg, 1998. Integrating
fi sheries and agriculture to enhance fi sh production
and food security. FAO Aquaculture Newsletter, Dec.
1998, No. 20:3-8.
Wu, L. 1995. Methods of rice-fi sh culture and their
ecological effi ciency, p. 91-96. In K.T. MacKay (ed.)
Rice-fi sh culture in China. International Development
Research Centre (IDRC), Ottawa, Canada, 276 p.
Xiao, P. 1995. Fish culture in rice fi elds: rice-fi sh symbiosis,
p. 183-188. In K.T. MacKay (ed.) Rice-fi sh culture in
China. International Development Research Centre
(IDRC), Ottawa, Canada, 276 p.
Xiao, Q.Y. 1992. Role of fi sh in pest control in rice farming,
p. 235-244. In C.R. De la Cruz, C. Lightfoot, B.A. Costa-
Pierce, V.R. Carangal and M.P. Bimbao (eds.) Rice-fi sh
research and development in Asia. ICLARM Conf. Proc.
24, 457 p.
Xu, S. 1995b. The development of rice-fi sh farming in
Chongqing City, p. 43-48. In K.T. MacKay (ed.) Rice-fi sh
culture in China. International Development Research
Centre (IDRC), Ottawa, Canada, 276 p.
Yan, D., J. Ping, Z. Wenliang, Z. Chuanlu and W. Yingduo.
1995a. Ridge-cultured rice integrated with fi sh
farming in trenches, Anhui Province, p. 97-102. In K.T.
MacKay (ed.) Rice-fi sh culture in China. International
Development Centre (IDRC), Ottawa, Canada, 276 p.
Yang, G., Q. Xiao and T. He. 1995. Techniques adopted
in rice-azolla-fi sh system with ridge culture, p. 107-
116. In K.T. MacKay (ed.) Rice-fi sh culture in China.
International Development Research Centre (IDRC),
Ottawa, Canada. 276 p.
Yap, W.G. 1998. Who’s afraid of El Niño? Aquaculture Asia
3(1) 45-49.
Yu, S.Y., W.S. Wu, H.F. Wei, D.A. Ke, J.R. Xu, and Q.Z. Wu.
1995. Ability of fi sh to control rice diseases, pests, and
weeds, p. 223-228. In K.T. MacKay (ed.) Rice-fi sh culture
in China. International Development Research Centre
(IDRC), Ottawa, Canada, 276 p.
Yunus, M., A. Hardjamulia, I. Syamsiah and S. Suriapermana.
1992. Evaluation of rice-fi sh production systems in
Indonesia, p. 131-138. In C.R. De la Cruz, C. Lightfoot,
B.A. Costa-Pierce, V.R. Carangal and M.P. Bimbao (eds.)
Rice-fi sh research and development in Asia. ICLARM
Conf. Proc. 24, 457 p.
Zhang, R. 1995. Scientifi c and technological development
of rice-fi sh culture in China, p. 23-30. In K.T. MacKay (ed.)
Rice-fi sh culture in China. International Development
Research Centre (IDRC), Ottawa, Canada, 276 p.
